Construction and Characterization of a Highly Efficient Francisella Shuttle Plasmid

Bacterial strains, plasmids, and growth conditions.Bacterial strains and plasmids used in this study are listed in Table 1. F. tularensis LVS and F. novicida U112 were routinely grown at 37°C in modified Mueller-Hinton (MH) broth or on MH agar (Difco Laboratories, Detroit, Mich.) containing 0.1% glucose, 0.025% ferric pyrophosphate, and 2% IsoVitaleX (Becton Dickinson, Cockeysville, Md.); MH agar also contained 2.5% donor calf or fetal bovine serum (Invitrogen, Carlsbad, Calif.). Alternatively, cysteine heart agar (Difco) with 5% defibrinated horse blood (Becton Dickinson) was used. When required, the growth medium for F. tularensis LVS and F. novicida U112 was supplemented with kanamycin (10 μg ml⁻¹ for LVS and 30 μg ml⁻¹ for U112) or streptomycin (100 μg ml⁻¹) for selection of rpsL alleles. All manipulations with Francisella spp. were performed in a class II biological safety cabinet (The Baker Company, Sanford, Maine). E. coli DH5α and HB101 were grown at 37°C aerobically in Luria-Bertani (LB) medium (Difco) supplemented with kanamycin (50 μg ml⁻¹), streptomycin (100 μg ml⁻¹), or ampicillin (100 μg ml⁻¹) when required. All antibiotics were purchased from Sigma-Aldrich (St. Louis, Mo.) or United States Biochemical Corporation (Cleveland, Ohio).

DNA manipulations.Restriction enzyme digestion, cloning, subcloning, and DNA electrophoresis were done by using standard techniques (32). PCR was performed by using Deep Vent DNA polymerase (New England Biolabs, Beverly, Mass.), Platinum Pfx DNA polymerase (Invitrogen), or the Failsafe PCR system (Epicentre, Madison, Wis.). Ligations were performed by using T4 DNA ligase (Invitrogen or New England Biolabs). Plasmid DNA was prepared with a QIAprep spin miniprep kit or a QIAGEN plasmid midi kit (QIAGEN, Valencia, Calif.). Francisella genomic DNA was isolated from cell pellets treated with 10% sodium dodecyl sulfate and proteinase K, followed by precipitation with cetyltrimethylammonium bromide (39), or by using a QIAGEN Genomic-Tip kit (QIAGEN). DNA fragments were purified with either a QIAquick gel extraction kit or a QIAquick PCR purification kit (QIAGEN). DNA sequencing was performed by using an ABI PRISM BigDye terminator cycle sequencing Ready Reaction kit and an automated long capillary method (ABI PRISM 3100 genetic analyzer; Applied Biosystems, Foster City, Calif.). Computer-assisted sequence analysis was performed with Sequencher (Gene Codes Corporation, Ann Arbor, Mich.). Custom oligonucleotide primers were synthesized by QIAGEN, and the primers used in this study are listed in Table 1. DNA maps were constructed by using MacPlasmap Pro (CGC Scientific, Inc, Ballwin, Mo.).

Electroporation and chemical transformation.Aerobically grown mid-logarithmic-phase electrocompetent E. coli cells were prepared by using 10% glycerol, and electroporation was performed by using a Bio-Rad Gene Pulser II with a 0.2-cm cuvette as recommended by the manufacturer. The gene pulser settings included a voltage of 2.5 kV, a capacitance of 25 μF, and a resistance of 200 Ω. After electroporation, cells were immediately placed into 1 ml of SOC medium (32), incubated with shaking for 1 h at 37°C, and plated on LB agar with the appropriate antibiotic(s). Chemical transformation of E. coli was performed as described previously (19).

