Protective and therapeutic efficacy study of divalent fusion protein rL7/L12-Omp25 against B. abortus 544 in presence of IFNγ

Brucella as intracellular pathogen requires a coordinate interaction between Th1 subset of gamma interferon-secreting CD4 T cells and CD8 T cells for optimal protective immunity. It was previously recognized that L7/L12 as T cell-reactive antigen from the pathogen. On other hand, Omp25 was found as another antigen to provide protection against the Brucella infection by eliciting both Th1 and Th2 type of immune responses in mice. Here, we analyzed the prophylactic and therapeutic efficacy of a divalent fusion protein (rL7/L12-Omp25) comprising these two promising immunogens of Brucella in the presence of murine IFN-gamma in mice against B. abortus 544 challenge. rIFN-gamma with rL7/L12-Omp25 resulted in superior immune response when compared to the animal vaccine strain B. abortus S19. The vaccine candidate caused dominance of IgG1 over IgG2a and upregulated cytokine secretion (IFN-gamma, TNF-α, and IL-10) among immunized mice. Moreover, the antigen in combination with murine IFN-gamma elicited stronger cell-mediated immune response among the immunized animals when compared to standard vaccine (S19). The registered log protection unit among challenged mice with B. abortus 544 pathogen was 2.16, p = 0.0001 when rL7/L12-Omp25 was administered alone and 2.4, p = 0.0001 when it was administered along with rIFN-gamma. However, the molecule upon administration with murine IFN-gamma imparted very minimal or no therapeutic effect against brucellosis. To conclude, our study demonstrates the potential of rL7/L12-Omp25 as an immunogen of prospective and efficient prophylaxis as it is capable of eliciting both cell-mediated and humoral immune responses against brucellosis.

     

Lipocalin 2 (Lcn2) interferes with iron uptake by Brucella abortus and dampens immunoregulation during infection of RAW 264.7 macrophages

Lipocalin 2 (Lcn2) is an important innate immunity component against bacterial pathogens. In this study, we report that Lcn2 is induced by Brucella (B.) abortus infection and significantly contributes to the restriction of intracellular survival of Brucella in macrophages. We found that Lcn2 prevented iron uptake by B. abortus through two distinct mechanisms. First, Lcn2 is secreted to capture bacterial siderophore(s) and abrogate iron import by Brucella. Second, Lcn2 decreases the intracellular iron levels during Brucella infection, which probably deprives the invading Brucella of the iron source needed for growth. Suppression of Lcn2 signalling resulted in a marked induction of anti‐inflammatory cytokine, interleukin 10, which was shown to play a major role in Lcn2‐induced antibrucella immunity. Similarly, interleukin 6 was also found to be increased when Lcn2 signalling is abrogated; however, this induction was thought to be an alternative pathway that rescues the cell from infection when the effective Lnc2 pathway is repressed. Furthermore, Lcn2 deficiency also caused a marked decrease in brucellacidal effectors, such as reactive oxygen species and nitric oxide but not the phagolysosome fusion. Taken together, our results indicate that Lcn2 is required for the efficient restriction of intracellular B. abortus growth that is through limiting iron acquisition and shifting cells to pro‐inflammatory brucellacidal activity in murine macrophages.     

Exosomes released by Brucella-infected macrophages inhibit the intracellular survival of Brucella by promoting the polarization of M1 macrophages

Exosomes, membrane vesicles released extracellularly from cells, contain nucleic acids, proteins, lipids and other components, allowing the transfer of material information between cells. Recent studies reported the role of exosomes in pathogenic microbial infection and host immune mechanisms. Brucella-invasive bodies can survive in host cells for a long time and cause chronic infection, which causes tissue damage. Whether exosomes are involved in host anti-Brucella congenital immune responses has not been reported. Here, we extracted and identified exosomes secreted by Brucella melitensis M5 (Exo-M5)-infected macrophages, and performed in vivo and in vitro studies to examine the effects of exosomes carrying antigen on the polarization of macrophages and immune activation. Exo-M5 promoted the polarization of M1 macrophages, which induced the significant secretion of M1 cytokines (tumour necrosis factor-α and interferon-γ) through NF-κB signalling pathways and inhibited the secretion of M2 cytokines (IL-10), thereby inhibiting the intracellular survival of Brucella. Exo-M5 activated innate immunity and promoted the release of IgG2a antibodies that protected mice from Brucella infection and reduced the parasitaemia of Brucella in the spleen. Furthermore, Exo-M5 contained Brucella antigen components, including Omp31 and OmpA. These results demonstrated that exosomes have an important role in immune responses against Brucella, which might help elucidate the mechanisms of host immunity against Brucella infection and aid the search for Brucella biomarkers and the development of new vaccine candidates.     

