![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Drew Rholl,1
Ayush Kumar,1,
Ifor R. Beacham,2 and
Herbert P. Schweizer1*
Department of Microbiology, Immunology and Pathology, Rocky Mountain Regional Center of Excellence for Biodefense and Emerging Infectious Diseases Research, Colorado State University, Fort Collins, Colorado 80523-1682,1 School of Medical Science and Institute for Glycomics, Griffith University—Gold Coast Campus, Gold Coast, Queensland 9726, Australia2
Received 28 October 2007/ Accepted 13 December 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(amrRAB-oprA) efflux pump mutant. Exploiting natural transformation, chromosomal DNA fragments carrying this mutation marked with zeocin resistance were transferred between the genomes of two different B. pseudomallei strains. Lastly, the deletion mutation was complemented by a chromosomally integrated mini-Tn7 element carrying the amrAB-oprA operon. The new tools allow routine select-agent-compliant genetic manipulations of B. pseudomallei and other Burkholderia species. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Site-specific excision of selection markers allows recycling of the few precious markers available for many bacteria. Useful methods previously employed with bacteria for in vivo site-specific excision of unwanted selection markers from the chromosome involve the Saccharomyces cerevisiae Flp-FRT system (13, 47), the Cre-lox system of the P1 phage/plasmid (1, 55), the ParA-res system of RP4 (32), and the TnpR-res system of Tn1000 (or 
transposon) (7). Of these, the Flp-FRT and Cre-lox systems are perhaps the most versatile, functioning in bacteria, yeasts, plants, insects, and mammalian cells (27, 35, 42, 49, 54, 57). Site-specific recombination systems minimally require two components: (i) a target genetic element flanked by recombination sites and (ii) a genetic element encoding the recombinase. For bacteria, the recombinase source is typically encoded by a curable plasmid and plasmid curing is generally achieved by use of a counterselection marker, e.g., Bacillus subtilis sacB (reviewed in reference 52) or a temperature-sensitive (TS) replicon (37, 44). To date, neither one of these is applicable for B. pseudomallei, because most Burkholderia spp. contain endogenous sacB genes and no TS plasmid replicon is available.
The recent availability of the genomic sequences of several B. pseudomallei strains (29) (several GenBank entries) provides new opportunities for understanding the biology and pathogenesis of these opportunistic yet highly fatal pathogens. To take full advantage of the wealth of information contained in these genomes, however, new genetic tools that are in compliance with select-agent guidelines need to be developed. Specifically needed are recyclable selection markers and curable plasmids for mutant construction, transfer of mutations between chromosomes, regulatable and constitutive promoters, and site-specific chromosome integration systems. The genetic tools and accompanying methods described in this paper address all of these needs. In this study, we report the isolation of a TS replicon and its use for establishment of in vivo Flp-FRT and Cre-loxP site-specific recombination systems for B. pseudomallei and other Burkholderia spp. Our results demonstrate that together with corresponding Gm, Km, and Zeo marker cassettes and exploitation of natural transformation, these new tools allow routine and select-agent-compliant genetic manipulation of B. pseudomallei with the few antibiotic selection markers currently approved for use with these bacteria. Examples are provided for the usefulness of these tools for isolation of an unmarked efflux pump mutant, transfer of this mutation between chromosomes of different strains, and establishment of a site-specific chromosomal integration system.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(amrAB-oprA) derivative DD503 (39), and 1710b (Table 1). The Escherichia coli strains used for routine cloning experiments were DH5
(34), HPS1 (51), and strains provided with specific cloning kits. All bacteria were routinely grown at 37°C. Strains containing TS plasmid derivatives were grown at 30°C (permissive temperature) or 42°C (nonpermissive temperature). LB broth Miller and LB agar Miller from EMD Chemicals (Gibbstown, NJ) were used as the rich media for all bacterial strains, except for propagation and selection of Zeo-resistant strains, for which low-salt (5 g liter–1 NaCl) Lennox LB broth (LSLB) and agar (MO BIO Laboratories, Carlsbad, CA) were used. For preparation of the competent B. pseudomallei cells used for transformation with linear DNA fragments, bacteria were grown in DM medium, which consists of 0.25x M63 (38) supplemented with 0.2% glucose, 0.4% glycerol, 1 mM MgSO4, 1 µg ml–1 thiamine-HCl, and 40 µg ml–1 each of leucine, isoleucine, valine, tryptophan, glutamic acid, and glutamine, as described previously (M. Thongdee, L. A. Gallagher, T. Dharakul, S. Songsivilai, and C. Manoil, presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand). Unless otherwise noted, antibiotics were added at the following concentrations: 100 µg ml–1 ampicillin (Ap), 35 µg ml–1 Km, 10 µg ml–1 Gm, 100 µg ml–1 trimethoprim (Tp), and 25 µg ml–1 Zeo for E. coli; 700 to 1,000 µg ml–1 Km and 2,000 µg ml–1 Zeo for wild-type B. pseudomallei and 100 µg ml–1 Km or 50 µg ml–1 Gm for (amrRAB-oprA) mutants; and 200 µg ml–1 Zeo, 100 µg ml–1 Tp, and 100 µg ml–1 Km for B. thailandensis. Antibiotics were purchased from Sigma, St. Louis, MO (Ap, Km, and Tp), EMD Biosciences, San Diego, CA (Gm), and Invitrogen, Carlsbad, CA (Zeo).
|
DNA methods and transformation procedures.