For electroporation of Francisella spp., MH broth cultures (50 ml) were grown to the mid-exponential phase (optical density at 550 nm, 0.3 to 0.6), washed two times with 0.5 M sucrose (Boehringer Mannheim Biochemicals, Indianapolis, Ind.), and suspended in ∼1 ml of 0.5 M sucrose to obtain a concentration of ∼1 × 10¹⁰ cells ml⁻¹. For electroporation, ∼1 μl of plasmid DNA (∼100 μg/ml) was mixed with 200 μl of electrocompetent cells, incubated at room temperature for 10 min, and electroporated in a 0.2-cm cuvette by using the following gene pulser settings: 2.5 kV, 25 μF, and 600 Ω. Immediately after electroporation, cells were suspended in 1 ml of MH broth and incubated at the appropriate temperature (30 or 37°C) for 4 h (F. tularensis LVS) or 2 h (F. novicida U112) before selection on MH agar containing the appropriate antibiotic(s).

F. tularensis LVS macrophage infection and fluorescent microscopy.F. tularensis LVS derivatives were grown to the mid-exponential phase in MH broth and suspended in prewarmed Dulbecco's modified Eagle's medium (DMEM) (Sigma) supplemented with 10% fetal bovine serum (Invitrogen) and 4 mM l-glutamine (Invitrogen). The murine BALB/c macrophage cell line J774A.1 (ATCC TIB-67) was maintained in supplemented DMEM at 37°C in humidified air containing 5% CO₂. For infection, 4 × 10⁵ macrophages were seeded into wells of a 12-well tissue culture plate and infected with F. tularensis LVS at a multiplicity of infection (MOI) of 1 bacterium per macrophage in DMEM. Macrophages were allowed to ingest bacteria for 2 h before extracellular bacteria were removed by two washes with DMEM. At various times (2 h and 1, 2, and 3 days), macrophage monolayers were lysed with distilled water, and the number of CFU of F. tularensis LVS per milliliter was determined by plating various dilutions prepared in phosphate-buffered saline (PBS). As previously reported (2, 13) and also observed by us (data not shown), DMEM is an unfavorable environment for F. tularensis LVS growth, which eliminated the need for gentamicin treatment for removal of extracellular bacteria.

For fluorescent microscopy, 8 × 10⁵ J774A.1 macrophages were seeded in 35-mm dishes containing 8-cm² coverslips, and infection was carried out as described above. Cells were fixed with 3.7% paraformaldehyde for 10 min at room temperature, washed twice with PBS, permeabilized with 0.2% Triton X-100 in PBS for 5 min at room temperature, and then washed with three changes of 0.2% Triton X-100 in PBS over 5 min. Mouse monoclonal F. tularensis anti-lipopolysaccharide (LPS) antibody (1:2,000 dilution; Advanced ImmunoChemical Inc., Long Beach, Calif.) diluted in PBS with 0.2% Triton X-100 and 3% bovine serum albumin was then added and incubated for 1 h at room temperature. The cells were washed three times with 0.2% Triton X-100 in PBS and incubated with Cy3-conjugated goat anti-mouse monoclonal immunoglobulin G (IgG) antibody (1:13,500 dilution; Jackson ImmunoResearch Laboratories, Inc., West Grove, Pa.) diluted in PBS with 0.2% Triton X-100 and 3% bovine serum albumin for 20 min at room temperature. Control slides treated with the Cy3 monoclonal IgG antibody but not the primary F. tularensis anti-LPS antibody were also analyzed to confirm the specificity of the signal. Finally, macrophages were washed three times with 0.2% Triton X-100 in PBS and mounted on slides by using a ProLong Anti-Fade kit (Molecular Probes, Eugene, Oreg.). Green fluorescent protein (GFP) and Cy3 fluorescent images were captured by using an Eclipse 600 epifluorescence microscope equipped with a ×100 N.A. 1.4 objective (Nikon, Melville, N.Y.) interfaced to a Micromax-Interline Transfer charge-coupled device camera (Princeton Instruments, Inc., Monmouth Junction, N.J.) and Metamorph imaging software (Universal Imaging Corp., Downingtown, Pa.).

