Brucella abortus cyclic beta-1,2-glucan mutants have reduced virulence in mice and are defective in intracellular replication in HeLa cells

Null cyclic beta-1,2-glucan synthetase mutants (cgs mutants) were obtained from Brucella abortus virulent strain 2308 and from B. abortus attenuated vaccinal strain S19. Both mutants show greater sensitivity to surfactants like deoxycholic acid, sodium dodecyl sulfate, and Zwittergent than the parental strains, suggesting cell surface alterations. Although not to the same extent, both mutants display reduced virulence in mice and defective intracellular multiplication in HeLa cells. The B. abortus S19 cgs mutant was completely cleared from the spleens of mice after 4 weeks, while the 2308 mutant showed a 1.5-log reduction of the number of brucellae isolated from the spleens after 12 weeks. These results suggest that cyclic beta-1,2-glucan plays an important role in the residual virulence of the attenuated B. abortus S19 strain. Although the cgs mutant was cleared from the spleens earlier than the wild-type parental strain (B. abortus S19) and produced less inflammatory response, its ability to confer protection against the virulent strain B. abortus 2308 was fully retained. Equivalent levels of induction of spleen gamma interferon mRNA and anti-lipopolysaccharide (LPS) of immunoglobulin G2a (IgG2a) subtype antibodies were observed in mice injected with B. abortus S19 or the cgs mutant. However, the titer of anti-LPS antibodies of the IgG1 subtype induced by the cgs mutant was lower than that observed with the parental S19 strain, thus suggesting that the cgs mutant induces a relatively exclusive Th1 response.     

Comparison of immune responses and resistance to brucellosis in mice vaccinated with Brucella abortus 19 or RB51.

Immune responses and resistance to infection with Brucella abortus 2308 (S2308) were measured in mice following vaccination with B. abortus 19 (S19) or the lipopolysaccharide (LPS) O-antigen-deficient mutant, strain RB51 (SRB51). Live bacteria persisted for 8 weeks in spleens of mice vaccinated with 5 x 10(6) or 5 x 10(8) CFU of SRB51, whereas bacteria persisted for 12 weeks in mice vaccinated with 5 x 10(6) CFU of S19. Mice vaccinated with 5 x 10(6) or 5 x 10(8) CFU of SRB51 had increased resistance to infection with S2308 at 12, 16, and 20 weeks after vaccination, but the resistance was lower than that induced by vaccinating mice with 5 x 10(6) CFU of S19. Spleen cells obtained from mice vaccinated with S19 or SRB51 generally exhibited similar proliferative responses to S2308 bacteria or bacterial proteins (106 to 18 kDa) following challenge of mice with S2308 at 12, 16, or 20 weeks after vaccination. Mice vaccinated with S19 had antibody to S2308 bacteria and S2308 smooth LPS at 4, 8, and 12 weeks after vaccination. In contrast, mice vaccinated with either dose of SRB51 did not produce antibody to S2308 smooth LPS. In addition, only mice vaccinated with the highest dose of SRB51 (5 x 10(8) CFU) had antibody responses to S2308 bacteria, although the responses were lower and less persistent than those in mice vaccinated with S19. Collectively, these results indicate that SRB51-vaccinated mice have similar cell-mediated immune responses to S2308 but lower resistance to infection with S2308 compared with S19-vaccinated mice. The lower resistance in SRB51-vaccinated mice probably resulted from a combination of rapid clearance of SRB51 and an absence of antibodies to S2308 LPS.     

Comparison of spleen cell proliferation in response to Brucella abortus 2308 lipopolysaccharide or proteins in mice vaccinated with strain 19 or RB51.

Mice vaccinated with Brucella abortus 19 (S19) or RB51 (SRB51) had spleen cells which proliferated in response to proteins of 32, 27, 18, and < 18 kDa but not in response to proteins of 106, 80, and 49 kDa from B. abortus 2308 (S2308) following vaccination and challenge infection with S2308. Spleen cells from mice vaccinated with S19 but not with SRB51 had increased proliferation in response to S2308 lipopolysaccharide (LPS) following challenge infection with S2308. We previously reported that mice vaccinated with S19 or SRB51, which were analyzed in the current study, have increased resistance to infection with S2308 and that only mice vaccinated with S19 produce antibody to S2308 LPS (M. Stevens, S. Olsen, G. Pugh, Jr., and D. Brees, Infect. Immun. 63:264-270, 1995). The results from our current and previous studies support the contention that vaccination of mice with S19 or SRB51 induces protection from infection with S2308 by cell-mediated immune responses to the same immunodominant (32, 27, 18, and < 18 kDa) protein antigens of S2308. In addition, the absence of S2308 LPS-responsive spleen cells and antibody to S2308 LPS in mice vaccinated with SRB51 suggests that immune responses to LPS have no role in SRB51-induced protective immunity.     

Comparative protection of mice against virulent and attenuated strains of Brucella abortus by passive transfer of immune T cells or serum.

