Abstract

We report a rapid and reliable two-step multiplex polymerase chain reaction (PCR) assay to identify the 10 Bacteroides fragilis group species –Bacteroides caccae, B. distasonis, B. eggerthii, B. fragilis, B. merdae, B. ovatus, B. stercoris, B. thetaiotaomicron, B. uniformis and B. vulgatus. These 10 species were first divided into three subgroups by multiplex PCR-G, followed by three multiplex PCR assays with three species-specific primer mixtures for identification to the species level. The primers were designed from nucleotide sequences of the 16S rRNA, the 16S–23S rRNA intergenic spacer region and part of the 23S rRNA gene. The established two-step multiplex PCR identification scheme was applied to the identification of 155 clinical isolates of the B. fragilis group that were previously identified to the species level by phenotypic tests. The new scheme was more accurate than phenotypic identification, which was accurate only 84.5% of the time. The multiplex PCR scheme established in this study is a simple, rapid and reliable method for the identification of the B. fragilis group species. This will permit more accurate assessment of the role of various B. fragilis group members in infections and of the degree of antimicrobial resistance in each of the group members.

Introduction

The Bacteroides fragilis group, anaerobic, bile-resistant, non-spore-forming, Gram-negative rods are part of the endogenous human bowel flora [1]. Species most frequently isolated from that flora are Bacteroides vulgatus, B. thetaiotaomicron, B. distasonis and, less frequently, B. eggerthii and B. fragilis[2]. The B. fragilis group is commonly associated with a variety of human infections, such as intra-abdominal abscesses, wound infections and bacteremia [3]. The B. fragilis group bacteremia contributes significantly to morbidity and mortality [4]. The distribution of individual species in clinical infections is as follows: B. fragilis accounts for 63% of all the group isolates, B. thetaiotaomicron for 14%, B. vulgatus and B. ovatus for 7% each, B. distasonis for 6% and B. uniformis for 2% [5]. The choice of antibiotics for therapy is limited because the species of the B. fragilis group are among the most resistant of all anaerobes to antimicrobial agents, and this resistance has increased recently [6,7]. Accordingly, there is need for rapid, accurate identification of clinical isolates to permit an early, effective management of infected patients.

Phenotypic methods have been used in clinical laboratories to differentiate Bacteroides spp. These are time-consuming, and it may be difficult to differentiate Bacteroides species with them [8,9]. Automated methods currently used are also unreliable for identifying isolates of the B. fragilis group [10,11]. Incorrect identification of strains of the B. fragilis group may result in inappropriate antibiotic therapy.

Recently, genotypic-based techniques are emerging as alternatives or complements to phenotypic methods. Hybridization assays using a DNA or an RNA probe [12,13], polymerase chain reaction (PCR) amplification using species-specific primers [14–16], rRNA restriction fragment length polymorphism [17], restriction endonuclease analysis [18], ribotyping [19], arbitrary primer PCR [20] and intergenic spacer region (ISR) PCR [21] have been used to study the B. fragilis species. However, for other species in the B. fragilis group, such as B. thetaiotaomicron, B. ovatus, B. vulgatus and B. distasonis, that are also clinically important, there has been little work carried out on molecular diagnosis.

With multiplex PCR, more than one locus is simultaneously amplified in the same reaction. It has the potential to save considerable time and effort for the laboratory without compromising test utility. Since its introduction, multiplex PCR has been successfully applied in many bacterial identifications [22–25]. In the present study, based on the sequence analysis of the 16S rRNA gene, the 16S–23S rRNA ISR and a variable region of the 23S rRNA gene of the 10 ATCC type strains representing the 10 B. fragilis group species, a two-step multiplex PCR identification scheme was established to rapidly and accurately identify the B. fragilis group species.

Materials and methods

Bacterial strains and culture conditions

Strains used in this study included 10 ATCC strains representing the 10 B. fragilis group species and 155 Bacteroides isolates previously recovered from clinical specimens and identified phenotypically (Tables 1 and 2). The clinical isolates were chosen to represent 10 commonly isolated Bacteroides species. In addition, four ATCC strains of Bacteroides species other than the B. fragilis group and 47 ATCC or NCTC strains of Gram-negative anaerobic bacilli other than Bacteroides species that are either phylogenetically related species or may grow on Bacteroides bile esculin agar were used to verify the specificity of the established multiplex PCR assay (Tables 1 and 2). All strains were cultured anaerobically overnight on Brucella blood agar (Anaerobe Systems, CA, USA) at 37°C and were characterized by a combination of conventional tests as described in the Wadsworth Anaerobe Manual [26] and the BD BBL Crystal™ Identification System (Becton Dickinson Microbiology Systems, MD, USA).

