Alanine Racemase Mutants of Burkholderia pseudomallei and Burkholderia mallei and Use of Alanine Racemase as a Non-Antibiotic-Based Selectable Marker

Burkholderia pseudomallei and Burkholderia mallei are category B select agents and must be studied under BSL3 containment in the United States. They are typically resistant to multiple antibiotics, and the antibiotics used to treat B. pseudomallei or B. mallei infections may not be used as selective agents with the corresponding Burkholderia species. Here, we investigated alanine racemase deficient mutants of B. pseudomallei and B. mallei for development of non-antibiotic-based genetic selection methods and for attenuation of virulence. The genome of B. pseudomallei K96243 has two annotated alanine racemase genes (bpsl2179 and bpss0711), and B. mallei ATCC 23344 has one (bma1575). Each of these genes encodes a functional enzyme that can complement the alanine racemase deficiency of Escherichia coli strain ALA1. Herein, we show that B. pseudomallei with in-frame deletions in both bpsl2179 and bpss0711, or B. mallei with an in-frame deletion in bma1575, requires exogenous d-alanine for growth. Introduction of bpsl2179 on a multicopy plasmid into alanine racemase deficient variants of either Burkholderia species eliminated the requirement for d-alanine. During log phase growth without d-alanine, the viable counts of alanine racemase deficient mutants of B. pseudomallei and B. mallei decreased within 2 hours by about 1000-fold and 10-fold, respectively, and no viable bacteria were present at 24 hours. We constructed several genetic tools with bpsl2179 as a selectable genetic marker, and we used them without any antibiotic selection to construct an in-frame ΔflgK mutant in the alanine racemase deficient variant of B. pseudomallei K96243. In murine peritoneal macrophages, wild type B. mallei ATCC 23344 was killed much more rapidly than wild type B. pseudomallei K96243. In addition, the alanine racemase deficient mutant of B. pseudomallei K96243 exhibited attenuation versus its isogenic parental strain with respect to growth and survival in murine peritoneal macrophages.


Introduction
Burkholderia pseudomallei and Burkholderia mallei are Gram-negative bacteria that cause meliodosis in humans and glanders in horses, respectively [1]. B. mallei is an equine pathogen [2,3] that occasionally causes infections in humans. In contrast, B. pseudomallei is frequently isolated from the environment in tropical and subtropical areas [4,5], and it is transmitted to humans by inhalation, ingestion, or direct contact [5]. The mortality rate for acute cases of human melioidosis exceeds 40%, and a percentage of the survivors may experience a relapse at a later time despite previous antibiotic treatment and apparent cure [5,6].
Due to their ability to cause fatal infections, their intrinsic antibiotic resistance, and the lack of an approved vaccine, B. pseudomallei and B. mallei are considered to be potential biological warfare agents and are classified as Category B select agents in theUnited States (http://www.selectagents.gov/exclusions. html#hhsAgents). Under current select agent guidelines, antibiotic resistance markers cannot be used for experimental studies in B. pseudomallei or B. mallei if they confer resistance to antibiotics used to treat infections caused by these bacteria in humans or animals. To date, only antibiotic resistance genes that confer resistance to gentamicin, kanamycin, or zeocin are approved for use in B. pseudomallei, and only genes that encode resistance to kanamycin or zeocin are approved for use in B. mallei. Few genetic tools have been developed for use in Burkholderia spp., and they typically employ the antibiotic resistance genes described above as selectable markers [7,8,9,10,11,12]. Because of these regulatory constraints, it is desirable to investigate the use of non-antibiotic-based selectable markers for the development of additional genetic tools for use in Burkholderia spp. Genes that encode alanine racemase, a key enzyme for bacterial cell wall biosythesis, have been used for this purpose in several other bacterial species.
Alanine racemases are pyridoxal 59 phosphate-containing, homodimeric proteins that catalyze interconversion of L-alanine and D-alanine. D-alanine is an essential building block for biosynthesis of peptidoglycan in bacterial cell walls, and it is also found in lipotechoic acids of some Gram-positive bacteria [13,14]. Alanine racemases have been well-studied in several bacteria, including Escherichia coli [15], Listeria monocytogenes [16], Mycobacterium smegmatis [17], Pseudomonas putida [18], Salmonella enterica serovar Typhimurium [19], Corynebacterium glutamicum [20,21], Lactobacillus plantarum [22], and Bacillus spp. [23,24,25,26]. Bacterial genomes usually contain either one or two alanine racemase genes. In bacteria with two alanine racemase genes, one is typically expressed constitutively and used for D-alanine biosynthesis, whereas the other is typically inducible and used for catabolism of D-alanine [27,28,29,30]. In E. coli, alr encodes an alanine racemase that is constitutively expressed and has apparent anabolic activity, while dadX encodes an alanine racemase that is involved in L-alanine catabolism [15,31]. A double knockout of both dadX and alr is required to produce an E. coli auxotroph that requires exogenous D-alanine for growth [31]. Bacteria that cannot produce alanine racemase exhibit a conditional lethal phenotype in the absence of exogenous D-alanine. An alanine racemase gene can therefore function as a selectable genetic marker in an alanine racemase deficient bacterial host growing in medium without exogenous D-alanine. Genes that encode alanine racemase have been used as an alternative to antibiotic resistance genes as selectable genetic markers in alanine racemase deficient variants of several bacterial species [20,32,33,34]. Additionally, an alanine racemase deficient mutant of L. monocytogenes that exhibits defective growth in phagocytic cells and reduced virulence in animals has been investigated for potential use as a live attenuated vaccine against L. monocytogenes [16].
The characteristics of alanine racemase deficient mutants of Burkholderia spp. have not previously been evaluated. In this study, we identified the genes that encode alanine racemase in B. pseudomallei and B. mallei, constructed alanine racemase deficient Figure 1. Alignment of amino acid sequences of alanine racemases from B. pseudomallei (BPSL2179 and BPSS0711), B. mallei (BMA1575), B. thailandensis (BTH_I2007 and BTH_II1715), E. coli (EcoliDadX and EcoliAlr), and P. aeruginosa (PAO1DadX and PAO1Alr). Residues that are identical in all of these alanine racemase enzymes are shown with a black background. Residues that are identical in 7 or 8 of these enzymes are shown with a dark grey background, and residues that are identical in 5 or 6 of these enzymes are shown with a light grey background. doi:10.1371/journal.pone.0021523.g001 mutant strains of both species, and compared the parental and alanine racemase deficient mutants of both species for growth and viability in the presence and absence of exogenous D-alanine, both in vitro and in murine peritoneal macrophages. Furthermore, we developed genetic tools with selectable alanine racemase markers to perform allelic exchange and gene complementation experiments in alanine racemase deficient host strains of B. pseudomallei and B. mallei. Finally, as proof of principle, we used these tools to construct an inframe deletion in the flgK gene of B. pseudomallei and to complement the flgK mutant with a wild type flgK + allele on a replicating plasmid, thereby avoiding the need for antibiotics as selective agents.