Western blot analysis.The total protein of F. tularensis LVS cell lysates was determined by using a bicinchoninic acid protein assay kit (Pierce, Rockford, Ill.) and a Molecular Devices Spectra Max 340 with the Softmax Pro 4.0 software (Molecular Devices Corporation, Sunnyvale, Calif.). Protein samples were boiled for 5 min in loading buffer, separated on 10% polyacrylamide-sodium dodecyl sulfate gels, and transferred to nitrocellulose membranes (Schleicher and Schuell BioScience, Inc., Keene, N.H.) by standard methods. The membranes were blocked with 5% skim milk for 30 min and then incubated with primary antibody overnight with gentle shaking at 4°C by using either a mouse monoclonal F. tularensis anti-LPS antibody (1:8,000 dilution; Advanced ImmunoChemical Inc.) or a mouse monoclonal anti-GFP antibody (1:5,000 dilution; Covance, Princeton, N.J.). The blots were washed with Tris-buffered saline before application of a mouse horseradish peroxidase secondary antibody (1:10,000 dilution; Roche Molecular Biochemical, Indianapolis, Ind.) for 2 h with gentle shaking at room temperature. The SuperSignal West Pico chemiluminescent substrate (Pierce) was used to detect bound peroxidase. The membranes were exposed to BX57 film (Midwest Scientific, St. Louis, Mo.). Quantitation of GFP was performed by using purified recombinant GFP (Clontech, Palo Alto, Calif.) as a standard, and the relative concentration of GFP was assessed by using NIH Image (http://rsb.info.nih.gov/nih-image/).

Isolation of pFNLTP1 derivatives with expanded multiple cloning sites or with a temperature-sensitive Francisella replicon.Several multiple cloning sites (designated MCS1 to MCS4) were designed as oligonucleotides and used to introduce new restriction enzyme sites into pFNLTP1, creating pFNLTP5 to pFNLTP8 (Table 1). Complementary oligonucleotides containing KpnI and BamHI restriction enzyme sites at either end were annealed at 95°C for 2 min and cooled to room temperature. pFNLTP1 was digested with KpnI and BamHI before ligation with the annealed oligonucleotides. All derivatives were confirmed by DNA sequencing. For isolation of temperature-sensitive derivatives of pFNLTP1, plasmid DNA was treated with 7 M hydroxylamine overnight at 37°C or for 1 h at 70°C and subsequently dialyzed. The hydroxylamine-treated pFNLTP1 was electroporated into F. tularensis LVS, and various dilutions were plated on MH agar with kanamycin (10 μg ml⁻¹) at the permissive temperature (32°C). Plates containing 100 to 200 colonies were then replica plated onto MH agar containing kanamycin and incubated either at the permissive temperature (32°C) or at the nonpermissive temperature (42°C). Individual transformants that grew well at 32°C but failed to grow at 42°C were recovered as temperature sensitive and were characterized further. One isolate, pFNLTP9, was chosen for further study, and DNA sequencing was performed to determine the mutation responsible for the temperature-sensitive phenotype.

GFP expression in F. tularensis LVS with pFNLTP1 derivatives.A promoterless gfp gene (25) was amplified with primers G1 and G2 and cloned into the BamHI site of pFNLTP6 (Table 1). Francisella promoters were amplified from the F. tularensis LVS genome by using primers P1 and P2 specific for the groE operon (11) or primers P3 and P4 specific for acpA (31) and were cloned upstream of gfp by using KpnI (Table 1). All constructs were verified by DNA sequencing.

Complementation of streptomycin-resistant E. coli HB101 and F. tularensis LVS strains with pFNLTP1.The wild-type rpsL gene and its promoter region were amplified from F. tularensis LVS genomic DNA (rpsL_LVS) by using primers R1 and R2 and were cloned into BamHI-digested pFNLTP1 (Table 1). Both pFNLTP1 and pFNLTP1 rpsL_LVS were electroporated into the E. coli HB101 and F. tularensis LVS rpsL strains, and their sensitivity to streptomycin was assessed by growth in the presence and absence of streptomycin. A spontaneous F. tularensis LVS rpsL mutant was isolated on MH agar containing 100 μg of streptomycin ml⁻¹, and the genotype of rpsL was confirmed by DNA sequencing.