1

List of strains used in this study and the multiplex PCR results

Strain Strain No. Accession No. Bac-F G23S-1 G23S-1 1392A 1392A 1392A 1392A 1392A Begg-F Buni-F Bste-F Bdis-F Bmer-F 
   Bac-R Bfr-G2 Bfr-G3 Bth-R Bvul-R Bfra-R Bcac-R Bova-R 23R4 23R4 23R4 Bdis-R Bmer-R 
B. thetaiotaomicron ATCC29148T AY155588 − − − − − − − − − − − 
B. vulgatus ATCC8482T AY155596 − − − − − − − − − − − 
B. fragilis ATCC25285T AY155459 − − − − − − − − − − − 
B. caccae ATCC43185T AY155590 − − − − − − − − − − − 
B. ovatus ATCC8843T AY155589 − − − − − − − − − − − 
B. eggerthii ATCC27754T AY155591 − − − − − − − − − − 
B. uniformis ATCC8492T AY155592 − − − − − − − − − − 
B. stercoris ATCC43183T AY155593 − − − − − − − − − − 
B. distasonis ATCC8503T AY155594 − − − − − − − − − − 
B. merdae ATCC43184T AY155595 − − − − − − − − − − 
B. splanchnicus ATCC29572T  − − − − − − − − − − − − 
B. tectus ATCC43331T  − − − − − − − − − − − − 
Pr. zoogleoformans ATCC33285T  − − − − − − − − − − − − 
Po. cansulci NCTC12858T  − − − − − − − − − − − − 
Po. endodontalis ATCC35406T  − − − − − − − − − − − − 
Po. gingivalis ATCC33277T  − − − − − − − − − − − − 
Strain Strain No. Accession No. Bac-F G23S-1 G23S-1 1392A 1392A 1392A 1392A 1392A Begg-F Buni-F Bste-F Bdis-F Bmer-F 
   Bac-R Bfr-G2 Bfr-G3 Bth-R Bvul-R Bfra-R Bcac-R Bova-R 23R4 23R4 23R4 Bdis-R Bmer-R 
B. thetaiotaomicron ATCC29148T AY155588 − − − − − − − − − − − 
B. vulgatus ATCC8482T AY155596 − − − − − − − − − − − 
B. fragilis ATCC25285T AY155459 − − − − − − − − − − − 
B. caccae ATCC43185T AY155590 − − − − − − − − − − − 
B. ovatus ATCC8843T AY155589 − − − − − − − − − − − 
B. eggerthii ATCC27754T AY155591 − − − − − − − − − − 
B. uniformis ATCC8492T AY155592 − − − − − − − − − − 
B. stercoris ATCC43183T AY155593 − − − − − − − − − − 
B. distasonis ATCC8503T AY155594 − − − − − − − − − − 
B. merdae ATCC43184T AY155595 − − − − − − − − − − 
B. splanchnicus ATCC29572T  − − − − − − − − − − − − 
B. tectus ATCC43331T  − − − − − − − − − − − − 
Pr. zoogleoformans ATCC33285T  − − − − − − − − − − − − 
Po. cansulci NCTC12858T  − − − − − − − − − − − − 
Po. endodontalis ATCC35406T  − − − − − − − − − − − − 
Po. gingivalis ATCC33277T  − − − − − − − − − − − − 

Genera are abbreviated as: B., Bacteroides; Pr., Prevotella; Po., Porphyromonas.

GenBank accession numbers for the partial 23S rRNA gene.

Sequences of the 16S–23S rRNA ISR were determined in this study. The GenBank accession numbers are AY153428, AY153429, AY153427 for B. caccae, B. stercoris and B. merdae, respectively.

2

Other strains used in this study

Strain Strain No. No. of strains 
Bacteroides putredinis ATCC29800T 
Bacteroides ureolyticus ATCC33387T 
Bilophila wadsworthia ATCC51581 
Bilophila wadsworthia ATCC49260T 
Campylobacter rectus ATCC33238T 
Eubacterium sulci ATCC35585T 
Fusobacterium gonidiaformans ATCC25563T 
Fusobacterium mortiferum ATCC9817 
Fusobacterium naviforme ATCC25832T 
Fusobacterium necrogenes ATCC25556T 
Fusobacterium necrophorum subsp. necrophorum ATCC25286T 
Fusobacterium nucleatum subsp. nucleatum ATCC25586T 
Fusobacterium nucleatum subsp. polymorphum ATCC10953T 
Fusobacterium perfoetens ATCC29250T 
Fusobacterium periodonticum ATCC33693T 
Fusobacterium pseudonecrophorum ATCC51644T 
Fusobacterium russii ATCC25533T 
Fusobacterium simiae ATCC33568T 
Fusobacterium ulcerans ATCC49185T 
Fusobacterium varium ATCC8501T 
Fusobacterium varium ATCC27725 
Mitsuokella multiacida ATCC27723T 
Mobiluncus curtisii subsp. curtisii ATCC35241T 
Mobiluncus mulieris ATCC35243T 
Porphyromonas asaccharolytica ATCC25260T 
Porphyromonas cangingivalis NCTC12856T 
Porphyromonas canoris NCTC12835T 
Porphyromonas gingivalis ATCC49417 
Porphyromonas levii ATCC29147T 
Porphyromonas macacae ATCC33141T 
Porphyromonas salivosus NCTC11632 
Prevotella bivia ATCC29303T 
Prevotella buccae ATCC33574T 
Prevotella buccalis ATCC35310T 
Prevotella corporis ATCC33547T 
Prevotella denticola ATCC33185 
Prevotella disiens ATCC29426T 
Prevotella intermedia ATCC25611T 
Prevotella loescheii ATCC15930T 
Prevotella melaninogenica ATCC25845T 
Prevotella nigrescens ATCC33563T 
Prevotella oralis ATCC33269T 
Prevotella oulorum ATCC43324T 
Prevotella veroralis ATCC33779T 
Sutterella wadsworthii ATCC51579T 
Bacteroides caccae Clinical isolates 12 
Bacteroides distasonis/merdae Clinical isolates 25 
Bacteroides eggerthii Clinical isolates 
Bacteroides fragilis Clinical isolates 51 
Bacteroides ovatus Clinical isolates 13 
Bacteroides stercoris Clinical isolates 
Bacteroides thetaiotaomicron Clinical isolates 20 
Bacteroides uniformis Clinical isolates 12 
Bacteroides vulgatus Clinical isolates 15 
Strain Strain No. No. of strains 
Bacteroides putredinis ATCC29800T 
Bacteroides ureolyticus ATCC33387T 
Bilophila wadsworthia ATCC51581 
Bilophila wadsworthia ATCC49260T 
Campylobacter rectus ATCC33238T 
Eubacterium sulci ATCC35585T 
Fusobacterium gonidiaformans ATCC25563T 
Fusobacterium mortiferum ATCC9817 
Fusobacterium naviforme ATCC25832T 
Fusobacterium necrogenes ATCC25556T 
Fusobacterium necrophorum subsp. necrophorum ATCC25286T 
Fusobacterium nucleatum subsp. nucleatum ATCC25586T 
Fusobacterium nucleatum subsp. polymorphum ATCC10953T 
Fusobacterium perfoetens ATCC29250T 
Fusobacterium periodonticum ATCC33693T 
Fusobacterium pseudonecrophorum ATCC51644T 
Fusobacterium russii ATCC25533T 
Fusobacterium simiae ATCC33568T 
Fusobacterium ulcerans ATCC49185T 
Fusobacterium varium ATCC8501T 
Fusobacterium varium ATCC27725 
Mitsuokella multiacida ATCC27723T 
Mobiluncus curtisii subsp. curtisii ATCC35241T 
Mobiluncus mulieris ATCC35243T 
Porphyromonas asaccharolytica ATCC25260T 
Porphyromonas cangingivalis NCTC12856T 
Porphyromonas canoris NCTC12835T 
Porphyromonas gingivalis ATCC49417 
Porphyromonas levii ATCC29147T 
Porphyromonas macacae ATCC33141T 
Porphyromonas salivosus NCTC11632 
Prevotella bivia ATCC29303T 
Prevotella buccae ATCC33574T 
Prevotella buccalis ATCC35310T 
Prevotella corporis ATCC33547T 
Prevotella denticola ATCC33185 
Prevotella disiens ATCC29426T 
Prevotella intermedia ATCC25611T 
Prevotella loescheii ATCC15930T 
Prevotella melaninogenica ATCC25845T 
Prevotella nigrescens ATCC33563T 
Prevotella oralis ATCC33269T 
Prevotella oulorum ATCC43324T 
Prevotella veroralis ATCC33779T 
Sutterella wadsworthii ATCC51579T 
Bacteroides caccae Clinical isolates 12 
Bacteroides distasonis/merdae Clinical isolates 25 
Bacteroides eggerthii Clinical isolates 
Bacteroides fragilis Clinical isolates 51 
Bacteroides ovatus Clinical isolates 13 
Bacteroides stercoris Clinical isolates 
Bacteroides thetaiotaomicron Clinical isolates 20 
Bacteroides uniformis Clinical isolates 12 
Bacteroides vulgatus Clinical isolates 15 

