Introduction

Brucellae are Gram negative, facultative, intracellular bacteria that cause an important zoonotic disease called brucellosis (Young 1994). The genus contains six-well recognized (Brucella abortus, B. melitensis, B. suis, B. canis, B. ovis, and B. neotomae) and four recently added (B. ceti, B. pennipedialis, B. microti, and B. inopinata) species (Corbel 1997a; Franco et al. 2007). 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). Brucellosis is one of the commonest zoonotic diseases, with more than 500,000 new cases annually (Pappas et al. 2006b). In humans, the major cause of brucellosis is B. melitensis (Corbel 1997a), although several cases have also been attributed to B. abortus (Young 1994; Ashford et al. 2004). Human brucellosis manifests itself as a chronic infection with undulant fever and general malaise; other clinical signs vary depending on the affected organ systems (Franco et al. 2007). It is a weakening disease that requires prolonged antibiotic treatment, lasting at least 6 weeks in moderate cases and may extend for years in complicated cases (Young 1994; Pappas et al. 2005). 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). Brucella abortus is considered potential bioterror agent and has been classified as NIAID Category B priority pathogens because of the ease of aerosolization, diverse symptoms, and chronic persistence (Pappas et al. 2006a). Brucellosis is also one of the most common laboratory acquired infections (Noviello et al. 2004).

Attenuated live Brucella strains, such as B. abortus RB51 and S19, and B. melitensis Rev1, are being used as vaccines to control brucellosis in domestic animals (Schurig et al. 2002). However, these vaccines are not suitable for humans since they can cause disease even in healthy individuals (Blasco and Díaz 1993; Perkins et al. 2010). The live attenuated B. abortus S19 and RB51 veterinary vaccines have limitations, as they may induce abortions when administered to pregnant animals (Schurig et al. 1991; Corbel 1997b) and show cross-reaction with natural infection during sero-diagnosis (Stevens et al. 1994, 1995). Therefore, there is an urgent need to develop efficacious vaccines to combat human form of infection and also to improve the veterinary vaccine on the same time.

In general, the use of live attenuated vaccines is a time tested approach, yet it possess some problems in terms of safety e.g. risk of reversion to the original pathogenic form and shedding in the environment with danger to immunocompromised individuals (Liljeqvist and Stahl 1999; Hansson et al. 2000). Hence, many laboratories, including ours, are working on the development of subunit vaccines. Many immunogenic Brucella proteins i.e. p39, Cu–Zn SOD, L7/L12, BLS, Omp31, GroEL, YacJ, DnaK, SurA, Mdh have been tried as recombinant subunit vaccines, but only a few have conferred significant protection (Xiang et al. 2008). In this study, we targeted cysteine synthase A (CysK) of B. abortus which was shown to be immunogenic protein in a proteomic analysis of B. abortus cell envelope (Connolly et al. 2006). Cysteine synthase A and cysteine synthase B are the key enzymes of cysteine biosynthesis in bacteria and they form a bienzyme complex, called cysteine synthase (Tanous et al. 2008). CysK is involved in various stress conditions as it was shown that its expression increased during bile stress in Lactobacillus casei (Wu et al. 2010) and during chromate stress in E. coli (Ackerley et al. 2006). In the line with the crucial role of cysteine synthesis in response to environmental changes, we studied immunogenicity and protective potential of cysteine synthase A.

In the present work, gene encoding cysteine synthase A (cysK) was cloned, expressed in E. coli and the recombinant protein was purified up to homogeneity by metal affinity chromatography. Immunogenicity and protective ability of the recombinant protein in formulation with two different adjuvants were evaluated in BALB/c mice. The rCysK was found to generate specific antibody response and release IFN-γ in spleen lymphocytes of immunized mice. Further, rCysK also conferred partial protection against challenge with virulent strain of B. abortus.

Materials and methods

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 Institutional Animal Ethics Committee of DRDE, Gwalior. The challenged mice were housed in biocontainment safety level -3 (BSL-3) facility.

Bacterial strains and plasmids

Escherichia coli host strains BL21 (DE3) and DH5α, and pET28a+ expression vector (Novagen, USA) were used for cloning and expression of gene. All E. coli strains were routinely grown at 37 °C in Luria–Bertani broth or agar with appropriate concentration of kanamycin whenever needed. B. abortus strain NCTC 10093 (544) was obtained from NCTC, UK. Pathogenic B. abortus strain S99 and vaccine strain S19 were obtained from our own culture collection. 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 plate count method. All live B. abortus manipulations were performed in BSL -3 facility.

Cloning and expression of cysK gene and purification of rCysK

Full length open reading frame of cysK gene was amplified by PCR from the genomic DNA of B. abortus S99. The primers with NdeI and XhoI restriction sites at the 5′ ends were designed in accordance to the sequence information available in GenBank (Gene ID: 3341091). The primers used were: CysKF- 5′ TATGCACATATGTTCAATTCGGTACTCGA3′ and CysKR- 5′ TAGCAGCTCGAGTTACCCCTCGAACGGTATGT3′. Plasmid pET28a+ and purified PCR product (purified using GenElute PCR Clean-up kit, Sigma, USA) were digested with NdeI and XhoI followed by setting up of ligation reaction. Ligated product was transformed in E. coli DH5α competent cells and the positive clones were selected according to standard protocol (Sambrook and Russell 2001). The integrity of cloned gene was confirmed by restriction digestion analysis and DNA sequencing. The plasmid of the clone containing the insert was used to transform competent cells of E. coli BL21 (DE3). The recombinant protein was expressed by induction with 0.5 mM isopropyl-β-D-thiogalactopyranoside (IPTG) in LB medium containing kanamycin. The rCysK was found to express in soluble fraction. The recombinant protein was purified by Ni+-NTA resin (Qiagen, Germany) using imidazole as the elution reagent according to manufacturer’s instructions. The lysate of IPTG induced cells and the purified protein were identified by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE). Confirmation of expressed rCysK was done by Western blotting using monoclonal anti-polyhistidine HRP conjugate (Sigma, USA). The purified rCysK was dialysed against buffer containing 50 mM NaH2PO4 (pH 8.0) with reduced concentration of imidazole (150, 100 and 50 mM) and NaCl (200, 100 and 50 mM) for 12 h; and finally with buffer lacking both. The endotoxin level in purified protein was determined by a chromogenic LAL (limulus amebocyte lysate) endotoxin assay kit (Lonza, Switzerland).

Immunization

Groups of BALB/c mice (n = 12) were injected with a total three doses (day 0, 14 and 21) of rCysK without any adjuvant (rCysK group), rCysK formulated with Freund’s complete/incomplete adjuvant (rCysK-FA group) and rCysK with aluminium hydroxide gel (rCysK-Al group) by subcutaneous route. Each animal was injected with 25 μg (100 μl) of recombinant protein in every dose. First dose in FA group was with Freund’s complete adjuvant and boosters were administered with Freund’s incomplete adjuvant. A group of mice was injected PBS (0.01 M, pH 7.2) subcutaneously and was included as negative control (PBS Control). Sera from all groups were collected on day 0 and 28 to determine the antibody response.

Indirect enzyme linked immunosorbent assay (ELISA)

Levels of anti-rCysK antibodies in the sera of the immunized mice were assayed by ELISA. Briefly, the polystyrene plates (Nunc-Immuno Plate MaxiSorp surface, Denmark) were coated with rCysK (10 μg/ml, 100 μl/well) in carbonate-bicarbonate buffer (0.05 M, pH 9.6) and were incubated overnight at 4 °C. Next morning the plates were washed twice with PBST (PBS, 0.01 M, pH 7.2 containing 0.05 % vol/vol Tween 20). Blocking was done with 1 % (wt/vol) BSA in PBS (200 μl/well) and incubated for 2 h at 37 °C. After three washings with PBST, the plates were incubated with 100 μl of test sera dilution (two fold) in PBS containing 0.1 % BSA for 1 h at 37 °C. The wells were washed five times with PBST. Goat anti- mouse (IgG) antibodies and isotypes (IgG1 and IgG2a) conjugated to horseradish peroxidase obtained from Sigma and Pharmingen, respectively, were diluted as recommended by manufacturers and were added to the wells. After incubation for 1 h at 37 °C and subsequent washes, plates were developed with ortho-phenylenediamine (0.4 mg/ml) and H2O2 (6 %, 0.4 μl/ml) in citrate phosphate buffer (pH 4.5). The absorbance was measured at 492 nm in an ELISA reader (BioTek, USA) and values are presented as end point titres. The reciprocal of highest dilution of test sera having OD more than the mean OD (+2 SD) of preimmune serum (1:100 dilution) in the same group was considered as the end point titre.

Lymphoproliferation and cytokine assays

For lymphoproliferation assay, four mice from each group were sacrificed on day 49 (28 days after last immunization) and their spleens were removed aseptically. Spleen cell proliferation was carried out by an earlier described method (Kumar et al. 2009). Briefly, single cell suspension of splenocytes was prepared by maceration of spleens and by lysing erythrocytes with ammonium chloride solution. The splenocytes were suspended in 96-well tissue culture plate at ca. 1 × 106 cells/ml (100 μl/well) along with one volume of antigen (1, 5 or 10 μg/ml) in RPMI1640 medium. Appropriate positive (Con A, 5 μg/ml) and negative controls (without antigen) were also included. After 48 h of incubation, 0.1 volume (20 μl) of alamar blue 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 results are expressed as mean specific absorbance, which was calculated by subtracting the delta absorbance of cells without antigen from the delta absorbance of cells with antigen. Delta absorbance is obtained by subtracting readings of cells at 600 nm from that at 570 nm.

Supernatants from parallel cultures stimulated with 10 μg/ml of recombinant protein were harvested after 72 h and stored at −70 °C until assayed for cytokine interferon-γ (IFN-γ). Appropriate positive (Con A, 5 μg/ml) and negative controls (without antigen) were also included. The levels of IFN-γ were determined in culture supernatants using sandwich ELISA (OptEIA, Pharmingen, USA) according to manufacturer’s instructions. The levels of cytokine were determined with the help of standard curves that were plotted using recombinant cytokine (Pharmingen) and are expressed as pg/ml.

Protection assay

A group (n = 6) of mice (vaccine group) was inoculated intraperitoneally with B. abortus S19 (5 × 104 CFU/mouse) on day 21 and included as positive control for protection studies. On day 51 (30 days after last injection), six immunized mice from vaccine group and the groups described above were challenged with 2 × 105 CFU/mouse of B. abortus strain 544 by intra-peritoneal injection. The challenged animals were sacrificed on day 81 (30 days after challenge) by cervical dislocation and their spleens were removed aseptically. Each spleen was homogenized, serially diluted in PBS and plated on Brucella agar. The plates were incubated at 37 °C with 5 % CO2. The numbers of colonies were counted after 48 h and the results were expressed as the mean log CFU/spleen ± SD per group.

Statistical analysis

CFU data was logarithmically transformed and statistically analysed by analysis of variance using multiple comparisons versus control group by Dunnett’s method. A non parametric Mann–Whitney rank-sum test was used to compare the cellular response (Sigma Stat, Jandel Scientific, USA).

Results

Production of purified rCysK

To obtain the recombinant CysK, E. coli BL21 (DE3) cells harbouring pETCysK were induced with 0.5 mM IPTG. SDS-PAGE analysis of the cell lysate showed band of ~39 kDa rCysK (Fig. 1a). Expressed rCysK was confirmed by Western blotting using monoclonal anti-polyhistidine antibody (Fig. 1b). The rCysK protein was purified by affinity chromatography with Ni+-NTA column. Purity of the recombinant protein (>95 %) was estimated by staining 12 % (wt/vol) SDS-PAGE with Coomassie blue stain. Figure 1c shows the SDS-PAGE profile of purified rCysK. Expression level of rCysK was found to be 18.6 mg/l of medium. The protein preparation contained <0.01 EU/μg of protein.

Fig. 1
figure 1

Expression and purification of rCysK. a A 12 % SDS-PAGE profile showing the expression of rCysK in E. coli BL21 (DE3) cells. Lane 1 Protein molecular mass marker (in kilodaltons), Lane 2 Un-induced E. coli BL21 (DE3) cell pellet, Lane 3 Induced E. coli BL21 (DE3) cell pellet. b Western blots showing the specific expression of rCysK using monoclonal anti-polyhistidine conjugate. Lane 1 Induced E. coli BL21 (DE3) cell pellet, Lane 2 Protein molecular mass marker. c 12 % SDS-PAGE analysis showing the affinity purified rCysK. Lane 1 Protein molecular mass marker, Lane 2 Purified protein by immobilized metal affinity chromatography

Full size image

Antibody response elicited by rCysK immunization

Immune response to rCysK was studied in BALB/c mice after subcutaneous inoculation. The generation of antigen-specific antibodies was measured in the sera of the immunized mice 1 week after the last dose by indirect ELISA. Maximum anti-rCysK IgG antibody titres were observed in rCysK-FA group followed by rCysK-Al, rCysK and PBS groups (Fig. 2a). Highest IgG titre of 32,000 was obtained in rCysK-FA group mice. The IgG titre in serum samples of rCysK-Al group mice was 16,000 (Fig. 2b). The sera were also analysed for presence of different IgG isotypes. IgG1 titres predominated over IgG2a in all immunized mice, and maximum IgG titres were obtained in rCysK-FA group (Fig. 2c).

Fig. 2
figure 2

Anti- rCysK IgG (a and b) and IgG isotypes (c) titres in the sera of rCysK immunized mice and PBS control (n = 12). The mice were bled on day 0 and 28 retero-orbitally and antibody titres were evaluated by ELISA. Data are the representative of three separate experiments. Each bar represents the mean end point titre ± SD

Full size image

Lymphocyte proliferation and IFN-γ response

The proliferative response of splenocytes harvested 28 days after the last dose from immunized mice was investigated. No significant proliferation was observed in spleen cells of rCysK (with or without adjuvants) immunized mice when compared to PBS group. The ConA mitogen was able to induce spleen cell proliferation in all cultures non-specifically. Splenocytes of only rCysK-Al group induced the significant (p < 0.05) production of IFN-γ (Fig. 3) as compared to PBS group. ConA was able to produce IFN-γ in various groups and ranged from 1,000 to 1,500 pg/ml (data not shown).

Fig. 3
figure 3

Release of IFN-γ in the spleen cells of rCysK immunized mice and PBS control. IFN-γ in culture supernatants were measured by sandwich ELISA. Results are expressed in pg/ml and represent the mean (±SE) of the 4 animals in each group. Data are the representative of three separate experiments. Significant differences of comparison with PBS group are determined by t test and are indicated by asterisk (p < 0.05)

Full size image

rCysK partially protects BALB/c mice against B. abortus infection

The potential of rCysK to impart protection was evaluated on the basis of their ability to reduce the bacterial load of virulent B. abortus strain in spleens of immunized mice. On day 51, mice were challenged with virulent B. abortus strain 544 and sacrificed after 30 days to determine the number of CFU in their spleens. Mice inoculated with rCysK with either of the adjuvants or without any adjuvant conferred protection against B. abortus infection. To compare the extent to which mice could be protected, a group of mice inoculated with S19 vaccine was included and 2.155 units of protection were obtained in this group. Among rCysK groups, maximum protection was observed with aluminium hydroxide gel formulation (rCysK-Al, 1.064 units) followed by Freund’s adjuvant formulation (rCysK-FA, 0.879 units) (Table 1).

Table 1 Protection against B. abortus strain 544 provided to BALB/c mice by vaccination with rCysK protein with different adjuvant formulations
Full size table

Discussion

Brucellosis is the world’s most common zoonosis, with no effective vaccine available against it. Several strategies are being sought to prevent this disease while avoiding the disadvantages of the currently used live vaccines. One of the attractive vaccine development approaches is the development of subunit vaccines. In this study, we investigated the immunogenicity and protective efficacy of recombinant cysteine synthase A. The gene encoding the protein was cloned in prokaryotic expression vector pET28a+ and the expressed protein was purified to near homogeneity by affinity chromatography.

The immunogenic and protective potential of rCysK formulated in Freund’s complete/incomplete and aluminium hydroxide adjuvant was studied in BALB/c mice. Freund’s adjuvant is the most commonly used adjuvant in research setting, whereas aluminium hydroxide is the only adjuvant licensed for human vaccines. The antibody response to rCysK was determined 1 week after the last dose. The results of this work show that rCysK is able to elicit significantly high levels of IgG antibodies in BALB/c mice when administered with FA and aluminium hydroxide adjuvants. Serum samples collected from immunized and control groups were analysed for the presence of different IgG subtypes. Titres of both IgG1 and IgG2a increased in adjuvant formulated immunized groups in comparison to control group, and IgG1 titres predominated over IgG2a. In Brucella, like many other intracellular pathogens, role of antibodies in protection is considered dispensable. Though, literature is available suggesting protective effects of antibodies in brucellosis (Araya and Winter 1990; Jaques et al. 1992). However, now it is becoming clearer that a humoral response may not be protective but it is indispensable in host defenses as opsonisation is required for successful uptake of Brucella by macrophages (Eze et al. 2000; Baldwin and Goenka 2006) and opsonization may result in increased bacterial uptake by macrophages (Eze et al. 2000).

IFN-γ is the characteristic cytokine of Th1 immune response. The Th1 immune response characterized by IFN-γ is associated with protection against brucellosis (Murphy et al. 2001; Paranavitana et al. 2005; Rafiei et al. 2006). 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 (Boehm et al. 1997). During a CMI response, there is gradual change in the predominant immunoglobulin class of the specific antibody produced. This isotype switch is controlled by T cells and their cytokines. In mice, IFN-γ generally switches activated B cells to the IgG2a and IgG3 isotypes (Roitt et al. 2001). Here, in this study, rCysK induced spleen cells to produce significant levels of IFN-γ in rCysK-Al immunized group, which may have influenced the B cells to secrete IgG2a.

Protective ability of rCysK formulations was studied in BALB/c mice. BALB/c mouse is a proven animal model to study the protection of Brucella vaccine candidates. Immunization with rCysK could confer significant (p < 0.05) protective immunity against B. abortus infection as compared to PBS group. Vaccination with rCysK-Al provided comparatively better protection (1.064 units) than rCysK-FA (0.879 units) or rCysK (0.713 units), though there was no statistical significant difference in protection provided by the different formulations. The protection offered by CysK, however, was lower than the B. abortus S19 vaccine (2.155 units). It is possible that the newer adjuvants which skew immune response to the Th1 type may provide better protection using CysK protein. Nonetheless, the results of this study show that rCysK is able to confer partial protection in mouse model. It is likely that CysK protective efficacy and immunogenicity may further be increased by using other delivery methods like liposome-mediated delivery, escheriosome-mediated delivery and co-stimulation with interleukins. Taken together, we believe that CysK can be a valuable adjunct with other candidates.