Burkholderia pseudomallei and Burkholderia mallei are closely related bacteria responsible for melioidosis mainly in humans, sheep, goats, and pigs and glanders in solipeds and humans. They can be mistaken for the recently described Burkholderia oklahomensis and for a nonpathogenic species, Burkholderia thailandensis (3, 4). From a strictly taxonomic point of view, B. mallei should be considered a subspecies of B. pseudomallei, while B. thailandensis and B. oklahomensis are genetically distinct and form separate species (4, 5). The fatal outcomes of the diseases caused by B. mallei and B. pseudomallei, as well as the potential use of these organisms as biowarfare agents, necessitate unequivocal and rapid detection and identification. Several molecular tests have been set up during the last decade. Most of these tests target portions of the genome with housekeeping functions (2, 6-9, 11, 13, 14, 17), while a few others target genes with presumable functions in pathogenicity (10, 12) or antibiotic resistance (15). In view of the high plasticities of their genomes, molecular identification of B. mallei and B. pseudomallei should rely ideally on sequence data validated for this purpose. Such sequence data have been made available from the multiple-locus sequence typing (MLST) database developed by Godoy et al. (5). It consists of seven gene fragments approximately 500 bp in size that have been sequenced in more than 1,000 strains (1,047 at the time of this writing). From the analysis of multiple-sequence alignments, PCR primers common to the four Burkholderia species and discriminating probes with subspecies specificity could be inferred. The present article describes the setup of PCR and real-time PCR assays that use primers and probes of the molecular beacon type to identify and distinguish unambiguously B. mallei from B. pseudomallei and from B. thailandensis/B. oklahomensis.

Short nucleotide stretches found to be 100% conserved in all different alleles of narK and gltB in the B. pseudomallei database (5) were chosen for PCR primer selection. The narK and gltB PCR primers (listed in Table ​Table1)1) generated 453-bp and 351-bp amplification products, respectively, and were subjected to BLASTN analysis (1). Homology was eventually noticed with the corresponding narK and gltB genes from related Burkholderiaceae. Therefore, to assess the specificities of the PCR assays, a number of strains of the B. pseudomallei group were tested together with Burkholderia cepacia and environmental Burkholderia and Ralstonia strains. B. cepacia complex type strains LMG1222^T, LMG13010^T, LMG14191^T, LMG14294^T, LMG16656^T, LMG18835^T, LMG18943^T, LMG19182^T, and LMG20980^T were obtained from the Belgian Coordinated Collections of Microorganisms (Ghent, Belgium). Ralstonia metallidurans AE104 and environmental isolates of Burkholderia, namely, B. hospita R-23325, B. phytofirmans R-23371, B. caledonica R-23338, B. xenovorazns R-23344, B. phenazinium R-23336, and Burkholderia sp. strain R-23316, were from the Flemish Institute for Technological Research (Mol, Belgium). B. pseudomallei strains Bengla 01, ID 1476, and UCL 467 were isolated in Belgium hospitals from sick travelers returning from South Asia; their MLST markers were sequenced and the results submitted to the B. pseudomallei MLST database (http://bpseudomallei.mlst.net). B. pseudomallei strains NCTC 1688, NCTC 4845, NCTC 4846, NCTC 6700, NCTC 7383, NCTC 7431, NCTC 8016, NCTC 8707, NCTC 8708, NCTC 10274, NCTC 10276, and NCTC 11642 and B. mallei strains NCTC 120, NCTC 3708, NCTC 3709, NCTC 10229, NCTC 10230, NCTC 10245, NCTC 10247, NCTC 10248, and NCTC 10260 originate from the Health Protection Agency (London, United Kingdom). B. mallei strains 05-548, 05-554, 05-562, 05-580, 05-586, 764-1, and 768 were from the German Army collection. B. pseudomallei ATCC 11668 was from the 1983 catalogue of the American Type Culture Collection. B. mallei Dubai7 was isolated from an infected horse in Al Ain, United Arab Emirates. B. thailandensis strains CIP106301 and CIP106302 were from the Pasteur Institute Collection (Paris, France).

PCR assays were performed in 50 μl on 50 ng template genomic DNA, using the four PCR primers gltB-FWD, gltB-REV, narK-FWD, and narK-REV (Table ​(Table1)1) at a final concentration of 0.4 μM, 0.4 mM deoxynucleoside triphosphates, 1.5 mM MgCl₂, 1.5 U Taq polymerase, and a buffer supplied by the manufacturer (Invitrogen, Carlsbad, CA). The PCR protocol was the following: 1 cycle of 5 min at 95°C, followed by 40 cycles of 30 s at 95°C, 30 s at 60°C, and 1 min at 72°C. Both narK and gltB PCR products were obtained when DNA from B. mallei, B. pseudomallei, and B. thailandensis was assayed. By contrast, no narK product could be amplified from the other Burkholderia species. gltB PCR products were amplified from all Burkholderia sp. strains except those of B. phenazinium and Ralstonia metallidurans (Fig. ​(Fig.11).

