DNA VACCINATION OF BISON TO BRUCELLAR ANTIGENS ELICITS ELEVATED ANTIBODY AND IFN-γ RESPONSES

Abstract

Brucella abortus remains a threat to the health and well-being of livestock in states bordering the Greater Yellowstone Area. During the past several years, cohabitation of infected wildlife with cattle has jeopardized the brucellosis-free status of Idaho, USA; Wyoming, USA; and Montana, USA. Current livestock B. abortus vaccines have not proven to be efficacious in bison (Bison bison) or elk (Cervus elaphus nelsoni). One problem with the lack of vaccine efficacy may stem from the failure to understand wildlife immune responses to vaccines. In an attempt to understand their immune responses, bison were vaccinated with eukaryotic DNA expression vectors encoding the Brucella periplasmic protein, bp26, and the chaperone protein, trigger factor (TF). These DNA vaccines have previously been shown to be protective against Brucella infection in mice. Bison were immunized intramuscularly at weeks 0, 2, and 4 with bp26 and TF DNA vaccines plus CpG adjuvant or empty vector (control) plus CpG. Blood samples were collected before vaccination and at 8, 10, and 12 wk after primary vaccination. The results showed that bison immunized with bp26 and TF DNA vaccines developed enhanced antibody, proliferative T cell, and interferon-gamma (IFN-γ) responses upon in vitro restimulation with purified recombinant bp26 or TF antigens, unlike bison immunized with empty vector. Flow cytometric analysis revealed that the percentages of CD4⁺ and CD8⁺ T lymphocytes from the DNA-vaccinated groups were significantly greater than they were for those bison given empty vector. These data suggest that DNA vaccination of bison may elicit strong cellular immune responses and serve as an alternative for vaccination of bison for brucellosis.

INTRODUCTION

Brucellae are Gram-negative facultative intracellular bacteria endemic in many areas of the world. Ten species of Brucella are recognized and classified based mainly on their preferred hosts and pathogenicity (Godfroid, 2005; Perkins et al., 2010). Animals, including humans, become infected with Brucella when mucosal membranes, open wounds, or skin abrasions come in contact with infected secretions (milk, blood, uterine discharge) or aborted fetuses (Ko and Splitter, 2003). The most common clinical manifestation of animal brucellosis is reproductive loss resulting from abortion, birth of weak offspring, or infertility (Seleem et al., 2010). In humans, brucellosis usually is associated with nonspecific flu-type symptoms, such as malaise, undulant fever, and joint aches (Olsen and Tatum, 2010).

The enormous cost of brucellosis to the livestock industry, as well as its effect on public health, has prompted many countries to adopt brucellosis control and eradication programs (Olsen and Stoffregen, 2005). In the United States, a brucellosis eradication program was established in 1954 with the aim of eliminating Brucella abortus infections from cattle. The program was successful, and currently, US cattle are essentially brucellosis-free. However, the threat from brucellosis remains; infected wild animals continue to serve as reservoirs, and B. abortus–infected bison (Bison bison) and elk (Cervus elaphus nelsoni) still pose a threat in the Greater Yellowstone Area (GYA; Olsen et al., 2009). Outbreaks of brucellosis in Idaho, Wyoming, and Montana in vaccinated cattle that coexist with infected wildlife have resulted in all three states temporarily losing their brucellosis-free status (Van Campen and Rhyan, 2010). Aside from being problematic for livestock and wildlife, free-ranging wildlife remains a concern as a source of emerging human pathogens. Because humans are susceptible to zoonotic exposure through infected livestock, the coexistence of wildlife with livestock can increase disease frequency and transmission potential to humans (Rhyan and Spraker, 2010).

Vaccination with live, attenuated B. abortus strains 19 and RB51 has been used to control brucellosis in cattle (Schurig et al., 2002). In most instances, the use of Brucella vaccines in wildlife species has been problematic. Vaccination with B. abortus strain RB51 has had little efficacy in bison (Davis and Elzer, 2002; Olsen et al., 2003). Strain 19 has been associated with chronic infections and abortions in bison and has been found to be ineffective as a calfhood vaccine for bison (Davis et al., 1991). Thus, the development of a more effective vaccine to protect susceptible wildlife and livestock is warranted.

