|Home | About | Journals | Submit | Contact Us | Français|
Coxiella burnetii is a ubiquitous zoonotic bacterial pathogen and the cause of human acute Q fever, a disabling influenza-like illness. C. burnetii's former obligate intracellular nature significantly impeded the genetic characterization of putative virulence factors. However, recent host cell-free (axenic) growth of the organism has enabled development of shuttle vector, transposon, and inducible gene expression technologies, with targeted gene inactivation remaining an important challenge. In the present study, we describe two methods for generating targeted gene deletions in C. burnetii that exploit pUC/ColE1 ori-based suicide plasmids encoding sacB for positive selection of mutants. As proof of concept, C. burnetii dotA and dotB, encoding structural components of the type IVB secretion system (T4BSS), were selected for deletion. The first method exploited Cre-lox-mediated recombination. Two suicide plasmids carrying different antibiotic resistance markers and a loxP site were integrated into 5′ and 3′ flanking regions of dotA. Transformation of this strain with a third suicide plasmid encoding Cre recombinase resulted in the deletion of dotA under sucrose counterselection. The second method utilized a loop-in/loop-out strategy to delete dotA and dotB. A single suicide plasmid was first integrated into 5′ or 3′ target gene flanking regions. Resolution of the plasmid cointegrant by a second crossover event under sucrose counterselection resulted in gene deletion that was confirmed by PCR and Southern blot. ΔdotA and ΔdotB mutants failed to secrete T4BSS substrates and to productively infect host cells. The repertoire of C. burnetii genetic tools now allows ready fulfillment of molecular Koch's postulates for suspected virulence genes.
The intracellular bacterium Coxiella burnetii causes the zoonotic disease Q fever. Symptomatic infections normally manifest as an acute, debilitating influenza-like illness with rare but serious long-term sequelae, including chronic endocarditis. The organism is highly infectious, environmentally stable, and usually transmitted to humans via inhalation of contaminated aerosols generated by animal husbandry operations. Sheep, goats, and dairy cattle are important animal reservoirs, with large-scale dairy goat farming in the Netherlands recently associated with the largest Q fever outbreak ever recorded (>4,000 cases) (21, 32).
C. burnetii has a unique intracellular lifestyle that involves acid-activated metabolism within a phagolyosome-like vacuole (15, 20). The abilities of C. burnetii to replicate to high numbers in the degradative confines of this compartment and to modulate host cell functions that promote replication vacuole biogenesis represent important pathogenic strategies (46). Many C. burnetii genes encoding putative virulence factors were revealed by genome sequencing, including a Dot/Icm type IVB secretion system (T4BSS) (6, 36). Several C. burnetii proteins have been shown to be translocated into the host cell cytoplasm in a T4BSS-dependent fashion that are predicted to have important effector functions (9, 10, 23, 29, 45, 47). However, the lack of methods for targeted gene inactivation has greatly hindered establishment of the functional roles of T4BSS effector proteins and other putative virulence factors in C. burnetii pathogenesis.
Genetic transformation of C. burnetii was first accomplished in 1996 by Suhan et al., who transformed the organism to ampicillin resistance using a shuttle vector containing a 5.8-kb C. burnetii autonomous replication sequence (40, 41). More than a decade elapsed before the next report of transformation, achieved using the mariner family transposon Himar1 to transform the organism to chloramphenicol resistance (4). These two genetic tools were developed using eucaryotic host cell-based propagation of C. burnetii, a limitation that imposes significant technical challenges. However, C. burnetii has recently been rescued from an obligate intracellular lifestyle by robust host cell-free (axenic) growth in a medium termed acidified citrate cysteine medium (ACCM) (27, 28). Axenic growth of C. burnetii in liquid medium and as clonal colonies on agarose plates substantially reduces the time of transformant isolation. Indeed, in the last 2 years axenic growth has enabled development of an improved Himar1 transposon system (5), RSF1010 ori-based shuttle vectors (3, 10, 27), a Tn7 system for single-copy, site-specific, in cis complementation (3, 5), and a system for inducible gene expression using an anhydrotetracycline (aTc)-inducible promoter (3).