Same as Sutterella wadsworthensis.

Sequencing of the 16S–23S rRNA ISR and partial 23S rRNA gene

The primer pair G23S-1 and G23S-2 (Table 3) [27], corresponding to the positions 1508–1525 and 2114–2132 of the 23S rDNA gene of Escherichia coli, was used to amplify partial 23S rRNA genes of 10 ATCC type strains of the B. fragilis group. PCR amplification was performed as follows: one or two colonies of bacterial strains were suspended in 50 µl of Tris–HCl–EDTA saline (pH 8.0), incubated for 10 min at 95°C and centrifuged at 18 600×g for 2 min to obtain the DNA as the PCR template. PCR amplification was performed in 50 µl of reaction mixture containing 1.25 U of Taq polymerase (Promega, WI, USA), 50 mM KCL, 10 mM Tris–HCl (pH 9.0), 0.1% Triton, 2.5 mM MgCl2, 0.5 mM (each) primer, 0.2 mM dNTPs, and 3 µl of the bacterial lysate as the DNA template. After initial denaturation for 2 min at 95°C, the reactions were subjected to 35 cycles comprising 30 s at 95°C, 30 s at 52°C, and 1 min at 72°C, with a final extension at 72°C for 5 min.

3

Oligonucleotide primers used in this study

Primer Sequence (5′–3′) Reference 
16S-2 AGTCGTAACAAGGTARCCGTA [21] 
23R4 GGGTTBCCCCATTCGG [21] 
G23S-1 GTTGGCTTAGAAGCAGC [27] 
G23S-2 CATTTTGCCGAGTTCCTT [27] 
Bfr-F CTGAACCAGCCAAGTAGCG present study 
Bfr-R CCGCAAACTTTCACAACTGACTTA present study 
Bfr-G2 ATCAGGTTCGACTCTTGCT present study 
Bfr-G3 CCGTCAGCTGGCAGGA present study 
1392A GTACACACCGCCCGT present study 
Bth-R ACCTATGAAATCGTTGTTACG present study 
Bvul-R GGCTTCTTACTTTCTCTCTTCCG present study 
Bfra-R GCTAATCCCCCAATCATAC present study 
Bcac-R TCGTTTCCCATTGCTGG present study 
Bova-R AATAATGCGTACTCGAACAC present study 
Begg-F GTCATATTAACGGTGGCG present study 
Buni-F TCCGTTTTCCACTTATAAGA present study 
Bste-F CTACGACATAGTCTTGGTGAG present study 
Bdis-F TGATCCCTTGTGCTGCT present study 
Bdis-R ATCCCCCTCATTCGGA present study 
Bmer-F GAGGTATGTAGCTCTCTGGTA present study 
Bmer-R TTTTTACCCCTTACGGAG present study 
Primer Sequence (5′–3′) Reference 
16S-2 AGTCGTAACAAGGTARCCGTA [21] 
23R4 GGGTTBCCCCATTCGG [21] 
G23S-1 GTTGGCTTAGAAGCAGC [27] 
G23S-2 CATTTTGCCGAGTTCCTT [27] 
Bfr-F CTGAACCAGCCAAGTAGCG present study 
Bfr-R CCGCAAACTTTCACAACTGACTTA present study 
Bfr-G2 ATCAGGTTCGACTCTTGCT present study 
Bfr-G3 CCGTCAGCTGGCAGGA present study 
1392A GTACACACCGCCCGT present study 
Bth-R ACCTATGAAATCGTTGTTACG present study 
Bvul-R GGCTTCTTACTTTCTCTCTTCCG present study 
Bfra-R GCTAATCCCCCAATCATAC present study 
Bcac-R TCGTTTCCCATTGCTGG present study 
Bova-R AATAATGCGTACTCGAACAC present study 
Begg-F GTCATATTAACGGTGGCG present study 
Buni-F TCCGTTTTCCACTTATAAGA present study 
Bste-F CTACGACATAGTCTTGGTGAG present study 
Bdis-F TGATCCCTTGTGCTGCT present study 
Bdis-R ATCCCCCTCATTCGGA present study 
Bmer-F GAGGTATGTAGCTCTCTGGTA present study 
Bmer-R TTTTTACCCCTTACGGAG present study 