Nucleotide sequence accession numbers.The entire pFNLTP1 and pFNLTP9 sequences have been deposited in the GenBank database under accession numbers AY622904 and AY622905, respectively. The wild-type rpsL gene of F. tularensis LVS and the mutant rpsL gene of F. tularensis LVS rpsL have been deposited in the GenBank database under accession numbers AY622906 and AY622907, respectively.

Isolation and characterization of pFNLTP1 and other deletion derivatives of pTOPO/FNL10.Several plasmid derivatives of the cryptic Francisella plasmid pFNL10 have been constructed (22, 27-30). These plasmids can replicate in Francisella but generally are not optimal for genetic studies because of their variable transformation efficiencies, failure to replicate in E. coli, or instability in vivo (27, 29). To develop a useful shuttle plasmid for Francisella, we fused pCR2.1-TOPO and pFNL10 (SpeI/NheI ligation) and selected for kanamycin-resistant transformants in E. coli DH5α (Fig. 1A). The resulting plasmid, pTOPO/FNL10, was then introduced into F. tularensis LVS by electroporation, and transformants were selected on MH agar containing kanamycin. A single kanamycin-resistant colony was obtained and was demonstrated to be F. tularensis LVS by Western blot analysis of cellular lysates performed with a mouse monoclonal antibody specific for F. tularensis LPS (data not shown). To confirm that antibiotic resistance was a result of pTOPO/FNL10 introduction and not due to a spontaneous mutation, plasmid DNA was isolated from the single transformant, and several restriction enzyme digestions were performed (Fig. 1B). This analysis confirmed the presence of a replicating plasmid in F. tularensis LVS. Restriction mapping indicated that the plasmid isolated from F. tularensis LVS (designated pFNLTP1) was genetically distinct from the original plasmid, pTOPO/FNL10 (Fig. 1). For example, digestion of pTOPO/FNL10 with SphI produced 4-, 2.4-, and 1.5-kb fragments, while digestion of pFNLTP1 with SphI resulted in 4- and 2.9-kb fragments. Similarly, digestion of pTOPO/FNL10 with BglII produced 5- and 3-kb fragments, while digestion of pFNLTP1 with BglII produced 5- and 2-kb fragments. Taken together, these restriction enzyme digestion results suggested that ∼1 kb of DNA had been spontaneously deleted in pFNLTP1. Next, the sequence of pFNLTP1 was determined to map the exact location of the deletion (Fig. 1A). This analysis confirmed that approximately 1 kb of DNA, corresponding to regions present in both pCR2.1-TOPO and pFNL10, was lost (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 1.

Isolation of E. coli and F. tularensis shuttle plasmid pFNLTP1 and additional deletion derivatives of pTOPO/FNL10. (A) Plasmid maps of pTOPO/FNL10 and the derivative pFNLTP1 isolated from F. tularensis LVS due to spontaneous deletion of pTOPO/FNL10. The KpnI (K), SacI (S), and BamHI (B) sites are unique restriction enzyme sites on the plasmids. Additional restriction sites on the maps include SphI (Sp) and BglII (Bg) sites. The genes or origins from pCR2.1-TOPO are indicated by light gray arrows and boxes, while those from pFNL10 are indicated by dark gray arrows. The β-lactamase gene (bla) encodes ampicillin resistance, while the neomycin phosphotransferase gene (npt) encodes kanamycin resistance. The deletion in pFNLTP1 is indicated by the striped box; orfm, orf4, a portion of orf5 (136 to 255 bp of the coding sequence), and a portion of f1 ori (1 to 99 bp) are included in the deletion. (B) Restriction enzyme digestion of pTOPO/FNL10 (lanes 1 and 3) and pFNLTP1 (lanes 2 and 4) with either SphI (lanes 1 and 2) or BglII (lanes 3 and 4). The sizes of the DNA markers (in kilobases) are indicated on the left. (C) Additional spontaneous deletion derivatives pFNLTP2, pFNLTP3, and pFNLTP4 obtained following electroporation of pTOPO/FNL10 into F. tularensis LVS. Regions deleted from the plasmids are indicated by the striped boxes. Note that the deletions in all the derivatives encompass orf4 and orfm from pFNL10 and fl ori from pCR2.1-TOPO; the pFNLTP2 deletion includes orfm, orf4, fl ori, and a portion of orf5 (76 to 255 bp of the coding sequence), the pFNLTP3 deletion includes orfm, fl ori, and a portion of orf4 (169 to 204 bp of the coding sequence), and the pFNLTP4 deletion includes orfm, orf4, a portion of orf5 (238 to 255 bp of the coding sequence), and a portion of fl ori (1 to 288 bp).