Routine procedures were employed for manipulation of DNA (48). Plasmid DNAs were isolated from E. coli and Burkholderia spp. by using a QIAprep mini-spin kit (Qiagen). Bacterial chromosomal DNA fragments (20 to 30 kb) were isolated using a QIAamp DNA mini kit, and the DNA was suspended in 200 µl of buffer AE (10 mM Tris-HCl, 0.5 M EDTA, pH 9). Plasmid DNA fragments were purified from agarose gels by utilizing a Qiagen QIAquick gel extraction kit. E. coli strains were transformed using chemically competent cells (48). Replicative plasmids were transformed into B. thailandensis and B. pseudomallei by using the rapid electroporation procedure previously described for Pseudomonas aeruginosa (16). Custom oligonucleotides were synthesized by Integrated DNA Technologies (Coralville, IA). DNA maps were constructed using Gene Construction Kit 2.5 (Textco, West Lebanon, NH) and then exported to Microsoft PowerPoint for final annotation.
Isolation of a TS pRO1600 replicon.
The pRO1600 wild-type replicon was mutagenized by error-prone PCR using pUCP28T containing the ori1600-rep sequences on a 1.2-kb fragment as the template and primers rep-UP (5'-TTCGTGAATTCGCCGCTGGT) and rep-DN (5'-GGCCTCTAGGCCAGATCCA). A 100-µl PCR mixture contained 100 ng of pUCP28T template DNA, 1x Taq buffer, 0.7 mM MgCl2, 0.3 mM MnCl2, 20 µM of dATP and dGTP, 0.1 mM of dCTP and dTTP, 0.51 µM of rep-UP and rep-DN, 5% dimethyl sulfoxide, and 5 units of Taq DNA polymerase (New England Biolabs). This reaction mixture was divided into 10-µl aliquots, which were then subjected to the following cycle conditions: one cycle of 95°C for 5 min; 30 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 1 min 25 s; and a final extension at 72°C for 10 min. The resulting PCR products were ligated into the pCR2.1 TA cloning vector (Invitrogen), and the ligation mixes were transformed into E. coli DH5 cells. Transformants were selected on LB plates containing 100 µg ml–1 Ap. Transformants were pooled, and plasmid DNA was extracted. The pooled plasmid DNAs were digested with EcoRI and blunt ended with T4 DNA polymerase. The 1.2-kb ori1600-rep fragment was ligated to ScaI-digested pFTP1 DNA and the transformation mixture transformed into E. coli HPS1. Transformants were selected at 37°C on LB plates containing 100 µg ml–1 Tp. For screening of plasmids containing a TS replicon, the plasmid pool obtained from these transformants was transformed into B. thailandensis E264, followed by selection of Tp-resistant (Tpr) transformants at 30°C. Individual transformants were patched onto replicate Tp-containing plates, which were then incubated at 30°C and 42°C. Plasmid DNA was isolated from several colonies which grew at 30°C but not at 42°C and was used to retransform E264, followed by selection of Tpr transformants and screening of TS phenotypes as described above. Five isolates were retained, and the entire sequence of the ori1600-rep insert contained on the respective plasmids was determined using primers rep-UP and rep-DN, but only one (pPS2158) was retained for further studies.
Determination of plasmid copy number.
The wild-type ori1600-rep and ori1600-rep(TsBt) regions were amplified from pUCP28T and pPS2158, respectively, using primers rep-UP (5'-TTCGTGAATTCGCCGCTGGT) and rep-DN (5'-GGCCTCTAGGCCAGATCCA). The resulting PCR fragments were cloned into TA cloning vector pCR2.1 (Invitrogen), generating pPS2194 and pPS2163. These two plasmids were electroporated into B. thailandensis E264, and transformants were selected on LB medium containing 100 µg ml–1 Km. Total DNA was isolated from transformants by using the QIAamp DNA mini kit (Qiagen). Plasmid copy number was determined as described by Lee et al. (33) with minor modifications. With this method, the separate detections of a plasmid and host chromosomal DNA were achieved using two separate primer sets, specific for the plasmid β-lactamase gene (bla) and for the chromosomal, single-copy aspartate-β-semialdehyde dehydrogenase gene (asd). To this end, the B. thailandensis asd gene was first cloned into the bla-containing pCR2.1 TA cloning vector to form pPS2191 and the copy number of this plasmid was calculated from the plasmid concentration by using the following equation: DNA copy number = (6.02 x 1023 [no. of copies/mol] x plasmid DNA amount [g])/(plasmid DNA length [bp] x 660 [g/mol/bp]). Tenfold serial dilutions of this plasmid and real-time quantitative reverse transcription-PCR were then employed for creating the standard curves for the asd and bla genes, where threshold cycle values were plotted against log copy number. Primers were designed using Primer 3 software and were as follows: for bla, bla-RT-UP (5'-ATAAATCTGGAGCCGGTGAG) and bla-RT-DN (5'-CTACGATACGGGAGGGCTTA), yielding an 81-bp product; and for asd, asd-RT-UP (5'-ACACGTCGTTCGTGTAGTCG) and asd-RT-DN (5'-AAAACGAGACCACGCTCAAG), yielding a 99-bp product. Real-time quantitative reverse transcription-PCR amplifications of the total DNA samples from two separate B. thailandensis E264 cultures harboring either pPS2194 or pPS2163 were performed simultaneously. The absolute copy numbers of bla and asd genes in the B. thailandensis total DNA samples were determined from the respective standard curves, using the threshold cycle values. The plasmid copy numbers of pPS2194 or pPS2163 were calculated by dividing the absolute copy number of bla by the copy number of asd. Experiments were repeated three times for calculation of copy number.
