OUP user menu

Genotypic characterization of Burkholderia cenocepacia strains by rep-PCR and PCR–RFLP of the fliC gene

Sang-Tae Seo, Kenichi Tsuchiya
DOI: http://dx.doi.org/10.1016/j.femsle.2005.02.020 19-24 First published online: 1 April 2005


Thirty-five strains of Burkholderia cenocepacia from clinical and environmental sources were characterized genotypically by repetitive sequence PCR (ERIC- and BOX-PCR) and polymerase chain reaction–restriction fragment length polymorphism (PCR–RFLP) analysis of the flagellin gene (fliC). In cluster analysis based on the repetitive PCR profiles the strains were composed of five clusters, of which clusters 1, 2 and 3 were more closely related to each other than to clusters 4 and 5. It has been reported that the majority of Burkholderia cepacia complex strains can be separated into two types on the basis of fliC size (types I and II correspond to 1.4 and 1.0 kb, respectively). When the strains were analysed by PCR of fliC, all strains yielded amplified products of 1.0 kb except for three strains. The latter strains gave PCR products of 0.7 kb (atypical type), which belonged to repetitive PCR cluster 5. These results indicated that the majority of B. cenocepacia strains belonged to flagellin type II. In the RFLP analysis of the large fliC amplicons with HaeIII, 10 patterns were observed indicating remarkable variation. Strains grouping in repetitive PCR cluster 4 had a unique fliC RFLP pattern. The results of repetitive PCR typing and PCR–RFLP analysis of fliC showed a strong correlation. Strains belonging to the repetitive PCR clusters 4 or 5 were distinctly different from other B. cenocepacia strains as shown by PCR–RFLP analysis of the fliC gene and phenotypic assays.

  • Burkholderia cenocepacia
  • Rep-PCR
  • RFLP
  • fliC

1 Introduction

Bacteria belonging to the Burkholderia cepacia complex are important opportunist pathogens, particularly in relation to patients suffering from cystic fibrosis (CF) [1,2]. B. cepacia complex strains could also be considered as a possible agent of biocontrol or bioremediation [35]. The B. cepacia complex comprises at least nine closely related species or genomovars: B. cepacia (genomovar I), Burkholderia multivorans (genomovar II), Burkholderia cenocepacia (genomovar III), Burkholderia stabilis (genomovar IV), Burkholderia vietnamiensis (genomovar V), B. cepacia genomovar VI, Burkholderia ambifaria (genomovar VII), Burkholderia anthina (genomovar VIII) and Burkholderia pyrrocinia (genomovar IX) [69]. Although all the species have been isolated from clinical specimens, infection with B. cenocepacia represents a significant clinical risk to patients with CF [10,11]. Two distinct recA-based phylogenetic subgroups were found within B. cenocepacia (III-A and III-B subgroups) [12]. Recent studies have shown that B. cenocepacia is also present in various environmental niches [13].

It is important to determine the diversity of B. cenocepacia from clinical and environmental origin for assessing the risk associated with the use of environmental isolates as an agent of biocontrol or bioremediation. However, little information is available on the degree of phenotypic and genetic diversity among B. cenocepacia strains of different origin, especially in Asia. A number of methods have been used to establish relationships within the B. cepacia complex, including phenotypic assays and PCR-based fingerprinting techniques including repetitive sequence PCR (rep-PCR) and restriction fragment length polymorphism (RFLP) [1418].

In the present study, we compared results obtained with rep-PCR (ERIC-PCR and BOX-PCR) and PCR–RFLP of the flagellin gene (fliC) for establishing relationships between 35 strains, identified as B. cenocepacia by RFLP analysis of 16S rRNA and recA genes, and recA-based genomovar specific-PCR, isolated from clinical and environmental sources [19]. Subsequently, strains were analyzed by phenotypic tests and the results were compared to each other.

2 Materials and methods

2.1 Bacterial strains

The bacterial strains used in this study are listed in Table 1. Strains were maintained by freezing in a medium containing 10% skimmed milk supplemented with 1.5% sodium glutamate at −70 °C. When required, each bacterial strain was cultured aerobically on YPDA (yeast extract 3 g, peptone 0.6 g, dextrose 3 g, agar 15 g, in 1 l distilled water, pH 7.2) for 2 days at 28 °C.

