|
|
||||||||
| ABSTRACT |
|
|
|---|
| INTRODUCTION |
|
|
|---|
| METHODS |
|
|
|---|
Isolation and selective enrichment of Burkholderia species. Water samples were vacuum-filtered through 0.2-µm filter papers. The filter papers were then placed into 20 mL of tryptone soya broth (TSB) amended with 10 µg/mL gentamicin. For soil samples, 10-g volumes of soil were added to 15-mL sterile, demineralized water and shaken at room temperature for 4 hrs. The soils were allowed to sediment for another hour without shaking, after which time 1 mL of water was aspirated and inoculated into 20 mL of TSB amended with 10 µg/mL gentamicin. Control broths and plates containing B. pseudomallei or distilled water were inoculated concurrently with test samples.
After incubation, 1-mL volumes of TSB enrichment broths were transferred to 30 mL of Ashdowns broth containing 50 mg/mL colistin. In addition, 10-µL amounts were streaked onto both Ashdowns agar and B. pseudomallei-selective agar (BPSA). After growth in Ashdowns broths, 10-µL volumes were streaked onto both types of selective agar. All broths and plates were allowed to grow at 37°C for 48 hrs.
Phenotypic characterization. Ashdowns and TSB broths were monitored for pellicle formation, while selective agars were scrutinized for colonies resembling B. pseudomallei after both the initial incubation period of 48 hrs and following a further 5-days growth at room temperature. Suspect colonies were assessed for resistance to gentamicin and colistin sulfate (10-µg discs on blood agar), and Gram stains and oxidase tests were performed. For putative Burkholderia isolates, API20NE assays (bioMérieux, Marcy lEtoile, France) were also performed. Growth on single-carbon sources were assessed using minimal-salts medium, 5x (pH 6.8; Difco, Detroit, MI), prepared according to the manufacturers instructions. Single-carbon sources (Sigma, St. Louis, MO) and agar were added.10 Nitrate utilization was determined according to a standard method11 by inoculating isolates into nitrate broth with Durham tubes (Excel Laboratory Products, Belmont, Western Australia). Isolates were also subjected to fatty acid methyl ester (FAME) analysis by fine-capillary column gas chromatography (MIDI Systems Inc., Newark, DE), according to the manufacturers instructions.12
Nucleic acid extraction, polymerase chain reaction, sequencing, and analysis. Crude deoxyribonucleic acid (DNA) preparations were generated by suspending 1 large (> 1 mm diameter) bacterial colony in 50 µL of a solution of 0.05 M sodium hydroxide and 0.025% sodium dodecyl sulfate (v/v).13 The suspension was heated at 100°C for 15 min before the addition of 950 µL of sterile DNase/RNase-free water. This was thoroughly vortexed and briefly centrifuged to pellet the cell debris. The extracts were diluted 1/10 to produce the inoculum used for all subsequent molecular assays.
Two separate polymerase chain reaction (PCR) assays were performed to exclude B. pseudomallei and B. mallei. Semi-nested 16s–23s spacer and real-time lpxO PCRs were performed as previously described.14–16 Amplification, sequencing, and analysis of the BUR3–BUR4 amplicon of the recA gene was carried out as previously described.4 The consensus sequence generated for each isolate was compared with the available recA sequences to confirm the gene identity and determine the highest percentage match using the basic local alignment search tool (BLAST; www.ncbi.nlm.nih.gov). The consensus sequences and any Burkholderia recA sequences publicly available were aligned and trimmed to the length of the BUR3–BUR4 amplicon using BioEdit (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). From alignment, the phylogenetic tree was produced using MEGA v2.1 (http://www.megasoftware.net/) with published computational settings.4 Pseudomonas aeruginosa PA01 recA sequence (NC002156) was used as the root for all trees.
