Brucella abortus Strain RB51 Leucine Auxotroph as an Environmentally Safe Vaccine for Plasmid Maintenance and Antigen Overexpression

Abstract

To avoid potentiating the spread of an antibiotic resistance marker, a plasmid expressing a leuB gene and a heterologous antigen, green fluorescent protein (GFP), was shown to complement a leucine auxotroph of cattle vaccine strain Brucella abortus RB51, which protected CD1 mice from virulent B. abortus 2308 and elicited GFP antibodies.

Brucella abortus strain RB51 is a USDA-approved vaccine used in the eradication of bovine brucellosis (15, 16). The strain has been investigated as a platform to overexpress and deliver homologous (Brucella derived) as well as heterologous (non-Brucella derived) antigens because of its ability to induce a strong Th1-type immune response (15, 16). In order to use strain RB51 as a platform vaccine for inducing specific immune responses, a plasmid containing the gene encoding the foreign antigen along with an antibiotic resistance gene as a plasmid marker has been employed (15, 16). This practice has been criticized as it has the potential to introduce the antibiotic resistance gene into normal flora, as well as pathogens, in the vaccinated animals (1, 7, 17). Expressing the protective antigens on a plasmid that is not dependent on an antibiotic resistance gene for maintenance would be an excellent alternative, as it would have a minimum of environmental risk. An intracellular pathogen, such as B. abortus, is known to survive and replicate in nutrient-limited environments inside a host. Therefore, a B. abortus mutant lacking a gene for the biosynthesis of an essential amino acid is unlikely to survive in that environment (2, 4). The complementation of this auxotroph with a plasmid carrying the wild-type gene (encoding the enzyme necessary for the amino acid synthesis) would provide the selection for the maintenance of this plasmid in B. abortus in those conditions. Markerless auxotrophic mutants may be used in place of genetically engineered conventional bacterial vaccine strains to avoid the introduction of antibiotic resistance genes into the bacterial flora of an animal population (17). The cre-lox system has been used successfully for creating antibiotic resistance-free mutants in various bacteria (8). In this study, cre-lox methodology was used to produce an unmarked leuB mutant in the cattle vaccine B. abortus strain RB51. The leuB gene, encoding isopropyl malate dehydrogenase, is one of the four genes essential for the biosynthesis of leucine in B. abortus (4). The resultant leuB auxotroph cannot grow in leucine-deficient conditions (3). Complementation of the leuB auxotroph with a plasmid carrying the wild-type leuB gene allows its survival in leucine-deficient minimal medium and under nutrient-limiting conditions in vivo, thus providing selective pressure for the maintenance of the plasmid. We propose this leuB auxotroph of the RB51 vaccine strain as an environmentally safe vector to overexpress homologous and heterologous antigens without potentiating the spread of antibiotic resistance markers.

Construction of strain RB51leuB.

An unmarked mutation was created in the leuB gene of strain RB51 by using cre-lox methodology (Fig. 1). Briefly, two regions of the leuB gene of B. abortus strain RB51 were amplified as separate fragments: a 750-bp (leuB_I) fragment containing 300 bp upstream of the leuB gene was amplified using the primers 5′ GGG GAA TTC AGT TTC GCT CGC GGT GAG TGG 3′ and 5′ GGG GGA TCC ATG ATT TCC TTC GGT TCG CCG 3′, and a 450-bp fragment (leuB_II) was amplified using the primer pair 5′ GGG GGA TCC TAT GCT GGC TGA TGC TGG CGG 3′ and 5′ GGG AAG CTT TCA GGC CGA AAG TGC CTT GAA 3′. These two fragments deliberately delete 216 bp of the leuB gene (Fig. 1) in order to eliminate any reversal of the deletion during subsequent auxotroph growth. The 1,200-bp amplicon containing a disrupted leuB gene was cloned into vector pGEM3Z. The aacC1 gene, coding for gentamicin resistance, flanked by loxP sites was cloned as a single BamHI fragment from the plasmid pUCGmlox (11). This fragment was cloned into the BamHI site of the disrupted leuB gene of the intermediate construct to produce the final suicide plasmid pLGL. The suicide plasmid pLGL was introduced into B. abortus strain RB51 by electroporation as described by McQuiston et al. (9). The final strain, RB51leuB, was obtained by plating the transformants on medium containing gentamicin and then on medium containing kanamycin. The leucine deficiency of the unmarked mutant was demonstrated in a leucine-deficient minimal medium as described by Plommet (10). Southern hybridization was performed as described previously (5) to demonstrate that only the leuB gene was disrupted, leaving the rest of the genome unaltered (Fig. 2). Standard laboratory procedures approved by the U.S. Centers for Disease Control and Prevention (CDC) were followed while handling live B. abortus cells in a biosafety level 3 facility of the Virginia-Maryland Regional College of Veterinary Medicine. The use of the antibiotic resistance markers Gm^R, Amp^R, and Kn^R in strain RB51 was approved by the CDC. The plasmids used in this study are described in Table S1 in the supplemental material.