Repetition of the F. tularensis LVS electroporation experiments with pTOPO/FNL10 yielded three additional plasmid derivatives (pFNLTP2 to pFNLTP4) (Fig. 1C). All transformants isolated in F. tularensis LVS possessed plasmid DNA whose restriction pattern differed slightly from that of pFNLTP1 (data not shown). This analysis indicated that similar to pFNLTP1, pFNLTP2 to pFNLTP4 were deletion products in which DNA surrounding the ligation junction between the original two plasmids was missing. To confirm this observation, the region between repA of pFNL10 and npt of pCR2.1-TOPO was sequenced to map the precise boundaries of each deletion. All plasmids derived from F. tularensis LVS passage had complete deletions of orfm (pFNL10) and partial or complete deletions of orf4 (pFNL10) and the fl ori (pCR2.1-TOPO).

Efficiency of transformation of pFNLTP1 and pTOPO/FNL10 into E. coli DH5α, F. tularensis LVS, and F. novicida U112.A comparative analysis was performed to measure the frequencies of transformation between the original pTOPO/FNL10 plasmid and one of the deletion derivatives isolated from F. tularensis LVS (pFNLTP1). pTOPO/FNL10 had a high frequency of transformation (∼5 × 10⁷ CFU μg of DNA⁻¹) in DH5α (Fig. 2, bar A). As in our initial experiments, few transformants were obtained with F. tularensis LVS (∼3 CFU μg of DNA⁻¹) or F. novicida U112 (no transformants) (Fig. 2, bars B and C). In contrast, pFNLTP1 was electroporated into both E. coli DH5α and F. tularensis LVS with similar high frequencies (∼3 × 10⁷ to 6 × 10⁷ CFU μg of DNA⁻¹) (Fig. 2, bars D and E). These frequencies of DNA inheritance are at least 2 logs higher than the previously reported frequency for chemical, cryo-, or electrotransformation into any Francisella species (27, 28, 30). In addition, only a 10-fold decrease in electroporation frequency with pFNLTP1 was observed in F. tularensis LVS when frozen competent cells were used instead of freshly prepared cells (data not shown). When F. novicida U112 was made electrocompetent, a transformation frequency of approximately 2 × 10³ CFU μg of DNA⁻¹ was observed for pFNLTP1 derived from F. tularensis LVS (Fig. 2, bar F). This is similar to the frequency observed following electroporation of F. novicida U112 with pFNLTP1 derived from E. coli DH5α (Fig. 2, bar J). Interestingly, this frequency increased substantially when plasmid DNA was isolated from F. novicida U112 and used in subsequent electroporation experiments with F. novicida U112 (∼2 × 10⁸ CFU μg of DNA⁻¹) (Fig. 2, bar H). In contrast, the frequency of transformation of pFNLTP1 from F. novicida U112 or E. coli DH5α into F. tularensis LVS remained unchanged (Fig. 2, bars G and I). From this series of experiments, we concluded that pFNLTP1 can be electroporated into F. tularensis LVS at a high frequency. In addition, F. novicida U112 may contain a restriction-modification system that limits the rate of transformation with pFNLTP1.

• Open in new tab
• Download powerpoint

FIG. 2.

Efficiency of electroporation of pTOPO/FNL10 and pFNLTP1 into E. coli DH5α or Francisella spp. (F. tularensis LVS or F. novicida U112). The transformation efficiency, expressed as log CFU per microgram of DNA (mean ± standard deviation), was determined from three individual preparations of competent cells. Competent cells of E. coli were prepared with 10% glycerol, while 0.5 M sucrose was used to prepare competent cells of F. tularensis LVS and F. novicida U112. Most electroporations were performed with 100 ng of plasmid DNA; the only exception was the electroporation of pTOPO/FNL10 into F. tularensis LVS, for which 500 ng was used. The strains indicated in parentheses are the hosts from which plasmid DNA was isolated.