Maintenance of brucellosis in Yellowstone bison: linking seasonal food resources,host–pathogen interaction,and life‐history trade‐offs

The seasonal availability of food resources is an important factor shaping the life‐history strategies of organisms. During times of nutritional restriction, physiological trade‐offs can induce periods of immune suppression, thereby increasing susceptibility to infectious disease. Our goal was to provide a conceptual framework describing how the endemic level bovine brucellosis (Brucella abortus) may be maintained in Yellowstone bison based on the seasonality of food resources and the life‐history strategies of the host and pathogen. Our analysis was based on active B. abortus infection (measured via bacterial culture), nutritional indicators (measured as metabolites and hormones in plasma), and carcass measurements of 402 slaughtered bison. Data from Yellowstone bison were used to investigate (1) whether seasonal changes in diet quality affect nutritional condition and coincide with the reproductive needs of female bison; (2) whether active B. abortus infection and infection intensities vary with host nutrition and nutritional condition; and (3) the evidence for seasonal changes in immune responses, which may offer protection against B. abortus, in relation to nutritional condition. Female bison experienced a decline in nutritional condition during winter as reproductive demands of late gestation increased while forage quality and availability declined. Active B. abortus infection was negatively associated with bison age and nutritional condition, with the intensity of infection negatively associated with indicators of nutrition (e.g., dietary protein and energy) and body weight. Data suggest that protective cell‐mediated immune responses may be reduced during the B. abortus transmission period, which coincides with nutritional insufficiencies and elevated reproductive demands during spring. Our results illustrate how seasonal food restriction can drive physiological trade‐offs that suppress immune function and create infection and transmission opportunities for pathogens.     

AIM2 senses Brucella abortus DNA in dendritic cells to induce IL-1β secretion,pyroptosis and resistance to bacterial infection in mice

Absent in melanoma 2 (AIM2) is a sensor of cytosolic dsDNA and is responsible for the activation of inflammatory and host immune responses to DNA viruses and intracellular bacteria. AIM2 is a member of the hematopoietic interferon-inducible nuclear proteins with a 200 amino-acid repeat (HIN200) family, containing a pyrin domain (PYD) at the N-terminus. Several studies have demonstrated that AIM2 is responsible for host defense against intracellular bacteria such as Francisella tularensis, Listeria monocytogenes and Mycobacerium tuberculosis. However, the role of AIM2 in host defenses against Brucella is poorly understood. In this study, we have shown that AIM2 senses Brucella DNA in dendritic cells to induce pyroptosis and regulates type I IFN. Confocal microscopy of infected cells revealed co-localization between Brucella DNA and endogenous AIM2. Dendritic cells from AIM2 KO mice infected with B. abortus showed impaired secretion of IL-1β as well as compromised caspase-1 cleavage. AIM2 KO mice displayed increased susceptibility to B. abortus infection in comparison to wild-type mice, and this susceptibility was associated with defective IL-1β production together with reduced IFN-γ responses. In summary, the increased bacterial burden observed in vivo in AIM2 KO animals confirmed that AIM2 is essential for an effective innate immune response against Brucella infection.     

Serologic evidence of Brucella infection in pinnipeds along the coast of Hokkaido,the northernmost main island of Japan