Results
Identification of genes in B. pseudomallei and B. mallei that encode alanine racemases Analysis of the genome sequence of B. pseudomallei strain K96243 [35] revealed two putative alanine racemase genes: bpsl2179 on chromosome 1 which encodes a 356 amino acid protein, and bpss0711 on chromosome 2 which encodes a 374 amino acid protein. Analysis of the genome sequence of B. mallei ATCC 23344 [36] revealed a single putative alanine racemase gene, designated bma1575. The coding sequences for bma1575 and bpsl2179 differ by only 3 nucleotides, and they encode identical proteins. We did not find a homologue of bpss0711 in the genome sequence of B. mallei ATCC 23344. Figure 1 compares and aligns the predicted amino acid sequences for these putative alanine racemases from B. pseudomallei and B. mallei with the two reported alanine racemases per species from Burkholderia thailandensis, E. coli and Pseudomonas aeruginosa [37,38,39]. The identical proteins encoded by bpsl2179 from B. pseudomallei and bma1575 from B. mallei have 73% amino acid sequence similarity with the protein encoded by bpss0711from B. pseudomallei and 60-99% amino acid sequence similarity with the other alanine racemases shown in Figure 1.
We used genetic complementation tests in E. coli to demonstrate that bpsl2179 and bpss0711 from B. pseudomallei encode functional alanine racemases. Toward that end, we first constructed an alanine racemase deficient strain of E. coli, called ALA1, as described in Materials and Methods. E. coli ALA1 was unable to grow on LB agar or in LB broth unless D-alanine was added at $1 mM. In contrast, the parental strain of E. coli did not require added Dalanine for growth. When E. coli ALA1 was grown overnight at 37uC in LB broth with D-alanine and washed bacteria were subcultured in LB broth with or without D-alanine, the bacteria resumed normal growth in the LB broth with D-alanine, but they failed to grow and viability decreased progressively by about 1,000,000-fold over 8 hours in the LB broth without D-alanine (data not shown). Finally, introducing plasmid pET17alr into E. coli ALA1 restored the ability of E. coli ALA1 to grow in LB medium without added D-alanine. These findings demonstrate that E. coli ALA1 is auxotrophic for D-alanine and provide genetic evidence that it does not produce functional alanine racemase. Next, we constructed plasmids containing bpsl2179 or bpss0711 from B. pseudomallei (designated pCR2.1-TOPOH -bpsl2179 and pCR2.1-TOPOH-bpss0711, respectively) and transformed them individually into E. coli ALA1. At 37uC on LB agar without D-alanine, visible growth of E. coli ALA1(pCR2.1-TOPOH -bpsl2179) transformants was detected after 15 hours, and visible growth of E. coli ALA1(pCR2.1-TOPOH-bpss0711) transformants was detected after 24 hours. We constructed a second bpss0711 clone with a longer upstream sequence (Tables 2 and 3), designated pCR2.1-TOPOH-bpss0711-F2, and demonstrated that the growth of E. coli ALA1(pCR2.1-TOPOH-bpss0711-F2) transformants on LB agar without D-alanine was comparable to that of E. coli ALA1(pCR2.1-TOPOH -bpsl2179) transformants. We did not investigate further the molecular basis for the different properties of the pCR2.1-TOPOH-bpss0711 and pCR2.1-TOPOH-bpss0711-F2 clones. We also showed that a pCR2.1-TOPOH-bma1575 clone from B. mallei was able to complement the alanine racemase deficiency of E. coli ALA1. In contrast, the E. coli ALA1(pCR2.1-TOPOH) vector control did not grow under these conditions. Taken together, these genetic data demonstrate that bpsl2179, bpss0711, and bma1575 encode functional alanine racemase proteins.
Construction and characterization of B. pseudomallei and B. mallei alanine racemase deficient mutants The arrangements of the bpsl2179 locus in the large chromosome and the bpss0711 locus in the small chromosome of B. pseudomallei K96243 are shown schematically in Fig. 2A and Fig. 2B, respectively. Both bpsl2179 and bpss0711 are predicted to be transcribed as single genes. The chromosomal loci carrying the corresponding Dbpsl2179 and Dbpss0711 in-frame deletion alleles described below are shown schematically in Fig. 2C and Fig. 2D, respectively.
We used in vitro methods to construct mutant alleles, designated Dbpsl2179 and Dbpss0711, with in-frame deletions corresponding to 99% of the coding regions of bpsl2179 and bpss0711, respectively. These mutant alleles were substituted, either singly or together, for the corresponding wild type alleles in the chromosomes of B. pseudomallei strain K96243 and B. pseudomallei strain 1026b by using the pMo130 allelic exchange system previously described [12]. In both of these B. pseudomallei strains, the resulting single mutants carrying either Dbpsl2179 or Dbpss0711 grew normally on LB agar without added D-alanine, but the Dbpsl2179/Dbpss0711 double mutant required exogenous D-alanine for growth. Fig. 3 illustrates the patch tests that we used to distinguish individual resolved co-integrants with the Dbpsl2179/Dbpss0711 double mutant genotype that grow only on LB agar containing D-alanine from individual resolved cointegrants with the Dbpsl2179 single mutant genotype that do not require D-alanine for growth. Our results showed that a single wild type allele of either bpsl2179 or bpss0711 is sufficient for normal growth of B. pseudomallei on LB agar without added D-alanine, and they showed that no gene except bpsl2179 or bpss0711 directs production of functional alanine racemase in B. pseudomallei strain K96243 or strain 1026b. In contrast, a single in-frame deletion at the bma1575 locus of B. mallei ATCC 23344, introduced by allelic exchange using the highly homologous pMo130D bpsl2179 clone, conferred a stringent growth requirement for exogenous D-alanine. This finding indicates that bma1575 is the only gene in B. mallei ATCC 23344 that directs production of alanine racemase. The presence of the appropriate wild type or in-frame deletion variant of bpsl2179 or bpss0711 in each newly constructed mutant was confirmed by PCR using primers located upstream and downstream of the appropriate gene, as illustrated in Fig. 4. For all subsequent experiments, the term ''alanine racemase deficient mutants'' refers to a Dbpsl2179/Dbpss0711 double mutant for B. pseudomallei strain K96243 or B. pseudomallei strain 1026b, and to a Dbma1575 single mutant for B. mallei strain ATCC 23344.
