Introduction

Brucellae are Gram-negative intracellular bacteria that cause an important zoonotic disease called brucellosis. The genus Brucella includes six classical species namely Brucella melitensis, Brucella abortus, Brucella suis, Brucella canis, Brucella ovis and Brucella neotomae (Corbel 1997). Some more species viz Brucella pinnipedialis, Brucella ceti, Brucella inopinata and Brucella microti were included later on. Brucellosis accounts for more than 500,000 new cases annually. Infection with B. abortus, a species that primarily affects bovines, often results in abortions and infertility in domestic and wild mammals (Franco et al. 2007). The persistence of brucellosis in domestic livestock reservoirs remains a continual source of significant numbers of human infections worldwide. Despite regulatory efforts, brucellosis remains endemic in many parts of the world and is re-emerging in many countries.

Vaccination against Brucella infections in animals is usually performed by administration of the live attenuated smooth Brucella strains: B. abortus strain S19 and B. melitensis strain Rev.1. The non-smooth strain B. abortus RB51 was later introduced in some countries. B. abortus S19 and B. melitensis Rev.1 are proven effective vaccines against B. abortus infection in cattle and against B. melitensis and B. ovis infection in sheep and goats, respectively (Nicoletti 1990). Both vaccines have the disadvantages of causing abortion in a proportion of pregnant animals and of being pathogenic for humans (Perkins et al. 2010; Schurig et al. 1991). Due to the epidemic potential of Brucella, absence of a human vaccine, drawback of current vaccine strains in terms of safety, the efficiency of aerosol infection and poor symptomatology of disease, this airborne pathogen is classified as a biosafety level 3 pathogen and considered as a potential biological warfare agent. Humans infected with B. abortus are treated with antibiotics such as doxycycline and rifampin; however, the survival of B. abortus in macrophages is responsible for the chronic nature of the disease that necessitates a prolonged course of treatment (Hall 1990). Even after treatment, low levels of bacteria have been detected by PCR and relapses have been detected in 5–30 % of cases (Franco et al. 2007; Vrioni et al. 2008).

The Th1 immune response characterized by interferon gamma (IFN-γ) is associated with protection against brucellosis (Murphy et al. 2001; Paranavitana et al. 2005). IFN-γ causes upregulation of macrophage anti-Brucella activity (Oliveira et al. 1998) which is the main component of protective response. IFN-γ also induces the expression of many IFN-γ- inducible genes, crucial for the development of innate and adaptive immunity against this pathogen. Clearance of B. abortus infection in mice coincided with an increase in IFN-γ-producing CD4+ and CD8+ T cells (Goenka et al. 2011) suggesting the involvement of both types of cells in protection.

Several strategies, e.g. subunit recombinant protein vaccines and vector vaccines based on E. coli, Salmonella enterica, Ochrobactrum anthropi, Semliki Forest virus and influenza viruses, are being sought to prevent this disease while avoiding the disadvantages of the currently used live vaccines (Jain et al. 2014a). One strategy for developing safe and efficacious vaccine is immunization with protective T cell epitopes. Determining the epitopes recognized by Brucella-specific CD8+ and CD4+ T cells and the Brucella genes encoding the protein containing these epitopes can be helpful to establish peptides critical for vaccine development. Since peptides are synthetic, there would be no risk of mutation or reversion and little or no risk of contamination by live bacteria. The synthetic epitope or peptide-based vaccines have been advocated as an attractive approach for prevention or treatment of infectious diseases and malignant disorders (Buteau et al. 2002; Kast and Melief 1991; Rothbard 1987). Epitope-based vaccines have also been suggested as quick method to design and produce vaccines with short turn-around time for biothreat agents (De Groot et al. 2013).

Any invading pathogen or delivered antigen is taken up by antigen-processing cells (APCs) which process it resulting in formation of peptides that are loaded into grooves of newly synthesized MHC class I and class II molecules that then pass to the cell surface (Germain 1994). T cells are stimulated via MHC–peptide complexes and costimulatory molecules. Therefore, combining class I MHC- and class II MHC-restricted peptides in vaccines can increase immunogenic response. In the present work, MHC-I- and MHC-II-restricted T cell peptides of protective proteins or virulence determinants of B. abortus were designed and evaluated in vivo for their ability to produce IFN-γ in immunized BALB/c mice spleen cells. We demonstrate the ability of these peptides formulated either with poly dl-lactide-co-glycolide (PLG) microparticles or as plasmid DNA vaccine to confer protection in mouse model.