DNA macrorestriction analysis. Pulsed-field gel electrophoresis (PFGE) was performed using the PulseNet (Centers for Disease Control, Atlanta, GA) Escherichia coli method,17 with the following amendments. Isolates were cultured on brain–heart infusion agar in a humidified environment at 37°C for 18 hrs. Chromosomal DNA was digested overnight with restriction endonuclease XbaI and run on a 1% SeaKem Gold (Cambrex Bio Science, Rockland, ME) agarose gel, after pre-electrophoresis, with a pulse time and ramp of 5.5–52 s for 20.2 hrs at 6 V/cm on a CHEF DRIII apparatus (Bio-Rad, Hercules, CA). The standard H49812 was used. After ethidium bromide staining, a gel image was obtained using the Gel Doc imaging system (Bio-Rad) and relatedness of PFGE patterns was analyzed with the Bionumerics software package Version 3.2 (Applied Maths, Kortrijk, Belgium). The dendrogram was calculated by using the un-weighted pair group method (UPGM) using average linkages and the Dice coefficient, with a position tolerance of 1% and an optimization of 0.5%. In addition to several newly isolated strains, a panel of B. thailandensis strains was also subjected to DNA macrorestriction analysis. The Western Australian B. pseudomallei outbreak strain, National Collection of Type Cultures (NCTC) 13177,18 was included for comparison purposes.
Whole-genome sequencing and in-silico multilocus sequence typing analysis. Burkholderia isolates 49639 and A1301 were submitted for whole-genome shotgun sequencing to Tim Reid at The Naval Medical Research Center Annex (Rockville, MD). Following public release of the genomes via DNA Data Bank of Japan (DDBJ)/European Molecular Biology Laboratory (EMBL)/GenBank, multilocus sequence typing (MLST) analysis was performed by extracting the target genes sequences from the sorted contigs using BioEdit software. Isolate 49639 was examined using the B. pseudomallei MLST website (http://bpseudomallei.mlst.net/misc/info.asp) developed by Daniel Godoy and Gaynor Randle (Imperial College).19 Isolate A1301 was examined using the Burkholderia cepacia complex MLST website (http://pubmlst.org/bcc/).20
| RESULTS |
|
|
|---|
Phenotypic characterization.
All 4 isolates were Gram-negative rods with examples of bipolar staining. Colonies grew well on the selective agars, developing distinctive appearances as colonies matured (Figure 1
). API20NE identified all 4 isolates as B. cepacia. All 4 isolates were oxidase-positive, gentamicin- and colistin-resistant, and capable of L-arginine and L-arabinose assimilation. Isolates A1301 and 17540 grew well on maltose or L-arabinose as single-carbon sources, whereas isolates A10A02 and 49639 grew on L-arabinose but not on maltose when supplied as single-carbon sources. Isolates A10A02 and 49639 were able to convert nitrate to gas, whereas isolates A1301 and 17540 were only able to convert nitrate to nitrite.
|
|
|
|
Burkholderia isolate 49639 returned MLST type sequences of 6 (ace), 10 (gltB), 15 (gmhD), 29 (lepA), 9 (lipA), 14 (narK), and 9 (ndh). This was most similar to 3 profiles in the database. In all 3 cases, the MLST profile of isolate 49639 differed at only one locus: lipA for strains 132/02 and E0177 (Sequence Type 352) and lepA for E0433 (Sequence Type 361). All 3 of these profiles were generated from isolates described as B. thailandensis. Further widening of the terms of the profile comparison to "5 or more loci matching" returned a greater number of results, all of which were identified as B. thailandensis. In the case of isolate Burkholderia A1301, none of the MLST sequences corresponded to known alleles, and a sequence type (ST) could not be determined. These sequences were submitted to the curator of the B. cepacia complex MLST website (Adam Baldwin) to create new ST and allele profiles.