To demonstrate complementation of the functional defect in the alanine racemase deficient mutants of B. pseudomallei and B. mallei, the wild type bpsl2179 gene was first cloned into the replicative plasmid pMo168 [12] in place of the aphA cassette, yielding pALRcomp. The pALR-comp plasmid was then introduced by conjugation into the alanine racemase deficient mutant strains of B. pseudomallei K96243, B. pseudomallei 1026b, and B. mallei ATCC 23344. Transconjugant colonies were recovered on LB agar containing zeocin or polymyxin B. The presence of pALR-comp in transconjugant colonies was confirmed by detecting the XylE reporter enzyme encoded by pALR-comp, and complementation of the alanine racemase deficiency was confirmed by the ability of the transconjugants to grow with or without the addition of D-alanine.
To determine the minimal concentration of D-alanine necessary to support growth of the Burkholderia alanine racemase deficient mutants, samples from log phase cultures of each mutant, wild type, or complemented strain were spread onto LB agar with Dalanine at various concentrations from 0 to 10 mM, and the plates were inspected for bacterial growth during subsequent incubation at 37uC. Fig. 5 shows the results for the B. pseudomallei K96243derived strains. After 24-36 hr, the wild type parental strain ( Fig. 5A) and the complemented mutant ( Fig. 5C) showed heavy confluent growth on medium without D-alanine. In contrast, the Dbpsl2179/Dbpss0711 mutant showed no growth on medium without D-alanine, scattered colonies on medium with 1.25 mM Dalanine, sub-confluent growth on medium with 2.5 mM D-alanine, and confluent growth on medium with 5 or 10 mM D-alanine Figure 4. PCR confirmation of deletions in Dbpsl2179 and Dbpss0711 mutants of B. pseudomallei. Primers BPSL2179-up and BPSL2179-down were used to generate a 3108 bp fragment from strains carrying the wild type bpsl2179 allele or a 2100 bp fragment from stains carrying the Dbpsl2179 allele. Primers BPSS0711-up and BPSS0711-down were used to generate a 3247 bp fragment from strains carrying the wild type bpss0711 allele or a 2188 bp fragment from strains carrying the Dbpss0711 allele. Lane labels and samples analyzed are as follows: MW, DNA ladder; Dbpss0711, amplicon from strain carrying the Dbpsl2179 mutant allele; bpss0711, amplicon from strain carrying the wild type bpss0711 allele; Dbpsl2179, amplicon from strain carrying the Dbpsl2179 mutant allele; and bpsl2179, amplicon from strain carrying the wild type bpsl2179 allele. doi:10.1371/journal.pone.0021523.g004 Individual colonies of resolved co-integrants were picked from YT agar plates containing 15% sucrose and 10 mM D-alanine and inoculated in patches at comparable locations on LB agar plates with and without 10 mM D-alanine. Resolvants with the Dbpsl2179/Dbpss0711 genotype grew only on the LB agar with D-alanine, and resolvants with the Dbpsl2179 genotype grew on LB agar with or without D-alanine. doi:10.1371/journal.pone.0021523.g003 (Fig. 5B). Comparable results were observed with the parental, mutant and complemented strains derived from B. pseudomallei 1026b and B. mallei ATCC 23344 (data not shown). It is noteworthy that the concentration of D-alanine needed to support growth of the alanine racemase deficient mutants of B. pseudomallei and B. mallei was significantly greater than the concentration of Dalanine required to support growth of E. coli ALA1.
To assess the effects of D-alanine deprivation on growth and viability of the wild type and alanine racemase deficient variants of B. pseudomallei and B. mallei as a function of time, we collected bacteria from cultures grown overnight in LB broth containing 10 mM D-alanine, washed the bacteria with LB broth, and transferred inocula into LB broth subcultures with or without 10 mM D-alanine. We measured turbidity and viable counts in samples taken periodically from each subculture during the first seven hours (Fig. 6) and again at 24 hours. The alanine racemase deficient mutants of B. pseudomallei K96243 and B. mallei ATCC 23344 exhibited growth arrest in LB broth without D-alanine ( Fig. 6A and Fig. 6C), but the growth of each mutant in LB broth with D-alanine was comparable to the growth of its wild type parental strain in LB broth with or without D-alanine. Under the conditions that permitted growth, the doubling time for the wild type and mutant B. mallei strains was about 48 minutes, and the doubling time for the wild type and mutant B. pseudomallei strains was about 36 minutes. Furthermore, each alanine racemase deficient mutant, but not its isogenic parental strain, lost viability progressively with increasing incubation time in the LB broth without D-alanine ( Fig. 6B and Fig. 6D). During the first two hours without D-alanine, the alanine racemase deficient mutant of B. pseudomallei K96243 lost viability more rapidly than the alanine racemase deficient mutant of B. mallei strain ATCC 23344 (,4 log 10 decrease vs. ,1 log 10 decrease, respectively). After 7 hours without D-alanine, viability declined by about 5 log 10 for both of the alanine racemase deficient mutants. After 24 hours without Dalanine, no viable bacteria were recovered from cultures of either of the alanine racemase deficient mutants (data not shown). Results with the wild type and alanine racemase deficient mutant of B. pseudomallei 1026b were comparable to those shown above for the wild type and alanine racemase deficient mutant of B. pseudomallei K96243, respectively (data not shown). To investigate how efficiently each of the alanine racemase deficient mutants of B. pseudomallei or B. mallei can recover from Dalanine deprivation for varying periods of time, we grew each mutant overnight in LB broth containing 10 mM D-alanine and then transferred washed bacteria into six identical subcultures containing LB broth alone. At time 0, 30, 60, 90, and 120 minutes post inoculation, 10 mM D-alanine was added to separate subcultures, and the sixth subculture served as a control without added D-alanine. Samples were withdrawn from each subculture periodically for a total of 7 hours, and each sample was examined for turbidity and viable counts (Fig. 7). For the alanine racemase deficient mutant of B. pseudomallei strain K96243, growth resumed after a few minutes when D-alanine was added back to the cultures at 30 or 60 minutes (Fig. 7A). In contrast, D-alanine deprivation for 90 minutes caused a lag period of at least 2 hours before growth resumed, and D-alanine deprivation for 120 minutes caused growth arrest to the end of the 7-hour observation period (Fig. 7A). D-alanine deprivation for 30 or 60 minutes did not cause a substantial decrease in viability of the alanine racemase deficient mutant of B. pseudomallei strain K96243 (Fig. 7B). In contrast, 90 minutes of D-alanine deprivation resulted in ,1 log 10 decrease in viability, and 120 minutes of D-alanine deprivation caused ,3 log 10 decrease in viability. Bacteria that were viable after each period of D-alanine deprivation resumed growth shortly after Dalanine was added back to the medium. Similar results were obtained with the alanine racemase deficient mutant of B. pseudomallei strain 1026b (data not shown). For the alanine racemase deficient mutant of B. mallei, prolonged growth retardation and a transient decrease in viability occurred with Dalanine deprivation for 120 minutes but not with D-alanine deprivation for 90 minutes or less (Figs. 7C and 7D).