View this table:
Table 1

Strains of Burkholderia cenocepacia used in this study

StrainaSourcebGeographical originrecA RFLPcesmRcrep-PCR clusterdfliC RFLPe
Pc C-61Clinical (Shigeta, S., FMU)JapanG+11
Pc 1712Clinical, urine (Yabuuchi, E., GUSM)JapanG+11
Pc 1751Clinical, pus (Yabuuchi, E.)JapanG+11
Pc C-88Clinical (Shigeta, S., FMU)JapanG+12
Pc 2046Clinical, spinal fluid (Yabuuchi, E.)JapanG+12
Pc 1115Unknown (Yabuuchi, E.)JapanG+12
Pc 1115RColony mutant of Pc 1115 (Tsuchiya, K.)JapanG+12
Pc 1151Clinical, sputum (Yabuuchi, E.)JapanG+13
Pc KF1Clinical (Nakazawa, T., YUSM)JapanH+24
Pc-1Lettuce rhizosphere (Tsuchiya, K.)JapanH+24
MAFF 302528Rice rhizosphereJapanH+24
Pc 3018Clinical, blood (Yabuuchi, E.)JapanH+25
Pc 3021Clinical, blood (Yabuuchi, E.)JapanH+25
Pc 3030Clinical, spinal fluid (Yabuuchi, E.)JapanH+25
nao 12Onion (Takikawa, Y., SU)JapanH25
nao 13Onion (Takikawa, Y.)JapanH25
Pc 342-43Onion (DOA)ThailandH26
Pc 342-45Onion (DOA)ThailandH26
Pc 342-46Onion (DOA)ThailandH26
Pc 342-47Onion (DOA)ThailandI+28
Pc 342-48Onion (DOA)ThailandI+28
Pc-3Lettuce rhizosphereJapanI+29
Pc 722Forest soil (Yabuuchi, E.)JapanI+210
Pc A1Cymbidium (Tsuchiya, K.)JapanH+35
Pc A2Cymbidium (Tsuchiya, K.)JapanH+35
Pc A4Cymbidium (Tsuchiya, K.)JapanH+35
Pc A10Cymbidium (Tsuchiya, K.)JapanH+35
Pc A11Cymbidium (Tsuchiya, K.)JapanH+35
Pc JN1Clinical (Nakazawa, T.)JapanH47
Pc JN6Clinical (Nakazawa, T.)JapanH47
Pc JN25Clinical (Nakazawa, T.)JapanH47
Pc 1211Clinical, thermal injury (Yabuuchi, E.)JapanH47
Pc 342-41Onion (DOA)ThailandH5
Pc 342-42Onion (DOA)ThailandH5
Pc 342-42WColony mutant of Pc342-42 (Tsuchiya, K.)ThailandH5
  • a MAFF, Ministry of Agriculture, Forestry and Fisheries Genebank, Japan.

  • b FMU, Fukushima Medical University (Fukushima, Japan); GUSM, Department of Microbiology, Gifu University (Japan); YUSM, Yamaguchi University School of Medicine (Ube, Japan); SU, Shizuoka University (Shizuoka, Japan); DOA, Bacteriology Division, Department of Agriculture (Bangkok, Thailand).

  • c Data from Seo and Tsuchiya [19].

  • d Cluster analysis on the basis of rep-PCR (ERIC- and BOX-PCR). See Fig. 1.

  • e RFLP analysis of flagellin gene (fliC). See Fig. 2.

2.2 Molecular characterization

Genomic DNA was prepared using InstaGene DNA purification matrix (Bio-Rad) according to manufacturer's instructions. The rep-PCR method employed was adapted from Louws et al. [20]. The Enterobacterial Repetitive Intergenic Consensus (ERIC) and BOXA subunit (BOX) primer sets were synthesized by Amersham Pharmacia Biotech (Tokyo, Japan). Amplification was performed in a total volume of 25 μl containing 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each dNTP, 50 pmol each primer, 2.5 units of DNA polymerase Takara Taq (Takara, Japan), and 2 μl of each DNA sample prepared with matrix. PCR reactions were performed with a DNA thermal cycler (GeneAmp 9600, Perkin–Elmer Applied Biosystems) under the following conditions: 95 °C for 5 min for the first cycle, followed by 30 cycles of 94°C for 1 min, 52°C for 1 min, and 65°C for 8 min with a final extension step of 65°C for 15 min. PCR products were separated on 1.5% (w/v) agarose gels in 0.5 × Tris–Borate–EDTA (pH 8.0). Cluster analysis of the rep-PCR profiles was conducted using the site http://aoki2.si.gunma-u.ac.jp/BlackBox.