Materials and methods

Bacterial strains, clones and vectors

B. abortus NCTC 10093 (544) and vaccine strain S19 were obtained from Central Public Health Laboratory, Colindale, London, UK, and Indian Veterinary Research Institute, Izatnagar, UP, India, respectively. B. abortus DB79BRAB4, a wild strain, was taken from Defence Research & Development Establishment (DRDE)’s culture collection. E. coli host strain DH5α (Novagen, USA) was used for cloning. pVax1 (Invitrogen, USA) eukaryotic expression vector was used for plasmid DNA (pDNA) vaccine preparation. All Brucella cultures were grown in Brucella broth (HiMedia, India) at 37 °C with 5 % CO2 for 2–3 days. For vaccination and challenge, the cultures were suspended in a sterile phosphate-buffered saline (PBS) (0.01 M, pH 7.2), and colony-forming units (CFU) were determined by the plate count method. All live B. abortus manipulations were performed in biocontainment safety level 3 (BSL-3) facility.

Animals

Female BALB/c mice (6–8 weeks old) were obtained from the Animal Facility of DRDE and were given water and food ad libitum. The mice were maintained and used in accordance with the recommendations of the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), Ministry of Forests and Environment, Govt. of India. The study had an approved animal protocol from the Institutional Animal Ethics Committee of DRDE, Gwalior (protocol no. MB-06/48/SK). The challenged mice were housed in animal biocontainment safety level 3 (ABSL-3) facility.

Prediction of MHC class I- and class II-restricted T cell epitopes by immunoinformatics tools

Various Brucella proteins earlier shown to be either protective proteins or virulence determinants were chosen for designing the peptides (Table 1). The amino acid sequence of these proteins was downloaded from Brucella bioinformatics portal (www.violinet.org). MHC-I-restricted epitopes were designed for H-2Dd, H-2Kd and H-2Ld haplotypes, while MHC-II-restricted epitopes were designed for H-2Ad and H-2Ed haplotypes. MHC-I-restricted epitopes were designed using tools available at Immune Epitope Data Base (IEDB), Rankpep and BIMAS. Lowest Ic50 values (ANN, <500) or top 1 percentile of high binders (consensus) was used as cut-off in IEDB method. Rankpep determines the binding threshold by itself and predicts the epitopes accordingly. Epitopes having highest half-life of dissociation (>100) were chosen by BIMAS. The epitopes picked up by most softwares were shortlisted for synthesis. Rankpep and IEDB tools were used to predict the MHC-II-restricted epitopes using similar cut-offs. Rankpep predicted 9-mer epitopes, whereas IEDB predicted 15-mer epitopes.

Table 1 Protective proteins and virulence determinants used for designing peptides by various softwares
Full size table

Selection of peptides

The predicted epitopes were custom synthesized as peptides and screened for shortlisting by animal experimentation in BALB/c mice. This was carried out by their ability to induce IFN-γ production (peptide immunogenicity) or be immune dominant during natural infection (natural processing). For peptide immunogenicity, a group BALB/c mice (n = 5) was immunized subcutaneously with pool of peptides (50 μg each peptide) in PBS emulsified 1:1 with incomplete Freund adjuvant (IFA). The pools of peptides used for immunization in various experiments have been shown in Supplementary data (Table S1). Control mice were injected PBS emulsified in 1:1 IFA. For natural processing, BALB/c mice were injected with 2 × 105 CFU of B. abortus strain 544 by intraperitoneal injection, and control mice were injected PBS. The mice were sacrificed on day 10, and the spleen of each mouse was separated. Determination of IFN-γ in spleen culture supernatant was determined by the earlier described method (Kumar et al. 2009). The splenocytes were stimulated with 25 and 50 μg ml−1 concentrations of individual peptide for 72 h at 37 °C and 5 % CO2. Appropriate positive (Con A, 5 μg ml−1) and negative controls (without antigen) were also included. The supernatant was collected, and level of IFN-γ was measured using commercially available ELISA kits (R & D Systems, USA). For final selection of the above-shortlisted peptides, another round of screening by peptide immunogenicity and natural processing was undertaken. The experiments were same as above except that 100 μg of each peptide was used for mice immunization instead of 50 μg for peptide immunogenicity and booster dose was given at day 7. For natural processing experiment, single dose with field isolate B. abortus DB79BRAB4 was administered instead of strain 544. The animals were sacrificed on day 14 for quantification of IFN-γ.