Promoter cloning.
The B. thailandensis ribosomal protein S12 gene promoter (PS12) or the P1 integron promoter (P1) were used for constitutive gene expression in Burkholderia spp. The two complementary synthetic oligonucleotides S12-UP (5'-GAGCTGTTGACTCGCTTGGGATTTTCGGAATATCATGCCGGGTGggcc) and S12-DN (5'-CACCCGGCATGATATTCCGAAAATCCCAAGCGAGTCAACAGCTCCtgca), specifying the PS12-containing region (predicted –35 and –10 boxes are underlined) from B. thailandensis E264 (62), were annealed. The resulting double-stranded oligonucleotide contains ApaI and PstI restriction enzyme overhangs (indicated by lowercase letters in the primer sequences) flanking the promoter sequences for direct subcloning into target plasmids by use of the same sites. The P1 integron promoter was PCR amplified from p34ETp1 (23) on a 383-bp PCR fragment by using primers P1-UP (5'-GAATTCACGAACCCAGTTGAC) and P1-Sma1-DN (5'-TACTTTGACCCGGGTCGAATC) (the EcoRI and SmaI sites are underlined), and the fragment was cloned into pCR2.1 to form pPS2225. The E. coli rhaBAD operon promoter (PrhaBAD) was used for regulated expression of genes in Burkholderia spp. (8). The region containing PrhaBAD and its upstream regulatory genes rhaS and rhaR was obtained by PCR amplification of a 2,066-bp fragment from E. coli DH5(
pir) chromosomal DNA by using primers PrhaB-UP (5'-CGACCATGGCATATGTGATGGTGCTGAATTTCA) and PrhaB-DN (5'-CATATGCATTTAATCTTTCTGCGAATTGAGATG). The resulting fragment was then cloned into pGEM-T Easy (Promega, Madison, WI), which yielded pPS2036. The integrities of all cloned promoter-containing fragments were verified by DNA sequencing.
Flp and Cre excision of chromosomally integrated antibiotic resistance markers.
Marker excision was performed using a previously described protocol (18) with minor modifications, as described below, for excision of a Zeor marker by use of either pFLPe4 or pCRE5. Briefly, Zeor strains were electroporated with 
150 ng of either pFLPe4 or pCRE5. Two aliquots (50 and 100 µl) and the concentrated rest of the mixture were plated on three LB plates containing 1,000 µg ml–1 Km (LB-Km1000) and incubated at 30°C for up to 48 h or until colonies were grown. Four Kmr transformants were struck for single colonies onto a four-sectored plate with LB-Km1000 plus 0.2% L-rhamnose, and the plate was incubated for 2 days at 30°C or until colonies were clearly discernible. Five colonies from each plate sector were then patched onto an LB-Km1000 plate and an LSLB plate containing 2,000 µg ml–1 Zeo (LSLB-Zeo2000), and the plates were incubated at 30°C for up to 48 h. Zeo-susceptible and Kmr cells were selected and streaked onto two or three LB plates for single colonies, and the plates were incubated at 42°C for 2 days or until colonies were discernible. Single colonies were tested at 30°C for susceptibility to Km and Zeo by patching them onto LB-Km1000, LSLB-Zeo2000, and LB plates. Km- and Zeo-susceptible cells were retained, and Flp or Cre-mediated marker excision was verified by colony PCR with ble-specific primers. The same protocol was used for Kmr marker excision with pFLPe2 and pCRE3, except that plasmid selection and recombinase induction steps were performed using LSLB-Zeo2000 plates. Rhamnose-inducible Flp and Cre gene expression can also be achieved in LB medium supplemented with 0.2% rhamnose and the appropriate antibiotics.
Isolation of an unmarked (amrRAB-oprA) deletion by PCR fragment transformation and Flp excision.
