Brucella melitensis 16MΔhfq attenuation confers protection against wild-type challenge in BALB/c mice



Chuangfu Chen, College of Animal Science and Technology, Shihezi University, Shihezi 832003, China.

Tel: +86 0993 2058002; Fax: +86 0993 2058612; email:


Brucellosis is a globally distributed zoonotic disease that causes animal and human diseases. Although effective, the current Brucella vaccines (Rev.1 and M5-90) have several drawbacks. The first involves residual virulence for animals and humans and the second is the inability to differentiate natural infection from that caused by vaccination. Therefore, Brucella melitensis 16M hfq mutant (16MΔhfq) was constructed to overcome these drawbacks. Similarly to Rev.1 and M5-90, 16MΔhfq reduces survival in macrophages and mice and induces strong protective immunity in BALB/c mice. Moreover, these vaccines elicit anti-Brucella-specific IgG1 and IgG2a subtype responses and induce secretion of gamma interferon and interleukin-4. The Hfq antigen also allows serological differentiation between infected and vaccinated animals. These results show that 16MΔhfq is an ideal live attenuated vaccine candidate against virulent Brucella melitensis 16M infection. It will be further evaluated in sheep.

List of Abbreviations



complement fixation test




Concanavalin A


Escherichia coli BL21

E. coli

Escherichia coli


interferon gamma


immunoglobulin G








multiplicity of infection


Rose Bengal plate test


standard tube agglutination test


Tris-buffered saline


tryptic soy agar


tryptic soy broth

Brucella spp are gram-negative facultative intracellular pathogens that can multiply within professional and nonprofessional phagocytes of both humans and animals, resulting in heavy economic losses and human suffering [1]. Brucella spp cause acute transmissible infections in animal reservoirs and abortion in pregnant animals [2]. Brucella melitensis, Brucella abortus and Brucella suis also cause human disease. Infection in humans can cause fever, arthritis, spondylitis and dementia and meningitis or endocarditis on rare occasions [3-5]. Currently, there is no effective or safe Brucella vaccine for humans and animals. New low virulence, strongly protective and marker vaccines are required to overcome these drawbacks.

The live vaccine strain Rev.1 was originally derived in 1957 from wild B. melitensis 6056 by sequential passaging on streptomycin-containing media until streptomycin-resistant clones had developed [6]. Rev.1 is the best available vaccine for prophylaxis against brucellosis in sheep and goats [7]. The other B. melitensis vaccine strain, M5-90, was derived from a virulent B. melitensis strain M28 isolated from a sheep and serially passaged for 90 generations in chicken embryo fibroblasts [8]. M5-90 is one of the key factors that decreased the incidence of brucellosis in animals and humans from the 1970s to 1990s in China [9], where it is currently administered to sheep and goats to prevent brucellosis. However, Rev.1 and M5-90 are not safe for pregnant animals because they may cause abortion. Moreover, serological tests cannot distinguish between natural infection and vaccination. One potential approach to the second of these problems is to develop a marker vaccine by deleting virulence or antigenic genes from parental vaccine strains with good immunogenicity and vaccine efficacy. In-depth research on deletion of virulence genes is required for the development of live vaccines against B. melitensis 16M infection that are superior to Rev.1 and M5-90.

Hfq is a protein that mediates RNA-RNA interactions and regulates gene expression at the post-transcriptional level [10, 11]. Hfq is required for the regular expression of some target genes, affects mRNA stability [12], and influences levels of virB transcription regulator [13]. That Hfq plays a role in the virulence of some pathogenic bacteria is clear and well documented [14, 15]; the B. abortus 2308 hfq mutant is severely attenuated in infected mice [14] and B. melitensis 16MΔhfq is attenuated in goats [15]. Previous studies have shown that Hfq is an important genetic determinant of Brucella virulence [16, 17]. Our results, which are consistent with previous results, show that Hfq has an important function in wild-type virulence. When the hfq gene is disrupted, there is a major decrease in virulence of 16MΔhfq. These findings indicate for the first time that 16MΔhfq may be useful as an attenuated live B. melitensis 16M vaccine with low virulence and a high protective effect.


1.1 Mice

All 8-week-old female BALB/c mice were obtained from the Experimental Animal Center of Academy of Military Medical Science (Beijing, China). All experimental procedures and animal care were performed in compliance with institutional animal care regulations.

1.2 Bacterial strains, plasmids, and growth conditions

Brucella melitensis strains 16M, Rev.1 and M5-90 were obtained from the Chinese Center for Disease Prevention and Control (Beijing, China). The Brucella were cultured in TSB or TSA (Sigma, St. Louis, MO, USA). E. coli strain DH5a was grown on LB medium. In lab medium, pGEM-7Zf+ was purchased from Promega (Madison, WI, USA). The media were supplemented with the appropriate antibiotics (100 μg/mL ampicillin for 16M and 50 μg/mL kanamycin for E. coli) when necessary.

1.3 Construction of 16MΔhfq

A pair of primers was designed for hfq DNA fragment amplification for construction of the Δhfq deletion mutant. The designations of the primers were based on 16M. The primer sequences were as follows: P-F, 5′-GCATGCGCATGTGCGTGTTCGTTTAT-3′ and P-R, 5′-GTCGACCAAATATCGGAAAGCGTCAG-3′. The restriction enzyme sites of SphI and SalI (underlined nucleotides in the above sequences) were integrated to the 5′ end of primers P-F and P-R; this pair of primers was designed to amplify hfq. In addition, a 316 bp PCR product was cloned into pMD18-T Simple Vector for sequencing.

For construction of the hfq deletion mutant, two pairs of primers with restriction sites at the 5′ ends were designed for amplification of the upstream (1003 bp) and downstream (1051 bp) arms of the 16M hfq ORF, in which the XhoI, KpnI, KpnI and SacI (underlined) sites were integrated into both PCR fragment ends. The primer sequences were as follows: up-F, 5′-CTCGAGCCTGGCATCAATGGTGATGAG-3′, up-R, 5′-GGTACCTGTTTTTATTCCTTTAATCGATCAGC-3′, dn-F, 5′-GGTACCGTGGCCTATCAGGACTTTACC-3′; and dn-R, 5′-GAGCTCTGCCGACCCAGTAATAGAAAT-3′. The two homologous arms were sequentially cloned into pGEM-7Zf+ to generate suicide plasmid pGEM-7Zf+-hfq. The following pair of primers was then designed for SacB DNA fragment amplification: S-F, 5′-GAGCTCGGGCTGGAAGAAGCAGACCGCTA-3′ (containing a SacI site) and S-R, 5′-GAGCTCGCTTATTGTTAACTGTTAATTGTCC-3′ (containing a SacI site). A 1500 bp fragment was amplified by PCR from Bacillus subtilis. The SacI SacI insert of this plasmid containing the PCR-amplified DNA was subcloned in plasmid pGEM-7Zf+-hfq, generating plasmid pGEM-7Zf+-hfq-SacB. Competent 16M was electroporated with pGEM-7Zf+-hfq-SacB and Brucella transformants selected in the presence of 100 μg/mL ampicillin for the first screening and 5% sucrose for the second screening. The deletion mutant was further confirmed by PCR amplification and RT-PCR sequencing, as described previously [18].

1.4 Complementation of 16MΔhfq

Polymerase chain reaction amplification fragments with primers P-F and P-R and the products were ligated with pMD18-T simple vector (Takara, Beijing, China) to generate pMhfq, which was then transformed into E. coli DH5a to restore hfq activity. The pMhfq plasmids from these transformants were isolated and analysed by agarose gel electrophoresis after restriction digestion. Subsequently, pMhfq plasmids were electroporated into competent 16MΔhfq. The transformants were selected on TSA containing 100 µg/mL ampicillin.

1.5 Evaluation of 16MΔhfq attenuation in RAW 264.7 macrophages

RAW 264.7 macrophages were used to assess 16MΔhfq survival by comparing it with that of complementary strain (16M-hfq), Rev.1, M5-90, and their wild type strain 16M. Briefly, 2 × 105 cells/well were cultured in a 24-well plate for 24 hrs at 37°C and infected with Brucella at 50 MOI. Culture plates were centrifuged for 5 mins at 350g in a Jouan centrifuge at room temperature, and then placed in a 5% CO2 atmosphere at 37 °C. At 1 hr postinfection, the cells were washed twice with medium without antibiotics and incubated with 50 µg/mL of gentamicin (Invitrogen, Carlsbad, CA, USA) for 30 mins to kill any extracellular bacteria. Subsequently, the culture was replaced with Dulbecco's modified Eagle's medium (DMEM) containing 25 µg/mL of gentamicin. At 0, 4, 24, 48 and 72 hrs postinfection, the cells were lysed and the live bacteria enumerated by plating on TSA plates. All assays were performed in triplicate and repeated at least three times.

1.6 Evaluation of 16MΔhfq attenuation in mice

Female BALB/c mice were used to evaluate the survival of 16MΔhfq. Briefly, 8-week-old mice (n = 10 per group) were inoculated i.p. with 1 × 106 CFU of 16MΔhfq, 16M-hfq, Rev.1 or M5-90. The survival or persistence of the bacteria in mice was evaluated by their enumeration in the spleens at different time points after the infection. At 1, 2, 4, 6 and 8 weeks after inoculation, the mice were killed and their spleens removed aseptically, weighed, and homogenized in 1 mL of PBS. The suspensions were diluted in sterile saline and plated on TSA. The plates were incubated at 37°C and the number of CFU counted after three days. The experiments were repeated twice with similar results.

1.7 Protection induced in BALB/c mice by 16MΔhfq

Protective activity was evaluated by comparing the incidence of spleen infection after a standardized virulence challenge of 16M in mice who had received 16MΔhfq (experimental vaccine group); Rev.1, M5-90 (reference vaccine control group) or PBS (unvaccinated control group). Groups of 8-week-old-female BALB/c mice (n = 10 per group) were vaccinated i.p. with 1 × 106 CFU of 16MΔhfq, Rev.1 or M5-90. A control group of 10 unvaccinated mice was injected i.p. with 200 µL of PBS. The mice were challenged i.p. with 1 × 106 CFU per mouse (200 µL) of virulent strain 16M at 8 weeks after vaccination. All mice were killed by cervical dislocation at 2 and 4 weeks after challenge to determine bacterial CFUs in their spleens. A mean value for each spleen count was obtained after logarithmic conversion. Log10 units of protection were obtained by subtracting the mean log10 CFU for the experimental group from the mean log10 CFU for the control group, as previously described [19]. The experiment was repeated twice.

1.8 Evaluation of antibody

Serum samples were obtained from immunized mice 6 weeks after immunization and IgG1 or IgG2a measured by ELISA to determine antibody titers in serum of inoculated mice. Briefly, heat-killed and sonicated B. melitensis 16M whole-cell antigen was used to coat 96-well plates at a concentration of 25 µg protein/well. After overnight incubation at 4°C, the plates were washed once with 100 mL PBST buffer (PBS containing 0.05% Tween-20) and blocked with 200 mL blocking buffer (10% heat-inactivated FBS in PBS, pH 7.4) for 2 hrs at 37°C. Serial dilutions of serum were added to wells in triplicate and incubated for 2hrs at 37°C. After incubation, horseradish peroxidase-conjugated rabbit anti-mouse IgG1 or IgG2a antibodies (SBA, Birmingham, Al, USA) were added, after which the plates were incubated at 37°C for 1 hr. After two washes with PBS, 100 mL per well of 3,3,5,5-tetramethyl benzidine substrate solution was added and incubated at 37°C in darkness for 15 mins. The reaction was terminated by adding 50 µL of H2SO4; absorbance was measured at 450 nm. All assays were performed in triplicate and repeated at least three times.

1.9 Cytokine detection

Six weeks after immunization, BALB/c mice (n = 10 per group) were killed and their spleens removed under aseptic conditions. Single cell suspensions were obtained from the spleens by homogenization. The cells were suspended in RPMI 1640 medium (Gibco BRL, Gaithersburg, MD, USA) supplemented with 2 mM L-glutamine and 10% (v/v) heat-inactivated FBS. Erythrocytes were eliminated with ACK lysis solution (150 mM NH4Cl, 1 mM Na2 − EDTA [pH 7.3]; Gibco). Splenocytes were cultured in 96-well microtiter plates with 4 × 105 cells/well; the cultures were stimulated by adding 25 µg of Rev.1, M5-90 or heat-killed B. melitensis 16M lysate/well, 0.5 µg of ConA (positive control), or medium alone (negative control). The cells were then incubated at 37°C with 5% CO2 for 72 hrs. IFN-γ and IL-4 levels in the supernatants were measured using an ELISA Quantikine Mouse kit (R&D Systems, Minneapolis, MN, USA).

1.10 Expression and purification of recombinant proteins

The Hfq and L7/L12 open reading frames were amplified by PCR from B. melitensis 16M genome. The amplified DNA fragments were then cloned into the pET32a vector (Novagen, Madison, WI, USA) and expressed in DE3 as N-terminally His-tagged fusion proteins. Expression of the recombinant protein was confirmed by SDS − PAGE. The Hfq protein was purified by affinity chromatography with Ni2-conjugated sepharose.

1.11 Hfq indirect enzyme-linked immunosorbent assay

Serum samples were obtained from Brucella-infected mice. Antibody responses to the purified recombinant Hfq protein were performed via Hfq-based indirect ELISA, as described previously [20].

1.12 Western blot analysis

For analysis of Hfq and L7/L12 expression in 16M, 16MΔhfq or 16M-hfq, 1 mL of each culture was pelleted and resuspended in SDS sample buffer at an optical density at 600 nm (OD600) of 1. The samples were boiled for 5 0mins and a 10 µL sample loaded on 12% SDS − PAGE gel. The protein was electrotransferred to a nitrocellulose membrane with a Mini Trans-Blot Cell (Bio-Rad, Hercules, CA, USA) at 200 mA for 1 hr. Unbound sites on the membrane were blocked overnight with TBS with the addition of 3% fish gelatin (Sigma, St Louis, MO, USA); the membrane was blocked overnight with 10% dry milk in TBST buffer (100 mM Tris − HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.2). After they had been blocked, the membranes were washed three times with TBST buffer and incubated with Brucella-vaccinated serum (diluted 1:300). The bands were incubated with rabbit anti-mouse IgG (peroxidase conjugated) for 1 hr at room temperature after being washed three times. Bound conjugate was visualized with a DAB substrate kit (Beyotime Biotech, Jiangsu, China) after further washing.


2.1 16MΔhfq is attenuated in macrophages

Intracellular survival and replication within macrophages is an indispensable characteristic of Brucella pathogenesis. RAW 264.7 macrophages were infected with 16MΔhfq and their ability to replicate intracellularly compared with other B. melitensis strains to evaluate whether loss of 16MΔhfq has any influence on intracellular growth. At 0 hr, no differences in numbers of bacteria were observed among these five different strains (Fig. 1). At 4 hrs after infection, only slight differences in the amount of replication in macrophages were evident, indicating that the bacteria's ability to invade RAW264.7 macrophages was similar. In addition, hfq did not affect entry of Brucella into the macrophages (Fig. 1). However, 24 hrs post-infection, 2.07-log, 1.66-log, 1.04-log and 1.13-log decreases (P < 0.001) in numbers of 16MΔhfq inside the macrophages compared to 16M, 16M-hfq, Rev.1 or M5-90, respectively, were observed; thus, there were significantly fewer 16MΔhfq than the other strains assessed. At 72 hrs post-infection, the decreases inside the macrophages were even greater at 3.93-log, 3.69-log, 1.96-lo and 2.42-log, respectively (P < 0.001; Fig. 1). These findings show that the ability of 16MΔhfq to replicate in RAW 264.7 macrophages is limited compared with virulent 16M or the vaccine Rev.1 and M5-90 strains, indicating that hfq is involved in chronic Brucella infection. In addition, these findings indicate that 16MΔhfq was attenuated in RAW 264.7 macrophages.

Figure 1.

Intracellular replication of 16MΔhfq within RAW264.7 macrophages. RAW 264.7 macrophages were infected at an MOI of 50 with 16M, 16M-hfq, 16MΔhfq, Rev.1 and M5-90, as described in Materials and Methods. The level of initial infection was the same for all B. melitensis strains. 16MΔhfq failed to achieve the level of colonization reached by the wild type and other B. melitensis strains. At 0, 4, 24, 48 and 72 hrs postinfection, infected macrophages were lysed and supernatants diluted for CFU enumeration. *, P < 0.001, mutant versus other B. melitensis strains.

2.2 16MΔhfq is attenuated in BALB/c mice

Compared with 16M or 16M-hfq, 16MΔhfq was significantly attenuated (P < 0.05) at all observation times. Significantly, splenic CFU could not be detected in 16MΔhfq-dosed mice by week 8 (Fig. 2a). In addition, as shown by comparative splenic weights, at most time points there was less splenic inflammation in the animals that had received 16MΔhfq than in those that had received 16M, 16M-hfq, Rev.1 and M5-90 (Fig. 2b). These results show that the 16MΔhfq strain was greatly attenuated in mice.

Figure 2.

Clearance of 16MΔhfq after infection. Ten BALB/c mice/group were infected with 1 × 106 CFU/mouse of wild-type 16M, 16M-hfq, 16MΔhfq, Rev.1 and M5-90. At 1, 2, 4, 6 and 8 weeks postinfection, mouse spleens were harvested and individual spleens assessed for (a) colonization and (b) weight. Values are expressed as means ± SD, and the significance of differences in splenic weight and colonization between 16MΔhfq and 16M or 16M–hfq were determined by Student's unpaired test. *, P < 0.05.

2.3 16MΔhfq induces immune protection against challenge with virulent 16M

Mice were vaccinated i.p. with 1 × 106 CFU of 16MΔhfq, PBS, Rev.1 or M5-90 to determine whether 16MΔhfq confers protection against wild-type infection. Eight weeks after vaccination, the animals were challenged with 1 × 106 CFU of strain 16M. Mice immunized with 16MΔhfq had significantly fewer Brucella in their spleens than did nonimmunized ones 2 (1.64 log units) and 4 weeks (2.06 log units) after challenge (P < 0.05; Table 1). The levels of protection at 4 weeks post-infection. were higher than at 2 weeks post-infection for the three strains. In addition, similar log units of protection were observed at 2 and 4 weeks post-challenge in BALB/c mice immunized with 16MΔhfq and those immunized with the commercial vaccine strain Rev.1 (P < 0.05). 16MΔhfq conferred stronger protection than did M5-90. Rev.1 and M5-90 also induced significant protection. These results indicate that the protection provided by 16MΔhfq is similar to that provided by the Rev.1 vaccine strain.

Table 1. Protection conferred by 16MΔhfq against 16M
Vaccinelog10CFU Brucella per spleen (mean ± SD) at 2 weeks post-challengeLog10 units of protection at 4 weeks post-challenge
  • *, P ≤ 0.05 compared with the PBS control group.
  • , Log10 units of protection = average of the log10 CFU in spleens of control unvaccinated mice minus the average of log10 CFU in spleens of vaccinated mice.
PBS7.06 ± 0.176.27 ± 0.27  
16hfq5.42 ± 0.244.21 ± 0.071.64*2.06*
Rev.15.13 ± 0.414.13 ± 0.191.93*2.14*
M5-905.65 ± 0.354.78 ± 0.111.41*1.49*

2.4 16MΔhfq immunization induces both humoral immune response and cytokine responses

Serum from mice inoculated with 16MΔhfq, Rev.1 or M5-90 were collected from immunized mice at selected intervals post-immunization to monitor amounts of IgG1 and IgG2a subclasses by ELISA. Figure 3 shows that a humoral immune response was detected at 6 weeks post-vaccination in the serum of mice immunized with 16MΔhfq vaccines. Mice inoculated with the mutant strain had significantly greater amounts of serum IgG1 and IgG2a than did control mice (P < 0.001). Immunization with 16MΔhfq and M5-90 induced a stronger IgG1 than IgG2a response, whereas Rev.1 induced similar amounts of IgG1 and IgG2a.

Figure 3.

Humoral immune responses in sera from mice immunized with 16MΔhfq. BALB/c mice were inoculated i.p. with 1 × 106 CFU of either 16MΔhfq, M5-90 or Rev.1. The control groups received PBS. Six weeks post-immunization, amounts of serum IgG1 and IgG2a were determined by ELISA. The values are expressed as means ± SD (n = 10) of the absorbance at 450 nm (OD450). *, P < 0.001 16MΔhfq versus PBS.

To determine the types of cytokines expressed by spleen cells from vaccinated mice when incubated in the presence of the homologous immunogen ConA and heat-inactivated 16M, the concentrations of IFN-γ (Fig. 4a) and IL-4 (Fig. 4b) in mouse splenocytes were assessed 6 weeks after immunization. On re-stimulation, cells from mice vaccinated with 16MΔhfq, Rev.1 and M5-90 produced higher amounts of IFN-γ and IL-4 than did cells from PBS mice (ConA was used as positive control; P < 0.05). However, there was no statistically significant difference in production of IFN-γ and IL-4 between 16MΔhfq, Rev.1 and M5-90. These results indicate that immunization with 16MΔhfq elicits classical Th1 and Th2 responses.

Figure 4.

IFN-γ and IL-4 production by spleen cells of BALB/c mice vaccinated with 16MΔhfq, M5-90 or Rev.1. BALB/c mice were inoculated i.p. with 1 × 106 CFU of either 16MΔhfq, M5-90 or Rev.1. Six weeks post-vaccination, splenocytes were recovered and stimulated with either heat-inactivated B. melitensis 16M, ConA or RPMI 1640. Splenocyte culture supernatants were harvested after 72 hrs of culture. (a) IFN-γ and (b) IL-4 concentrations in the supernatant were measured by ELISA. Significant differences from the same stimulus in PBS-immunized mice are indicated as follows: *, P < 0.001.

2.5 Differentiation of 16MΔhfq immunization from infection using the protein Hfq as test antigen

Sera from mice immunized with 16MΔhfq and 16M were collected to determine whether the protein Hfq could be used as a diagnostic antigen. Western blotting was performed using immunogenic L7/L12 protein as positive control to determine whether antibodies against Hfq and L7/L12 were induced in these sera. A L7/L12 reaction band was observed in the sera of 16M-infected mice and 16MΔhfq-vaccinated ones. However, a single reaction band of the Hfq protein was observed in the sera of 16M vaccinated mice; this was not present in 16MΔhfq-vaccinated ones, indicating that vaccination with 16MΔhfq does not induce antibodies against Hfq (Fig. 5a). The result of indirect ELISA was similar to that of western blotting (Fig. 5b). Thus, according to Hfq-iELISA using HIS-Hfq as solid-phase antigen, animals infected with 16M were positive, whereas those vaccinated with 16MΔhfq were negative. These findings indicate that Hfq-iELISA with Hfq protein can be used to differentiate immunization from natural infection after confirmation of brucellosis by LPS-based serological tests.

Figure 5.

Humoral immune responses to Hfq and L7/L12 were assessed in these sera by (a) western blotting and (b) Hfq-iELISA. Humoral immune responses against the Hfq protein were not detected in sera from 16MΔhfq immunized mice.


The development of an effective vaccine against brucellosis has been a challenge for scientists for several years. Most of the present licensed vaccines have various limitations, such as residual virulence, induction of splenomegaly and interference with serodiagnosis [21-23]. One of the factors hindering development of new Brucella vaccines is limited knowledge about the virulence factors. The recent deciphering of the complete genome sequence of B. melitensis will help identify the genes involved in virulence and in avirulent mutant candidates used as vaccines [9, 24]. Serological interference with classical vaccines is also a significant problem. Therefore, the ideal vaccine must be protective and should also carry a genetic marker that is neither antibiotic-resistant nor virulent for the host and gene involved [25]. 16MΔhfq was constructed and its virulence and protection efficacy assessed in cells and mice to investigate its ability to provide effective protection.

The deletion mutant of hfq and its complementary strain were constructed to confirm that the reduced ability of the mutant to survive is directly related to the deleted gene hfq. 16MΔhfq and 16M-hfq were confirmed by PCR and transcription analysis (data not shown). The present findings show that 16MΔhfq had impaired survival in RAW 264.7 macrophages and BALB/c mice and was cleared faster than Rev.1 or M5-90, being completely cleared within 8 weeks. The lack of splenomegaly in inoculated mice indicates 16MΔhfq is safer than the mutant in that it is significantly less virulent and induces less inflammatory response. This finding is consistent with previous results and substantiates that hfq is involved in the virulence of Brucella. Thus, wild-type and complemented strains had similar intracellular replication, whereas 16MΔhfq failed to replicate in macrophages and mouse models.

Based on assessment of the number of spleen CFUs, vaccination with 16MΔhfq offers significant protection against challenge inoculation. An ideal live vaccine strain combines efficient immunogenicity with minimal reactogenicity. Therefore, a protection experiment was performed in BALB/c mice; this showed that the three strains were still conferring significant protection at 4 weeks post-challenge and that the protection provided by 16MΔhfq was similar to that of Rev.1 and greater than that of M5-90. This study showed that 16MΔhfq can elicit good protective efficacy against subsequent challenge with the wild-type strain.

Cytokine and humoral immune responses were evaluated to ascertain the protection conferred by 16MΔhfq. IFN-γ that is induced by cytokines can gather at infection sites to enhance bacterial killing. Antigen-specific production of Th1 cytokines (IFN-γ) and Th2 cytokines (IL-4) was detected in spleen cells from 16MΔhfq-immunized mice. The Th1 immune responses characterised by IFN-γ production are associated with protective immunity to Brucella; these responses are best stimulated by live vaccines [26]. Previous studies have shown that the cytokine IFN-γ is critical for macrophage bactericidal activity [27]. IL-4, a Th2 cytokine, is considered an anti-inflammatory molecule. Thus, detection of IFN-γ and IL-4 indicates mixed Th1/Th2 responses. The present findings show that 16MΔhfq induces less IFN-γ and more IL-4 than do Rev.1 and M5-90. The isotype profile of an antibody response is a reflection of the T-helper-cell types [28, 29]. IgG2a antibody formation depends on IFN-γ secreted by Th1 cells, whereas IgG1 production depends on IL-4 secreted by Th2 cells [30]. Our findings show that 16MΔhfq induces high concentrations of anti-Brucella IgG1 and IgG2a, which may be associated with high concentrations of IFN-γ and IL-4, indicating that the mutant can induce a strong humoral immunological response. Our vaccination strategy should lead to mixed Th1/Th2 responses.

Current serological diagnostic tests, such as the RBPT, SAT, CFT, iELISA or cELISA, use hot saline extract and smooth LPS antigens. The LPS of smooth Brucella species is a much stronger antigen than other antigenic molecules [31]. However, LPS-based serological tests have difficulty differentiating between vaccinated animals and infected ones. Thus, the possibility of using Hfq protein as a diagnostic antigen was evaluated. It was found that a humoral immune response to Hfq can be detected in infected serum, but not in 16MΔhfq-immunized serum, indicating that immunization and vaccination sera can be differentiated by using Hfq as a diagnostic antigen. Thus, brucellosis is first confirmed by LPS-based serological tests, whereas 16MΔhfq vaccination is detected by Hfq-iELISA. Using Hfq-iELISA based on a recombinant HIS-fusion protein HIS-Hfq as the solid-phase antigen, it was found that mice infected with 16M were positive whereas animals vaccinated with 16MΔhfq were negative. Therefore, 16MΔhfq makes it possible to differentiate between infected and vaccinated animals.

In conclusion, our results show that 16MΔhfq may be another suitable live vaccine candidate for B. melitensis because it has low virulence in RAW 264.7 macrophages and BALB/c mice, while providing protection similar to that of Rev.1 and M5-90 vaccine strains. Postvaccination humoral responses indicated that the vaccine candidate can elicit an anti-Brucella-specific IgG response, providing an ideal diagnostic Hfq antigen for differentiation of immunization from infection by Hfq-iELISA. Thus, although the three vaccines appear to offer similar protection after challenge, the overall results indicate that 16MΔhfq offers better protection than M5-90 against virulent 16M in this study. It should therefore be considered as a new potential vaccine candidate against Brucellosis. In future studies, comprehensive protection experiments must be conducted via different routes to determine whether measurable immune responses in systemic compartments confer detectable protection against Brucella infection. In addition, further testing in livestock will determine whether the 16MΔhfq is a promising live vaccine candidate.


This work was supported by grants from the National Basic Research Program of China (973 Program; 2010CB530203), National Twelfth Five-Year Plan for Science and Technology Support Program (2013BAI05B05) and the National Natural Science Foundation of China (31201863).


None of the authors of this paper have any financial or personal relationship with other people or organizations that could inappropriately influence or bias the content of the paper.