Preparation of PLG microparticles with adsorbed/entrapped peptides

PLG microparticles for adsorption and entrapment by selected or irrelevant peptides were prepared by a solvent evaporation technique as described by Singh et al. (2004). Briefly, for preparation of PLG microparticles, 10 ml of 6 % w/v PLG solution in methylene chloride (Sigma-Aldrich, USA) was homogenized with 2.5 ml PBS using a 10-mm probe (Sonics) for 10 min, thus forming a water-in-oil emulsion. This emulsion was added to 50 ml of triple-distilled water (TDW) containing 6 μg ml−1 docusate sodium salt (DSS) and homogenized at very high speed using a homogenizer with a 20-mm probe for 5 min in an ice bath. This resulted in water-in-oil-in-water emulsion that was stirred at 1000 rpm for 12 h at room temperature, and the methylene chloride was allowed to evaporate. The resulting PLG content was lyophilized and weighed. For preparation of PLG microparticles with adsorbed peptides, a suspension containing 15 mg of PLG was incubated with 1 mg of peptide in 1 ml of PBS buffer and left overnight on a rocker at 4 °C. This was followed by centrifugation at 10,000g for 10 min. Pellet was re-suspended in PBS, and supernatant was collected to determine the concentration of peptide.

For preparation of PLG microparticles with entrapped peptides, 1 mg of peptide was added in the first step of PLG preparation and the resultant water-in-oil emulsion was added to 50 ml of 10 % w/v poly vinyl alcohol. The rest of the procedure was same as described for preparation of PLG microparticles. The microparticles were washed with TDW by centrifugation and lyophilized. The adsorption/entrapment efficiency of peptide to the microparticles was determined by subtracting the calculated peptide amount in supernatant of the final sample from initial concentration used. PLG microparticles were loaded individually with selected or irrelevant peptides and pooled later on. The microparticles were characterized by scanning electron microscopy (Quanta 400 ESEM-EDX, FEI).

Immunization with PLG microparticle formulations

Groups of mice (n = 10) were immunized subcutaneously with PLG microparticles containing either adsorbed (APLG-pep) or entrapped peptides (EPLG-pep). Each mouse in these groups received a dose of 25 μg of each peptide on days 0 and 60. Equivalent control groups of animals injected with microparticles of both adsorbed and entrapped types with irrelevant peptides (APLG-NP and EPLG-NP) were included. A group of mice was immunized subcutaneously with the pool of peptides (effective peptide/mouse was 25 μg) with IFA (Pep-Ad) with same immunization schedule. The negative control group animals were injected with PBS on days 0 and 60, and the positive control group of mice (n = 6) was inoculated intraperitoneally with B. abortus strain S19 vaccine strain (5 × 104 CFU/mouse) on day 60 and used for protection studies.

Splenocyte proliferation and cytokine assays

After 28 days of last immunization (day 88), the spleens from mice were removed aseptically and the single cell suspension of splenocytes was taken for splenocyte proliferation and cytokine assays as described earlier (Kumar et al. 2009). Briefly, the mice were sacrificed and the spleen of each mouse was separated. The splenocytes were suspended in 96-well tissue culture plate ca. 1 × 106 cells ml−1 (100 μl well−1) along with one volume of homologous peptide (50 μg ml−1) in RPMI 1640 medium. Appropriate positive (Con A, 10 μg ml−1) and negative controls were also included. After 48 h of incubation, 0.2 volume of alamarBlue dye (Biosource, USA) was added to each well and the plate was incubated for further 15–18 h. Experiments were carried out in triplicate wells for each mouse. The reading was taken at 570 and 600 nm, and the results are expressed as mean specific absorbance by subtracting the absorbance at 600 nm from 570-nm absorbance. Supernatants from parallel cultures were harvested after 72 h, and level of IFN-γ was measured using commercially available ELISA kits (R & D Systems, USA).