| DISCUSSION |
|
|
|---|
Maltose utilization was found to distinguish B. thailandensis from B. ubonensis, as described previously.3 Similarly, assimilation of L-arabinose effectively differentiated the B. thailandensis isolates from B. pseudomallei in our collection. These tests were performed by incorporating these sugars as sole carbon sources into agar plates. Oxidation–fermentation type reactions involving complex media, with acid production leading to a color change, were not found to be suitable for isolate discrimination based on maltose, L-arabinose, or L-arginine utilization. In these tests, a false-positive result could be recorded for an isolate that did not possess the ability to metabolize the single-carbon source in question. Nitrate utilization profiles of isolates A10A02 and 49639 were consistent with the established nitrate to gas phenotype of B. thailandensis.1 Similarly, nitrate metabolism of isolates A1301 and 17540 (nitrate to nitrite only) was in agreement with the description of B. ubonensis.3
The FAME peak 14:0 2OH, corresponding to the 2-hydroxymyristic acid commonly detected in B. pseudomallei12 was not present in any of these 4 isolates. Some fatty acids corresponded with type strains, whereas others differed markedly from published fatty acid profiles.3,12 These discrepancies could be due to small variations in fatty acid content between assays performed at different laboratories under varying growth conditions or may reflect genuine differences between Australian and southeast Asian Burkholderia species.
Of the 4 organisms described in this report, none was found to be positive by either of the B. pseudomallei-specific PCR assays used. Both of these assays have previously been shown to discriminate between B. pseudomallei and B. thailandensis.14,16 However B. ubonensis was not available during validation of these assays, so potential cross-reactivity could not be excluded. Sequencing of a portion of the recA gene4 enables differentiation of closely related Burkholderia species. This approach was shown to distinguish B. thailandensis and B. ubonensis both from each other and from all other known Burkholderia species. Both pairwise and multiple sequence comparisons demonstrated that isolates A10A02 and 49639 were most closely related to B. thailandensis, and isolates A1301 and 17540 were equally well matched to B. ubonensis. DNA macrorestriction analysis was performed on the 2 B. thailandensis-like organisms to further evaluate their relationship to other B. thailandensis strains and to B. pseudomallei. These 2 isolates were more closely grouped with the B. thailandensis strains from Thailand than with an Australian B. pseudomallei isolate. Whole-genome sequencing followed by insilico MLST indicated that isolate 49639 is closely related to B. thailandensis. MLST of isolate A1301 did not produce a match to any isolates in the B. cepacia MLST database. However, the database did not include any B. ubonensis profiles.
On the basis of these results, we propose the following provisional designation for the isolated organisms: Burkholderia thailandensis A10A02, Burkholderia thailandensis 49639, Burkholderia ubonensis A1301, and Burkholderia ubonensis 17540.
The recognition that both B. thailandensis and B. ubonensis are present means that systematic surveys in northern Australia are needed to map the extent of their distribution in the tropical rhizosphere and other habitats. In addition, culture collections of Australian Burkholderia should be reevaluated for the presence of these species. The 2 B. thailandensis strains described here were not isolated from the same location in Northern Australia but were separated by more than 500 km. Likewise, the 2 B. ubonensis strains were isolated from the same distinct areas. Discrimination between B. pseudomallei and B. thailandensis requires at least an additional series of substrate utilization tests, as previously documented.10 The well-recognized limitation of routine laboratory identification tests for B. pseudomallei argues in favor of a more formal polyphasic identification process, including molecular methods. In view of reports claiming melioidosis-like infection caused by strains identified as B. thailandensis,5,6 this species may not retain its status as a strict non-pathogen.
The argument that B. thailandensis and B. ubonensis did not exist in Australia made identification of Burkholderia species from clinical samples a little easier. Recognition of the presence of these species on this continent places a burden on the clinical laboratory to use more-discriminating bacterial identification methods. L-Arabinose utilization is one such simple preliminary test which can discriminate between B. pseudomallei and closely related species.10 Without such tests we cannot now be certain whether B. thailandensis isolates are missed either as a cause of pulmonary disease or as a contaminant of surface lesions. The additional significance of these species to the diagnostic laboratory is in the interpretation of serological tests, particularly the indirect hemagglutination assay (IHA), which relies on a crude antigen preparation. The IHA is widely used because of its low cost and speed, despite its low specificity and sensitivity.21 In Australia, the specificity of the IHA has not been questioned because B. thailandensis was not previously known to be present. The assumption that B. thailandensis was absent has possibly contributed to an over-reliance on serodiagnostic methods for melioidosis. In view of our observations, this specificity of melioidosis serodiagnosis needs to be reevaluated in Australia. It is possible that some borderline IHA results may reflect recent or high-level exposure to B. thailandensis rather than to B. pseudomallei.8,9 In a recent study in Australia, 88% of serum samples from patients with culture-confirmed melioidosis contained antibodies reactive with B. thailandensis antigen.22 However, there is a lack of data on the proportion of individuals with low-to-medium B. pseudomallei IHA titers but with high B. thailandensis IHA titers. If borderline melioidosis IHA results can reflect exposure to B. thailandensis, then serological surveys in some remote Australian communities may overestimate subclinical melioidosis disease load.