Intracellular survival of wild type and alanine racemase deficient strains of B. pseudomallei and B. mallei in periodate-elicited murine peritoneal macrophages B. pseudomallei is able to survive in non-phagocytic cells and phagocytic cells [40] and B. mallei can survive in macrophages [41]. Due to the rapid decrease in viability of the alanine racemase deficient mutants under conditions of D-alanine deprivation and the requirement for high concentrations of D-alanine to support normal growth of these mutants, we tested whether the alanine racemase deficient mutant strains of B. pseudomallei and B. mallei would exhibit decreased survival in phagocytic cells in comparison with their isogenic parental strains. First, we infected murine peritoneal macrophages with wild type or alanine racemase deficient strains of B. pseudomallei K96243. Bacteria were added to the macrophage cultures in RPMI + medium supplemented with 5 mM D-alanine, and the infected cultures were incubated for 3 hours to permit phagocytosis of B. pseudomallei. The cultures were then washed with media containing kanamycin plus or minus 5 mM D-alanine, and the viable intracellular bacteria were enumerated at times 0, 3, 4, and 5-hours in sets of 5 replicate cultures. The % survival was expressed as (cfu at t n /cfu at t 0 ) 6100. For wild type B. pseudomallei K96243, survival of intracellular bacteria was approximately 1% after 5 hours, and survival was not affected by the presence or absence of 5 mM Dalanine in the culture medium (Fig. 8A). In contrast, intracellular survival of the alanine racemase deficient mutant of B. pseudomallei K96243 after 5 hours in medium without D-alanine was only 0.01%, compared to about 0.5% in medium containing 5 mM Dalanine (Fig. 8B). To confirm the internalization of the wild type and mutant bacteria, we collected macrophages after the initial 3hour period of contact with bacteria and analyzed them by transmission electron microscopy ( Fig. 9). Comparable numbers of intracellular bacteria were observed for the wild type and mutant bacterial strains (Fig. 9, Panels A and B). Taken together, these findings demonstrate a substantially decreased ability of the alanine racemase deficient mutant of B. pseudomallei K96243 to survive within murine peritoneal macrophages unless a high concentration of D-alanine was present during the killing assay.
Next, we compared the wild type and alanine racemase deficient strains of B. mallei in macrophage killing assays. Murine peritoneal macrophages were infected with the wild type or alanine racemase deficient mutant strain of B. mallei ATCC23344 under conditions similar to those described above for B. pseudomallei. After 2 hours of phagocytosis in D-alanine-supplemented RPMI + medium, the macrophages were washed with prewarmed RPMI + medium containing 6 mg/ml gentamicin with or without 5 mM D-alanine. Electron microscopy demonstrated comparable numbers of wild type and mutant bacteria within macrophages immediately after the 2-hour phagocytosis phase of the killing assays (Fig. 9, Panels C and D). Survival of the intracellular bacteria was determined after an additional 4 hours of incubation. In contrast to the previous results for B. pseudomallei, both wild type B. mallei and the isogenic alanine racemase deficient mutant were equally susceptible to killing by macrophages, and enhanced killing of the alanine racemase deficient mutant was not observed in medium without D-alanine (Fig. 10). In conclusion, the alanine racemase deficient mutant of B. mallei did not exhibit an increased susceptibility to intracellular killing in macrophages under these experimental conditions.
The use of alanine racemase as a selectable genetic marker To demonstrate the use of alanine racemase as selectable genetic marker for construction of mutant strains of Burkholderia spp., we first replaced the aphA gene in the allelic exchange vector pMo130 [12] with bpsl2179 from B. pseudomallei K96243 to generate pAlr-allex. Then, as proof of principle, we used the pAlrallex vector system to construct a DflgK mutant without use of antibiotics for selection. This experiment was based on results of a previous study demonstrating that deletion of flgK, which encodes the flagellar protein FlgK, causes loss of motility in B. pseudomallei [12]. By cloning DflgK from pMo146 [12] into pAlr-allex, we targeted the flgK gene of B. pseudomallei for deletion. The resulting plasmid designated pAlr-allex-DflgK was introduced by conjugation into the alanine racemase deficient B. pseudomallei strains K96243 Dbpss0711/Dbpsl2179 and 1026b Dbpss0711/Dbpsl2179. Transconjugants were selected by growth on LB agar medium without added D-alanine and confirmed by detection of XylE activity. Following counter-selection on sucrose, some of the resolved transconjugants of B. pseudomallei strains K96243 and 1026b exhibited both a loss of motility and a growth requirement for D-alanine, consistent with their DflgKDbpss0711Dbpsl2179 genotypes (data not shown).   To demonstrate use of alanine racemase as a selectable genetic marker for complementation tests, we cloned the wild type flgK + allele into pAlr-comp and introduced the resulting pAlr-comp-flgK + plasmid by conjugation into the DflgK/Dbpss0711/Dbpsl2179 mutants of B. pseudomallei strains K96243 and 1026b described above. Transconjugants were selected by growth on LB agar medium lacking D-alanine and confirmed by detection of XylE activity. Introduction of pAlr-comp-flgK + into these non-motile mutants of B. pseudomallei restored motility to wild type levels in addition to restoring the ability of the mutants to grow without added D-alanine (data not shown). These experiments using pAlrallex and pAlr-comp established proof of principle that alanine racemase can be used as a selectable marker for construction of inframe mutants by allelic exchange and for in trans complementation in pathogenic Burkholderia spp.

Discussion
In the United States, the study of B. pseudomallei and B. mallei is highly regulated because they are classified as category B select agents. Furthermore, few antibiotic resistance determinants are approved as selectable markers for use in genetic studies of B. pseudomallei and B. mallei [5,42]. Previous studies have identified several non-antibiotic selectable markers for use in genetic studies of bacteria. Most are based on the lethal consequences of inactivating an essential gene, and preserving viability by providing a wild type allele of the inactivated essential gene to complement the lethal phenotype. One such system used a thyA mutant of Lactococcus lactis that cannot survive without an exogenous source of thymidine or thymine unless thyA + is introduced on a replicating plasmid or by another genetic method [43,44]. In Salmonella enterica serovar Typhimurium and P. aeruginosa, a comparable system was based on inactivation of the asd gene encoding aspartate b-semialdehyde dehydrogenase, an enzyme required for diaminopimelic acid production and peptidoglycan synthesis in these bacterial species [45,46]. Without diaminopimelic acid, asd mutants will lyse, whereas asd mutants complemented with an asd + allele will survive and grow normally. In C. glutamicum and L. monocytogenes [20,33,34], inactivation of the gene encoding alanine racemase, an enzyme required for production of D-alanine and for peptidoglycan biosynthesis, was shown to be lethal unless D-alanine was added to the growth medium or the genetic defect was complemented by the presence of a wild type allele of the alanine racemase gene.