Gene disruption by deletion was performed using a previously described method (20). Briefly, using 5 to 50 ng of chromosomal or plasmid DNA templates, three partially overlapping DNA fragments representing flanking DNA segments and the Zeor marker from pFZE1 were amplified separately and then spliced together by an overlap extension PCR. To do this, the following fragments were amplified in a first-round PCR using the following primers: a 892-bp amrR upstream fragment using primers amrRAB-oprA-UP-For (5'-agggtgtccacatccttgaa) and amrRAB-oprA-UP-Rev-GM (5'-TCAGAGCGCTTTTGAAGCTAATTCGggacacttcaacggcaagat), a 828-bp oprA downstream fragment using amrRAB-oprA-DN-For-GM (5'-AGGAACTTCAAGATCCCCAATTCGgtcgccgaatacgagaagac) and amrRAB-oprA-DN-Rev (5'-gaaatacgccttgacgcact), and a 774-bp ble fragment using GmFRT-UP (5'-CGAATTAGCTTCAAAAGCGCTCTGA) and GmFRT-DN (5'-CGAATTGGGGATCTTGAAGTACCT). (Lowercase letters denote chromosome-specific sequences and uppercase letters FRT cassette-specific sequences.) These fragments were combined in a second PCR, and after gel purification, the resulting recombinant 2.45-kb DNA fragment was cloned into pGEM-T Easy, which yielded pPS2254. The plasmid-borne deletion was transferred to the B. pseudomallei 1026b chromosome by transformation of competent cells, using an unpublished procedure developed in Colin Manoil's laboratory at the University of Washington (Thongdee et al., presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand). Briefly, strain 1026b cells were grown in DM medium to an optical density at 600 nm of 0.5. Cells from 1 ml of culture were harvested by centrifugation, and the cell pellet was resuspended in 50 µl of DM medium. One hundred nanograms of pPS2254 insert DNA purified from an agarose gel was added to the concentrated cells, and the mixture was incubated without shaking for 30 min at room temperature. After addition of 2 ml DM medium, the cell suspension was incubated with shaking at 30°C for 24 h. Cells were harvested by centrifugation, washed once with 1 ml DM medium, and then resuspended in 0.25 ml of fresh DM medium. Fifty-microliter aliquots were then plated on LSLB-Zeo2000 plates, and the plates were incubated at 37°C for 48 h. Deletion of the amrR-amrAB-oprA genes was confirmed by colony PCR using a previously described procedure (18). Finally, an unmarked (amrRAB-oprA)::FRT deletion mutant was obtained by Flp-mediated excision of the Zeor marker, using pFLPe4. The presence of (amrRAB-oprA)::FRT was verified using colony PCR and primers amrRAB-oprA-UP-For and amrRAB-oprA-DN-Rev.
Transfer of mutations between B. pseudomallei chromosomes.
For transfer of antibiotic resistance-marked mutations, chromosomal DNA fragments (20 to 30 kb) were isolated using the QIAamp DNA mini kit, and the DNA was suspended in 200 µl of buffer AE (10 mM Tris-HCl, 0.5 M EDTA, pH 9). Approximately 150 ng of this DNA was added to competent cells, and bacteria were treated and plated as described above for recombinant DNA fragment transfer.
Site-specific transposition of mini-Tn7 elements.
Mini-Tn7 elements were transposed into the B. thailandensis and B. pseudomallei chromosomes, following the protocol previously described for Burkholderia mallei, using tri-parental mating for suicide vector delivery (14) and the helper plasmid pTNS3. pTNS3 was constructed by ligating a 375-bp EcoRI-SmaI fragment from pPS2225 between the same sites of pTNS2 (15). Insertion events were verified by colony PCR as previously described (14), utilizing the following primers. For B. thailandensis, the two PCR primers glmS1-DN (5'-GTTCGTCGTCCACTGGGATCA) and glmS2-DN (5'-AGATCGGATGGAATTCGTGGAG) were used to determine Tn7 insertions downstream of either glmS1 or glmS2 in combination with primer Tn7L (PTn7L; 5'-ATTAGCTTACGACGCTACACCC). For B. pseudomallei, insertions at the glmS1, glmS2, and glmS3 linked insertion sites were verified utilizing primers BPGLMS1 (5'-GAGGAGTGGGCGTCGATCAAC), BPGLMS2 (5'-ACACGACGCAAGAGCGGAATC), and BPGLMS3 (5'-CGGACAGGTTCGCGCCATGC) in combination with PTn7L. PCRs for B. thailandensis consisted of an initial denaturation step at 95°C for 2 min and then 30 cycles of 95°C for 45 s, 56°C for 30 s, and 72°C for 30 s, followed by a final extension at 72°C for 10 min. The PCR cycle conditions were the same for B. pseudomallei except that the annealing temperature was 61°C.
Nucleotide sequence accession numbers.
The plasmid sequences determined in this study were deposited in GenBank and assigned the following accession numbers: for pCRE3, EU215439; for pCRE5, EU215440; for pFGM1, AY597270; for pFKM2, EU215433; for pFLPe2, EU215437; for pFLPe4, EU215438; for pFRT2, AY597269; for pFRT3, EU240901; for pFTP1, AY712951; for pFZE1, DQ124230; for pLOX1, EU215436; for pLKM1, EU215434; for pLZE1, EU215435; for pTNS2, AY884833; for pTNS3, EU215432; for pUC18T-mini-Tn7T, AY599230; for pUC18T-mini-Tn7T-LAC, EU240902; for pUC18T-mini-Tn7T-Km-FRT, EU223384; for pUC18T-mini-Tn7T-Zeo-FRT, DQ493876; for pUC18T-mini-Tn7T-Km-lox, EU215442; for pUC18T-mini-Tn7T-Zeo-lox, EU215441; and for pUCP28T, U33751.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We therefore decided to isolate a new pRO1600(Ts) replicon. To do this, the wild-type replicon (consisting of ori1600 and the rep coding sequence) was mutagenized by error-prone PCR, and a plasmid pool containing the cloned, mutagenized PCR fragments was directly transformed into B. thailandensis for screening of the TS phenotype. TS replicon mutants were obtained at a frequency of 8%. Five representative TS plasmids were retained and their ori1600-rep sequences determined. Compared to the wild-type ori1600-rep sequence, all plasmids contained several mutations in the ori1600 or rep sequence. For example, TS replicon no. 1, which was retained for further studies, contained one mutation in the ori1600 region (an A-to-T change at nucleotide 986 of the published sequence; GenBank accession no. L30112) (60) and eight mutations in the rep gene sequence. Of the latter, three mutations did not change the Rep protein amino acid sequence, whereas five other mutations led to R19C, C31S, D145G, M157T, and G245R amino acid changes. We do not know whether all of these mutations are required for the observed TS phenotype. Intriguingly, the TS plasmids isolated in B. thailandensis (hereafter named TsBt) replicated in P. aeruginosa at 30°C and did not exhibit a TS phenotype. Plasmid stability at 30°C was not assessed.