A key remaining challenge in C. burnetii genetic manipulation is development of reliable protocols for targeted gene disruption. Conventional allelic-exchange methods generally rely on homologous recombination between an introduced mutant allele and a wild-type copy present on the chromosome. An important discovery made by Suhan et al. (41) is that homologous recombination occurs between the C. burnetii 5.8-kb autonomous replication sequence present on plasmid transformation DNA and the corresponding region of the chromosome. Subsequent genome sequencing further supported functional homologous recombination machinery in C. burnetii by revealing chromosomal rearrangements mediated by recombination between abundant insertion sequence elements as small as 1.1 kb in size (6, 36).
Several strategies for targeted bacterial gene inactivation exist that would exploit the homologous recombination ability of C. burnetii. Inactivation could occur via recombination between a wild-type gene and a linear DNA fragment containing a mutated allele (13). However, this strategy requires a low-frequency double-crossover event. Insertional duplication via single-crossover recombination with a suicide plasmid carrying an internal target gene fragment can result in target gene copies with 5′ and 3′ deletions (42, 44). Deficiencies of this method include spontaneous deletion of the integrated plasmid, resulting in reversion back to wild type, and difficultly in inactivating small genes due to suboptimal recombination substrate. A more sophisticated method of gene deletion exploits both single-crossover plasmid integration and a counterselectable marker (31). The first step involves chromosomal integration of a suicide plasmid carrying upstream and downstream regions of a target gene, and a counterselectable marker. In the next step, the “cointegrant” is resolved by a second recombination event between the plasmid encoded flanking region and the reciprocal region in the chromosome, resulting in removal of the wild-type gene. This “loop-in/loop-out” strategy (17, 25) frequently uses the counterselectable sacB, conferring sucrose sensitivity, which allows positive selection of the second crossover event, and can be configured to generate unmarked mutations (31, 34).
Protocols that exploit both bacterial homologous recombination and the activities of the heterologous, site-specific recombinases Flp and Cre have also proven successful in mutating/deleting bacterial genes (8, 22, 35, 38). Flp recombinase from Saccharomyces cerevisiae promotes recombination between two 34-bp Flp recombinase target sites (22, 35), whereas Cre recombinase from P1 bacteriophage promotes recombination between two 34-bp loxP sites (8, 38). The Cre-lox system was recently used to delete multiple genes of the Borrelia burgdorferi plasmid lp54 (8). Suicide plasmids containing kanamycin or streptomycin resistance genes, and a loxP site, were integrated into lp54 by single crossover events. In one deletion mutant, eight plasmid genes encompassing approximately 4,000 bp were completely or partially deleted following transformation with a third suicide plasmid encoding Cre recombinase.
Results of the present study expand the repertoire of C. burnetii genetic tools to include two protocols for targeted gene inactivation. As proof of principle, we describe deletion of dotA using Cre-lox-mediated recombination and deletion of dotA and dotB using a loop-in/loop-out system. Both strategies used sacB-mediated sucrose counterselection to select for mutants that had undergone the desired deletion event. Mutant strains did not secrete characterized T4BSS substrates and showed severe growth defects in mammalian host cells. Both phenotypes were rescued upon complementation.
The plasmids used in the present study are listed in Table 1. C. burnetii Nine Mile phase II (NMII) clone 4 was used in all transformation experiments and was propagated microaerobically in ACCM-2 or ACCM-2 agarose as previously described (27). Escherichia coli TOP10 (Invitrogen, Carlsbad, CA) was used for recombinant DNA procedures and cultivated in Luria-Bertani (LB) broth. E. coli transformants were selected on LB agar plates containing 50 μg of kanamycin/ml or 10 μg of chloramphenicol/ml. THP-1 cells, a human acute monocytic leukemia cell line (TIB-202; American Type Culture Collection [ATCC]), and African green monkey kidney (Vero) cells (CCL-81; ATCC) were maintained in RPMI 1640 medium (Invitrogen) containing 10% fetal calf serum (Invitrogen) at 37°C and 5% CO2. C. burnetii replication in host cells or ACCM-2 was measured by quantitative PCR of genome equivalents (GE) as previously described (20, 28) using a probe specific to CBU1206.
The plasmids used here are depicted in Fig. 1 and detailed descriptions of their construction are found in Table S1 in the supplemental material. Restriction enzymes were obtained from New England Biolabs (Ipswich, MA). PCR was performed using Accuprime Pfx or Accuprime Taq (Invitrogen). PCR primers were obtained from Integrated DNA Technologies (San Diego, CA), and their sequences are listed in Table S2 in the supplemental material. All cloning procedures were conducted using an In-Fusion PCR cloning system (BD Clontech, Mountain View, CA).