FIG. 1.

Illustration of the recombination events. The leuB gene is disrupted between base pairs 450 and 666, and the final unmarked mutant lacks 216 bp. WT, wild type; ΔGmlox, intermediate marked mutant; ΔleuB, unmarked leuB mutant.

FIG. 2.

Southern hybridization of KpnI-digested genomic DNA using pLGL as a probe to demonstrate that the integration occurred only at the leuB site of the genome. Molecular sizes are shown on the left. Lane, DNA source (expected band size[s]): 1, pLGL (4.8 kb); 2, molecular mass standards; 3, RB51 (3 kb); 4, RB51Gmlox (2.7 kb and 1 kb); 5, RB51leuB (1.8 kb and 1 kb).

Construction of pNS4/GFP and leuB complementation.

The leuB gene (1,412 bp) of strain RB51, along with its promoter, was amplified by PCR using the following primers: 5′ GGG-AAG- CTT-GGG-TCT-AGA-AGT-TTC-GCT-CGC-GGT-GAG- TGG-CGA 3′ and 5′ GGG-ACT-AGT-TCA-GGC-CGA-AAG-TGC-CTT-GAA 3′. The “origin of replication” (1,700 bp) and a 259-bp expression segment (Brucella groE promoter plus the multiple cloning site [MCS] plus a His₆ tag) of the plasmid pNSGroE were amplified using primers described previously, as they have cloning sites as well as the minimal sequence necessary for plasmid replication (14). After restriction enzyme digestion, the three fragments were purified and ligated to form plasmid pNS4 (Fig. 3). The marker leuB gene in place of the antibiotic resistance gene acts to complement any leuB auxotrophic strains in minimal medium deficient in leucine. The green fluorescent protein (GFP) gene, which is used as a model heterologous antigen, was cloned into the MCS of pNS4 by using the BamHI and XbaI sites, and the plasmid was designated pNS4/GFP. The complementing plasmid was electroporated into the competent strain RB51leuB as described earlier by McQuiston et al. (9), and the transformants were selected by being plated on leucine-deficient Brucella minimal medium (BMM) plates. The complemented strain RB51leuB expressing GFP appeared as green fluorescent colonies when observed under UV light; the colonies were later screened for the presence of pNS4/GFP by plasmid extraction. The expression of GFP was confirmed by immunoblot using GFP antibodies (data not shown).

FIG. 3.

Illustration of plasmid pNS4.

In vitro growth.

A single colony of a particular B. abortus leuB clone was inoculated in liquid BMM and grown for 72 h at 37°C and 200 rpm to create a starter culture that was used to inoculate the minimal medium and adjusted to 10 to 12 Klett units. At different time points the Klett units were measured by using a Klett-Summerson colorimeter, and the corresponding CFU were determined. The doubling time in minimal medium of strain RB51 or the complemented leuB auxotrophs was observed to be approximately 7 h. Leucine-deficient BMM did not support the growth of the leuB auxotrophs. Complementation of the leuB auxotrophs with pNS4 restored their growth in leucine-deficient BMM. The expression of GFP in B. abortus/pNS4/GFP did not affect the strain's ability to be complemented with leuB (Fig. 4).

FIG. 4.

Growth in CFU of the different strains RB51, RB51leuB, RB51leuB/pNS4, RB51leuB/pNS4/GFP, and RB51leuB/leu+ (leucine supplemented) at different time points when grown in a leucine-deficient BMM.