Stability of pFNLTP1 in F. tularensis grown in vitro and in J774A.1 macrophages.The stability of plasmids in hosts grown in vitro and in vivo is an important consideration when genetic studies in which virulence determinants will be characterized are undertaken. pFNLTP1 was maintained in F. tularensis LVS for 3 days during growth in J774A.1 macrophages at a 99.9% efficiency in the absence of selection (data not shown) and also grew and replicated in J774A.1 macrophages to an extent similar to that seen with wild-type strain LVS. We observed a 4-log increase in the number of CFU recovered from the macrophages over a 3-day incubation period (Fig. 3). Growth of F. tularensis LVS containing pFNLTP1 in MH medium was also studied to determine the stability of the plasmid in vitro. pFNLTP1 was maintained at 100% efficiency in F. tularensis LVS for 10 repeated passages on MH agar without kanamycin selection (data not shown). The presence of pFNLTP1 also did not significantly affect the growth (doubling time) of LVS in MH broth (Fig. 3, inset). Thus, pFNLTP1 is stable and does not significantly affect the growth of F. tularensis LVS in vitro or in J774A.1 macrophages.

• Open in new tab
• Download powerpoint

FIG. 3.

Growth of wild-type F. tularensis LVS and F. tularensis LVS carrying pFNLTP1 in the murine BALB/c macrophage cell line J774A.1 and in MH broth (inset). Macrophage monolayers containing 4 × 10⁵ cells per well in 12-well tissue culture plates were infected at an MOI of 1 and allowed to undergo phagocytosis for 2 h. Cells were lysed at different times by addition of distilled water, and bacteria were enumerated by plating serial dilutions on MH agar. The results are the results (mean ± standard deviation) for experiments performed in triplicate. Growth of F. tularensis LVS derivatives in vitro in MH broth was monitored by measuring the optical density at 550 nm (O.D. 550 nm) for triplicate cultures.

Isolation of a temperature-sensitive pFNLTP1 replicon.Conditionally replicating plasmids are useful vehicles for constructing chromosomal mutations by allelic exchange or for delivering transposons for mutagenesis studies. pFNLTP1 was treated with hydroxylamine and electroporated into F. tularensis LVS, and kanamycin-resistant transformants capable of growing at 32°C (permissive temperature) but not at 42°C (nonpermissive temperature) were isolated. One plasmid derivative exhibiting this phenotype, pFNLTP9, was selected for further analysis, including DNA sequencing to determine the mutation responsible for the temperature-sensitive phenotype and growth studies to determine maintenance of the plasmid. Sequence analysis of pFNLTP9 resulted in identification of a single base pair mutation in repA (G to A), altering amino acid 120 (M120I), that was responsible for the temperature-sensitive phenotype. As expected, this mutation did not result in a temperature-sensitive phenotype when the plasmid was introduced into E. coli DH5α (data not shown). To determine the severity of the temperature-sensitive phenotype, F. tularensis LVS carrying pFNLTP9 was subjected to in vitro growth analysis at the permissive and nonpermissive temperatures. Maintenance of pFNLTP9 in F. tularensis LVS was assessed by picking colonies isolated on MH agar at the permissive temperature and patching them onto MH agar with or without kanamycin that was incubated at 32 or 42°C. pFNLTP9 was maintained at 99 to 100% efficiency in F. tularensis LVS at 32°C in the presence or absence of kanamycin (data not shown). This result is consistent with the result obtained previously for the parent plasmid pFNLTP1, which was stably maintained without selection for several generations. In contrast, pFNLTP9 was not maintained in F. tularensis LVS grown on MH agar at 42°C in the presence of selection and exhibited reduced growth rates when the organism was grown at 42°C in the absence of selection (data not shown). These results indicate that pFNLTP9 carries a strong temperature-sensitive mutation that effectively prevents maintenance of this plasmid at 42°C, the nonpermissive temperature.