In the present study, we constructed and characterized alanine racemase mutants of B. pseudomallei and B. mallei and demonstrated that they exhibit a stringent requirement for exogenous D-alanine for growth and viability. In B. pseudomallei, it was necessary to inactivate both bpsl2179 and bpss0711 to produce an alanine racemase deficient phenotype, whereas in B. mallei inactivation of bma1575 alone was sufficient to produce an alanine racemace deficient phenotype. Other bacterial species whose genomes encode two different but functional alanine racemases include S. enterica serovar Typhimurium, P. aeruginosa, and E. coli [19,27,47,48,49]. In contrast, C. glutamicum [20] and Lactobacillus plantarum [22] are representative of bacterial species whose genomes encode only one functional alanine racemase.
In this study, we showed that a gene encoding alanine racemase can be used effectively as a selectable marker for genetic manipulations in alanine racemase deficient parental strains of pathogenic Burkholderia spp. We constructed an allelic exchange vector with a selectable alanine racemase gene and used it to generate a DflgK mutation in two different alanine racemase deficient reference strains of B. pseudomallei. In addition, we constructed a replicative plasmid carrying wild type alleles of the alanine racemase gene bpsl2179 and the flgK gene and showed that it could complement the DflgK mutation in an alanine racemase deficient strain and restore its ability to grow in the absence of exogenous Dalanine. The pAlr-allex and pAlr-comp plasmids described in this study are significant additions to the current repertoire of tools for use in genetic studies of pathogenic Burkholderia spp.
We also demonstrated that an alanine racemase deficient mutant of B. pseudomallei K96243 lost viability at a much faster rate than its isogenic parental strain in murine peritoneal periodateelicited macrophages growing in cell culture medium without added D-alanine. However, the addition of 5 mM D-alanine to the cell culture medium allowed the alanine racemase deficient mutant to survive intracellularly as well as its wild type parental strain. In another published study, an alanine racemase deficient (dal dat) mutant of L. monocytogenes also exhibited growth attenuation in macrophages and required exogenous D-alanine in the culture medium to achieve levels of intracellular survival comparable to the wild type parental strain [16]. L. monocytogenes lacking alanine racemase failed to elicit an immune response in mice when it was inoculated without added D-alanine. In contrast, when it was inoculated together with D-alanine, the mice developed a protective immune response and survived a subsequent challenge with a dose of wild type L. monocytogenes that was lethal for unimmunized mice [16]. The authors proposed that the addition of D-alanine enabled the attenuated strain to undergo limited replication in the inoculated mice that was sufficient to elicit a protective immune response [16]. They also showed that few viable bacteria remained in the mouse tissues after 1-2 days [16], presumably because depletion of the inoculated D-alanine by Damino acid oxidases of the mouse [50,51,52,53,54] resulted in rapid death of the inoculated bacteria and their progeny. Our alanine racemase deficient mutant of B. mallei did not exhibit attenuation under the conditions used for our experiments. Because wild type B. mallei grows more slowly than B. pseudomallei and is killed more rapidly than B. pseudomallei in macrophages, it seems likely that the B. mallei mutant did not grow enough within the macrophages to permit D-alanine deprivation to contribute significantly to the overall loss of viability. Since there is currently no vaccine for B. pseudomallei or B. mallei, the use of an attenuated strain as a potential vaccine is appealing. Additional studies will be needed to examine the degree of attenuation of our alanine racemase deficient mutants of B. pseudomallei and B. mallei in normal and immunocompromised animals and to test their ability to stimulate protective immune responses after they are inoculated with or without added D-alanine, as described above for alanine racemase deficient mutants of L. monocytogenes. Such experiments are beyond the scope of this study and will require animal facilities approved for use with select agents that are not available at our institution. Recently, Propst et al. [55] reported that a DpurM mutant of B. pseudomallei is attenuated both in immunocompetent and immunodeficient animals and could be considered for possible exclusion from select agent classification. Additional studies will be needed to demonstrate whether alanine racemase deficient mutants of B. pseudomallei and B. mallei may also be appropriate candidates for possible exclusion from select agent classification. Strongly attenuated mutants of Bacillus anthracis, Brucella abortus, Coxiella burnetii, Francisella tularensis, and Yersinia pestis have already been excluded from select agent classification (http://www. selectagents.gov/exclusions.html#hhsAgents) [55].
There are many ways that purine-or D-alanine requiring mutants such as those described above could be used to facilitate future genetic studies with B. pseudomallei or B. mallei. Possible exclusion from select agent classification would enable such strains to be used for genetic manipulations under BSL2 containment rather than BSL3 containment. This would greatly simplify the technical procedures required for construction of mutants with targeted inactivating mutations in specific non-essential genes, or construction of libraries of mutants with random inactivating mutations in any non-essential genes. Mutants of interest could then be placed under BSL3 conditions, and the effects on virulence of inactivating mutations in individual genes could be tested after the purine or D-alanine requirement of the parental strain was eliminated by introduction of an appropriate complementing gene.

Materials and Methods
Bacterial strains, growth conditions, and media All bacterial strains and plasmids used in this study are listed in Table 1 and Table 2, respectively, and all primers used for cloning in this study are listed in Table 3. All primers were purchased from Integrated DNA Technologies (Coralville, IA). All E. coli strains were grown in Luria-Bertani (LB) media at 30uC or 37uC as described in the text. All experiments involving live cultures of B. pseudomallei or B. mallei were performed in the BSL3 facility at the University of Colorado School of Medicine. All Burkholderia strains were grown in LB broth at 37uC unless otherwise stated. Kanamycin, when appropriate for plasmid selection, was added at a concentration of 50 mg/ml unless otherwise noted. D-alanine, when needed, was added to LB media at a concentration of 1 mM for experiments with E. coli and at 10 mM for experiments B. pseudomallei or B. mallei. For resolution of co-integrants by sucrose counter-selection, Burkholderia co-integrants were grown in YT broth, which was made by dissolving 10 g yeast extract and 10 g tryptone in 1 l of H 2 O containing 15% sucrose as described by Hamad et al [12].