Generation of ‘bead of string’-based vaccine construct

DNA encoding BP238, Zn317, L923, CSA124, E2o343 and Orf136 peptides was custom synthesized without linker in pBluescript SK vector and subcloned into eukaryotic expression vector pVax1. Vaccine-grade pDNA (pVaxPep) was prepared by Qiagen maxi kit as per manufacturer’s instruction and was used to immunize BALB/c mice (n = 8) by in vivo electroporation by the earlier described method (Jain et al. 2014a). Mice were given two doses of pDNA (100 μg) on days 0 and 28. The control mice were injected with PBS or empty pVax1 vector. A positive control group (n = 6) vaccinated intraperitoneally with B. abortus strain S19 as described above was also included and was given a single dose on day 28.

Protection assay

Protection studies were carried out by previously described method (Jain et al. 2014b). Briefly, 30 days after last injection (day 90 for PLG groups and day 58 for DNA vaccine groups), six immunized mice from all groups were challenged with 2 × 105 CFU mouse−1 of B. abortus strain 544 by intraperitoneal injection. The challenged animals were sacrificed 30 days after being challenged by cervical dislocation, and their spleens were removed aseptically. The numbers of colonies were counted after 48 h, and the results were expressed as the mean log CFU spleen−1 ± SD per group. Units of protection were obtained by subtracting the mean log CFU spleen−1 of the vaccinated group from the mean log CFU spleen−1 of the control (PBS immunized) group.

Statistical analysis

The cytokine, splenocyte proliferation and protection study data of all experiments was analyzed statistically by one-way ANOVA followed by Tukey’s multiple comparison test (GraphPad Prism 6.0 USA). Peptide immunogenicity and natural processing data was analyzed by Student’s t test between peptide and PBS groups. Difference between two groups (p < 0.05 or lesser) was considered as significant.

Results

In silico prediction of MHC-I- and MHC-II-restricted T cell epitopes of B. abortus proteins

A total of 93 MHC-I-restricted epitopes were predicted, out of which 37 highly scored binding epitopes by maximum number of prediction tools were custom synthesized for in vivo experiments (Table 2). Epitopes predicted by single tool were not considered. Four negative control (NC) epitopes having very low binding values were also included. By Rankpep, 118 9-mer MHC-II epitopes were predicted, while by IEDB, about 80–100 15-mer epitopes were predicted for each protein. Since no rational criterion to shortlist the peptide synthesis was available, hence, we chose to randomly select 32 peptides for in vivo screening. Here, also, four peptides having very low binding values were also included as NC (Table 3).

Table 2 MHC-I-restricted T cell epitopes shortlisted for in vivo studies
Full size table
Table 3 MHC-II-restricted T cell peptides shortlisted for in vivo studies
Full size table

Determining the peptide immunogenicity in mice

The predicted epitopes were screened on the basis of their ability to induce IFN-γ production in mice (peptide immunogenicity) or their ability to get naturally processed for IFN-γ production during infection with live culture of B. abortus (natural processing). Peptides with vaccine potential were expected to be positive for both peptide immunogenicity and natural processing.

Twelve MHC-I-restricted peptides, e.g. Bp238, Zn316, Gap71, Pur71, Zn317, Dna563, Mp130, Hp65, L923, CSA124, O2b90 and E2o343, were found to release significant amount of IFN-γ in both peptide immunogenicity and natural processing (Fig. 1a, b). Surprisingly, the predicted low affinity binder, Zn317, which was included as NC, produced high level of IFN-γ. To further validate the results, the above-shortlisted MHC-I peptides were taken for second round of in vivo screening. Seven of the 12 peptides induced IFN-γ (Fig. 3a) and were taken further for natural processing experiment.

Fig. 1
figure 1

In vivo peptide screening (round 1) for MHC-I-restricted T cell peptides by peptide immunogenicity (a) and natural processing (b). Mice were injected with pool of peptides as shown. Release of IFN-γ in culture supernatant of splenocytes was determined after stimulation with individual peptide of the pool. Significant differences of comparison with PBS group were determined by Student’s t test and are indicated by asterisk (p < 0.05). Results shown at stimulating concentration of 50 μg ml−1 of peptide

Full size image

MHC-II-restricted peptides were relatively poor produces of IFN-γ. Only three peptides Orf136, Hp55 and Rpsl108 produced statistically significant amount of IFN-γ in peptide immunogenicity and natural processing (Fig. 2a, b). In the second round, only one peptide, Orf136, was found immunogenic (Fig. 3b). All seven MHC-I and one MHC-II shortlisted peptides were tested by natural processing, and five MHC-I and one MHC-II peptides were selected for studying the vaccine potential in mice (Fig. 3c, Table 4). Some of the peptides that did not release IFN-γ in immunogenicity or natural processing experiments on repeated occasions were randomly selected as irrelevant peptides (as NC) for immune response and protection studies (Table S2).