Plasmids containing the TsBt replicon were efficiently (>99.9%) cured from cells grown at either 37 or 42°C. Their copy numbers were 1 (0.86 ± 0.23) in B. thailandensis cells grown at 30°C. By comparison, the copy numbers of plasmids containing the wild-type pRO1600 replicon were 3 to 4 (3.6 ± 0.85) in cells grown at the same temperature.
Construction of Flp and Cre recombinase-producing plasmids.
Previous experience with a broad-host-range P. aeruginosa Flp-FRT system indicated that when Flp recombinase was transiently expressed from a nonreplicative plasmid, excision of the FRT cassette was observed only at very low frequencies (0.5 to 1%) (50). In contrast, excision frequencies were >99.9% when Flp recombinase was expressed from a replicative plasmid which could be cured from the cells after marker excision by sacB-mediated sucrose counterselection (28). Because Burkholderia spp. contain endogenous sacB genes (40), this previously developed broad-host-range system could unfortunately not easily be adapted for use with Burkholderia spp.
We therefore used the newly developed TsBt replicon to derive curable plasmids producing Flp and Cre recombinases. The two most useful Flp-producing plasmids, pFLPe2 and pFLPe4 (Fig. 1), possess several desirable features: (i) they produce an Flp recombinase (Flpe) with enhanced recombinase activity over a wider temperature range (5); (ii) FLPe gene expression is under the control of the rhamnose-regulated PrhaBAD promoter, which was previously used for regulated gene expression in Burkholderia cenocepacia (8), and we reasoned that it could be used for the same purpose with other Burkholderia spp; (iii) they contain either a Km or a Zeo selection marker so that they are compatible with currently approved B. pseudomallei and B. mallei markers present on FRT cassettes; (iv) the ble and nptI genes are transcribed from the constitutive PEM7 and PS12 promoters, respectively, which enable expression of these genes in Burkholderia spp. (PEM7 is a synthetic bacterial promoter derived from the T7 promoter [Invitrogen]); (v) the plasmids contain an origin of conjugal transfer (oriT) which allows efficient introduction into strains or species recalcitrant to commonly used transformation methods; and (vi) the combination of low copy number (1 per cell) and the TsBt replicon allows for efficient plasmid curing.
|
To further extend the usefulness of site-specific excision technology with Burkholderia spp., we also constructed two Cre recombinase-producing plasmids, pCRE3 and pCRE5 (Fig. 1). With the exception of encoding Cre recombinase, these two plasmids possess the same useful features as those outlined above for the pFLPe plasmids. Diverse experiments demonstrated that Cre recombinase production was rhamnose inducible and that Cre-mediated site-specific excision was extremely efficient. Cre-mediated excision efficiencies in B. thailandensis were quite often significant, even in uninduced cells (ranging from 4 to 93%), and quite variable in rhamnose-induced cells (18 to 89%). Excision frequencies with pCRE5 were generally higher than those with pCRE3, approaching 90% in B. pseudomallei.
Excision efficiencies are probably not as high as those previously observed in P. aeruginosa (where they routinely approach 100%), because despite the high Zeo and Km concentrations used in these studies, we routinely found that not all colonies growing on the selection plates were true transformants but rather were antibiotic-tolerant cells which did not contain Cre- or Flp-producing plasmids.
Versatile Km, Tp, and Zeo resistance FRT and loxP cassettes.
The pFLPe vectors described in this study can be used with FRT Gm, Tp, and Zeo resistance cassettes contained on pFGM1, pFTP1, and pFZE1 (Table 1). To complement these cassettes and maximize the applications of the newly established Cre site-specific recombination systems, a number of new cassette vectors were constructed (Fig. 2). Plasmid pFKM2 contains a Kmr FRT cassette, and pLKM1 and pLZE1 contain Kmr and Zeor loxP cassettes. For subcloning into other plasmids or transposons or for insertional mutagenesis, individual markers can easily be excised from the respective plasmids by using unique restriction sites in the flanking polylinkers. Alternatively, because their priming sites are conserved on all plasmids, the GmFRT-UP and GmFRT-DN primers can be utilized to PCR amplify the individual markers from the respective FRT plasmids, and similar strategies can be employed for amplification of loxP resistance cassettes.