Electroporation of NMII was conducted as previously described (27). Selection of NMII “loop-in” transformants with chromosomal integration of a suicide plasmid was conducted by culture of bacteria in ACCM-2 containing kanamycin (final concentration, 350 μg/ml) and chloramphenicol (final concentration, 3 μg/ml). Resolution of sacB-encoding plasmid cointegrants was accomplished by subculture of transformants for 3 days in ACCM-2 supplemented with 1% sucrose and kanamycin. NMII strains containing gene deletions were subsequently expanded by culture in ACCM-2 containing kanamycin. For Cre-lox mediated recombination of transformants harboring cat-loxP and sacB-kan-loxP encoding plasmid cointegrants, transformants were electroporated with a cre-encoding suicide plasmid, followed by culture for 4 days in ACCM-2 to allow expression of Cre recombinase. The medium was then supplemented with sucrose (1% final concentration), and growth of the transformants continued for an additional 3 days to counterselect against organisms that had not undergone Cre-lox mediated recombination. Mutant NMII strains were cloned by picking colonies propagated on ACCM-2 agarose as previously described (27) or by limiting dilution in ACCM-2.
Genes conferring resistance to chloramphenicol, kanamycin, or ampicillin are approved for C. burnetii genetic transformation studies by the Rocky Mountain Laboratories Institutional Biosafety Committee and the Centers for Disease Control and Prevention, Division of Select Agents and Toxins Program.
Deletion of dotA and dotB was verified using PCR and Southern blotting. Verification by PCR was accomplished by amplifying an internal gene fragment with specific primer pairs (see Table S2 in the supplemental material), with wild-type and mutant genomic DNAs as a template. Southern blot verification was conducted by digesting wild-type and mutant genomic DNAs with BglII and EcoRI (dotA) or PstI and SalI (dotB), followed by separation of fragments by electrophoresis in a 0.8% agarose gel. DNA was transferred by blotting to Hybond N+ membranes (GE Healthcare, Piscataway, NJ) as described by Sambrook et al. (33), except that the transfer medium contained 0.4 M NaOH. Probe DNA specific to regions flanking either dotA or dotB were generated by PCR using specific primer pairs (see Table S1 in the supplemental material). A probe specific to the 1-kb Plus DNA marker (Invitrogen) was also used. Probe DNA (200 ng) labeling and subsequent blot hybridizations were conducted using instructions and reagents provided by a Gene Images AlkPhos Direct Labeling and Detection kit (GE Healthcare).
Cre-lox mediated deletion of Tn7 sequences bounded by loxP sites was verified by PCR using primers specific to the chloramphenicol acetyltransferase (CAT) gene and CBU1788 (see Table S2 in the supplemental material).
Expression of adenylate cyclase (CyaA) fusion proteins or DotB was assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting. Membranes were incubated with a rabbit polyclonal antibody directed against L. pneumophila (Philadelphia-1 strain) DotB or a mouse monoclonal antibody directed against Bordetella pertussis CyaA (clone 3D1; Santa Cruz Biotechnology, Santa Cruz, CA). C. burnetii and L. pneumophila DotB proteins display 63% amino acid identity. Reacting proteins were detected using anti-rabbit or anti-mouse IgG secondary antibodies conjugated to horseradish peroxidase (Pierce, Rockford, IL) and chemiluminescence using ECL Pico reagent (Pierce).
CyaA translocation assays were performed as previously described using a cyclic AMP (cAMP) enzyme immunoassay (GE Healthcare) (45, 47). Construction of the CyaA reporter plasmids pJB-CAT-CyaA, pJB-CAT-CyaA-A15, and pJB-CAT-CyaA-A16 has been reported elsewhere (45).
Vero cells infected with mutant or wild-type NMII were fixed for 20 min in 4% paraformaldehyde plus phosphate-buffered saline (PBS; 1 mM KH2PO4, 155 mM NaCl, 3 mM Na2HPO4 [pH 7.4]), followed by permeabilization for 5 min in 0.1% saponin in PBS. Cells were stained for indirect immunofluorescence as previously described (4, 19). Guinea pig anti-C. burnetii serum and a mouse monoclonal antibody directed against LAMP-3 (CD63) (clone H5C6; BD Biosciences) were used as primary antibodies. Alexa Fluor 488 and Alexa Fluor 594 IgG (Invitrogen) were used as secondary antibodies. Coverslips were mounted using ProLong Gold containing DAPI (4′,6′-diamidino-2-phenylindole; Invitrogen) to visualize nuclei. Epifluorescence microscopy images were acquired with a TE-2000 microscope equipped with a CoolSNAP HQ digital camera (Roper Scientific, Tucson, AZ). Images were obtained using Metamorph software (Molecular Devices, Inc., Downingtown, PA) and processed with ImageJ software (written by W. S. Rasband at the U.S. National Institutes of Health, Bethesda, MD [http://rsb.info.nih.gov/ij/]).