The 16S–23S rDNA ISR regions of B. merdae, B. ovatus, and B. stercoris were amplified as described previously [21], as there were no sequence data of these species in the GenBank. The PCR was carried out as described above except the annealing temperature was 58°C.

The 16S rRNA gene fragments were amplified as previously described [28]. Briefly, almost the full length of the 16S rRNA gene was amplified by using two pairs of primers (8UA and 907B; 774A and 1485B). The PCR was carried out as described above except the annealing temperature was 45°C.

The major PCR products were excised from a 1% agarose gel after electrophoresis, were purified using a QIAquick Gel Extraction kit (Qiagen Inc., Chatsworth, CA, USA) and were sequenced directly with a Biotech Diagnostic (Biotech Diagnostic, CA, USA) Big Dye Sequencing kit on an ABI 377 sequencer (Applied Biosystems, Foster City, CA, USA). Sequences obtained have been registered at the GenBank; the accession numbers are listed in Table 1. All sequences were analyzed by multialignment using CLUSTAL-W (http://genome.kribb.re.kr).

Development of group-specific and species-specific primers

Based on the multialignment analysis data, a potential B. fragilis group-specific primer pair, Bfr-F and Bfr-R, was selected from the 16S rRNA gene, and two potential subgroup-specific downstream primers, BFR-G2 (for subgroup-II which includes B. eggerthii, B. stercoris and B. uniformis) and BFR-G3 (for subgroup-III which includes B. distasonis and B. merdae), were selected from the 23S rRNA gene. In addition, 10 species-specific primer pairs were designed from the 16S–23S rRNA ISR regions (Table 3). The primer sequences were analyzed for secondary structure formation, G+C content, and primer–dimer formation with the NetPrimer analysis software (http://www.premierbiosoft.com/netprimer/netprlaunch/netprlaunch.html). The specificities of these primers were predicted by comparison to the aligned SSU_rRNA database of the Ribosomal Database Project using the CHECK_PROBE utility [29]. These primers were designed with minimal differences in their annealing temperature within each primer set, and to yield amplification products that ranged between 200 and 700 bp and differed by at least 50 bp. The relative locations of the primers in the E. coli rRNA gene sequence are indicated in Fig. 1.

1

The alignment of the approximate location of the PCR primers and amplicon sizes of each PCR amplification system; multiplex PCR-G for subgrouping the B. fragilis group species, multiplex PCR-I, multiplex PCR-II, and multiplex PCR-III for identification to the species level. The primer set used for multiplex PCR-G comprised primers Bfr-F, Bfr-R, G23S-1, Bfr-G2 and Bfr-G3. The primer set used for multiplex PCR-I comprised primers 1392A, Bth-R, Bvul-R, Bfra-R, Bcac-R and Bova-R. The primer set used for multiplex PCR-II comprised primers Begg-F, Buni-F, Bste-F and 23R4. The primer set used for multiplex PCR-III comprised primers Bdis-F, Bdis-R, Bmer-F and Bmer-R. Arrows indicate the direction of primers.

1

The alignment of the approximate location of the PCR primers and amplicon sizes of each PCR amplification system; multiplex PCR-G for subgrouping the B. fragilis group species, multiplex PCR-I, multiplex PCR-II, and multiplex PCR-III for identification to the species level. The primer set used for multiplex PCR-G comprised primers Bfr-F, Bfr-R, G23S-1, Bfr-G2 and Bfr-G3. The primer set used for multiplex PCR-I comprised primers 1392A, Bth-R, Bvul-R, Bfra-R, Bcac-R and Bova-R. The primer set used for multiplex PCR-II comprised primers Begg-F, Buni-F, Bste-F and 23R4. The primer set used for multiplex PCR-III comprised primers Bdis-F, Bdis-R, Bmer-F and Bmer-R. Arrows indicate the direction of primers.

Identification of the B. fragilis group species by two-step multiplex PCR assays

Ten species of the B. fragilis group were first grouped by multiplex PCR (designated multiplex PCR-G) and then further identified to the species level by three multiplex PCR assays (named multiplex PCR-I, multiplex PCR-II and multiplex PCR-III).