Brucella infection in Hokkaido was serologically surveyed in four species of pinnipeds inhabiting Cape Erimo during 2008–2013 and the Shiretoko Peninsula in 1999 by ELISA using Brucella abortus and B. canis as antigens. Anti‐Brucella positive sera showed higher absorbance to B. abortus than B. canis in almost all samples. Anti‐B. abortus antibodies were detected in serum samples from 24% (n = 55) of Western Pacific harbor seals (Phoca vitulina stejnegeri) in Cape Erimo and from 66% (n = 41) of spotted seals (P. largha), 15% (n = 20) of ribbon seals (Histriophoca fasciata) and 18% (n = 17) of Western Steller's sea lions (Eumetopias jubatus jubatus) in the Shiretoko Peninsula. Anti‐Brucella antibodies were detected at higher absorbance in 1‐ to 4‐year‐old harbor seals than in the pups and mature animals, suggesting either that Brucella infection mainly occurs after weaning or that it is maternally transmitted to pups with premature or suppressed immunity. Anti‐Brucella antibodies were detected in both immature and mature spotted seals and ribbon seals, with higher absorbance in the former. The antibodies were detected only in mature Western Steller's sea lions. Western blot analysis of the serum samples showed some differences in band appearances, namely discrete versus smeary, and in the number of bands, indicating that multiple different Brucella may be prevalent in pinnipeds in Hokkaido. Alternatively, the Brucella of pinnipeds may have some intra‐species diversity.     

Immunization with a combination of recombinant Brucella abortus proteins induces T helper immune response and confers protection against wild-type challenge in BALB/c mice

Protective efficiency of a combination of four recombinant Brucella abortus (B. abortus) proteins, namely, ribosomal protein L7/L12, outer membrane protein (OMP) 22, OMP25 and OMP31, was evaluated as a combined subunit vaccine (CSV) against B. abortus infection in RAW 264.7 cell line and murine model. Four proteins were cloned, expressed and purified, and their immunocompetence was analysed. BALB/c mice were immunized subcutaneously with single subunit vaccines (SSVs) or CSV. Cellular and humoral immune responses were determined by ELISA. Results of immunoreactivity showed that these four recombinant proteins reacted with Brucella-positive serum individually but not with Brucella-negative serum. A massive production of IFN-γ and IL-2 but low degree of IL-10 was observed in mice immunized with SSVs or CSV. In addition, the titres of IgG2a were heightened compared with IgG1 in SSV- or CSV-immunized mice, which indicated that SSVs and CSV induced a typical T-helper-1-dominated immune response in vivo. Further investigation of the CSV showed a superior protective effect in mice against brucellosis. The protection level induced by CSV was significantly higher than that induced by SSVs, which was not significantly different compared with a group immunized with RB51. Collectively, these antigens of Brucella could be potential candidates to develop subunit vaccines, and the CSV used in this study could be a potential candidate therapy for the prevention of brucellosis.     

STING regulates metabolic reprogramming in macrophages via HIF-1α during Brucella infection

Macrophages metabolic reprogramming in response to microbial insults is a major determinant of pathogen growth or containment. Here, we reveal a distinct mechanism by which stimulator of interferon genes (STING), a cytosolic sensor that regulates innate immune responses, contributes to an inflammatory M1-like macrophage profile upon Brucella abortus infection. This metabolic reprogramming is induced by STING-dependent stabilization of hypoxia-inducible factor-1 alpha (HIF-1α), a global regulator of cellular metabolism and innate immune cell functions. HIF-1α stabilization reduces oxidative phosphorylation and increases glycolysis during infection with B. abortus and, likewise, enhances nitric oxide production, inflammasome activation and IL-1β release in infected macrophages. Furthermore, the induction of this inflammatory profile participates in the control of bacterial replication since absence of HIF-1α renders mice more susceptible to B. abortus infection. Mechanistically, activation of STING by B. abortus infection drives the production of mitochondrial reactive oxygen species (mROS) that ultimately influences HIF-1α stabilization. Moreover, STING increases the intracellular succinate concentration in infected macrophages, and succinate pretreatment induces HIF-1α stabilization and IL-1β release independently of its cognate receptor GPR91. Collectively, these data demonstrate a pivotal mechanism in the immunometabolic regulation of macrophages during B. abortus infection that is orchestrated by STING via HIF-1α pathway and highlight the metabolic reprogramming of macrophages as a potential treatment strategy for bacterial infections.     