Fig. 2
figure 2

In vivo peptide screening (round 1) for MHC-II-restricted T cell peptides by peptide immunogenicity (a) and natural processing (b). Mice were injected with pool of peptides as shown. Release of IFN-γ in culture supernatant of splenocytes was determined after stimulation with individual peptide of the pool. Significant differences of comparison with PBS group were determined by Student’s t test and are indicated by asterisk (p < 0.05). Results shown at stimulating concentration of 50 μg ml−1 of peptide. Pool 4 peptides did not release IFN-γ and, hence, have not been shown

Full size image
Fig. 3
figure 3

In vivo peptide screening (round 2) by peptide immunogenicity for MHC-I (a)- and MHC-II-restricted T cell peptides (b). The shortlisted peptides were tested by natural processing (c). Mice were injected with pool of peptides as shown. Release of IFN-γ in culture supernatant of splenocytes was determined after stimulation with individual peptide of the pool. Significant differences of comparison with PBS group were determined by Student’s t test and are indicated by asterisk (p < 0.05). The results are shown at two stimulating concentrations of 25 and 50 μg ml−1

Full size image
Table 4 Peptides selected for immunogenicity and protection studies
Full size table

Evaluation of epitope-based vaccine formulated in PLG microparticles

PLG microparticles were prepared as described in “Materials and methods” section, and their sizes were determined by scanning electron microscope (Fig. S1). Size range of microparticles of adsorbed type (APLG) with or without peptide was found out to be 0.25–1.25 or 0.3–1.0 μm, respectively. PLG microparticles with entrapped (EPLG) peptides were of size range of 0.3–1.21 μm. Adsorption and entrapment efficiency of PLG microparticles were determined to be 89.6 %.

Mice were immunized on day 0 and day 60 with APLG-pep/EPLG-Pep formulations. Splenocytes from immunized mice were prepared 28 days after the last dose to investigate the cellular immune response. Splenocytes from EPLG-pep-, APLG-pep- and Pep-Ad-vaccinated mice had significant proliferative response to all peptides when compared with APLG-NP, EPLG-NP and PBS control groups (p < 0.0001). No proliferation occurred in cultures of PBS group mice by the selected peptides and of EPLG-NP (except vjbR30) and APLG-NP group mice by irrelevant peptides. The proliferative response to most peptides in EPLG-pep or APLG-Pep groups was higher than in the Pep-Ad group (Fig. 4a). The ConA mitogen was able to induce spleen cell proliferation in all cultures non-specifically.

Fig. 4
figure 4

Immunogenicity determination of peptide PLG formulations by splenocyte proliferation response (a) or IFN-γ production in the spleen cells (b). Splenocyte proliferation in response to peptides Bp238, Zn317, L923, CSA124, E2o343 and Orf136 is shown. APLG-NP and EPLG-NP group was stimulated with irrelevant peptides. Significant differences of comparison among groups were determined by one-way ANOVA and are indicated by ***p < 0.0001

Full size image

Release of IFN-γ in the culture supernatants of spleen cells of immunized mice was evaluated by ELISA. A significant (p < 0.0001) production of IFN-γ in spleen cells from mice of APLG-pep, EPLG-pep and Pep-Ad groups in comparison to APLG-NP, EPLG-NP and PBS groups was observed. Splenocytes of mice from groups APLG-NP, EPLG-NP and PBS did not produce IFN-γ (Fig. 4b). Except for peptide L923, stimulation by all other peptides induced higher levels of IFN-γ in EPLG-Pep or APLG-Pep groups than Pep-Ad group (p < 0.05 to p < 0.0001). Between EPLG-Pep and APLG-Pep groups, peptides Bp238, Zn317 and L923 produced higher IFN-γ in the former, whereas E2o343 and Orf136 produced higher IFN-γ in the latter (p < 0.001).