PCR amplification was carried out in a total volume of 50 µl containing 1.25 U of Taq polymerase (Promega, WI, USA), 50 mM KCL, 10 mM Tris–HCl (pH 9.0), 0.1% Triton, 2.5 mM MgCl2, 0.5 mM (each) primer, 0.2 mM dNTPs, and 3 µl of bacterial lysate as the DNA template. PCR was carried out for 35 cycles. Each cycle consisted of 95°C for 20 s for denaturation, annealing for 1 min at 52°C for multiplex PCR-G, 62°C for multiplex PCR-I, 60°C for multiplex PCR-II, and 55°C for multiplex PCR-III; extension was performed at 72°C for 30 s. A cycle of 72°C for 5 min was added to the final extension. PCR products were analyzed by electrophoresis on a 6% polyacrylamide gel followed by ethidium bromide staining.

Sensitivity of multiplex PCR assays

The sensitivities of the multiplex PCR assays were evaluated by titrating cultures of 10 ATCC strains of the B. fragilis group species (CFU 106). We made serial 10-fold dilutions of cultures with Tris–HCl (pH 7.5) and plated equal volumes (100 ml) of dilutions onto Brucella blood agars. The cultures and the dilutions were taken for DNA preparation and subsequent multiplex PCR assays. Colonies were counted after 3 days of incubation. The detection limits of multiplex PCR assays were determined with known numbers of bacteria diluted in Tris–HCl (pH 7.5).

Results and discussion

Members of the B. fragilis group are the anaerobes most commonly recovered from clinical specimens and the most resistant to antimicrobials. The members of this group exhibit species-to-species variability in both virulence and drug resistance. PCR assays have been developed to identify various microorganisms, including Bacteroides species [14–16]. However, they target only one species; therefore, a large number of individual PCR assays may be required. In this study, using the sequences of rRNA genes as targets for specific primer selection, we developed a two-step multiplex PCR that allows rapid detection of the 10 B. fragilis group species; first, the 10 species of the B. fragilis group are differentiated into three subgroups by one multiplex PCR (multiplex PCR-G), followed by one multiplex PCR assay for each subgroup for species identification.

Although Bacteroides species have been classified phylogenetically on the basis of the 16S rRNA sequence similarity, it is not useful for differentiation of Bacteroides species because sequence diversity, except for B. distasonis, is less than 10%. The 16S–23S rRNA ISRs are more variable within Bacteroides species. The present study showed that the ISR sequence similarity ranged from 56.4 to 78.2%, lower than the 16S rRNA sequence similarity (84.2–95.4%) among Bacteroides species. The ISR sequence expresses genetic diversity among Bacteroides species better than the 16S rRNA sequence. Therefore, the ISR sequences of Bacteroides species that are available from the GenBank database and those determined in this study were used for species-specific primer selection. A part of the 23S rRNA gene was sequenced in this study, and the sequences were used for designing subgroup-specific primers that were unique to each subgroup as well as common within the species of each subgroup. In addition, a Bacteroides genus-specific primer pair was selected from the16S rRNA gene.

Based on our two-step multiplex PCR strategy, the 10 B. fragilis group species were first differentiated into three subgroups by multiplex PCR-G using a primer set comprising Bfr-F, Bfr-R, Bfr-G2, Bfr-G3, and G23S-1. Multiplex PCR-G produced a DNA fragment of about 230 bp from all 10 ATCC strains, as expected (Fig. 2, lanes 1–10). B. caccae, B. fragilis, B. ovatus, B. thetaiotaomicron and B. vulgatus were identified as subgroup-I by producing only this common DNA fragment of 230 bp (Fig. 2, lanes 1–5). B. eggerthii, B. stercoris, and B. uniformis were identified as subgroup-II by yielding a specific DNA fragment 450 bp in size in addition to the common band of 230 bp (Fig. 2, lanes 6–8), and similarly, B. distasonis and B. merdae were identified as subgroup-III by the occurrence of an additional 400-bp PCR product (Fig. 2, lanes 9 and 10). The slight size differences of the PCR products are due to length polymorphism found in amplified gene portions. The specificity of the multiplex PCR-G was evaluated by testing four ATCC strains of the non-B. fragilis group Bacteroides species and 47 ATCC strains representing closely or more distantly related Gram-negative anaerobic bacilli. Of the 51 ATCC or NCTC strains tested, six strains: B. splanchnicus ATCC 29572T, B. tectus ATCC 43331T, Porphyromonas endodontalis ATCC 35406T, Prevotella gingivalis ATCC 33277T, Prevotella cansulci NCTC 12858T and Prevotella zoogleoformans ATCC 33285T, also produced amplicons corresponding in size to the common band of 230 bp, but no others did (data not shown).

2

Polyacrylamide gel electrophoresis of PCR products from multiplex PCR assays. Lane M: 50-bp DNA ladder; lanes 1–5: subgroup-I species with an amplicon of ca. 230 bp (B. caccae ATCC 43185T, B. fragilis ATCC 25285T, B. ovatus ATCC 8483T, B. thetaiotaomicron ATCC 29148T and B. vulgatus ATCC 8482T, respectively); lane 6–8: subgroup-II species with an amplicon of ca. 450 bp (B. eggerthii ATCC 27754T, B. stercoris ATCC 43183T, and B. uniformis ATCC 8492T, respectively); lanes 9 and 10: subgroup-III species with an amplicon of ca. 400 bp (B. distasonis ATCC 8503T and B. merdae ATCC 43184T); lanes 11–15: PCR products from multiplex PCR-I, B. thetaiotaomicron ATCC 29184T (ca. 180 bp), B. vulgatus ATCC 8482T (ca. 250 bp), B. fragilis ATCC 25285T (ca. 420 bp), B. caccae ATCC 43185T (ca. 500 bp) and B. ovatus ATCC 8483T (ca. 610 bp), respectively; lanes 16–18: PCR products from multiplex PCR-II, B. eggerthii ATCC 27754T (ca. 250 bp), B. uniformis ATCC 8492T (ca. 350 bp) and B. stercoris ATCC 43183T (ca. 400 bp), respectively; lanes 19 and 20: PCR products from multiplex PCR-III, B. distasonis ATCC 8503T (ca. 220 bp) and B. merdae ATCC 43184T (ca. 310 bp), respectively.