Statistical analyses were performed using a one-way analysis of variance and Prism software (GraphPad Software, Inc., La Jolla, CA).
As an initial step toward our goal to develop efficient methods of targeted gene deletion in C. burnetii, we tested the Cre-lox recombination system. This method relies on Cre recombinase catalysis of recombination between directly repeated 34-bp loxP recognition sites that flank a region targeted for deletion (1, 18).
To test whether the Cre-lox system functions in C. burnetii, we constructed a Tn7 derivative (pMiniTn7T-CAT::311P-MC-sacB) that contains a CAT gene (Cmr), and loxP sites flanking the mCherry red fluorescent protein-encoding gene (MC) and sacB driven as a single transcriptional unit by the P1 porin (CBU0311) promoter (311P) (5) (Fig. 1). We previously demonstrated that, consistent with other Gram-negative bacteria, Tn7 inserts as a single copy into an intergenic region of the C. burnetii chromosome immediately downstream from glmS (CBU1787) (5). The Bacillus subtilis sacB gene was included in the transposon as a counterselectable marker (31). The sacB gene product is the secreted enzyme levansucrase (sucrose: 2,6,-β-d-fructan 6-β-d-fructosyltransferase; EC 188.8.131.52) that converts sucrose to levans (high-molecular-weight-fructose polymers). Production of levans is toxic to most Gram-negative bacteria, including Legionella pneumophila, a close relative of C. burnetii (11, 14).
NMII was cotransformed with pMiniTn7T-CAT::311P-MC-sacB and pTnS2::1169P-tnsABCD (3), encoding the Tn7 transposase, and chloramphenicol-resistant transformants cloned from ACCM-2 agarose plates (Fig. 2a). Sequence analysis confirmed correct insertion of the Tn7 transposon. This strain was then transformed with a suicide plasmid encoding constitutively expressed cre (pUC19::1169P-cre) or cre under the control of an aTc-inducible promoter (pUC19::tetRAP-cre) (3) (Fig. 1). Organisms were cultivated for 4 days in ACCM-2, with aTc present in cultures of pUC19::tetRAP-cre transformants. PCR was conducted on Cre-expressing and -nonexpressing control organisms using primers specific to the CAT gene and CBU1788. Without Cre recombinase, a single 3,356-bp PCR product was amplified, indicating the Tn7T-CAT::311P-MC-sacB sequence was intact in these organisms (Fig. 2). With Cre recombinase, the 3,356-bp PCR product was the primary amplicon. However, a modest amount of a 639-bp PCR product was also produced, indicating some Cre-lox mediated excision of the mCherry-sacB gene cassette had occurred (Fig. 2).
The overall poor recovery of the desired deletion strain prompted us to test whether sucrose counterselection would improve recovery of bacteria having undergone Cre-lox-mediated recombination. NMII displayed no obvious growth defect in ACCM-2 containing up to 5% sucrose (data not shown), a concentration commonly used in sacB-based sucrose counterselection schemes (14, 16, 34). After transformation of NMII/Tn7T-CAT::311P-MC-sacB with pUC19::1169P-cre or pUC19::tetRAP-cre, organisms were cultivated in ACCM-2 for 4 days, and then sucrose was added to a final concentration of 0.5, 1, 2, 3, 4, or 5%, and the cultures were allowed to grow for 3 days. Growth was obvious only in medium containing 0.5 or 1% sucrose. Organisms subjected to 1% sucrose counterselection were then PCR genotyped. Complete excision of mCherry-sacB cassette, originally bounded by loxP sites, was observed with organisms transformed with pUC19::1169P-cre or pUC19::tetRAP-cre (Fig. 2). Thus, sacB-based sucrose counterselection substantially increased recovery of organisms having undergone Cre-lox mediated recombination by eliminating transformants that retained the intact Tn7 transposon.