Immunization with naked DNA is an attractive alternative approach for immunizing against infectious diseases. Intramuscular (IM) delivery of DNA permits host synthesis of vaccines, stimulating both humoral and cellular immune responses specific to the expressed proteins (Robinson and Torres, 1997). Furthermore, DNA vaccines may have many advantages over traditional vaccines, including induction of long-lived immune responses, better stability, ease of preparation, and low cost (Oñate et al., 2003). Previous studies have proven that DNA vaccination with sodC (Oñate et al., 2003), lumazine synthase gene (Velikovsky et al., 2002), and P39 (Al-Mariri et al., 2001) can elicit partial protection against Brucella challenge in the mouse. In our previous work, we applied a search strategy to screen the Brucella melitensis 16M genome for potential vaccine candidates. We found that the periplasmic protein, bp26, and the chaperone protein, trigger factor (TF), are protective antigens when delivered as peripheral DNA vaccines (Yang et al., 2005).

Because most efforts have relied mostly on small-animal laboratory models, methods remain to be shown as translatable to wildlife. We evaluated the immunogenicity of plasmid DNA carrying the Brucella bp26 and TF genes as a possible vaccine candidate for use in bison. The construction and preparation for vaccination of pcDNA3.1-bp26 and pcDNA3.1-TF vaccines have been described (Yang et al., 2005) as has the production of recombinant bp26 and TF in Pichia pastoris (Yang et al., 2007).

MATERIALS AND METHODS

Animal vaccination and blood collections

Eight 10-mo-old bison heifers were obtained from a brucellosis-free herd not previously vaccinated with RB51. After acclimation for 4 wk, bison were randomly assigned to two groups (n=4 animals/group) for IM vaccination with 600 μg empty pcDNA3.1 vector (negative control) or a combination of pCMVbp26 (300 μg) and pCMVTF (300 μg), coadministered with 50 μg of CpG as adjuvant. The sequence of CpG (the same as in the mouse studies; Yang et al., 2005) was 5’-TCCATGACGTTCCTGACGTT-3’. Vaccination was performed at wk 0, 2, and 4. Blood samples were collected by jugular venipuncture into 10-ml syringes preloaded with 0.2 ml of sodium heparin before vaccination and at 8, 10, and 12 wk postvaccination. For serologic evaluation, blood was allowed to clot for 12 hr at 4 C and then centrifuged. Serum was aliquoted, frozen, and stored at –20 C.

Antibody enzyme-linked immunosorbent assay (ELISA)

To determine induced antibodies, bp26-and TF-specific ELISAs (Yang et al., 2005) were used to measure immune serum immunoglobulin (Ig) G, IgG1, and IgG2 levels. Purified bp26 and TF proteins (5 μg/ml) were used to coat Maxisorp microtiter plates (Nunc, Roskilde, Denmark) at 100 μl/well overnight at 4 C. Plates were washed four times in a wash buffer (Tris-buffered saline [pH 7.4] with 0.05% Tween 20) and blocked with 1% albumin from chicken egg Grade II (Sigma-Aldrich, St. Louis, Missouri, USA) in Trisbuffered saline for 2 hr at 37 C, incubated with serial dilutions of the sera from bison for 3 hr at room temperature (RT), and washed three times. Horseradish peroxidase–conjugated mouse anti-bovine IgG monoclonal antibody (mAb; clone IL-AR; Serotec, Raleigh, North Carolina, USA) and sheep anti-bovine IgG1 and IgG2 polyclonal antibodies (Novus Biologicals, Littleton, Colorado, USA) were used for detection. After a 90 min incubation at 37 C and washing, specific reactivity was determined by the addition of an enzyme substrate, ABTS [2,2_azinobis(3-ethylbenzthiazolinesulfonic acid)] diammonium (Moss, Inc., Pasadena, California, USA) at 100 ml/well. The absorbance was measured at 415 nm on a Kinetics Reader model ELx808 (Bio-Tek Instruments, Winooski, Vermont, USA). Endpoint titers were defined as the reciprocal of the highest dilution of a sample giving an optical density at 415 nm of 0.100 U greater than negative controls after 1 hr of incubation at 25 C.