Analysis of nucleotide and encoded amino acid sequences of putative alanine racemase genes of B. pseudomallei and B. mallei
We searched the annotated genomes of B. pseudomallei K96243 and B. mallei ATCC 23344 for previously identified putative alanine racemase genes. To search for potentially unannotated alanine racemase genes, we used the nucleotide sequences of the alr and dadX genes from E. coli strain DH10B [37] in the tBLASTx program on the NCBI website (http://www.ncbi.nlm.hin.gov) to probe the sequenced genomes of B. pseudomallei K96243 and B. mallei ATCC 23344. An analysis of amino acid sequence homology between the putative alanine racemase proteins encoded by B. pseudomallei bpsl2179 and bpss0711, B. mallei bma1575, and several other known alanine racemases was performed using http://www. ch.embnet.org/software/ClustalW.html.

Construction of the alanine racemase deficient mutant strain E. coli ALA1
The genome of E. coli has two known alanine racemase genes, alr and dadX. We obtained the following constructs from Michael Benedik: a clone of the alr gene designated pET17alr; E. coli MB1910, an endA derivative of E. coli TG1 with a dadX allele that is insertionally inactivated by a kanamycin resistance (Km R ) determinant flanked by frt sites; and E. coli MB2786, which contains an alr allele insertionally inactivated by a tetracycline resistance (Tc R ) determinant flanked by frt sites. To construct E. coli ALA1, we first transferred the Tc R -marked alr allele from MB2786 into MB1910 by phage P1 transduction and demonstrated that the resulting transductant was both Km R and Tc R and also able to form colonies on LB agar supplemented with DLalanine but not on un-supplemented LB agar, consistent with inactivation of both alr and dadX. Next, we removed the Km R and Tc R determinants by introducing an Ap R temperature-sensitive plasmid pCP20 [56] that encodes flp recombinase, which catalyzes recombination between the frt sites flanking the Km R and Tc R determinants. To complete the construction of strain ALA1, we eliminated pCP20 by overnight growth at a non-permissive temperature (37uC), leaving the alr and dadX alleles with single inactivating frt sites but without the Tc R or Km R determinants.
Complementation of E. coli ALA1 with cloned alanine racemase genes from B. pseudomallei We designed the primer pairs BPSL2179F2/BPSL2179R2 and BPSS0711F/BPSS0711R to PCR amplify bpsl2179 and bpss0711, respectively, along with their putative promoters. Since the region of the B. mallei chromosome that contains bma1575 is nearly identical to the region containing bpsl2179 in B. pseudomallei, the primer pair BPSL2179F2/BPSL2179R2 was also used to amplify bma1575 with its putative promoter. Additionally, primer pair BPSS0711-F2/BPSS0711-R was used to amplify a fragment that contained a 250 bp longer upstream sequence than that amplified by primer pair BPSS0711-F/BPSS0711-R. Touchdown PCR was performed using Ex Taq DNA polymerase as previously described [12]. PCR amplicons were cloned into pCR2.1-TOPOH using conditions specified by the manufacturer (Invitrogen, Carlsbad, CA). The bpsl2179, bpss0711, bpss0711-F2, and bma1575 genes from these clones were sequenced at the University of Colorado Cancer Center DNA Sequencing and Analysis Core Facility and compared with the annotated genome sequence of B. pseudomallei strain K96243 and B. mallei ATCC 23344, respectively. To assess the function of the putative alanine racemase proteins encoded by bpsl2179, bpss0711, and bma1575, we transformed the pCR2.1-TOPOH-bpsl2179, pCR2.1-TOPOH-bpss0711, pCR2.1-TOPOH-bpss0711-F2, and pCR2.1-TOPOH-bma1575 clones into E. coli ALA1 made competent by chemical treatment, as described previously [57]. E. coli ALA1 transformants containing pCR2.1-TOPOH -bpsl2179, pCR2.1-TOPOH-bpss0711, pCR2.1-TOPOH-bpss0711-F2, or pCR2.1-TOPOH-bma1575 were selected on LB agar without added D-alanine. Representative colonies were picked and retested for their ability to grow on LB agar with and without D-alanine at 1 mM.

Construction of alanine racemase deficient mutants of B. pseudomallei and B. mallei
To generate unmarked, in-frame deletions in the bpsl2179 and bpss0711 chromosomal genes of B. pseudomallei, we first used PCR to construct clones with the desired Dbpsl2179 and Dbpss0711 alleles flanked by approximately 1000 bp segments of the contiguous upstream and downstream flanking regions from the Suicide vector for allelic exchange in Burkholderia; pUC19 ori, RK2 oriT, xylE, sacB, Km R , used to construct pAlr-allex [12] pMo146 Source of DflgK [12] pMo168 Replicative vector for Burkholderia; oripBBR1, mob+, xylE, Km R , used to construct pAlr [12] pMo173 Source of flgK for complementation [12] pAlr-allex Suicide vector for allelic exchange in Burkholderia alr mutants; pUC19 ori, RK2 oriT, xylE, sacB, bpsl2179  B. pseudomallei K96243 chromosome. We cloned each of these constructs separately into the mobilizable suicide vector pMo130, and we used the resulting clones with previously described allelic exchange protocols to substitute the Dbpsl2179 and Dbpss0711 alleles for their corresponding wild type alleles in the chromosomes of B. pseudomallei strains K96243 and 1026b [12]. In addition, we used the pMo130 clone carrying Dbpsl2179 with similar methods to substitute the Dbpsl2179 allele for the wild type bma1575 allele in B. mallei strain ATCC 23344. This was possible because the chromosomal region in B. mallei that contains bma1575 has almost complete nucleotide sequence identity with the corresponding chromosomal segment from B. pseudomallei K96243. Additional experimental details are provided in the following paragraphs.
To construct an in-frame deletion of bpsl2179, we used the primer pair BPSL2179-up and BPSL2179-orfdelR with chromosomal DNA from B. pseudomallei K96243 to PCR amplify a fragment containing approximately 1000 bp from the upstream flanking region of bpsl2179 linked to the first 30 bp from the 59 end and the last 30 bp from the 39 end of the bpsl2179 coding region. We used primers BPSL2179-down and BPSL2179-orfdelF (which is complementary to BPSL2179-orfdelR) in a similar manner to PCR amplify a second fragment containing the first 30 bp from the 59 end and the last 30 bp from the 39 end of the bpsl2179 coding region linked to approximately 1000 bp from the downstream flanking region of bpsl2179. The fragments produced by these first two PCR reactions were then used as templates in a second PCR with primers BPSL2179-up and BPSL2179-down to generate Dbpsl2179, an amplicon containing the 60 nucleotide long, internally in-frame deleted variant of the bpsl2179 coding sequence flanked by approximately 1000 bp long wild type upsteam and downstream sequences. The resulting Dbpsl2179 amplicon was cloned into pCR2.1-TOPOH, generating pCR2.1-TOPOH-Dbpsl2179. pCR 2.1-TOPOH-Dbpsl2179 was digested with HindIII and XbaI to release the Dbpsl2179 fragment, which was Klenow-treated as described by the manufacturer (Invitrogen, Carlsbad, CA) and cloned into SmaI digested pMo130 (15). The resulting plasmid pMo130-Dbpsl2179 was introduced into B. pseudomallei strains K96243 and 1026b by biparental mating as previously described [12]. Co-integrants were selected on LB agar containing 100 mg/ ml kanamycin, 10 mM D-alanine, and 50 mg/ml zeocin (to counterselect the E. coli donor strain). Co-integrants that expressed the xylE reporter gene from pMo130 were recognized as yellow colonies after being sprayed with a mist of pyrocatechol, as previously described [12]. For resolution of the co-integrants, individual colonies were inoculated into YT broth, grown for a minimum of 4 hours, diluted, and plated on YT agar containing 15% sucrose, as previously described [12], plus 10 mM D-alanine. A similar procedure was also performed with B. mallei; however, for selection of co-integrants, 50 mg/ml polymyxin B was used instead of zeocin to counterselect the E. coli donor.