Expression of GFP in macrophages.

Murine J774.A1 macrophage cells (American Type Culture Collection, Manassas, VA) were incubated in Dulbecco's modified Eagle's medium (Cellgro; Mediatech, Inc., Herndon, VA) containing 10% fetal bovine serum for 24 h at 37°C in 5% CO₂ on a 6-well plate to make them adherent on coverslips. A 48-h culture of the pNS4/GFP-complemented strain RB51leuB was resuspended in phosphate-buffered saline and used to infect the macrophages at a 100:1 (bacteria/macrophage) multiplicity of infection. After 45 min, the macrophages were washed three times with phosphate-buffered saline and then incubated in Dulbecco's modified Eagle's medium containing 100 μg/ml of streptomycin-penicillin. At 24, 36, and 72 h postinfection, the coverslips were washed, fixed in formalin, and mounted on glass slides. The slides were observed under a Zeiss LSM 510 laser-scanning microscope (Carl Zeiss, Thornwood, NY) in fluorescence mode to detect the expression of GFP. J774.A1 macrophages infected with B. abortus cells containing the leuB-complementing plasmid pNS4/GFP appeared green when observed under the fluorescent confocal microscope. A representative picture of a macrophage containing B. abortus cells expressing GFP at 36 h postinfection is shown (Fig. 5).

FIG. 5.

Panel 1, uninfected J774.A1 macrophages (negative control); panel 2, J774.A1 macrophages infected with an leuB mutant of strain RB51 expressing GFP viewed under confocal microscopy at 36 h postinfection; panel 3, single macrophage containing GFP-expressing RB51leuB strain under higher magnification at 36 h postinfection.

Strain RB51leuB protects CD1 mice against virulent challenge.

The protective efficacies of the B. abortus RB51leuB strain and the leuB-complemented strain were evaluated using 5- to 6-week-old female CD1 mice (Charles River Laboratories, Wilmington, MA). Four groups of 10 mice per group were vaccinated intraperitoneally with ∼3 × 10⁸ to 5 × 10⁸ CFU in 100 μl of strain RB51, RB51leuB, RB51leuB/pNS4, or RB51leuB/pNS4/GFP. Another group of 10 mice were vaccinated with saline to serve as the negative control. Three mice from each group were bled at 5 weeks postvaccination to harvest serum. The sera were screened for GFP-specific antibodies by immunoblot. At 6 weeks postvaccination, all groups of mice were challenged intraperitoneally with 4 × 10⁴ CFU of B. abortus 2308. At 2 weeks postchallenge, mice were euthanized by CO₂ asphyxiation and their spleens were recovered, homogenized, serially diluted, and plated on Trypticase soy agar plates to estimate CFU. The leuB auxotroph and the complemented auxotroph of strain RB51 were able to protect the CD1 mice against a virulent B. abortus 2308 challenge (Fig. 6). There was no significant difference in the protection levels (i.e., splenic clearance) afforded by the leuB auxotroph when the protection levels in the mice vaccinated with the complemented leuB auxotroph and with the complemented leuB auxotroph expressing GFP were compared to the protection level in the mice vaccinated with strain RB51. There was, however, a significant difference in protection afforded between the mice vaccinated with any of the RB51 strains and the saline control (P < 0.005). Only sera from the group inoculated with RB51leuB/pNS4/GFP possessed GFP-specific antibodies (Fig. 7). In a separate experiment, the leuB auxotroph and the complemented leuB auxotroph were found to be cleared from CD1 mouse spleens by 4 to 5 weeks, at the same rate as the parent vaccine strain RB51 (data not shown). The splenic clearance rate of the challenge strain B. abortus 2308 was subjected to analysis of variance; the means and variances were compared by Tukey's range procedure (honestly significant difference) method. All procedures involving mice were done following Virginia Polytechnic Institute and State University's animal care committee recommendations.

FIG. 6.