CD4+ T Cell-derived IL-10 Promotes Brucella abortus Persistence via Modulation of Macrophage Function

Evasion of host immune responses is a prerequisite for chronic bacterial diseases; however, the underlying mechanisms are not fully understood. Here, we show that the persistent intracellular pathogen Brucella abortus prevents immune activation of macrophages by inducing CD4⁺CD25⁺ T cells to produce the anti-inflammatory cytokine interleukin-10 (IL-10) early during infection. IL-10 receptor (IL-10R) blockage in macrophages resulted in significantly higher NF-kB activation as well as decreased bacterial intracellular survival associated with an inability of B. abortus to escape the late endosome compartment in vitro. Moreover, either a lack of IL-10 production by T cells or a lack of macrophage responsiveness to this cytokine resulted in an increased ability of mice to control B. abortus infection, while inducing elevated production of pro-inflammatory cytokines, which led to severe pathology in liver and spleen of infected mice. Collectively, our results suggest that early IL-10 production by CD25⁺CD4⁺ T cells modulates macrophage function and contributes to an initial balance between pro-inflammatory and anti-inflammatory cytokines that is beneficial to the pathogen, thereby promoting enhanced bacterial survival and persistent infection.     

Brucella spp noncanonical LPS: structure, biosynthesis, and interaction with host immune system

Brucella spp. are facultative intracellular pathogens that have the ability to survive and multiply in professional and non-professional phagocytes, and cause abortion in domestic animals and undulant fever in humans. Several species are recognized within the genus Brucella and this classification is mainly based on the difference in pathogenicity and in host preference. Brucella strains may occur as either smooth or rough, expressing smooth LPS (S-LPS) or rough LPS (R-LPS) as major surface antigen. This bacterium possesses an unconventional non-endotoxic lipopolysaccharide that confers resistance to anti-microbial attacks and modulates the host immune response. The strains that are pathogenic for humans (B. abortus, B. suis, B. melitensis) carry a smooth LPS involved in the virulence of these bacteria. The LPS O-chain protects the bacteria from cellular cationic peptides, oxygen metabolites and complement-mediated lysis and it is a key molecule for Brucella survival and replication in the host. Here, we review i) Brucella LPS structure; ii) Brucella genome, iii) genes involved in LPS biosynthesis; iv) the interaction between LPS and innate immunity.     

Smooth Brucella strains invade and replicate in human lung epithelial cells without inducing cell death

Inhalation is a common route for Brucella infection. We investigated whether Brucella species can invade and replicate within alveolar (A549) and bronchial (Calu-6 and 16HBE14o-) human epithelial cells. The number of adherent and intracellular bacteria was higher for rough strains (Brucella canis and Brucella abortus RB51) than for smooth strains (B. abortus 2308 and Brucella suis 1330). Only smooth strains exhibited efficient intracellular replication (1.5–3.5 log increase at 24 h p.i.). A B. abortus mutant with defective expression of the type IV secretion system did not replicate. B. abortus internalization was inhibited by specific inhibitors of microfilaments, microtubules and PI3-kinase activity. As assessed with fluorescent probes, B. abortus infection did not affect the viability of A549 and 16HBE14o- cells, but increased the percentage of injured cells (both strains) and dead cells (RB51) in Calu-6 cultures. LDH levels were increased in supernatants of Calu-6 and 16HBE14o- cells infected with B. abortus RB51, and to a lower extent in Calu-6 infected with B. abortus 2308. No apoptosis was detected by TUNEL upon infection with smooth or rough B. abortus. This study shows that smooth brucellae can infect and replicate in human respiratory epithelial cells inducing minimal or null cytotoxicity.     

Periplasmic protein EipA determines envelope stress resistance and virulence in Brucella abortus

Molecular components of the Brucella abortus cell envelope play a major role in its ability to infect, colonize and survive inside mammalian host cells. In this study, we have defined a role for a conserved gene of unknown function in B. abortus envelope stress resistance and infection. Expression of this gene, which we name eipA, is directly activated by the essential cell cycle regulator, CtrA. eipA encodes a soluble periplasmic protein that adopts an unusual eight‐stranded β‐barrel fold. Deletion of eipA attenuates replication and survival in macrophage and mouse infection models, and results in sensitivity to treatments that compromise the cell envelope integrity. Transposon disruption of genes required for LPS O‐polysaccharide biosynthesis is synthetically lethal with eipA deletion. This genetic connection between O‐polysaccharide and eipA is corroborated by our discovery that eipA is essential in Brucella ovis, a naturally rough species that harbors mutations in several genes required for O‐polysaccharide production. Conditional depletion of eipA expression in B. ovis results in a cell chaining phenotype, providing evidence that eipA directly or indirectly influences cell division in Brucella. We conclude that EipA is a molecular determinant of Brucella virulence that functions to maintain cell envelope integrity and influences cell division.     