Open in a separate window

FIG. 1.

Representative PCR results. Genomic DNA (50 ng) from B. pseudomallei ATCC 1668 (lane 1), B. mallei BW05-554 (lane 2), B. thailandensis CIP106301 (lane 3), B. cepacia LMG13010 (lane 4), B. phenazinium R-23336 (lane 5), and Ralstonia metallidurans AE104 (lane 6) were PCR amplified according to the procedure described in the text. Ten microliters was loaded on a 1.5% agarose gel stained with ethidium bromide. Bars in the right margin indicate 450- and 280-bp markers.

To allow differentiation between B. mallei and B. pseudomallei, polymorphic nucleotides (nt) in narK (nt 139) and gltB (nt 144) were selected. All 37 B. mallei strains displayed the same two-locus allelic profile (narK nt 139 [A] and gltB nt 144 [T]), while three different specific profiles were displayed by B. pseudomallei (945 strains). The hallmark of B. thailandensis (48 strains) and B. oklahomensis (4 strains) was a “C” at position 285 of narK, which was also found in two environmental strains in the database with unclear species attribution (Burkholderia sp. strains 2002721687 and 1992/2572). Three pairs of molecular beacons were selected to discriminate the critical nucleotides (Table ​(Table1).1). The presence of a stable, GC-rich hairpin structure in the DNA flanking gltB nt 144 was overcome by selecting a customized beacon pair whose properties are described elsewhere (16).

Real-time PCR assays were conducted in 50 μl on an IQ5 instrument, using premixed reagents supplied by the manufacturer (Bio-Rad Laboratories, La Jolla, CA). Primers and probes were high-performance liquid chromatography purified and were purchased from Eurogentec (Liège, Belgium). Final concentrations were 0.1 μM for molecular beacons and 0.4 μM for PCR primers. Each pair of molecular beacons was assayed separately using the same thermal cycling conditions: 3 min at 95°C, followed by 45 cycles of 30 s at 95°C, 15 s at 60°C, 30 s at 65°C, and 20 s at 72°C. Fluorescence was measured during the temperature step at 65°C, using the following optical filters for excitation and emission: 485 ± 10 nm and 530 ± 15 nm (FAM [6-carboxyfluorescein]), 530 ± 15 nm and 575 ± 10 nm (HEX [hexachlorofluorescein]/Yakima Yellow), and 575 ± 15 nm and 625 ± 15 nm (Texas Red). When assayed by real-time PCR, all tested B. pseudomallei, B. mallei, and B. thailandensis strains gave fluorescence signals matching the expected two-locus allelic profiles. Representative examples are shown in Fig. ​Fig.2.2. B. pseudomallei ID 1476 displayed the rare profile no. II (depicted in Table ​Table1),1), while the other B. pseudomallei strains displayed profile no. I. B. mallei strains displayed profile no. IV, and B. thailandensis strains displayed profile no. V. Signal-to-background ratio and allelic discrimination capacity differed from beacon to beacon. Therefore, allelic discrimination plots were drawn to strengthen the fluorescence data (Fig. 2C, F, and I). The assays yielded comparable results whenever pure DNA or raw-boiled cell suspensions were tested (data not shown).

Open in a separate window

FIG. 2.

Typical discriminative real-time PCR results. DNA (50 ng) from strains B. pseudomallei NCTC 8016 (○), B. pseudomallei ID 1476 (□), B. mallei NCTC 10248 (×), B. mallei NCTC 10245 (+), B. thailandensis CIP106301 (), and B. thailandensis CIP106302 () was assayed by real-time PCR, using probes of the molecular beacon type. Results for no-template controls are shown by plain lines. Fluorescence signal is given as a function of the PCR cycle for the amplification of gltB monitored in optical windows optimized for HEX (A) and Texas Red (TxRd) (B) using beacons BG0075 and BG0106 and of narK in optical windows optimized for FAM and HEX using either beacons BG0076 and BG0185 (D and E) or beacons BG0152 and BG0194 (G and H). Allelic discrimination plots are drawn in panels C, F, and I using the fluorescence values from the preceding graphs collected at the 40th PCR cycle. Each point of the plot falls within one out of two clusters matching a defined allele.

In conclusion, the PCR and real-time PCR assays described here can be used to identify and differentiate B. pseudomallei from B. mallei and B. thailandensis/B. oklahomensis. These assays are based on conserved sequences and critical single-nucleotide polymorphisms validated on more than 1,000 strains, making them probably the most validated rapid molecular assays aimed at recognizing bacteria of this group.