Protection studies were carried out by challenging APLG-Pep, EPLG-Pep, Pep-Ad, APLG-NP, EPLG-NP, PBS and B. abortus strain S19 group mice with B. abortus strain 544. Mice of EPLG-Pep, APLG-Pep and Pep-Ad groups exhibited a significant degree (1.69, 1.49 and 1.42 log units of protection, respectively) of protection compared to control receiving PBS (p < 0.001; Table 5). The mice immunized with S19 vaccine conferred 2.23 log units of protection. Results indicate that EPLG-Pep, APLG-Pep and Pep-Ad provided protection against B. abortus infection.

Table 5 Protection against B. abortus strain 544 in BALB/c mice by vaccination with peptides in formulation with PLG microparticles
Full size table

Evaluation of epitope-based pDNA vaccine

Synthetic gene of six selected epitopes was custom synthesized, was subcloned in pVax1 vector and was injected in the anterior tibialis of mice by in vivo electroporation. The mice were challenged 30 days after the last dose. Mice injected with pVaxPep exhibited a significant degree (1.66 log units of protection) of protection compared to control receiving PBS (p < 0.001; Table 6). The mice immunized with S19 vaccine induced 2.05 log units of protection.

Table 6 Protection against B. abortus strain 544 in BALB/c mice by vaccination with plasmid DNA construct encoding epitope sequences
Full size table

Discussion

The focus of the present work was to identify the T cell epitopes of B. abortus that can be relevant for enhancing the Brucella-specific cellular immunity of BALB/c mice. In order to increase the likelihood of getting large numbers of high-affinity binding epitopes as well as to keep the generated epitope design data to manageable levels, instead of using the full Brucella proteome, we chose 23 earlier reported protective proteins or virulence determinants to design the T cell epitopes. MHC-I- and MHC-II-restricted T cell epitopes were designed by tools available online, and the peptides that were predicted to be high binders were chosen. The prediction of MHC-II peptides appeared to be less accurate, as large numbers of high-scoring peptides were predicted by the two used softwares.

Since IFN-γ is a key cytokine for providing protection against Brucella infection (Jiang and Baldwin 1993; Paranavitana et al. 2005), therefore, the release of IFN-γ in splenocyte culture supernatant of mice was taken as a criterion for in vivo screening of T cell peptides. Further, screening of the peptides in the present study was decided not only by their ability to be immunogenic but also by their ability to be presented by MHC-I or MHC-II after host infection with intact B. abortus. Therefore, those peptides that released IFN-γ in peptide immunogenicity as well as in natural processing were selected. Similar approach was earlier used in MHC-I-restricted CD8 Brucella T cell peptide screening (Durward et al. 2010). In this study, a total of 37 high-affinity MHC-I- and 32 MHC-II-restricted T cell peptides were tested in mouse model, and after multiple rounds of screening, five MHC-I peptides, i.e. Bp238, Zn317, L923, CSA124 and E2o343, and one MHC-II peptide, i.e. Orf136, were finally selected for evaluation as epitope-based vaccine. Of the five MHC-I peptides, three were predicted by four softwares and one was predicted by two softwares, suggesting that prediction by MHC-I prediction tools is good and comparable. One peptide Zn317, however, was initially taken as NC yet showed release of IFN-γ in all experiments and was one of the finally selected peptides. This indicates that immunoinformatics tools for MHC-I prediction though may be good, they are yet to mature. Similar results were shown earlier by Durward et al. (2010) who also found that one of the NC peptides released significant amount of IFN-γ in mouse model. We did not find the prediction for MHC-II T cell epitopes to be that good. Although a large number of high-scoring epitopes were predicted, yet of the 32 synthesized peptides, only one could finally be selected.