2

Polyacrylamide gel electrophoresis of PCR products from multiplex PCR assays. Lane M: 50-bp DNA ladder; lanes 1–5: subgroup-I species with an amplicon of ca. 230 bp (B. caccae ATCC 43185T, B. fragilis ATCC 25285T, B. ovatus ATCC 8483T, B. thetaiotaomicron ATCC 29148T and B. vulgatus ATCC 8482T, respectively); lane 6–8: subgroup-II species with an amplicon of ca. 450 bp (B. eggerthii ATCC 27754T, B. stercoris ATCC 43183T, and B. uniformis ATCC 8492T, respectively); lanes 9 and 10: subgroup-III species with an amplicon of ca. 400 bp (B. distasonis ATCC 8503T and B. merdae ATCC 43184T); lanes 11–15: PCR products from multiplex PCR-I, B. thetaiotaomicron ATCC 29184T (ca. 180 bp), B. vulgatus ATCC 8482T (ca. 250 bp), B. fragilis ATCC 25285T (ca. 420 bp), B. caccae ATCC 43185T (ca. 500 bp) and B. ovatus ATCC 8483T (ca. 610 bp), respectively; lanes 16–18: PCR products from multiplex PCR-II, B. eggerthii ATCC 27754T (ca. 250 bp), B. uniformis ATCC 8492T (ca. 350 bp) and B. stercoris ATCC 43183T (ca. 400 bp), respectively; lanes 19 and 20: PCR products from multiplex PCR-III, B. distasonis ATCC 8503T (ca. 220 bp) and B. merdae ATCC 43184T (ca. 310 bp), respectively.

After successfully grouping 10 ATCC strains of the B. fragilis group into three subgroups by multiplex PCR-G, these 10 strains were identified to the species level by using one of the three two-step multiplex PCR assays. Five strains of subgroup-I were identified to the species level by producing major specific DNA fragments (B. thetaiotaomicron 180 bp; B. vulgatus 250 bp; B. fragilis, 420 bp; B. caccae, 500 bp; B. ovatus, 610 bp), as expected by multiplex PCR-I (Fig. 2, lanes 11–15). B. fragilis produced an additional amplicon (ca. 650 bp) (Fig. 2, lane 13) other than the expected amplicon. Three strains of subgroup-II were identified by yielding signature DNA fragments (B. eggerthii, 250 bp; B. uniformis 350 bp; B. stercoris, 400 bp) by multiplex PCR-II (Fig. 2, lanes 16–18). B. eggerthii also produced an amplicon corresponding in size to that of B. stercoris, but was distinguished from B. stercoris by its unique amplicon (250 bp) (Fig. 2, lane 16). Two strains of subgroup-III were identified as their corresponding species by yielding signature DNA fragments (B. distasonis, 220 bp, and B. merdae, 310 bp, by multiplex PCR-III) (Fig. 2, lanes 19 and 20). B. distasonis produced an additional amplicon (ca. 250 bp) other than the expected amplicon (Fig. 2, lane 19). The specificities of these two-step multiplex PCR assays were verified by PCR amplification with DNA samples of the 51 ATCC or NCTC type strains or reference strains mentioned above. All these multiplex PCR assays generated no amplicons with DNA other than the target organisms. Although six ATCC strains of the non-B. fragilis group species were differentiated into subgoup-I by multiplex PCR-G, they generated no amplicon by multiplex PCR-I for species identification.

All DNA templates used in PCR reactions were obtained by heating cells at 95°C. The multiplex PCR assays detected between 50 and 500 CFU of each species. By decreasing the annealing temperature and increasing the magnesium concentration, the sensitivities of our procedures could be increased to 10 cells but resulted in weak cross-reactivity and non-specific amplification bands (data not shown).

The established scheme for identification of human clinical isolates was evaluated with 155 clinical isolates of the B. fragilis group that were identified to the species level by phenotypic tests previously. One hundred and thirty-one (84.5%) of the strains had concordant results between the original and PCR-based identifications, but the other 24 strains showed discordance (Table 4). B. fragilis (n=51), B. distasonis (n=19), B. eggerthii (n=4), B. stercoris (n=2) were identified correctly by both methods. However, only 45.5% of B. caccae, 53.3% of B. ovatus, 66.7% of B. merdae and 78.6% of B. thetaiotaomicron were correctly identified phenotypically. Isolates with discrepant results were further analyzed by 16S rDNA sequencing. The identification obtained from 16S rDNA sequencing showed 100% agreement with the multiplex PCR-based identification. The comparison of the two methods showed that the initial phenotypic identification was accurate only 84.5% of the time, whereas the multiplex PCR-based scheme always gave clear-cut results. It even distinguished bacteria with similar phenotypic characteristics, such as B. distasonis and B. merdae.