The increased diversity of Burkholderia species we have observed in northern Australia raises questions as to the evolutionary origins of these isolates and, indeed, their process of speciation. Our observations do not indicate an obvious evolutionary source of B. thailandensis or B. ubonensis either in southeast Asia or Australia. If the geographic distribution of B. thailandensis includes North America5 and Australia, and the distribution of B. ubonensis also includes Australia, then have these species been transported between continents or has there been a series of parallel speciation events? Whole-genome sequencing of additional isolates may help to clarify their phylogenetic origin.
In conclusion, our observations of B. thailandensis and B. ubonensis in northern Australia indicate that the geographic distribution of these species is much wider than previously thought. This recognition of an expanded biodiversity of Burkholderia species in the topical north of Australia highlights the value of further studies into the distribution of Burkholderia species in our region.
Received November 9, 2007. Accepted for publication January 26, 2008.
Acknowledgments: We thank Vanaporn Wuthiekanun for kindly supplying the B. thailandensis strains used for comparison purposes. The authors thank Tim Reid at The Naval Medical Research Center Annex, Rockville, MD, for sequencing the B. thailandensis and B. ubonensis genomes. We also appreciate the advice offered by Steve Munyard in phenotypic characterization of isolates and Lyn OReilly in pulsed-field gel electrophoresis.
* Address correspondence to Avram Levy, Queen Elizabeth II Medical Centre, Nedlands, 6009 Western Australia, Australia. E-mail: a-levy{at}cyllene.uwa.edu ![]()
Authors addresses: Avram Levy, Adam J. Merritt, Max Aravena-Roman, Meredith M. Hodge, and Timothy J. J. Inglis, Division of Microbiology & Infectious Diseases, PathWest Laboratory Medicine, Western Australia, Telephone: +61 (08) 9346 1357, Fax: +61 (08) 9381 7139, E-mail: a-levy{at}cyllene.uwa.edu.au.
Reprint requests: Avram Levy, Queen Elizabeth II Medical Centre, Nedlands, 6009 Western Australia, Australia, Telephone: +61 (08) 9346 1357, Fax: +61 (08) 9381 7139, E-mail: a-levy{at}cyllene.uwa.edu.
| REFERENCES |
|
|
|---|
This article has been cited by other articles:
![]() |
J. R. Chandler, B. A. Duerkop, A. Hinz, T. E. West, J. P. Herman, M. E. A. Churchill, S. J. Skerrett, and E. P. Greenberg Mutational Analysis of Burkholderia thailandensis Quorum Sensing and Self-Aggregation J. Bacteriol., October 1, 2009; 191(19): 5901 - 5909. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. A. Duerkop, J. Varga, J. R. Chandler, S. B. Peterson, J. P. Herman, M. E. A. Churchill, M. R. Parsek, W. C. Nierman, and E. P. Greenberg Quorum-Sensing Control of Antibiotic Synthesis in Burkholderia thailandensis J. Bacteriol., June 15, 2009; 191(12): 3909 - 3918. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. J. J. Inglis, A. Levy, A. J. Merritt, M. Hodge, R. McDonald, and D. E. Woods Melioidosis Risk in a Tropical Industrial Environment Am J Trop Med Hyg, January 1, 2009; 80(1): 78 - 84. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |