The currently available commercial live Brucella vaccines have serious intrinsic drawbacks, which is why vaccination is restricted in many countries prone to brucellosis. This situation has motivated many researchers to develop new generations of vaccines which do not possess such drawbacks, but which can induce the same protection efficacy as commercial vaccines.
In our opinion, of the strategies currently being proposed, live vector vaccines are the most likely replacement for existing commercial vaccines against bovine brucellosis, given the long-term experience of their successful use in veterinary practice .
In this study, we first developed recombinant influenza A viruses of the subtypes Н5N1 and H1N1 expressing Brucella ribosomal protein L7/L12 or Omp16, and then demonstrated that these vaccines could be used as new candidates for a brucellosis vaccine. The choice of the Brucella proteins L7/L12 and Omp16 as antigenic determinants was based on their immunodominance in eliciting Th1 CD4+ and CD8+ T-cell immune responses [8–10, 13, 14, 16, 20, 22]. In addition, it was previously shown that these recombinant proteins, both when pure, combined with adjuvants, or expressed by DNA vaccines, provided mice with good protection when challenged with virulent B. abortus 544 [8, 10, 13, 14, 20].
A large body of data [31–34] has confirmed the ability of influenza viruses to infect cattle and elicit a serological reaction and, in some cases clinical disease, which provided the basis for choosing influenza A viruses as the vaccine vector in this study. Thus, the attenuated influenza A viruses selected as the vector should be able to infect cattle and express the recombinant Brucella proteins. The potential of the influenza A NS vector was confirmed in our previous studies of the development of a tuberculosis vaccine .
On the basis of previous study , and in order to achieve maximum expression of the Brucella proteins in vivo and elicit an increased T-cell immune response, the laboratory animals were immunized using a double vaccination schedule with viral constructs of the Н5N1 subtype (prime vaccination) and H1N1 subtype (booster vaccination). This immunization strategy effectively overcomes the immune background elicited against the viral vector during primary vaccination.
Since influenza viruses are primarily transmitted within droplets, and reproduce in the mucous membranes of the respiratory tract, the i.n. route was chosen for immunizing the animals, and the c. and s.c. routes were used for comparison. The mucosal routes of administration (i.n. and c.) were chosen as the mucous membranes are the main gateway for brucellosis infection, and these routes have been shown to provide immunized animals with a high level of protection against the virulent B. abortus strain [13, 37].
The first series of experiments established that despite the truncated nature of the NS1 gene, the viral constructs Flu-NS1-124-L7/L12-H5N1, Flu-NS1-124-Omp16-H5N1, Flu-NS1-124-L7/L12-H1N1, and Flu-NS1-124-Omp16-H1N1 had good reproductive properties in CE over five consecutive passages and retained their Brucella inserts.
In our previous studies, it was shown that as the size of the NS1 gene decreased in viral vectors, the degree of attenuation of the influenza A viruses increased ; however, it is well known that attenuation of influenza viruses may be dependent on the properties of the foreign insert in the C-terminal part of the truncated NS1 protein . Therefore, we considered it necessary to study the safety or degree of attenuation of the constructed recombinant influenza A viruses in mice, which provide the most sensitive model for testing the safety of influenza vectors. Both the mono- and bivalent vaccine formulations containing Brucella inserts (L7/L12 or Omp16) in the NS1 gene were safe in mice when administered by the i.n., c. or s.c. routes. No deaths, loss of body weight or pathomorphological changes occurred in the mice over the entire observation period, which provided evidence for the attenuation of the influenza A viruses.
Moreover it has to be noted that all guinea pigs vaccinated with mono and bivalent vaccine formulation by i.n. method did not shed the vaccine viruses through their upper respiratory tract the entire week after the prime and booster vaccination. These results proved the replication-deficient properties of the virus vectors and confirm no danger of viral transmission from vaccinated to non-vaccinated animals or people. Interestingly, one vaccine vector is based on the prepandemic flu A/H5N1 delNS1 vaccine. It was shown previously that this vaccine is completely safe and immunogenic when tested in a variety of laboratory models (chickens, ferrets and rhesus macaques)  and humans . We can assume that this vaccine in the future can be used not only for cattle but also for humans.
Furthermore, the mono- and bivalent vaccine formulations of the subtypes Н5N1 and H1N1 elicited antigen-specific humoral and T-cell immune responses after prime-boost immunization via all of the tested routes of administration. The greatest antibody response (by ELISA) was obtained with the viral constructs expressing the Brucella protein L7/L12; and the greatest T-cell immune response by ELISPOT assay was obtained with viral constructs expressing the Omp16 protein (c. immunization), not inferior (P > 0.05) to commercial vaccine B. abortus 19. It should be noted that, as the humoral and cellular immune responses indicate, the bivalent vaccine formulation were in no way inferior (P > 0.05) to their monovalent variants, despite the lower (by half) dose injected; the bivalent vaccine formulation even somewhat surpassed their monovalent variants. We also demonstrated that single or combined injection of mice with viral constructs expressing two different Brucella proteins did not lead to interference between the constructs, as the mice immunized with bivalent vaccine formulation formed a humoral and cellular immune responses to both (L7/L12 or Omp16) protein of B. abortus. The optimal T-cell immune responses were achieved by the c. immunization route, despite the fact that influenza viruses primarily reproduce in respiratory tract organs. In our opinion, this is mainly due to the ability of influenza viruses to reproduce not only in the conjunctival mucous membrane, but also in the cells of the cornea thus priming the conjunctiva-associated lymphoid tissue (CALT) and eye-associated lymphoid tissue (EALT). CALT can detect antigens from the ocular surface, present the antigens and generate protective effector cells; together, these properties signify the presence of a mucosal immune system at the conjunctiva [42–44]. Theoretically, the administration of antigens into the conjunctival sac would additionally drain to the nasal-associated lymphoid tissue (NALT). It was shown previously that conjunctival or intraocular infection with influenza viruses stimulates the local as well as systemic immune response . Another explanation for the higher cellular immunity obtained using viral constructs expressing Omp16 is that it Omp16 is able to act as an adjuvant and activate dendritic cells and macrophages in vitro, according to Pasquevich et al. .
Data in mice have found a logical reflection in experiments with guinea pigs. Guinea pigs were chosen as model animals for studying the protective efficacy of the viral constructs in this study due to their higher resistance to influenza infection than mice . In this case, the use of a more resistant animal model was a key condition for studying protective efficacy, since the vaccine is ultimately intended for cattle. Furthermore, according to Silva et al., guinea pigs are the most susceptible model for evaluating protective efficacy  for the commercial live brucellosis vaccine produced from strain B. abortus 19, This strain was used as a reference in the present study. The protective efficacy of the vaccines was evaluated by assessing both the bacterial load of the virulent strain B. abortus 544 in spleens of vaccinated and unvaccinated animals, and by other parameters such as the effectiveness of vaccination and I.I. In our opinion, when taken together, these parameters provide a more complete and objective characterization of the protective efficacy of vaccines.
In this study, guinea pigs were challenged with B. abortus 544 at a dose of 5 × 10 CFU/animal, while in other similar studies, mice were used as the model animals and higher doses of 4 × 104 - 5 × 105 CFU were administered [6, 13, 20]. Our previous studies (unpublished data) showed that s.c. administration of five B. abortus 544 microbial cells caused generalized infection (minimum infective dose) in guinea pigs. On this basis, a dose of ten times the minimum infective dose, i.e., 5 × 10 CFU/animal, was used to evaluate the protective efficacy of the vaccines. This dose enabled a more objective comparative evaluation of the protective efficacy of the vaccines.
In terms of the effectiveness of vaccination, I.I., and isolation rate, the highest levels of protection were achieved in the guinea pigs which were c. vaccinated with monovalent viral constructs expressing the Omp16 protein and bivalent vaccine formulation expressing the L7/L12 and Omp16 proteins. It is noteworthy that the groups revealing the highest protective efficacy also induced also the best cell-mediated immunity. These experiments demonstrated that the monovalent viral constructs expressing the Omp16 protein and bivalent vaccine formulation expressing the L7/L12 and Omp16 proteins, when administered in prime-boost c. immunization mode, were comparable in terms vaccine effectiveness and protective efficacy to the commercial live vaccine produced from strain B. abortus 19 in guinea pigs. It should be noted that the results obtained in the present study was successfully used in evaluating the immunogenicity and efficacy of our vector vaccine in cattle. It has been established that the administration of the vector vaccine via the c. method of vaccination promoted formation of IgG antibodies (with a predominance of antibodies of isotype IgG2a) in cattle against Brucella L7/L12 and Omp16 proteins in ELISA. Moreover, these vaccines in cattle induced a strong antigen-specific T-cell immune response and provided a high level of protection efficacy comparable to those of the commercial B. abortus S19 vaccine .