Peripheral blood mononuclear cells (PBMCs) for lymphocyte proliferation assays

Subsequent to vaccination, blood was obtained from the jugular vein of all bison and placed into an acid-citrate dextrose solution. Peripheral blood mononuclear cells were enriched by density centrifugation using a Histopaq gradient (Sigma Diagnostics, Inc., St. Louis, Missouri, USA). The PBMCs were cultured at 37 C with 5% CO₂ in 96-well flat-bottom plates at a concentration of 5×10⁵ viable cells per 200 μl/well in the presence of 20 μg/ml of bp26, TF, or both; and 5.0 μg/ml concanavalin A (Sigma-Aldrich), 1.0 mg/ml ovalbumin (OVA), or no additives (unstimulated control). RPMI 1640 medium (GIBCO BRL, Grand Island, New York, USA) was supplemented with 2 mM l-glutamine, 10% heat-inactivated horse serum (GIBCO BRL), and 100 U/ml penicillin, 100 μg/ml streptomycin, referred to as complete medium (CM), for culturing PBMCs. The cells were cultured for 4 days and pulsed for 12 hr with 0.5 μCi of thymidine (50 Ci/mmol; NEN-Dupont, Boston, Massachusetts, USA) per well, and the radioactivity incorporated in the DNA was measured as mean counts per minute and obtained from triplicate cultures from each bison. In addition, a stimulation index was calculated for each animal by dividing the counts per minute of cells with antigen by the counts per minute of cells without antigen.

Interferon-gamma (IFN-γ) production

In vitro production of IFN-γ by PBMCs was measured at all sampling times after immunization. Briefly, PBMCs from each animal were isolated and resuspended in CM. Cells were cultured in 24-well tissue plates at 5×10⁶ cells/ml in CM alone or with purified recombinant bp26 or TF (20 μg/ml) for 3 days at 37 C. The supernatants were collected by centrifugation and stored at −80 C. A bovine-capture ELISA was used to quantify the levels of IFN-γ from triplicate sets of samples. Microtiter wells were coated with 1 μg/ml of purified mouse anti-bovine IFN-γ mAb (clone CC302; Serotec), which we found recognized recombinant bison IFN-γ that shares >96% gene homology with bovine IFN-γ (data not shown). After blocking with PBS +1% bovine serum albumin for 2 hr at 37 C, washed wells were incubated with cell culture supernatants at 4 C for 24 hr. After washing, 0.5 μg/ml biotinylated mouse anti-bovine IFN-γ mAb (clone CC302; Serotec) was added for 90 min at 37 C. Following washing, 1:500 HRP-goat anti-biotin antibody (Vector Laboratories, Burlingame, California, USA) was added for 1 hr at RT. After washing, ABTS peroxidase substrate (Moss, Inc. Pasadena, California, USA) was added to develop the reaction. Production of IFN-γ by unstimulated cells set as a background was subtracted from all measurements.

Flow cytometry

Peripheral blood mononuclear cells were isolated from each blood sample by Ficoll-Hypaque density gradient centrifugation, adjusted to 1×10⁷ cells/ml and were then fluorescently labeled with fluorochrome-conjugated mAbs for bovine γδ T cells (GD3.8) and CD4 (clone CC30; Serotec, and CD8 T cells (clone CC63; Serotec). These mAbs have been widely used in bovine studies (Smyth et al., 2001; Wilson et al., 1998). To assess whether these mAbs would recognize bison T cells, cell surface staining was compared with mAbs previously shown to react against bison T cells: anti-bovine CD4 (clone ILA11A; VMRD, Pullman, Washington, USA), -CD8 (clone CACT80C; VMRD; Simon et al., 2003; Nelson et al., 2010), and anti-bovine γδ T cells (clone ILA29; VMRD; Nelson et al., 2010), and thus, GD3.8, CC30, and CC63 are cross-reactive for bison T cells. Fluorescence was acquired on an LSR II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) using BD FACSDiva software, and samples were analyzed using Flowjo (TreeStar Inc., Ashland, Oregon, USA) software.

Statistical analysis

Analysis of variance, followed by Tukey’s method, was used to evaluate differences among antibody titers, IFN-γ production, T-cell proliferative responses, and the percentages of T-cell subsets between immunized and control bison discerned to the 95% confidence interval.

RESULTS

Immunization with DNA vaccines induces elevated IgG and IgG1 anti-bp26 and anti-TF antibody titers

To determine whether pCMVbp26 and pCMVTF DNA vaccines could stimulate bison antibody responses, bison were immunized with IM pCMVbp26 and pCMVTF with CpG, and control group bison were immunized with pcDNA3.1 (empty vector) plus CpG three times at 2 wk intervals. Beginning at 8 wk, after primary immunization, serum samples were obtained to measure antigen-specific endpoint antibody titers (Figs. 1A, B). The bp26- and TF-specific IgG and IgG1 were detected in the immunized bison with a peak IgG titer of 2¹⁰ and IgG1 titer of 2⁹ at wk 12. The IgG2 antibodies were not detected throughout the experiment. No detectable IgG and low IgG1 antibody titers were observed for the vector-immunized bison (Figs. 1A, B).

Figure 1.

Immunization with DNA vaccines encoding for bp26 and trigger factor (TF) elicits serum immunoglobulin G (IgG) antibody responses. Bison (four/group) were immunized intramuscularly with pcDNA3.1 vector (600 μg) or a combination of pCMVbp26 (300 μg) and pCMVTF (300 μg), together with 50 μg CpG as adjuvant on days 0, 14, and 28. Serum IgG anti-bp26 and anti-TF titers were measured by enzyme-linked immunosorbent assay on wk 8, 10, and 12 after primary immunization. The IgG, IgG1 and IgG2 (A) anti-bp26 and (B) anti-TF endpoint titers peaked between wk 8 and 12 after primary immunization. Control bison immunized with control vector, pcDNA3.1 plus CpG failed to elicit antibodies to bp26 and TF. Asterisks represent significant differences in IgG responses versus pcDNA3.1 vector-immunized bison (P≤0.05).

DNA-immunized bison show enhanced PBMC proliferative and IFN-γ responses

To assess their cellular responses, PBMC proliferative assay was performed. Lymphocytes from pCMVbp26- plus pCMVTF-vaccinated bison showed greater proliferative responses (P<0.001; P<0.05) to purified recombinant bp26, TF, or their combination than did lymphocytes from animals immunized with pcDNA3.1 vector only at 8, 10, and 12 wk after primary immunization (Fig. 2). Only low levels of spontaneous proliferation occurred in cultures with no antigen (medium control) or irrelevant antigen (OVA control) stimulation. Throughout the study, PBMCs from bison showed elevated proliferative responses to the T-cell mitogen, ConA (Fig. 2).

Figure 2.

Immunization with DNA vaccines encoding for bp26 and trigger factor (TF) enhances antigen-specific bison T cell proliferative responses. Bison were immunized, as described in Fig. 1. T-cell proliferation responses were measured at 8, 10, and 12 wk after the primary immunization. Peripheral blood mononuclear cells (PBMCs) from each bison (5×10⁵ cells/well) were prepared and stimulated in vitro with purified 20 μg/ml of recombinant bp26, TF, both, 1 mg/ml ovalbumin (OVA) or 5 μg/ml of concanavalin A (Con-A). Results are expressed as mean±SE of triplicate cultures of cells obtained from each bison. ** P≤0.001 and * P≤0.05 represent the significant differences within the sampling time in proliferative T-cell responses between bison immunized with pcDNA3.1 vector versus bison immunized with bp26 and TF DNA vaccines.

Because cell-mediated immunity, particularly Th1 cell-dependent, is considered crucial in protection against B. abortus infection, IFN-γ production by antigen-restimulation was measured. The PBMCs were cultured with either bp26 or TF for 3 days and were then evaluated for IFN-γ production by cytokine ELISA. Upon restimulation with bp26 (Fig. 3A) or TF (Fig. 3B), the PBMCs from pCMVbp26 plus pCMVTF-vaccinated bison showed significantly greater levels of IFN-γ (P≤0.05) than those lymphocytes from pcDNA3.1 vector-vaccinated bison beginning at 10 wk postimmunization. In samples obtained on wk 6 and 8 postimmunization, mean IFN-γ production by either immunization group did not significantly differ (Fig. 3A, B).

Figure 3.

Immunization with bp26 and trigger factor (TF) DNA vaccines show enhanced interferon (IFN)-γ production by bison peripheral blood mononuclear cells (PBMCs). At 8, 10, and 12 wk after initial immunization, PBMCs from bison immunized with bp26 and TF DNA vaccines or with pcDNA3.1 vector were incubated for 72 hr in the presence or absence of 20 γg/ml (A) bp26 or (B) TF and were evaluated for IFN-γ production by cytokine enzyme-linked immunosorbent assay. Results are expressed as net mean±SE IFN-γ production (production in wells containing bp26 or TF minus production in wells without antigen) obtained from each bison. * P≤0.05 represents significant differences within the sampling time in IFN-γ production between bison immunized with pcDNA3.1 versus those immunized with pCMVbp26 plus pCMVTF.