|
(amrRAB-oprA) deletion.(amrRAB-oprA)::FRT-ble-FRT fragment was derived by splice overlap extension (SOE) PCR as illustrated in Fig. 3A and was cloned into pGEM-T Easy. For transfer of the plasmid-borne deletion, the (amrRAB-oprA)::FRT-ble-FRT fragment was gel purified and transferred to the B. pseudomallei 1026b chromosome by transformation of competent cells. Finally, the unmarked (amrRAB-oprA)::FRT deletion strain Bp50 was obtained by Flp-mediated excision of the Zeor marker (ble), using pFLPe4. The presence of the (amrRAB-oprA)::FRT-ble-FRT and (amrRAB-oprA)::FRT mutation in the respective strains was verified using colony PCR, which yielded the expected 2.45- and 1.76-kb fragments. The (amrRAB-oprA)::FRT mutant Bp50 had deleted 49 of the 223 codons of amrR, the entire amrA and amrB genes, and 406 of the 513 codons of oprA. Flp-mediated excision was shown to leave behind a scar of 86 nucleotides (Fig. 3B). The (amrRAB-oprA)::FRT mutant exhibited the same drug resistance phenotypes as the previously isolated (amrRAB-oprA) B. pseudomallei strain DD503 (39) (Table 2).
|
|
(amrRAB-oprA) between B. pseudomallei strains.(amrRAB-oprA)::FRT-ble-FRT strain Bp44 were used to transform strains 1026b and 1710b. Zeor 1026b transformants were obtained at a frequency of about 180 colonies per µg of DNA. The same procedure yielded fewer transformants with 1710b. Of six Zeor 1026b and 1710b colonies analyzed by PCR, four 1026b transformants and one 1710b transformant contained the ble gene. The five ble-containing colonies showed the same antibiotic resistance profile as the parental Bp44 strain and thus contained the (amrRAB-oprA) deletion.
Establishment of a mini-Tn7-based single-copy gene integration system.
Mini-Tn7 elements are inserted into bacterial chromosomes at defined attTn7 sites (15, 43). With few exceptions (17), most of these sites lie in a neutral, intergenic region downstream of the essential glmS gene encoding glutamine-6-phosphate synthase. Burkholderia spp. are unique in that they contain multiple glmS genes and thus multiple attTn7 sites (14, 15), but insertions at the different sites can easily be distinguished by PCR utilizing insertion site-specific primers.
We previously described conjugation strategies for mini-Tn7 delivery into B. thailandensis and B. mallei by use of helper plasmid pTNS2 (14, 15). To increase mini-Tn7 transposition efficiencies in Burkholderia spp., pTNS3, in which the tnsABCD genes, encoding the site-specific Tn7 transposition pathway, are transcribed from tandem promoters, the P1 integron promoter (P1) and the E. coli lac operon promoter (Plac), was constructed. Mini-Tn7 transposition efficiencies with pTNS3 in conjugation transfer experiments (14) were substantially increased to several thousand colonies per plate, compared to those observed with pTNS2, which yielded several hundred colonies per plate.
Utilizing pTNS3, a mini-Tn7 element with a Zeor marker was inserted into the chromosome of B. pseudomallei 1026b, and insertion sites were determined utilizing primers BPGLMS1, BPGLMS2, and BPGLMS3, designed to yield PCR fragments of 218, 263, and 309 bp, respectively (Fig. 4). Insertions were readily obtained at all sites, although in >65% of all instances, the mini-Tn7 elements transposed to the glmS2 attTn7 site on chromosome 1. There was no obvious preference for either of the other two sites in the remainder of the transformants. Some transformants contained insertions at more than one attTn7 site, but while double insertions in two separate attTn7 sites were fairly common (10 to 20% with some strains), triple insertions were rarely observed. Sequence analysis revealed that all three attTn7 sites were located 25 bp downstream of the respective glmS genes. It should be noted that the C-terminal sequences of B. pseudomallei glmS2 and its immediate downstream sequences differ from those of B. mallei glmS1 and its immediate downstream sequences, the preferred mini-Tn7 insertion site in B. mallei, by only a single nucleotide. The same insertion sites and site preferences were observed when mini-Tn7 elements were transposed into strains K96243 and 08 (data not shown).
|
(amrRAB-oprA) strain.(amrRAB-oprA) strain Bp50 (Fig. 5). Since Bp50 is Gm susceptible, mini-Tn7-containing transformants could be selected on Gm-containing selective medium. As expected, mini-Tn7 insertions were observed at all three attTn7 sites, with the majority occurring downstream of glmS2 on chromosome 1, the same chromosome harboring the (amrRAB-oprA) mutation. We selected transformants that had mini-Tn7-LAC-amrA+B+-oprA+ integrated at glmS2 or glmS3 (for unknown reasons, we were unable to obtain insertions at glmS1 in this particular experiment). AmrAB-OprA efflux pump activity was restored to wild-type levels in all strains, irrespective of the location of the complementing amrA+B+-oprA+ operon, but only when amrAB-oprA gene expression was induced by addition of IPTG (isopropyl-β-D-thiogalactopyranoside) to the growth medium (Table 2).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In bacteria such as Burkholderia spp. that exhibit high intrinsic resistance to many antimicrobials or where the use of a resistance marker is restricted by select-agent regulations, one can envision recycling the same marker for isolation of consecutive chromosomal mutations and/or extrachromosomal element insertions in the same strain. As previously discussed (28, 49), one possible drawback to such an approach is that each Cre or Flp excision event leaves behind a loxP or FRT site, which could lead to undesirable recombination events upon reintroduction of the respective site-specific recombinases, causing deletions or inversion of chromosomal segments. However, these are extremely rare (during the many instances in which we have used the Flp-FRT system with P. aeruginosa, we have observed a single large chromosomal inversion between recombination sites that were 1.6 Mb apart [2]) but may be more common when recombination sites are spaced closely together on the chromosome, e.g., within the same operon. Such undesirable recombination events can easily be detected by PCR across recombination sites, which should be performed as a matter of routine. The 33-bp loxP and 86-bp FRT scars left behind after Cre and Flp excision of our cassettes are not substrates for the bacterial recombination machinery. We anticipate that the loxP and FRT cassettes will be valuable tools for micro- and macromanipulations of entire Burkholderia spp. chromosomes.