Brucella abortus induces intracellular retention of MHC‐I molecules in human macrophages down‐modulating cytotoxic CD8+ T cell responses

Brucella abortus elicits a vigorous Th1 immune response which activates cytotoxic T lymphocytes. However, B. abortus persists in its hosts in the presence of CD8⁺ T cells, establishing a chronic infection. Here, we report that B. abortus infection of human monocytes/macrophages inhibited the IFN‐γ‐induced MHC‐I cell surface expression. This phenomenon was dependent on metabolically active viable bacteria. MHC‐I down‐modulation correlated with the development of diminished CD8⁺ cytotoxic T cell response as evidenced by the reduced expression of the activation marker CD107a on CD8⁺ T lymphocytes and a diminished percentage of IFN‐γ‐producing CD8⁺ T cells. Inhibition of MHC‐I expression was not due to changes in protein synthesis. Rather, we observed that upon B. abortus infection MHC‐I molecules were retained within the Golgi apparatus. Overall, these results describe a novel mechanism based on the intracellular sequestration of MHC‐I molecules whereby B. abortus would avoid CD8⁺ cytotoxic T cell responses, evading their immunological surveillance.     

Yip1A,a Novel Host Factor for the Activation of the IRE1 Pathway of the Unfolded Protein Response during Brucella Infection

Brucella species replicate within host cells in the form of endoplasmic reticulum (ER)-derived vacuoles. The mechanisms by which the bacteria are sequestered into such vacuoles and obtain a continuous membrane supply for their replication remain to be elucidated. In the present study, we provided several lines of evidence that demonstrate the mechanism by which B. abortus acquires the ER-derived membrane. First, during Brucella infection, the IRE1 pathway, but not the PERK and ATF6 pathways, of the unfolded protein response (UPR) was activated in a time-dependent manner, and the COPII vesicle components Sar1, Sec23, and Sec24D were upregulated. Second, a marked accretion of ER-derived vacuoles was observed around replicating bacteria using fluorescent microscopy and electron microscopy. Third, we identified a novel host factor, Yip1A, for the activation of the IRE1 pathway in response to both tunicamycin treatment and infection with B. abortus. We found that Yip1A is responsible for the phosphorylation of IRE1 through high-order assembly of Ire1 molecules at ER exit sites (ERES) under the UPR conditions. In Yip1A-knockdown cells, B. abortus failed to generate the ER-derived vacuoles, and remained in endosomal/lysosomal compartments. These results indicate that the activation of the IRE1 pathway and the subsequent formation of ER-derived vacuoles are critical for B. abortus to establish a safe replication niche, and that Yip1A is indispensable for these processes. Furthermore, we showed that the autophagy-related proteins Atg9 and WIPI1, but not DFCP1, were required for the biogenesis of the ER-derived membrane compartments. 　On the basis of our findings, we propose a model for intracellular Brucella replication that exploits the host UPR and ER-derived vacuole formation machineries, both of which depend on Yip1A-mediated IRE1 activation.     

Helicobacter pylori-pulsed dendritic cells induce H. pylori-specific immunity in mice

Background: The growing concern over the emergence of antibiotic‐resistant Helicobacter pylori infection is propelling the development of an efficacious vaccine to control this highly adaptive organism. Aim: We studied the use of a dendritic cell (DC)‐based vaccine against H. pylori infection in mice. Methods: The cellular immune responses to murine bone marrow‐derived DCs pulsed with phosphate‐buffered saline (PBS‐DC) or live H. pylori SS1 (HP‐DC) were assessed in vitro and in vivo. The protective immunity against H. pylori SS1 oral challenge was compared between HP‐DC or PBS‐DC immunized mice. The effect of regulatory T‐cell (Treg) depletion by anti‐CD25 antibody on HP‐DC vaccine efficacy was also evaluated. Results: HP‐DC induced a Th1‐dominant response in vitro. In vivo, HP‐DC immunized mice were characterized by a mixed Th1/Th2 peripheral immune response. However, in the stomach, HP‐DC immunized mice expressed a higher level of IFN‐γ compared to PBS‐DC immunized mice; no difference was found for interleukin‐5 expressions in the stomach. A lower bacterial colonization post‐H. pylori challenge was observed in HP‐DC immunized mice compared to PBS‐DC immunized mice with no significant difference in gastritis severity. H. pylori‐specific Th1 response and protective immunity were further enhanced in vivo by depletion of Treg with anti‐CD25 antibody. Conclusion: DC‐based anti‐H. pylori vaccine induced H. pylori‐specific helper T‐cell responses capable of limiting bacterial colonization. Our data support the critical role of effector cellular immune response in the development of H. pylori vaccine.     

FixL-like sensor FlbS of Brucella abortus binds haem and is necessary for survival within eukaryotic cells

Replication of Brucella inside eukaryotic cells is essential for pathogenesis, and successful infection requires rapid adaptation to the intracellular milieu. Close relatives of Brucella use the two-component system FixLJ to survive inside the host. We aimed to identify a homologous sensor in Brucella abortus. A predicted protein with transmembrane and conserved histidine kinase domains was identified as the Fix-like Brucella sensor, FlbS. Although it lacks the PAS domain, recombinant FlbS binds haem in vitro. An internal in-frame deletion in flbS severely decreased B. abortus survival inside professional and non-professional phagocytes. This phenotype was reverted by genetic complementation. These results indicate the critical role of this haemoprotein in the intracellular lifestyle of Brucella.     

RNAi Screen of Endoplasmic Reticulum–Associated Host Factors Reveals a Role for IRE1α in Supporting Brucella Replication

Brucella species are facultative intracellular bacterial pathogens that cause brucellosis, a global zoonosis of profound importance. Although recent studies have demonstrated that Brucella spp. replicate within an intracellular compartment that contains endoplasmic reticulum (ER) resident proteins, the molecular mechanisms by which the pathogen secures this replicative niche remain obscure. Here, we address this issue by exploiting Drosophila S2 cells and RNA interference (RNAi) technology to develop a genetically tractable system that recapitulates critical aspects of mammalian cell infection. After validating this system by demonstrating a shared requirement for phosphoinositide 3-kinase (PI3K) activities in supporting Brucella infection in both host cell systems, we performed an RNAi screen of 240 genes, including 110 ER-associated genes, for molecules that mediate bacterial interactions with the ER. We uncovered 52 evolutionarily conserved host factors that, when depleted, inhibited or increased Brucella infection. Strikingly, 29 of these factors had not been previously suggested to support bacterial infection of host cells. The most intriguing of these was inositol-requiring enzyme 1 (IRE1), a transmembrane kinase that regulates the eukaryotic unfolded protein response (UPR). We employed IRE1α^−/− murine embryonic fibroblasts (MEFs) to demonstrate a role for this protein in supporting Brucella infection of mammalian cells, and thereby, validated the utility of the Drosophila S2 cell system for uncovering novel Brucella host factors. Finally, we propose a model in which IRE1α, and other ER-associated genes uncovered in our screen, mediate Brucella replication by promoting autophagosome biogenesis.     

The Glyceraldehyde-3-Phosphate Dehydrogenase and the Small GTPase Rab 2 Are Crucial for Brucella Replication

The intracellular pathogen Brucella abortus survives and replicates inside host cells within an endoplasmic reticulum (ER)-derived replicative organelle named the “Brucella-containing vacuole” (BCV). Here, we developed a subcellular fractionation method to isolate BCVs and characterize for the first time the protein composition of its replicative niche. After identification of BCV membrane proteins by 2 dimensional (2D) gel electrophoresis and mass spectrometry, we focused on two eukaryotic proteins: the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and the small GTPase Rab 2 recruited to the vacuolar membrane of Brucella. These proteins were previously described to localize on vesicular and tubular clusters (VTC) and to regulate the VTC membrane traffic between the endoplasmic reticulum (ER) and the Golgi. Inhibition of either GAPDH or Rab 2 expression by small interfering RNA strongly inhibited B. abortus replication. Consistent with this result, inhibition of other partners of GAPDH and Rab 2, such as COPI and PKC ι, reduced B. abortus replication. Furthermore, blockage of Rab 2 GTPase in a GDP-locked form also inhibited B. abortus replication. Bacteria did not fuse with the ER and instead remained in lysosomal-associated membrane vacuoles. These results reveal an essential role for GAPDH and the small GTPase Rab 2 in B. abortus virulence within host cells.