DNA immunization enhances the numbers of CD4⁺ and CD8⁺ T cells

The CD4⁺, CD8⁺, and γδ T-cell subsets were evaluated for their changes as a consequence of vaccination. The PBMCs from immunized bison were analyzed by flow cytometry at 8, 10, and 12 wk after primary immunization. The percentages of CD4⁺ and CD8⁺ T cells from the pCMVbp26 plus pCMVTF-vaccinated group were significantly greater than those bison immunized with pcDNA3.1 vector at wk 8, 10, and 12 postimmunization (Table 1). The percentage of γδT cells in the bison immunized with pCMVbp26 plus pCMVTF was significantly reduced in the peripheral blood compared with that of the bison immunized with pcDNA3.1 vector at 8 and 12 wk postimmunization (Table 1).

Table 1.

Flow cytometric analysis of T-cell responses to recombinant bp26 and trigger factor (TF) after DNA immunization. Peripheral blood mononuclear cells (PBMCs) from bison immunized intramuscularly with pcDNA3.1, or with a combination of pCMVbp26 plus pCMVTF, in presence of CpG were labeled with monoclonal antibodies specific for CD4⁺ T, CD8⁺ T, and TCRγδ⁺ cells and were analyzed by flow cytometry.

Time postimmunization and immunization group	% PBMC subpopulations
	CD4+ T cells	CD8+ T cells	TCRγδ+ cells
8 wk
Control	22.86±1.4	9.35±0.3	33.0±2.5
bp26+TF	33.86±0.3*	26±1.5*	22.03±0.98*
10 wk
Control	21.78±2.5	8.6±1.04	43.1±3.6
bp26+TF	32.1±0.23*	19.7±0.91*	41.2±1.1
12 wk
Control	24.8±0.50	8.2±1.39	39.1±1.1
p26+TF	39.8±3.9*	18.1±2.1*	23.2±4.4**

^*P≤0.001 and

^**P≤0.05 indicate a significant difference within the sampling time in percentages of T-cell subsets in bison immunized with pcDNA3.1 versus the combination of the two DNA vaccines.

DISCUSSION

Because infected wildlife in the GYA serves as a reservoir for B. abortus, the presence of infected wildlife impedes efforts to eliminate brucellosis in the United States and will continue to be problematic in the affected areas of Wyoming, USA; Idaho, USA; and Montana, USA. Although numerous studies evaluating brucellosis vaccines in livestock have been conducted during the past decades, only a limited number of vaccines have been studied in wildlife (Davis and Elzer, 2002; Olsen and Tatum, 2010). Additionally, the work has largely been done with live S19 and RB51 vaccines (Davis et al., 1991; Olsen et al., 1997, 2003, 2009), which were originally developed for livestock. However, neither of these vaccines conferred complete protection against abortion and infection in wildlife, as evidenced by the incidences of infected livestock (Olsen et al., 2009). Therefore, a safe and effective brucellosis vaccine for free-ranging bison in the GYA might be beneficial in reducing the risk of transmission and in increasing herd protection against infection from bison as a reservoir for brucellosis.

Cellular immune responses play a major role in protection against Brucella (Gonzalez-Smith et al., 2006), and for this reason, an effective vaccine against brucellosis must be based on its capacity to generate strong, cell-mediated immunity. Immunization with DNA vaccines encoding protective epitopes represents a means to generate and test various vaccines (Gurunathan et al., 2000). Because DNA vaccines can stimulate both cellular and humoral immunity (Liu, 2003), we have recently reported that, when bp26 is combined with TF, partial protection is obtained (Yang et al., 2005), suggesting that a subunit vaccine approach against brucellosis may be feasible. The combination of bp26 and TF shows significant protection against B. melitensis 16M challenge, either as a DNA vaccine (Yang et al., 2005) or as recombinant proteins (Yang et al., 2007) in BALB/c mice.

Despite a significant effort to identify a single protein that confers protection (Oñate et al., 1999; Baloglu et al., 2000; Al-Mariri et al., 2001), no such candidate has been described. Therefore, we hypothesize that perhaps a combination of various antigens is required for successful protection. The use of recombinant protein technology and monoclonal antibodies has shown that the major outer membrane proteins (OMPs) appear to be of limited use as vaccines against smooth B. abortus or B. melitensis infections (Cloeckaetr et al., 2002). Recently, however, Omp25 has been shown to be involved in the virulence of B. melitensis, B. abortus, and B. ovis, and mutants lacking Omp25 are indeed attenuated in animal-infection models (Edmonds et al., 2001, 2002a, b; Jubier-Maurin et al., 2001). Pasquevich et al. have shown that immunization with OMPs of recombinant Brucella species omp16 or omp19 induces protection against B. abortus infection (Pasquevich et al., 2009). Brucella heatshock proteins appear to be ineffective as recombinant protein vaccines. HtrA and GroEL + GroES + HtrA, formulated in Ribi adjuvant, fail to protect against B. abortus 2308 despite evidence that those proteins are able to elicit antigen-specific immunity (Bae et al., 2002). Until suitable, protective vaccine candidates are identified, we have elected to test the immunogenicity of our bp26 and TF DNA vaccines in bison.

Protection against brucellosis is thought to be dependent on cell-mediated immunity and, to a lesser extent, on antibodies specific to membrane proteins (Velikovsky et al., 2002). As with immunity to other intracellular pathogens, immunity to B. abortus depends on antigen-specific T-cell–mediated activation of macrophages, which are the major effectors facilitating killing and inhibiting replication of Brucellae. The Th1 cell-induced cytokines, like IFN-γ, are important for the activation of macrophages and in resistance to in vivo and in vitro Brucella infections (Zhan et al., 1993). In the present study, the induction of T-cell immunity following DNA immunization was evaluated by measuring T-lymphocyte proliferation and IFN-γ production following in vitro stimulation with purified recombinant bp26 or TF. Both proteins induced elevated T-cell proliferative responses and elevated levels of IFN-γ. Such potent IFN-γ responses suggest that DNA immunization stimulates polarization toward a Th1-type phenotype, which is typically correlated with protection against intracellular pathogens, such as Brucella (Yingst and Hoover, 2003).

Flow cytometry analysis revealed an increase in the percentages of CD4⁺ and CD8⁺ T cells in immunized bison compared with that of control bison. The percentage of γδ T cells was significantly lower in DNA-vaccinated bison. To our knowledge, this study is the first to evaluate γδ T-cells responses subsequent to DNA immunization. The γδ T cells represent a major lymphocyte subset in cattle and bison and can constitute up to 60–70% of the circulating T-cell pool in calves (Jutila et al., 2008). The γδ T cells have also been shown to respond and participate in host defense responses in a variety of pathogen-induced diseases, including brucellosis (Hayday, 2000). However, the decreased percentage of γδ T cells in vaccinated animals was surprising. The γδ T cells have long been documented to recognize unprocessed or nonpeptide antigens, independent of antigen processing and presentation through major histocompatibility complex molecules (Sireci et al., 1997). This means, unlike the CD4 or CD8 T cells, γδ T cells can recognize bacterial antigens directly in the absence of classic antigen presentation. In addition, γδ T cells express innate pathogen-recognition receptors and respond directly to pathogen-associated molecular patterns (Hedges et al., 2005; Jutila et al., 2008). Additional studies will elucidate the importance and role of γδ T cells subsequent to DNA vaccination.

Studies also revealed DNA immunization can elicit modest levels of IgG and IgG1 anti-bp26 and anti-TF responses. This does not necessarily suggest that pCMVbp26 and pCMVTF vaccines are poor immunogens because many investigators have also demonstrated weak antibody responses following DNA vaccination of mice (Cassataro et al., 2005; Gonzalez-Smith et al., 2006). Rather, antibody responses are generally suboptimal following DNA vaccination, often requiring a protein boost or, alternatively, they may be in part attributed to the dampening effect by Th1 cell bias used by this immunization method (Commander et al., 2007). Interestingly, IgG1 subclass responses are induced in response to both bp26 and TF, which suggest preferential bias by the DNA vaccines toward Th2-type immune response.

In summary, bison are responsive to DNA immunization, as demonstrated here using pCMVbp26 and pCMVTF DNA vaccines. Importantly, this method allows the stimulation of a potent Th1 cell bias evident by the elevated IFN-γ production following antigen restimulation of bison PBMCs. Although additional vaccine candidates need to be identified, this work demonstrates the feasibility of using a DNA vaccine approach to stimulate Brucella-specific immunity.