Our optimized sacB-based Cre-lox system was then tested for efficacy in generating a targeted gene deletion. As proof of concept, dotA (CBU1648), which encodes a structural component of the T4BSS of C. burnetii, was targeted for deletion. The homologous protein in L. pneumophila is essential for type IVB secretion of effector proteins and productive infection of macrophages (7, 26).
NMII was sequentially transformed with the suicide plasmids pUC19-Kan-loxP-sacB::DotA5′flank and pJC-CAT-loxP::DotA3′flank (Fig. 1) that contain approximately 2,000 bp of chromosomal DNA immediately flanking the 5′ and 3′ regions of dotA, respectively. Correct chromosomal integration of both plasmids was confirmed by PCR (data not shown). This strain was subsequently transformed with pUC19::1169P-cre and cultured in ACCM-2 using sucrose counterselection as described above. Southern blotting of a mutant clone revealed 0.7-kb BglII/EcoRI and 1.85-kb EcoRI fragments as opposed to a 3.9-kb BglII/EcoRI fragment present in wild-type genomic DNA, indicating the deletion of dotA via Cre-mediated recombination between loxP sites (Fig. 3A and andB).B). PCR analysis also confirmed the absence of dotA (Fig. 3C). A schematic representation of Cre-lox mediated deletion of dotA is shown in Fig. S1 in the supplemental material.
To our knowledge, the NMII/ΔdotA-loxP mutant strain is the first example of a C. burnetii mutant generated by targeted gene deletion. In addition to deleting individual genes, the Cre-lox system has proven effective in deleting bacterial chromosomal regions as large as 67.3 kb (43). Such a procedure may prove useful in constructing a new generation of attenuated C. burnetii strains as vaccine candidates.
Gene deletions in C. burnetii can be generated using Cre-lox-based recombination. However, the procedure is somewhat cumbersome in requiring sequential transformation with three different suicide plasmids. Therefore, we developed a simplified loop-in/loop-out system that uses a single suicide plasmid and sacB counterselection. Using this approach, dotA and dotB (CBU1645) genes were targeted for deletion. The DotB homolog in L. pneumophila is a cytoplasmic ATPase that is essential for secretion of T4BSS substrates and productive infection of macrophages (37). Suicide plasmids were constructed containing a kanamycin resistance gene bounded by approximately 2,000 bp of 5′ and 3′ regions flanking the dotA gene (pJC-CAT::DotA5′3′-Kan) or the dotB gene (pJC-CAT::DotB5′3′-Kan) (Fig. 1). Plasmids also contained a cat-sacB gene cassette driven by the CBU1169 (hsp20) promoter (4). Resistance to chloramphenicol and kanamycin was used to select for cointegrant transformants (loop-in), and plasmid integration into the chromosome confirmed by PCR (data not shown). Cointegrant transformants were subcultured for 3 days in ACCM-2 containing 1% sucrose and kanamycin to select for transformants with resolved (loop-out) plasmid sequences. Selection for both sucrose and kanamycin resistance was conducted to promote recovery of a Kanr-marked deletion of dotA or dotB. Without selection for kanamycin resistance, there was roughly equal probability of recovering the wild-type chromosome by simple in toto excision of introduced plasmid DNA (17) (see Fig. S2 in the supplemental material).
The putative dotA and dotB deletion mutants were cloned using ACCM-2 agarose, expanded in ACCM-2, and then their genomic DNA was examined by Southern blotting. Replacement of dotA with the Kan cassette resulted in a predicted 2.5-kb BglII/EcoRI fragment as opposed to a 3.9-kb fragment present in wild-type DNA (Fig. 4A and andB).B). Replacement of dotB with the Kan cassette resulted in predicted 1.2- and 4.1-kb PstI/SalI fragments as opposed to the 2.0- and 3.4-kb fragments present in wild-type DNA (Fig. 4D and andE).E). The availability of DotB antibody allowed confirmation by immunoblotting of defective protein production by the ΔdotB mutant (Fig. 4G). PCR analysis also showed a complete absence of dotA (Fig. 4C) or dotB (Fig. 4F) in mutant strains.
L. pneumophila dotA and dotB mutants are defective in secretion of Dot/Icm T4BSS substrates, a deficiency that correlates with an inability to productively infect mammalian host cells. The same phenotypes are associated with C. burnetii strains harboring transposon insertions in icmD (3) and icmL (9). To test whether deletion of dotA or dotB is associated with defects in T4BSS secretion and intracellular growth, CyaA translocation (3) and Vero cell growth assays (12), respectively, were conducted.