4

Comparison of multiplex PCR-based identification with phenotypic identification

Multiplex PCR identification No. of strains Phenotypic identification No. of strains No. of strains that match 
B. caccae 22 B. caccae 10 10 (45.5%) 
  B. distasonis  
  B. ovatus  
  B. stercoris  
  B. thetaiotaomicron  
B. distasonis 19 B. distasonis/B. merdae 19 19 (100%) 
B. eggerthii B. eggerthii 4 (100%) 
B. fragilis 51 B. fragilis 51 51 (100%) 
B. merdae B. distasonis/B. merdae 2 (66.7%) 
  B. caccae  
B. ovatus 15 B. ovatus 8 (53.3%) 
  B. distasonis  
  B. caccae  
  B. thetaiotaomicron  
B. stercoris B. stercoris 2 (100%) 
B. thetaiotaomicron 14 B. thetaiotaomicron 11 11 (78.6%) 
  B. uniformis  
B. uniformis 10 B. uniformis 9 (90%) 
  B. thetaiotaomicron  
B. vulgatus 15 B. vulgatus 15 15 (100%) 
   Total 131/155 (84.5%) 
Multiplex PCR identification No. of strains Phenotypic identification No. of strains No. of strains that match 
B. caccae 22 B. caccae 10 10 (45.5%) 
  B. distasonis  
  B. ovatus  
  B. stercoris  
  B. thetaiotaomicron  
B. distasonis 19 B. distasonis/B. merdae 19 19 (100%) 
B. eggerthii B. eggerthii 4 (100%) 
B. fragilis 51 B. fragilis 51 51 (100%) 
B. merdae B. distasonis/B. merdae 2 (66.7%) 
  B. caccae  
B. ovatus 15 B. ovatus 8 (53.3%) 
  B. distasonis  
  B. caccae  
  B. thetaiotaomicron  
B. stercoris B. stercoris 2 (100%) 
B. thetaiotaomicron 14 B. thetaiotaomicron 11 11 (78.6%) 
  B. uniformis  
B. uniformis 10 B. uniformis 9 (90%) 
  B. thetaiotaomicron  
B. vulgatus 15 B. vulgatus 15 15 (100%) 
   Total 131/155 (84.5%) 

Identification results obtained from PCR assays that were different from phenotypic identification were reconfirmed by 16S rDNA sequencing.

In conclusion, a simple, rapid and reliable multiplex PCR-based method for identification of the B. fragilis group species has been established. This multiplex PCR-based identification scheme is easy, reliable, and inexpensive. It is a powerful tool for routine identification of clinical isolates of the B. fragilis group. Further studies are required to determine conditions needed to detect the B. fragilis group species directly from clinical specimens.

References

[1]
Moore
W.E.C.
Cato
E.P.
Holdeman
L.V.
(
1978
)
Some current concepts in intestinal bacteriology
.
Am. J. Clin. Nutr.
 
31
(
10 Suppl.
),
S33
S42
.
[2]
Holdeman
L.V.
Good
I.J.
Moore
W.E.
(
1976
)
Human fecal flora: variation in bacterial composition within individuals and a possible effect of emotional stress
.
Appl. Environ. Microbiol.
 
31
,
359
375
.
[3]
Finegold
S.M.
(
1977
)
Anaerobic Bacteria in Human Disease
 .
Academic Press
,
New York
.
[4]
Redondo
M.C.
Arbo
M.D.
Grindlinger
J.
Snydman
D.R.
(
1995
)
Attributable mortality of bacteremia associated with the Bacteroides fragilis group
.
Clin. Infect. Dis.
 
20
,
1492
1496
.
[5]
Brook
I.
(
1989
)
Pathogenicity of the Bacteroides fragilis group
.
Ann. Clin. Lab. Sci.
 
19
,
360
376
.
[6]
Finegold
S.M.
(
1995
)
Anaerobic infections in humans: an overview
.
Anaerobe
 
1
,
3
9
.
[7]
Snydman
D.R.
Jacobus
N.V.
McDermott
L.A.
Ruthazer
R.
Goldstein
E.J.C.
Finegold
S.M.
Harrell
L.J.
Hecht
D.W.
Jenkins
S.G.
Pierson
C.
Venezia
R.
Rihs
J.
Gorbach
S.L.
(
2002
)
National survey on the susceptibility of Bacteroides fragilis group: report and analysis of trends for 1997–2000
.
Clin. Infect. Dis.
 
35
(
Suppl. 1
),
S126
S134
.
[8]
Holdeman
L.V.
Cato
E.P.
Moore
W.E.C.
, Eds. (
1977
)
Anaerobic Laboratory Manual
 ,
4
th edn.
VPI Anaerobe Laboratory, Virginia Polytechnic Institute and State University
,
Blacksburg, VA
.
[9]
Shah
H.N.
Collins
M.D.
(
1983
)
Genus Bacteroides: a chemotaxonomical perspective
.
J. Appl. Bacteriol.
 
55
,
403
416
.
[10]
Arzese
A.
Minisini
R.
Botta
G.A.
(
1994
)
Evaluation of an automated system for identification of anaerobic bacteria
.
Eur. J. Clin. Microbiol. Infect. Dis.
 
13
,
135
141
.
[11]
Cavallaro
J.J.
Wiggs
L.S.
Miller
J.M.
(
1997
)
Evaluation of the BBL crystal anaerobe identification system
.
J. Clin. Microbiol.
 
35
,
3186
3191
.
[12]
Grobes
D.J.
Clar
V.
(
1987
)
Preparation of ribonucleic acid probes specific for Bacteroides fragilis
.
Diagn. Microbiol. Infect. Dis.
 
7
,
273
278
.
[13]
Kuritza
A.P.
Getty
C.E.
Shaughnessy
P.
Hesse
R.
Salyers
A.A.
(
1986
)
DNA probes for identification of clinically important Bacteroides species
.
J. Clin. Microbiol.
 