The selected peptides were evaluated for immunogenicity and protective ability in mouse model. The peptides were formulated with PLG microparticles for administration. Biodegradable lactide polymers, such as PLG microparticles, have been used extensively in research pertaining to delivery of novel vaccine and drug candidates. The encapsulation of antigen in a polymer matrix limits access of the biological fluid into the antigen until the time of degradation. The adjuvant effect achieved through encapsulation and adsorption of antigens to PLG microparticles was demonstrated by several groups (O’Hagan et al. 2004). The microparticles used in this study were of size smaller than 1.5 μm. Particle size was shown to be an important parameter affecting the immunogenicity of microparticles, and smaller particles (<10 μm) were found to be significantly more immunogenic than the larger ones (O’Hagan et al. 2004). Immunogenicity of peptides with PLG microparticles was evaluated in BALB/c mice. After stimulation with individual peptide, a significant proliferation of splenocytes and release of IFN-γ were observed in mice of APLG-Pep and EPLG-Pep groups when compared with equivalent NC (APLG-NP, EPLG-NP) or PBS groups. A similar response but of lower intensity was observed in Pep-Ad group. Similar to our work, more potent responses of PLG formulations than the established aluminium-based adjuvant have been shown earlier (Singh et al. 2004). Comparable immunogenicity between Freund’s adjuvant formulated and microparticle-entrapped staphylococcal B enterotoxoid has also been shown (Eldridge et al. 1991). Delivery of hepatitis B virus peptides in PLG microparticles has earlier been shown by Moynihan and Howard (2001) to elicit T cell response. For determination of protective ability of any Brucella vaccine in mouse model, the reduction in the number of bacteria in spleen is observed. Expectedly, APLG-Pep, EPLG-Pep and Pep-Ad immunization conferred protection against B. abortus infection (1.49, 1.69 and 1.42 units of protection). EPLG-Pep conferred best protection in terms of protection units, which suggests that peptides entrapped in PLG microparticles can be most suitable as vaccine delivery method. However, the protection conferred by S19 vaccine strain was superior to PLG-entrapped peptides.

The roles of both CD8+ and CD4+ T cells have been described in protection against Brucella infection in mice. There are reports that suggest either in vivo depletion of CD8+ T cells results in higher bacterial load of B. abortus in infected BALB/c mice or CD8+ T cells kill Brucella specifically (He et al. 2001; Murphy et al. 2001). Five MHC-I-restricted peptides used in this study are likely to be responsible for conferring protection in mice. Earlier, two MHC-I restricted CD8+ T cells peptides were not tested for protection (Durward et al. 2010). It is difficult to comment to what extent one MHC-II-restricted CD 4+ T cell peptide contributed for protection in mice in this study. It has been suggested earlier that IFN-γ-producing CD4+ T cells have a major role in clearing the Brucella bacteria and CD8+ T cells and humoral response have only modest role to play (Vitry et al. 2012). We do not know whether only one MHC-II-restricted T cell peptide is one of the reasons for lesser protection by PLG-formulated epitope vaccine than S19 vaccine in this study. Nonetheless, our results show that MHC-I- and MHC-II-restricted CD8+ and CD4+ T cell peptides contribute to release of IFN-γ and protection in mouse model.

The pDNA vaccines seem to offer the best approach to activate both cellular components of the immune response (Th1 and CD8+ T cell), owing to the intrinsic feature of DNA vaccine to produce endogenous antigen in professional antigen-presenting cells (Liu et al. 2004). Therefore, to further evaluate the usefulness of selected peptides as vaccine, a pDNA vaccine construct was made and injected to mice using in vivo electroporation. In vivo electroporation-enhanced delivery of pDNA has been used in many animals and has resulted in increased DNA uptake in muscle, leading to robust trans-gene expression levels and enhanced humoral and cellular responses (Estein et al. 2009). It was observed that the protection conferred by pDNA vaccine delivered by in vivo electroporation was significantly higher when compared with PBS control or naked vector (pVax1) control groups. Protection offered by pDNA vaccine was 1.66 units and was comparable to S19 vaccine conferred protection.

Overall, results of this study prove that screened peptides are immunogenic and protective against B. abortus infection in mouse model. It is evident that the protection is mediated through release of IFN-γ cytokine. However, the protection conferred by selected peptides is lesser than the available S19 vaccine. Presence of more antigenic epitopes in S19 or CpG DNA in pDNA vaccine construct may have been the factors for better protection by S19 live or pDNA vaccine than the PLG-formulated epitope vaccine. It is also likely that cytokine other than IFN-γ or any other factors may also have a role in protection. New research on immunology of Brucella can give us better insight on this. Nonetheless, the results of this study support the feasibility of epitope-based brucellosis vaccine.