An in-frame deletion of bpss0711 was constructed in the same manner as Dbpsl2179; however, primer pair BPSS0711-up and BPSS0711-orfdelR and primer pair BPSS0711-down and BPSS0711-orfdelF were used in PCR with chromosomal DNA to amplify the upstream and downstream fragments linked to the internally deleted bpss0711 allele, respectively. A second PCR, using primers BPSS0711-up and BPSS0711-down and the fragments produced in the first PCR as templates produced Dbpss0711, which was cloned into pCR2.1-TOPOH. Following digestion by HindIII and XbaI, the Dbpss0711 fragment was Klenow-treated as described above and subcloned into SmaI digested pMo130 (15). The resulting pMo130-Dbpss0711 was then introduced into the Dbpsl2179 single mutants of B. pseudomallei K96243 and 1026b, and also into wild type B. mallei ATCC 23344, by biparental matings, and co-integrants were selected and resolved as described above. Resolved mutants were tested for growth on LB agar with and without 10 mM D-alanine to identify the putative Dbpsl2179/Dbpss0711 double mutants of B. pseudomallei K96243 and 1026b and the putative Dbma1575 single mutant of B. mallei ATCC 23344, all of which were shown to require the exogenous D-alanine for growth. Presence of the Dbpsl2179 and Dbpss0711 alleles in the B. pseudomallei double mutant, presence of the Dbpsl2179 allele in the B. mallei single mutant, and absence of the corresponding wild type alleles was confirmed by PCR tests using either primer sets BPSL2179-up and BPSL2179-down or BPSS0711-up and BPSS0711 down, as appropriate.
Complemention tests with alanine racemase deficient mutants of B. pseudomallei and B. mallei To perform complementation tests, we first constructed a derivative of the mobilizable, replication competent plasmid pMo168 [12] with a copy of bpsl2179 from B. pseudomallei K96243. Briefly, pMo168 was digested with SpeI and XbaI to remove the aphA cassette that confers resistance to kanamycin, yielding the digested pMo168DaphA fragment. Plasmid pCR2.1-TOPOH -bpsl2179 was digested with SpeI and XbaI, and the fragment containing bpsl2179 was ligated with the digested pMo168DaphA fragment, yielding pALR-comp. The resulting plasmid was transformed into E. coli ALA1 made competent by chemical treatment and plated on LB agar to confirm that the cloned bpsl2179 directed production of active alanine racemase. Following incubation at 37uC, the resulting colonies were tested to confirm the presence of pALR-comp, and the bpsl2179 allele from pALR-comp was sequenced to confirm its identity with the wild type bpsl2179 allele. Plasmid pALR-comp was then transferred from E. coli ALA1(pAlr-comp) into the Dbpsl2179/Dbpss0711 double mutants of B. pseudomallei K96243 and 1026b and the Dbma1575 single mutant of B. mallei ATCC 23344 by triparental matings with DH5a(pRK2013) [58]. Putative transformants were selected on LB agar containing 50 mg/ml zeocin for matings with B. pseudomallei recipients or 50 mg/ml polymyxin B for matings with B. mallei recipients, and the presence of pALA1 in transformant colonies was confirmed by spraying them with pyrocatechol to detect the xylE reporter and sub-culturing them onto LB agar and LB agar supplemented with 50 mg/ml zeocin or 50 mg/ml polymyxin B to confirm the intrinsic antibiotic resistance phenotypes of the B. pseudomallei and B. mallei parental strains, respectively.
Quantitative determination of D-alanine requirement for growth of alanine racemase deficient mutants of B. pseudomallei and B. mallei on solid medium Wild type and isogenic alanine racemase deficient mutant strains of B. pseudomallei and B. mallei were grown in LB broth supplemented with 10 mM D-alanine overnight at 37uC. Overnight cultures were subcultured into LB broth supplemented with 10 mM D-alanine at an initial OD 600 = 0.05 and grown to log phase (OD 600 nm = 0.2). Bacteria from log phase cultures were collected by centrifugation, washed with LB broth, resuspended in LB broth to the original volume, and inoculated on LB agar supplemented with 0 mM, 1.25 mM, 2.5 mM, 5 mM, or 10 mM D-alanine. Plates were incubated at 37uC for 36-48 hours and growth was observed.
Effects of D-alanine on growth and viability of alanine racemase deficient mutants of B. pseudomallei and B. mallei in liquid medium Wild type and isogenic alanine racemase deficient mutant strains were grown overnight at 37uC in LB containing 10 mM Dalanine. Cultures were centrifuged, supernatants removed, and the pellets were washed with LB broth and inoculated at an initial OD 600 nm = 0.1 into LB broth with or without 10 mM D-alanine. Cultures were incubated at 37uC with shaking, and samples were removed every 30 minutes for the first 4 hours, then hourly up to 7 hours and again at 24 hours for measurement of OD 600 nm and determination of viable counts on LB agar containing 10 mM D-alanine. To assess the consequences of shorter periods of Dalanine deprivation, each the alanine racemase deficient mutant strains of B. pseudomallei and B. mallei was inoculated into replicate cultures in LB medium without D-alanine as described above, and at 30 minute intervals from time zero up to 120 minutes D-alanine at 10 mM was added back to single cultures. During further incubation, samples were removed from all of the replicate cultures at 30 minute intervals up to 180 minutes and then at hourly intervals up to 7 hours for measurements of growth (OD 600 nm ) and of viability on LB agar with 10 mM D-alanine.