Insertion of multiple cloning sites into pFNLTP1.To take full advantage of the transformation efficiency and stability of pFNLTP1, additional derivatives with expanded multiple cloning sites for cloning were constructed. Oligonucleotides containing unique restriction enzyme sites were cloned into pFNLTP1 between the unique BamHI and KpnI sites. A variety of sites (SmaI, SalI, SpeI, HpaI, NdeI, NotI, NdeI, EcoRI, NheI, XhoI, StuI, PacI) were included to ensure flexibility for cloning manipulations (pFNLTP5 to pFNLTP8) (Table 1).

Ectopic expression of gfp in F. tularensis LVS with pFNLTP1 derivatives.To demonstrate the utility of pFNLTP1 for gene expression, gfp-containing derivatives of pFNLTP6 were constructed. First, the gfp gene was amplified and cloned into pFNLTP6. Francisella promoter sequences specific to the groE operon and the acpA genes were amplified and cloned into pFNLTP6 upstream of gfp. The groE operon and acpA promoters were selected based on established expression in Francisella (11, 31). After growth on MH agar, GFP expression from the reporter constructs was assessed by Western blot analysis of total bacterial proteins or by fluorescence microscopy.

Western blot analysis allowed steady-state quantitation of gfp. Titrations of purified recombinant GFP indicated that the chemiluminescent signal was linear when 10 to 100 ng of protein was loaded per lane (Fig. 4A, lanes 6 to 9). Titration of total F. tularensis LVS extracts (4 to 12 μg per lane) provided signals within the linear range of the Western blot analysis. As expected, F. tularensis LVS containing the promoterless gfp in pFNLTP6 did not express GFP even with the largest amount of total protein sampled (Fig. 4A, lane 1). In contrast, pFNLTP6 constructs containing gro-gfp or acp-gfp reporter fusions expressed GFP when 4 to 12 μg of total protein was loaded per lane (Fig. 4A, lanes 2 to 5). These data also indicated that GFP expression from F. tularensis LVS(pFNLTP6 gro-gfp) extracts was approximately four- to fivefold greater than GFP expression from extracts of F. tularensis LVS(pFNLTP6 acp-gfp) (Fig. 4A, lanes 2 to 5).

• Open in new tab
• Download powerpoint

FIG. 4.

Utilization of pFNLTP1 derivatives for ectopic expression of gfp in F. tularensis LVS. (A) Western blot of total cell extracts from F. tularensis LVS derivatives expressing gfp reporter fusions from pFNLTP6. The pFNLTP6 constructs carried a promoterless gfp fusion (lane 1) or gfp fused to promoter regions from the groE operon (lanes 2 and 3) or acpA (lanes 4 and 5) genes of F. tularensis LVS. Purified recombinant GFP (rGFP) (lanes 6 to 9) was titrated to demonstrate the relative levels of GFP detected by the monoclonal anti-GFP antibody. The total cellular protein loaded for each F. tularensis LVS strain was determined by a bicinchoninic acid protein assay. Approximately 4 to 12 μg of total cellular protein was loaded per lane to detect GFP in the range of the recombinant GFP standard curve. Lane 2, 4 μg; lanes 3 and 4, 8 μg; lanes 1 and 5, 12 μg. The Western blot is a representative example of three trials. (B) J774A.1 macrophage monolayers grown on coverslips were infected with F. tularensis LVS strains expressing GFP at an MOI of 1 as described in Materials and Methods. Cells were fixed after 1 day and incubated with a monoclonal F. tularensis anti-LPS antibody and subsequently with the Cy3-conjugated monoclonal IgG antibody before they were mounted onto slides. F. tularensis LVS strains carried pFNLTP1 (panels a to d), pFNLTP6 expressing gro-gfp (panels e to h), or pFNLTP6 expressing acp-gfp (panels i to l). Infected macrophages were visualized by Nomarski optics (panels a, e, and i), GFP fluorescence (panels b, f, and j), Cy3 fluorescence (panels c, g, and k), or GFP and Cy3 fluorescence (panels d, h, and l).