Clearance rate of the challenge strain B. abortus 2308 in CD1 mice. Mice vaccinated with the leuB mutant and the complemented RB51leuB strains were able to clear the challenge strain B. abortus 2308 at significant rates compared with clearance in the mice vaccinated with saline (P < 0.005). No significant difference was observed in the rates of clearance of the challenge strain in the mice vaccinated with auxotrophic vaccine strains and the standard strain RB51 and also between the RB51leuB strains when analysis of variance and Tukey's range procedure were used. *, significant difference (P < 0.005).

FIG. 7.

Immunoblot for GFP. A crude extract (7 μg) from Escherichia coli cells expressing GFP was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. The membrane was split into individual lanes and treated with immunized mouse or control sera. Lane, serum group: 1, strain RB51leuB/pNS4/GFP (band size, 27 kDa); 2, molecular mass standard; 3, strain RB51; 4, strain RB51leuB; 5, strain RB51leuB/pNS4; 6, anti-GFP serum (positive control); 7, no serum (negative control). Results shown are those from the 25- to 30-kDa region of the gel; no other positive reactions were seen in the lanes incubated with the sera from mice immunized with GFP.

Conclusion.

Because the USDA-approved vaccine strain RB51 is a good inducer of a Th1 response (15), it has been investigated as an antigen delivery platform for vaccination against intracellular pathogens. Recently, we have demonstrated both protective-cell-mediated immunity and antibody-mediated immunity against the obligate intracellular protozoan Neospora caninum in mice by using strain RB51 as a platform for the expression of N. caninum antigens (12, 13). The U.S. Food and Drug Administration discourages the use of antibiotic resistance markers for plasmid-based antigen expression in live vaccines (17). The plasmid pNS4 was constructed by replacing the antibiotic resistance gene with a wild-type B. abortus leuB gene. We have successfully used this plasmid not only to complement the leucine deficiency of strain RB51leuB but also to overexpress a heterologous antigen (GFP) both in vitro (pure culture) and in vivo (in mice). Compared to its parent plasmid pBBR1MCS, the refined group of pNS plasmids (14) is more stable in B. abortus under nonselective conditions in vitro and in vivo. Even after 11 subcultures in an enriched medium (nonselective conditions), we were able to recover the plasmid pNS4 from 10 random colonies of leuB-deficient strains of B. abortus. This suggests that the complementing plasmid pNS4 is stable inside the auxotroph under leucine-sufficient conditions also, thus expressing the antigen in both selective and nonselective conditions. When viewed under UV light, all the colonies of strain RB51leuB complemented with pNS4/GFP and recovered from CD1 mouse spleens (4 weeks after vaccination) were fluorescent. Combined with the GFP expression noted in macrophages, these data strongly suggest that pNS4 is able to express a heterologous antigen in B. abortus following immunization of mice. For this study, we successfully utilized cre-lox methodology to create a leucine auxotroph of B. abortus strain RB51. B. abortus now joins the list of other gram-negative bacteria (8) that can be mutated using the cre-lox approach. The advantage of using the cre-lox methodology is that multiple unmarked mutations can be made in the genome by marker recycling (8). The protective efficacies of the RB51leuB vaccine strain and the leuB-complemented version in CD1 mice were as good as that found for strain RB51 vaccine in BALB/c mice (Fig. 6). Unlike previous experiments, we used the CD1 strain of mice, an outbred strain (6), to more closely model outbred genetic backgrounds subjected to vaccination under field conditions, e.g., cattle. Both the leucine auxotroph and the complemented version of B. abortus strain RB51 were cleared in CD1 strain mice at the same rate as they were in inbred BALB/c mice (data not shown). Thus, the leuB knockout does not appear to further attenuate strain RB51 but provides a selective pressure to retain the pNS4 plasmid when residing in a nutrient-limited stressful environment. The results of immunoblotting using purified GFP as the antigen showed that the mice vaccinated with strain RB51leuB/pNS4/GFP developed a GFP-specific antibody response (Fig. 7). This observation suggests that, in principle, any protective homologous or heterologous antigen can replace the GFP gene and that the RB51leuB strain overexpressing the antigen would induce a protective response against both B. abortus and other infectious agents. The new vaccine strain RB51leuB and the complementing pNS4 plasmid developed in this study would be a good combination as an environmentally safe platform vaccine that will not contribute to the spread of antibiotic resistance genes to other flora.