Isolation of murine peritoneal macrophages C57BL/6 mice were bred and murine peritoneal macrophages were prepared under protocols 56409(05)1B and 56410(05)1E, which were approved on 5/18/10 and 5/5/10, respectively, by the University of Colorado Denver Animal Care and Use Committee. Peritoneal macrophages were harvested from mice 4 days after intraperitoneal inoculation of 1 mg/ml sodium periodate as described [59]. The peritoneal exudate cells were re-suspended in RPMI 1640 medium (Sigma-Aldrich, St. Louis, MO) supplemented with 10% heat-inactivated fetal bovine serum (BioWhittaker, Walkersville, MD), 15 mM Hepes, 2 mM Lglutamine, 1 mM sodium pyruvate (Sigma-Aldrich), and penicillin/streptomycin at 100 U/ml and 100 mg/ml, respectively (RPMI + ; Cellgro, Manassas, VA). This medium was supplemented with D-alanine at 5 mM where specified. The peritoneal exudate cells were seeded in flat-bottom 96-well plates at a density of 3 6 10 5 cells per well for macrophage killing assays. The macrophages were selected by adherence after 24 hours of culture at 37uC in a 5% CO 2 incubator. Just prior to infection, the macrophages were washed in pre-warmed RPMI containing 5 mM D-alanine.

Macrophage killing assays
Macrophage killing capacity was assessed by a gentamicin or a kanamycin protection assay using a modification of a protocol described for B. mallei [60]. B. mallei strains were grown overnight in LB and sub-cultured to OD 600 nm = 0.6 in LB. B. pseudomallei strains were grown overnight in LB to an OD 600 nm = 11. The bacteria were collected by centrifugation and opsonized by suspending them for 20 minutes at 37uC in RPMI + medium containing 10% normal mouse serum and 5 mM D-alanine. The opsonized wild type or isogenic alanine racemase deficient strains of B. mallei and B. pseudomallei were added at a multiplicity of infection of 200 to macrophage monolayers in RPMI + containing D-alanine at 5 mM, and the infected monolayers were incubated for 2 hours with the B. mallei strains or for 3 hours with the B. pseudomallei strains to permit phagocytosis of the opsonized bacteria by the macrophages [60,61]. Extracellular bacteria were removed by washing the monolayers with pre-warmed RPMI + medium containing 6 mg/ml of gentamicin for experiments with B.mallei or 250 mg/ml of kanamycin for experiments with B. pseudomallei. After removal of the extracellular bacteria, the average multiplicity of infection was determined to be 10. Fresh RPMI + medium, with or without D-alanine at concentrations indicated in the text, and with 6 mg/ml of gentamicin for experiments with B.mallei or 250 mg/ml of kanamycin for experiments with B. pseudomallei was added to the monolayers. After incubation for 0 hours, 3 hours, 4 hours, or 5 hours for experiments with B. pseudomallei and 0 hours or 4 hours for experiments with B. mallei, five replicate cultures of the infected macrophages were lysed with 1% Triton X-100 in phosphate buffered saline (PBS), and the numbers of viable intracellular bacteria were determined by plating on LB agar for the wild type stains or on LB agar supplemented with 5 mM D-alanine for the alanine racemase deficient strains. The percent survival was calculated as (cfu at t n /cfu at t 0 ) 6100.

Electron Microscopy
Following phagocytosis of wild type or isogenic alanine racemase deficient strains of B. pseudomallei and B. mallei, the macrophages were plated as previously described [59] at a density of 4610 5 per chamber of a 8-well Permanox Labtek chamber slide system (Nalgene Nunc International, Rochester, NY). The cells were fixed in 2.5% glutaraldehyde in phosphate buffer, pH 7.4. The specimens were post-fixed in 1% osmium tetraoxide, treated with uranyl acetate, dehydrated in ascending ethanol series, and infiltrated with Embed 812. Ultrathin sections were examined for the presence and morphological integrity of intracellular bacteria in a FEI Technai 62 electron microscope operated at 80 kV.
Construction and complementation of a B. pseudomallei flagellar mutant using alanine racemase as the selectable genetic marker To demonstrate the utility of alanine racemase as a selection marker for genetic manipulations in Burkholderia spp., we constructed the allelic exchange vector pAlr-allex by replacing aphA of pMo130 [12] with bpsl2179 from B. pseudomallei K96243. Briefly, pMo130 [12] was digested with SpeI and XbaI to remove the aphA cassette, yielding the pMo130DaphA fragment. To clone the wild type bpsl2179 allele into pMo130DaphA, pCR2.1-TOPOH -bpsl2179 was digested with SpeI and XbaI and the resulting fragment containing bpsl2179 was ligated with the restrictiondigested pMo130DaphA fragment, yielding pALR-allex. The pALR-allex plasmid was transformed into chemically competent alanine racemase deficient E. coli strain ALA1 and plated on LB agar without added D-alanine. Growth of the transformants at 37uC demonstrated expression of the bpsl2179 gene in pALRallex. The presence of pALR-allex in the transformants was confirmed, and the region flanking bpsl2179 was sequenced to confirm the predicted nucleotide sequence of the wild type bpsl2179 allele.
We wished to demonstrate directly that pALR-allex can be used for allelic exchange protocols in pathogenic Burkholderia spp. Toward that end, we first subcloned the DflgK allele from pMo146 [12] into pALR-allex. Briefly, we PCR amplified DflgK using primers Dflgk-US-NheI and DflgK-DS-HindIII, digested the resulting amplicon with NheI and HindIII, and ligated it with into linearized pALR-allex generated by treating pALR-allex with NheI and HindIII. The resulting plasmid, pALR-allexDflgK, was transformed into E. coli strain ALA1 made competent by chemical treatment, and transformants were selected by growth on LB agar without added D-alanine. Next, pALR-allexDflgK was introduced into the alanine racemase deficient Dbpss0711Dbpsl2179 double mutants of B. pseudomallei K96243 and B. pseudomallei 1026b by triparental matings, as described above. Co-integrants were selected by growth on LB agar without added D-alanine, and resolved co-integrants were subsequently selected by growth on LB agar containing 15% sucrose plus 10 mM D-alanine. We determined the DflgK or flgK+ genotype of resolved co-integrants by screening individual colonies for their non-motile or motile phenotypes, respectively.
For in-trans complementation of DflgK B. pseudomallei mutants, we subcloned the wild type flgK gene from pMo173 [12] into the pALR-comp plasmid described previously. The flgK gene with its predicted promoter were excised from pMo173 [12] by digestion with NheI and HindIII and the resulting fragment was ligated into NheIand HindIII-digested pALR-comp to generate pALR-comp-flgK. pALR-comp-flgK was introduced by triparental matings into the DflgK mutants of B. pseudomallei K96243 and 1026b strains described above, and the resulting transconjugant colonies were selected by growth on LB agar with added D-alanine and screened for restoration of the motile phenotype.