THP-1 macrophages were infected with NMII or Δdot mutants transformed with the plasmids pJB-CAT-CyaA-A15 or pJB-CAT-CyaA-A16 that encode CyaA fusions to the C. burnetii T4BSS effector proteins CpeD and CpeE, respectively (45). CpeD and CpeE fusion proteins were secreted by NMII as indicated by a >100-fold increase in cAMP levels relative to organisms expressing CyaA alone (Fig. 5). Conversely, the AMP levels generated by the Δdot mutants did not exceed the CyaA-alone negative control level, indicating no secretion (Fig. 5). Negative secretion was not due to a lack of CyaA fusion protein since immunoblotting revealed equal amounts of fusion protein produced by NMII and the Δdot mutants (data not shown).
To test whether the lack of secretion of two defined C. burnetii Dot/Icm substrates correlated with an inability to productively infect mammalian cells, growth measurements of the Δdot mutants were made at 6 days postinfection of Vero cells. Axenic growth of NMII and the Δdot mutants was indistinguishable, with all strains attaining roughly 1,000-fold increases in GE after 6 days of incubation (Fig. 6A). Conversely, ΔdotA and ΔdotB mutants exhibited only 3.3- and 4.6-fold increases, respectively, in GE at 6 days postinfection of Vero cells compared to a 1,000-fold increase by NMII (Fig. 6B). This modest amount of replication was previously observed for a NMII icmD mutant in THP-1 macrophages and was attributed to the mutant's ability to undergo a few rounds of genomic replication in the acidic but nutritionally deficient phagolysosome (3).
To confirm that deficient intracellular replication of the Δdot mutants was due to their respective gene deletions, complementation studies were performed. C. burnetii dotB is the third gene in a predicted operon with dotD and dotC, while dotA is the second gene in a predicted operon with icmV (24). Thus, rescue of the intracellular growth defects of Δdot mutants was attempted by complementation with Tn7 constructs encoding dotD, dotC, and dotB (pMiniTn7-CAT::dotDP-dotDCB) or icmV-dotA (pMiniTn7-CAT::icmVP-icmV-dotA) (Fig. 1). Constructs contained approximately 200 bp upstream of their respective operon predicted to encode an endogenous promoter. Transformation of the ΔdotA and ΔdotB mutants with Tn7 constructs resulted in functional complementation as scored by GE increases similar to NMII at 6 days postinfection (Fig. 6B).
The lack of intracellular replication by ΔdotA and ΔdotB mutants correlated with failed production of the large and spacious replication vacuole typical of C. burnetii infection (Fig. 7). Rather, mutant bacteria were randomly dispersed in LAMP-3-postive, tight-fitting vacuoles. The normal vacuole phenotype was rescued when mutants were transformed with the complementing Tn7 constructs (Fig. 7).
In summary, C. burnetii's historic obligate intracellular lifestyle has thwarted the development of systems for pathogen genetic manipulation. The present study details two efficient methods for targeted gene disruption, the development of which was facilitated by new axenic culture techniques (27). Although this work generated marked gene deletions, the Cre-lox and loop-in/loop-out approaches described here are adaptable to generation of markerless deletions or mutations. A markerless approach is desirable when making several mutations in a single genome and eliminates potential polar effects of marker genes. The loop-in/loop-out strategy also has the added advantage of generating “scarless” mutations completely devoid of cointegrant plasmid sequences (31), although the 34-bp loxP site scar remaining after Cre-mediated recombination is reported to not exert polar effects on downstream genes (30). Markerless approaches would necessitate additional screening of the 0.5-mm C. burnetii colonies that form on ACCM-2 agarose, a procedure that is still technically challenging. Nonetheless, the targeted gene inactivation techniques described here now permit routine mutation and complementation strategies for C. burnetii virulence factor discovery.
We thank Philip Stewart for critical review of the manuscript, Aaron Bestor for pBSV25-flgBp-cre, and Joseph Vogel for anti-DotB antibody.
This study was supported by the Intramural Research Program of the National Institutes of Health, National Institute of Allergy and Infectious Diseases
Published ahead of print 20 April 2012
Supplemental material for this article may be found at http://aem.asm.org/.