Many bacterial genes are organized in operons, and when promoter-proximal genes are deleted in such instances, it is desirable to isolate unmarked in-frame deletions to avoid possible unwanted effects on distal genes. This is usually achieved by merodiploid formation, resolution to either the wild-type or a mutant copy by use of counterselection against unwanted plasmid sequences, and screening for wild-type versus mutant alleles by phenotype and/or PCR. Aside from relying on spontaneous merodiploid resolution, such strategies are not feasible with wild-type bacteria, because no reliable counterselection marker that works with wild-type B. pseudomallei currently exists. Although the use of sacB-mediated counterselection has been reported by a few groups (10, 24), it is not known whether the deletion events were due to sacB-mediated sucrose toxicity rather than to spontaneous resolution events. Routine use of sacB-based counterselection in the closely related species B. mallei is possible only after deletion of the chromosomal sacB gene (40), and B. pseudomallei strains contain chromosomal sacB genes with 99 to 100% amino acid sequence identity to B. mallei sacB (e.g., BPSS0543 in K96243 and BURPS1710b_A2102 in 1710b). However, a recent report indicates that sacB-mediated counterselection in some B. pseudomallei strains may be possible under certain conditions (C. Logue, I. R. Peak, and I. R. Beacham, presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand).
Even though Cre and Flp excision events leave behind scars, these methodologies can also be used for isolation of in-frame deletion mutants. Besides serving as recombination sites, these scars are benign, i.e., they encode no promoter or transcriptional terminator sequences. However, they do contain translational stop codons; one TAG codon is found in the loxP sequence (5'-CTACTTCGTATAGCATACATTATACGAAGTTAT), and five stop codons are found in two separate frames in the FRT scar (Fig. 3A). In some instances, therefore, translation will be terminated within the scar sequences. Additionally, the scar can easily be placed in-frame during PCR-mediated assembly of the mutagenic DNA fragment by designing the gene-specific portions of the SOE PCR primers accordingly.
The one-step DNA fragment mutagenesis method is a rapid method which does not require counterselection and thus has widespread utility in Burkholderia wild-type strains. However, it has some drawbacks compared to the traditional method using a nonreplicative suicide plasmid. First, mutated fragments need to be tagged with an antibiotic resistance selection marker so that recombinational events can be selected. With the availability of Cre and Flp excision systems, however, this is not a serious impediment. Second, the method does not allow for transfer of single mutations, for example, TS alleles, into the chromosome without linking them to a selection marker, which, by definition, would introduce a secondary mutation. Such experiments will have to await development of counterselection methods which work with wild-type Burkholderia spp. Third, it may not work with all B. pseudomallei strains, e.g., K96243 (Thongdee et al., presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand). While we routinely use cloned and purified DNA fragments for mutagenesis, purified PCR products can be directly used for transformation (Thongdee et al., presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand).
In the absence of phage transduction, electroporation or natural transformation of bacteria with chromosomal DNA fragments containing a selectable marker is the only means of transferring mutations between chromosomes. A report in the Russian literature in the mid-1980s indicated that spontaneous transformation can be used to transfer DNA into and between B. pseudomallei strains (6), which was corroborated by recent observations (Thongdee et al., presented at the 5th World Melioidosis Congress, Khon Kaen, Thailand). We also demonstrated that natural transformation can be utilized to transfer antibiotic resistance-marked mutations between B. pseudomallei chromosomes, most efficiently within the same strain background but with reduced efficiency also between different strains. Natural transformation with linear fragments, as well as transfer of chromosomal DNA fragments, both containing a selectable marker, now allows combining of diverse mutations for rapid construction of isogenetic mutants, mutations in different strain backgrounds, etc. Unlike live strains, sterile DNA is exempt from select-agent rulings, and transfer between different laboratories does not require permits. "Mutants" can thus easily be shared between laboratories in the form of fragmented chromosomal DNA and reconstructed using linear fragment transfer, provided that the recipient uses an identical or compatible strain background and is approved for work with B. pseudomallei.
The unmarked (amrAB-oprA) strains generated in this study allow utilization of the Gmr marker for genetic manipulation in these strains, for example, complementation with mini-Tn7 elements (this study). Wild-type B. pseudomallei strains are highly streptomycin resistant because this antibiotic is an AmrAB-OprA substrate. As (amrAB-oprA) mutants are streptomycin susceptible (39), they allow use of streptomycin resistance markers for genetic manipulations. First, though to our knowledge they are currently not yet approved for use in B. pseudomallei, streptomycin resistance cassettes (45) could be utilized similarly to the Km and Zeo cassettes employed in this study. Second, spontaneous streptomycin-resistant rpsL mutant derivatives of the (amrAB-oprA) strains would allow utilization of streptomycin-based counterselection (56), which has previously been possible with strain DD503 only (58). However, there is evidence that B. pseudomallei rpsL mutants are attenuated in a murine model of pneumonic melioidosis (30) but not in the murine inhalation and hamster models (58), and such mutants may thus not be suitable for pathogenesis studies involving certain animal models.