PCR–RFLP of the fliC gene was performed using primers BVF and fliCR [18]. The volume and content of amplification reactions were as described above. Amplification was performed under the following conditions; 30 cycles of 1 min at 94 °C, 1 min at 58 °C, and 2 min at 72 °C. The amplification cycles were preceded by a denaturation step of 5 min at 95 °C and followed by an elongation step of 10 min at 72 °C. Amplified product samples were digested with the restriction enzyme HaeIII under the conditions recommended by the supplier (TOYOBO, Japan). The RFLP products were analyzed by electrophoresis in 2% (w/v) agarose gels.

2.3 Phenotypic assays

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) of the whole-cell proteins was performed in a vertical slab gel with Laemmli's discontinuous buffer system [21] and 12% polyacrylamide gels (Perfect NT Gel, DRC, Tokyo, Japan). Preparation of whole-cell proteins and SDS–PAGE were performed as described previously [22]. Tests in the API 20NE were performed according to manufacturer's instructions (BioMerieux, Marcy l'Etoile, France).

3 Results and discussion

To our knowledge, this is the first extensive study in which B. cenocepacia from clinical and environmental sources in Asia has been subjected to genetic fingerprinting to assess intraspecific diversity. Repetitive sequence PCR (rep-PCR) has increasingly been used for typing the B. cepacia complex [16]. In the present study, a dendrogram of the 35 B. cenocepacia strains was constructed based on the rep-PCR (ERIC- and BOX-PCR) profiles (Fig. 1). Cluster analysis separated the strains into five clusters at squared distance 100 (Fig. 1 and Table 1). Cluster 1 was composed of all 8 strains belonging to recA RFLP pattern G. Cluster 2 contained 15 strains, which shared recA RFLP patterns H and I. Cluster 3 was composed of all five strains isolated from Cymbidium. The five strains were first reported as a causal bacterium of bacterial brown spot disease of Cymbidium species [23]. Clusters 4 and 5 comprised 4 and 3 strains, which were isolated from clinical and environmental sources, respectively. All strains of clusters 3, 4 and 5 belonged to recA RFLP pattern H. Clusters 1, 2 and 3 were more closely related to each other than to clusters 4 and 5 (Fig. 1). There was some concordance between recA RFLP patterns and rep-PCR clusters; for example, all strains of recA RFLP pattern G belonged to rep-PCR cluster 1.

Figure 1

Illustration of genotyping methods with representative B. cenocepacia strains. ERIC-PCR (a), BOX-PCR (b) and dendrogram (c) were produced as described in the text. Scale bar indicates squared distance. Lanes: 1, Pc C-61; 2, Pc 1712; 3, Pc 1751; 4, Pc 2046; 5, Pc 1151; 6, Pc C-88; 7, Pc 1115; 8, Pc 1115R; 9, Pc KF1; 10, Pc 3018; 11, Pc 3021; 12, MAFF 302528; 13, nao 12; 14, nao 13; 15, Pc 342-43; 16, Pc 342-45; 17, Pc 342-46; 18, Pc 3030; 19, Pc-1; 20, Pc A2; 21, Pc A4; 22, Pc A10; 23, Pc A1; 24, Pc A11; 25, Pc JN1; 26, Pc JN6; 27, Pc JN25; 28, Pc 1211; 29, Pc 342-41; 30, Pc 342-42; 31, Pc 342-42W; 32, Pc 342-47; 33, Pc 342-48; 34, Pc 722 and 35, Pc-3. M, DNA size standard (100-bp ladder, Bayou Biolabs). The Roman number above the lane numbers indicates the recA RFLP pattern of each strain.

The fliC gene encoding the flagellin protein is a highly variable biomarker and has been used in epidemiologic studies [15,18]. Hales et al. [24] reported that the majority of isolates of B. cepacia could be separated into two types (types I and II) on the basis of flagellin protein size. Amplicon sizes of types I and II were 1.4 and 1.0 kb, respectively. Atypical amplicon sizes (0.7, 1.6 and 1.9 kb) of the fliC genes have also been reported [18,24]. In this study, when the 35 strains were analyzed by PCR using primers BVF and fliCR, all strains yielded amplified products of ca. 1.0 kb (type II) with the exception of three strains (Pc 342-41, Pc 342-42 and Pc 342-42W), which gave PCR products of ca. 0.7 kb (atypical type) (Table 1). These three strains all belonged to rep-PCR cluster 5. The results of the fliC PCR indicated a predominance of type II flagellins in B. cenocepacia strains used in this study. RFLP analysis of the Type II amplicons with HaeIII separated 32 B. cenocepacia strains into 10 patterns (Fig. 2). Although only one endonuclease was used in this study, there was a remarkable degree of variation in the fliC gene of B. cenocepacia. These results supported the idea that flagellin genotyping is a highly discriminatory method and can be used as an epidemiological tool for identification of isolates in the B. cepacia complex [25].