23
,
343
349
.
[14]
Jotwani
R.
Kato
N.
Kato
H.
Watanabe
K.
Ueno
K.
(
1995
)
Detection of Bacteroides fragilis in clinical specimens by polymerase chain reaction amplification of the neuraminidase gene
.
Curr. Microbiol.
 
31
,
215
219
.
[15]
Kuwahara
T.
Akimoto
S.
Ugai
H.
Kamogashira
T.
Kinouchi
T.
Ohnishi
Y.
(
1996
)
Detection of Bacteroides fragilis by PCR assay targeting the neuraminidase-encoding gene
.
Lett. Appl. Microbiol.
 
22
,
361
365
.
[16]
Yamashita
Y.
Kohno
S.
Koga
H.
Tomono
K.
Kaku
M.
(
1994
)
Detection of Bacteroides fragilis in clinical specimens by PCR
.
J. Clin. Microbiol.
 
32
,
679
683
.
[17]
Smith
C.J.
Callihan
D.R.
(
1992
)
Analysis of rRNA restriction fragment length polymorphisms from Bacteroides spp. and Bacteroides fragilis isolates associated with diarrhea in humans and animals
.
J. Clin. Microbiol.
 
30
,
806
812
.
[18]
Kleivdal
H.
Hofstad
T.
(
1995
)
Chromosomal restriction endonuclease analysis and ribotyping of Bacteroides fragilis
.
APMIS
 
103
,
180
184
.
[19]
Podglajen
I.
Breuil
J.
Casin
I.
Collatz
E.
(
1995
)
Genotypic identification of two groups within the species Bacteroides fragilis by ribotyping and by analysis of PCR-generated fragment patterns and insertion sequence content
.
J. Bacteriol.
 
177
,
5270
5275
.
[20]
Moraes
S.R.
Goncalves
R.B.
Mouton
C.
Seldin
L.
Ferreira
M.C.
Domingues
R.M.
(
1999
)
Bacteroides fragilis isolates compared by AP-PCR
.
Res. Microbiol.
 
150
,
257
263
.
[21]
Kuwahara
T.
Norimatsu
I.
Nakayama
H.
Akimoto
S.
Kataoka
K.
Arimochi
H.
Ohnishi
Y.
(
2001
)
Genetic variation in 16S–23S rDNA internal transcribed spacer regions and the possible use of this genetic variation for molecular diagnosis of Bacteroides species
.
Microbiol. Immunol.
 
45
,
191
199
.
[22]
Wisselink
H.J.
Joosten
J.J.
Smith
H.E.
(
2002
)
Multiplex PCR assays for simultaneous detection of six major serotypes and two virulence-associated phenotypes of Streptococcus suis in tonsillar specimens from pigs
.
J. Clin. Microbiol.
 
40
,
2922
2929
.
[23]
Yeboah-Manu
D.
Yates
M.D.
Wilson
S.M.
(
2001
)
Application of a simple multiplex PCR to aid in routine work of the mycobacterium reference laboratory
.
J. Clin. Microbiol.
 
39
,
4166
4168
.
[24]
Tran
S.D.
Rudney
J.D.
(
1999
)
Improved multiplex PCR using conserved and species-specific 16S rRNA gene primers for simultaneous detection of Actinobacillus actinomycetemcomitans, Bacteroides forsythus, and Porphyromonas gingivalis
.
J. Clin. Microbiol.
 
37
,
3504
3508
.
[25]
Song
Y.
Kato
N.
Liu
C.
Matsumiya
Y.
Kato
H.
Watanabe
K.
(
2000
)
Rapid identification of 11 human intestinal Lactobacillus species by multiplex PCR assays using group- and species-specific primers derived from the 16S–23S rRNA intergenic spacer region and its flanking 23S rRNA
.
FEMS Microbiol. Lett.
 
187
,
167
173
.
[26]
Jousimies-Somer
H.
Summanen
P.
Citron
D.M.
Baron
E.J.
Wexler
H.M.
Finegold
S.M.
(
2002
)
Wadsworth-KTL Anaerobic Bacteriology Manual
 ,
6
th edn.
Star Publishing, Belmont
,
CA
.
[27]
Roller
C.
Ludwig
W.
Schleifer
K.H.
(
1992
)
Gram-positive bacteria with a high DNA G+C content are characterized by a common insertion within their 23S rRNA genes
.
J. Gen. Microbiol.
 
138
,
1167
1175
.
[28]
Finegold
S.M.
Molitoris
D.
Song
Y.
Liu
C.
Vaisanen
M.L.
Bolte
E.
McTeague
M.
Sandler
R.
Wexler
H.
Marlowe
E.M.
Collins
M.D.
Lawson
P.A.
Summanen
P.
Baysallar
M.
Tomzynski
T.J.
Read
E.
Johnson
E.
Rolfe
R.
Nasir
P.
Shah
H.
Haake
D.A.
Manning
P.
Kaul
A.
(
2002
)
Gastrointestinal microflora studies in late-onset autism
.
Clin. Infect. Dis.
 
35
(
Suppl. 1
),
S6
S16
.
[29]
Maidak
B.L.
Cole
J.R.
Lilburn
T.G.
Parker
C.T.
Jr.
Saxman
P.R.
Farris
R.J.
Garrity
G.M.
Olsen
G.J.
Schmidt
T.M.
Tiedje
J.M.
(
2001
)
The RDP-II (Ribosomal Database Project)
.
Nucleic Acids Res.
 
29
,
173
174
.