To determine whether the gfp reporter constructs could be used to visualize bacteria during intracellular growth in macrophages, strains containing control and gfp expression constructs were used to infect J774A.1 macrophages. Bacteria and macrophages were cocultured for 24 h, washed, fixed, and incubated with the appropriate antibodies before analysis by fluorescence microscopy. A monoclonal F. tularensis anti-LPS antibody was used to provide a positive control for the fluorescent signal and to specifically localize the bacteria. Strains possessing either promoter construct (gro-gfp or acp-gfp) displayed GFP fluorescence (Fig. 4B, panels f and j). In contrast, a strain containing a plasmid without a gfp reporter exhibited only background fluorescence (Fig. 4B, panel b). The presence of F. tularensis LVS in all macrophage cultures was confirmed with the monoclonal antibody specific for F. tularensis LPS (Fig. 4B, panels c, g, and k). Importantly, GFP fluorescence localized to the same area within the macrophage that was detected by F. tularensis LPS fluorescence (Fig. 4B, panels h and l). Consistent with the Western blot analysis, the GFP fluorescence was lower for F. tularensis LVS carrying the pFNLTP6 derivative expressing acp-gfp than for F. tularensis LVS carrying the pFNLTP6 derivative expressing gro-gfp (Fig. 4B, panels f and j). Thus, pFNLTP1 derivatives can efficiently express exogenous genes in F. tularensis LVS. These experiments confirmed the intracellular expression of gfp within J774A.1 macrophages by F. tularensis LVS, which is consistent with previous studies performed with J774A.1 macrophages, as well as with Acanthamoeba castellanii (1, 15).

Complementation of E. coli and F. tularensis LVS streptomycin-resistant (Sm^r) mutants by using pFNLTP1.The ability to complement chromosomal mutations is an essential step for confirming the phenotypes of constructed mutants. Since streptomycin sensitivity is dominant over resistance (24), the utility of pFNLTP1 for complementation analysis was assessed. A spontaneous streptomycin-resistant isolate of F. tularensis LVS was selected on MH agar containing 100 μg of streptomycin ml⁻¹. Amplification and DNA sequence analysis of rpsL from this streptomycin-resistant F. tularensis LVS strain demonstrated that a single base pair mutation (G to T), altering amino acid 43 (K43N), was responsible for the streptomycin-resistant phenotype. Similar nucleotide mutations in rpsL are responsible for streptomycin resistance in other organisms (33, 37). E. coli HB101, a known rpsL mutant, and F. tularensis LVS rpsL were complemented by using pFNLTP1 expressing the wild-type rpsL gene from F. tularensis LVS (rpsL_LVS) (Fig. 5). E. coli HB101 or F. tularensis LVS rpsL strains or strains carrying the parent plasmid pFNLTP1 grew in the absence or presence of streptomycin (compare Fig. 5A, sector I, and Fig. 5B, sector I, as well as Fig. 5A, sector II, and Fig. 5B, sector II). In contrast, E. coli HB101 and F. tularensis LVS rpsL strains containing pFNLTP1 expressing wild-type rpsL_LVS were streptomycin sensitive when they were grown in the presence of the antibiotic (Fig. 5A, sector III, and Fig. 5B, sector III). Thus, expression of rpsL_LVS from pFNLTP1 complemented both E. coli HB101 and F. tularensis LVS rpsL mutants.

• Open in new tab
• Download powerpoint

FIG. 5.

Utilization of pFNLTP1 for complementation analysis. E. coli HB101 and F. tularensis LVS rpsL were complemented in trans with wild-type rpsL from F. tularensis LVS (rpsL_LVS) expressed from pFNLTP1. E. coli HB101 (A) or F. tularensis LVS rpsL (B) carrying no plasmid (sector I), pFNLTP1 (sector II), or pFNLTP1 expressing wild-type rpsL_LVS (sector III) was cultured on LB or MH agar with or without streptomycin (100 μg ml⁻¹). E. coli HB101 was grown for 1 day, while F. tularensis LVS rpsL was grown for 3 days. Note the lack of growth for E. coli HB101 and F. tularensis LVS rpsL carrying pFNLTP1 rpsL_LVS.