Figure 2

RFLP analysis of the fliC gene amplified from B. cenocepacia strains. The DNA products of the fliC gene were digested with restriction enzyme HaeIII. Lanes: 1, Pc C-61; 2, Pc C-88; 3, Pc 1751; 4, Pc KF1; 5, Pc 3018; 6, Pc 342-43; 7, Pc JN1; 8, Pc 342-47; 9, Pc-3 and 10, Pc 722. M, DNA size standard (100-bp ladder, Invitrogen). The alphabetical recA RFLP types are shown above the lanes.

In general, RFLP patterns of the fliC were largely in agreement with the grouping obtained by cluster analysis of rep-PCR (Table 1). Strains grouping in rep-PCR cluster 1 belonged to fliC RFLP patterns 1, 2 and 3. Strains of rep-PCR cluster 4 showed a unique fliC RFLP pattern (pattern 7). The four strains grouping in rep-PCR cluster 4 had very similar random amplified polymorphism DNA patterns (data not shown). Interestingly, of all tested strains, only three strains (Pc JN1, Pc JN6 and Pc JN25) belonging to cluster 4 were positive for assimilation of maltose. Subsequently, these strains showed a unique pattern on the basis of comparison of SDS–PAGE profiles of whole-cell proteins (lanes 25, 26 and 27 in Fig. 3). On the basis of the protein profiles, no significant differences were detected between clinical and environmental B. cenocepacia strains (Fig. 3). Strains of rep-PCR clusters 2 and 3 shared fliC RFLP pattern 5. Winstanley [18] reported that relationships shown by RFLP typing based on recA and fliC did not generally correlate with each other. However, our results showed a good relationship between RFLP typing of recA and fliC (Table 1); eight strains of recA RFLP pattern G belonged to fliC RFLP pattern 1, 2 and 3; 20 strains of recA RFLP pattern H belonged to fliC RFLP pattern 4, 5, 6 and 7; and 4 strains of recA RFLP pattern I belonged to fliC RFLP pattern 8, 9 and 10. From phenotypic and genotypic characterization of the rep-PCR cluster 4 and 5 strains, it became clear that these strains could belong to a distinct group within B. cenocepacia. Further studies are required to clarify the status of these strains within the B. cepacia complex.

Figure 3

SDS–PAGE of whole-cell proteins of B. cenocepacia strains. Lanes: 1, Pc C-61; 2, Pc 1712; 3, Pc 1751; 4, Pc 2046; 5, Pc 1151; 6, Pc C-88; 7, Pc 1115; 8, Pc 1115R; 9, Pc KF1; 10, Pc 3018; 11, Pc 3021; 12, MAFF 302528; 13, nao 12; 14, nao 13; 15, Pc 342-43; 16, Pc 342-45; 17, Pc 342-46; 18, Pc 3030; 19, Pc-1; 20, Pc A2; 21, Pc A4; 22, Pc A10; 23, Pc A1; 24, Pc A11; 25, Pc JN1; 26, Pc JN6; 27, Pc JN25; 28, Pc 1211; 29, Pc 342-41; 30, Pc 342-42; 31, Pc 342-42W; 32, Pc 342-47; 33, Pc 342-48; 34, Pc 722 and 35, Pc-3. Molecular mass markers are indicated in kDa (Protein ladder, Invitrogen). The alphabetical recA RFLP patterns are shown above the lanes.

Previous studies indicated that strains of B. cepacia isolated from the rhizosphere differ markedly from their clinical counterparts in various respects [14,17]. In our study, however, no obvious differences were detected between B. cenocepacia from clinical and environmental sources. For example, rep-PCR cluster 2 comprised both clinical and environmental strains. Similarly, fliC RFLP patterns 4 and 5 contained strains representing different sources. Moreover, no distinct characteristics were detected among them based on the results of the API 20NE tests (data not shown). Therefore, their use as an agent of biocontrol or bioremediation should be avoided, until accurate relationships between environmental and clinical strains are conclusively established.


This work was supported in part by a Grant-in-Aid for Scientific Research (B) (No.16380224) from the Japan Society for the Promotion of Science.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
View Abstract