Evaluation of human trophoblasts and ovine testis cell lines for the study of the intracellular pathogen Brucella ovis.

Since pathogenic Brucella survive and replicate inside phagocytes, cellular models of infection constitute important tools in brucellosis research. We describe the behavior of B. ovis PA (which causes a type of ovine brucellosis mainly affecting the male reproductive tract) and representative attenuated mutants in two commercially available cell lines of non-professional phagocytes related to Brucella tissue preference: OA3.Ts ovine testis cells and JEG-3 human trophoblasts. In comparison with J774.A1 macrophages and HeLa cells, intracellular bacteria were enumerated at several post-infection time points and visualized by confocal microscopy. Replication of B. ovis in OA3.Ts and JEG-3 cells was equivalent to that observed in J774.A1 macrophages-despite the more efficient internalization in the latter-and better than in HeLa cells. Multiplication and/or survival in all phagocytes was dependent on virB2 and vjbR but independent of cgs, despite the attenuation in mice of the Δcgs mutant. However, Omp25c was required for B. ovis internalization only in HeLa cells, and removal of Omp31 increased bacterial internalization in human HeLa and JEG-3 cells. The results presented here demonstrate variability in the interaction of B. ovis with different host cells and provide advantageous models of non-professional phagocytes to study the intracellular behavior of B. ovis.

Characterization of Cell Envelope Multiple Mutants of Brucella ovis and Assessment in Mice of Their Vaccine Potential.

Brucella ovis is a non-zoonotic Brucella species lacking specific vaccine. It presents a narrow host range, a unique biology relative to other Brucella species, and important distinct surface properties. To increase our knowledge on its peculiar surface and virulence features, and seeking to develop a specific vaccine, multiple mutants for nine relevant cell-envelope-related genes were investigated. Mutants lacking Omp10 plus Omp19 could not be obtained, suggesting that at least one of these lipoproteins is required for viability. A similar result was obtained for the double deletion of omp31 and omp25 that encode two major surface proteins. Conversely, the absence of major Omp25c (proved essential for internalization in HeLa cells) together with Omp25 or Omp31 was tolerated by the bacterium. Although showing important in vitro and in vivo defects, the Δomp10Δomp31Δomp25c mutant was obtained, demonstrating that B. ovis PA survives to the simultaneous absence of Omp10 and four out seven proteins of the Omp25/Omp31 family (i.e., Omp31, Omp25c, Omp25b, and Omp31b, the two latter naturally absent in B. ovis). Three multiple mutants were selected for a detailed analysis of virulence in the mouse model. The Δomp31Δcgs and Δomp10Δomp31Δomp25c mutants were highly attenuated when inoculated at 106 colony forming units/mouse but they established a persistent infection when the infection dose was increased 100-fold. The Δomp10ΔugpBΔomp31 mutant showed a similar behavior until week 3 post-infection but was then totally cleared from spleen. Accordingly, it was retained as vaccine candidate for mice protection assays. When compared to classical B. melitensis Rev1 heterologous vaccine, the triple mutant induced limited splenomegaly, a significantly higher antibody response against whole B. ovis PA cells, an equivalent memory cellular response and, according to spleen colonization measurements, better protection against a challenge with virulent B. ovis PA. Therefore, it would be a good candidate to be evaluated in the natural host as a specific vaccine against B. ovis that would avoid the drawbacks of B. melitensis Rev1. In addition, the lack in this attenuated strain of Omp31, recognized as a highly immunogenic protein during B. ovis infection, would favor the differentiation between infected and vaccinated animals using Omp31 as diagnostic target.