Although a previous report in the Russian literature indicated the use of an RP1 TS replicon in B. pseudomallei (36), these plasmids have to our knowledge not been characterized in molecular detail and are not readily available. These shortcomings are overcome by the newly described pRO1600(Ts) replicon. It allows for rapid curing of plasmids from bacterial cells and will facilitate development of other genetic tools for these bacteria, for example, more-efficient transposon mutagenesis systems. While we do not know which of the five amino acid substitutions of the ori1600 Rep protein are responsible for the TS phenotype, we have not yet observed spontaneous revertants, indicating that more than one mutation is needed to convey a TS phenotype. Plasmids containing either the wild-type or the TS ori1600 replicon are low copy number in Burkholderia spp., which makes them useful for cloning of genes whose products may be detrimental to host cells when the genes are cloned in higher-copy-number plasmids.
In many instances, development of genetic tools for any given bacterium is hampered by lack of suitable promoters for either constitutive or regulated gene expression. In this study, we successfully used the P1 integron and the synthetic bacterial PEM7 and B. thailandensis S12 ribosomal gene (PS12) promoters for constitutive expression of selection markers in Burkholderia spp. Furthermore, we demonstrated the utility of the E. coli rhaBAD operon and tac promoters for regulated gene expression in B. pseudomallei and B. thailandensis. Although it is not shown in this study, we obtained evidence that the E. coli araBAD promoter can also be used for arabinose-regulated gene expression in B. thailandensis and, by extension, probably also in B. pseudomallei. In many instances, significant Cre-mediated marker excision was observed in uninduced cells carrying the pCRE5 plasmid, indicating that the PrhaBAD promoter may be leaky in LSLB-grown cells. It has been reported that PrhaBAD is responsive to low concentrations of rhamnose and can be effectively repressed with 0.2% glucose in B. cenocepacia (8), but we have not yet explored glucose repression in B. pseudomallei.
Construction of an improved transposase delivery and expression vector in conjunction with marker excision systems also enabled development of a mini-Tn7-based site-specific genome integration system for B. pseudomallei. The mini-Tn7 system has many advantages compared to other integration systems: (i) it is easy and efficient to use, and detailed user protocols have been published (14, 17, 18); (ii) mini-Tn7 elements can be readily transferred into different mutant backgrounds, thus facilitating comparative gene expression analyses; (iii) chromosomally integrated mini-Tn7 elements replicate stably without continued antibiotic selection for at least 100 bacterial generations; and (iv) they are highly versatile and can be used instead of plasmids for gene cloning and complementation, regulated expression from the E. coli tac promoter, isolation of transcriptional or translational β-galactosidase (lacZ) fusions, transcriptional bacterial luciferase (lux) fusions, and tagging with green fluorescent protein or its cyan, yellow, and red variants (18). In comparison to other bacterial species, Burkholderia spp. possess multiple glmS-linked attTn7 sites. As indicated by the occasional occurrence of multiple insertions in the same transformant, multiple insertion sites can be exploited to tag the same bacterium with different mini-Tn7 elements, for example, a gene fusion tag and a complementing expression element, allowing gene expression and complementation analyses using single-copy constructs, as nature intended it. In this paper, we demonstrated the use of the mini-Tn7 system for complementation of a (amrRAB-oprA) mutant. For this purpose, we selected transformants that had a mini-Tn7-LAC-amrA+B+-oprA+ expression vector integrated at glmS2 or glmS3. AmrAB-OprA efflux pump activity was restored to wild-type levels, irrespective of the genomic location—chromosome 1 or chromosome 2—of the complementing amrA+B+-oprA+ operon.
In summary, the genetic tools described in this paper allow routine select-agent-ruling-compliant manipulations of B. pseudomallei and thus facilitate much-needed studies of the biology and pathogenesis of this understudied yet fascinating bacterium. While they were not explicitly tested in this study, it can be anticipated that some of these tools will also be useful for studies of other Burkholderia spp., for example, B. mallei, for which similar select-agent restrictions apply. Though the methods described here were developed to allow compliance with guidelines in the United States, we also recommend their implementation in settings not governed by these rulings to avoid usage of selection markers encoding resistance to clinically significant antibiotics and to facilitate resource sharing.
|
ACKNOWLEDGMENTS |
|---|
We are indebted to Metawee Thongdee and Colin Manoil (University of Washington) for providing pCre1 and the protocol for PCR fragment transformation and preparation of Burkholderia competent cells prior to publication. We thank Don Woods (University of Calgary) and Mark Schell (University of Georgia) for the gift of B. pseudomallei strains, Phoebe Rice (University of Chicago) for providing pET24d-FLPe, and Tung Hoang (University of Hawaii) for pointing out use of the S12 promoter for constitutive gene expression in Burkholderia spp.
|
FOOTNOTES |
|---|
Published ahead of print on 21 December 2007.
Present address: Departamento de Biología Estructural, Instituto Venezolano de Investigaciones Científicas, Caracas, Venezuela.
Present address: Faculty of Health Sciences, 2000 Simcoe St. N, University of Ontario Institute of Technology, Oshawa, Ontario L1H 7K4, Canada.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
H, SacI-KpnI-SmaI-BamHI-XbaI-SalI-PstI-SphI-HindIII; K