Passively transferred immune serum provided significantly greater protection to BALB/c mice against attenuated Brucella abortus 19 than against virulent strain 2308, whether serum donors had been infected with strain 19 or 2308. In contrast, immune T cells conferred better protection upon recipients challenged with the homologous strain of B. abortus. It is hypothesized that strain 2308, but not strain 19, can survive in macrophages after opsonization and that epitopes which induce protective cell-mediated immunity may differ between strains 19 and 2308.     

Penetration and intracellular growth of Brucella abortus in nonphagocytic cells in vitro.

In pregnant ruminants, Brucella abortus localizes and replicates within the rough endoplasmic reticulum of trophoblastic epithelial cells. In this study, Vero cells were exposed to B. abortus to investigate its internalization and intracellular growth in nonphagocytic cells. A new double-fluorescence staining procedure to discriminate between extracellular and intracellular bacteria was developed. Studies with the double-fluorescence staining procedure and quantitative bacteriologic culture of disrupted host cells showed that various B. abortus strains replicated within Vero cells, including smooth virulent (strains 2308S and 544), smooth attenuated (strain 19), and rough (strains 45/20 and 2308R) strains. Rough brucellae were more adherent and entered a greater number of Vero cells. Intracellular replication occurred in a larger percentage of cells with smooth virulent (2308S and 544) strains than with smooth attenuated (19) or rough (45/20 and 2308R) strains. Differences in adhesiveness and invasiveness were correlated to hydrophobicity of the organism, as measured by hydrocarbon adherence. Ultrastructurally, intracellular smooth (2308S) and rough (45/20) brucellae were consistently found within cisternae of the rough endoplasmic reticulum and nuclear envelope. The results suggest that transfer to the rough endoplasmic reticulum is the limiting step in the infection of nonphagocytic cells by B. abortus.     

Antibody responses to Brucella abortus 2308 in cattle vaccinated with B. abortus RB51.

Cattle vaccinated with Brucella abortus rough strain RB51 (SRB51) produced small amounts of serum immunoglobulin G (IgG) but no IgM antibody to smooth strain 2308 (S2308) bacteria and produced no IgG or IgM antibody to S2308 lipopolysaccharide (LPS). Western immunoblot analysis revealed that antiserum from SRB51-vaccinated cattle contained IgG antibody that reacted with S2308 proteins of 84 to <20 kDa. However, antiserum from the vaccinated cattle did not contain agglutinating B. abortus antibody in the tube agglutination test for brucellosis. These results suggest that SRB51-vaccinated cattle produced no antibody to S2308 LPS, although they did produce nonagglutinating IgG antibody that reacted with S2308 bacteria and bacterial proteins of 84 to <20 kDa.     

Immunological response to the Brucella abortus GroEL homolog.

Western blot (immunoblot) analysis of sera from cattle vaccinated with Brucella abortus S19 exhibit an elevated serologic response to Hsp62, the GroEL homolog (BaGroEL). Serologic screening of individual cows vaccinated with B. abortus S19 revealed no correlation between the immune response to BaGroEL and protection against a challenge with virulent organisms. The humoral immune response to BaGroEL was restricted to a region of the mature protein which mapped to amino acids 317 to 355 and may represent a useful diagnostic tool for monitoring exposure to B. abortus. Immunity to a challenge with virulent B. abortus S2308 was not observed in the BaGroEL vaccinated mouse model.     

Complement fixation test to assess humoral immunity in cattle and sheep vaccinated with Brucella abortus RB51.

The live attenuated Brucella abortus strain RB51 is a rifampin-resistant, lipopolysaccharide (LPS) O-chain-deficient mutant of virulent B. abortus 2308. The reduced O-chain content in RB51 prevents this bacterium from inducing antibodies detectable by the conventional serologic tests for bovine brucellosis diagnosis that mainly identify antibodies to LPS. The absence of available serologic tests for RB51 also complicates the diagnosis of possible RB51 infections in humans exposed to this strain. The purpose of this study was to evaluate the suitability of a complement fixation (CF) test performed with the rough strain B. abortus RB51, previously deprived of anticomplementary activity, in detecting anti-B. abortus RB51 antibodies in cattle and sheep experimentally vaccinated with this strain. The results of this study showed that a CF test with RB51 as the antigen is able to specifically detect antibodies following RB51 vaccination in cattle and sheep. In addition, this method could be a useful tool for detecting B. abortus RB51 infection in humans.     

Effectiveness of natural and synthetic complexes of porin and O polysaccharide as vaccines against Brucella abortus in mice.

A single vaccination of mice with a complex of porin and smooth lipopolysaccharide (porin-S-LPS) extracted from virulent Brucella abortus 2308 provided significant protection (P less than 0.01 to P less than 0.001) against challenge with the same strain, equivalent to that achieved by vaccination with living attenuated B. abortus 19. The porin-S-LPS vaccine given without adjuvant or in several adjuvants (trehalose dimycolate and muramyl dipeptide; the pluronic polymer L-121 and muramyl dipeptide; or complexed with Quil A in immunostimulating complexes) provided equivalent protection. In contrast, one vaccination with porin complexed with rough LPS (porin-R-LPS) from a rough mutant of strain 2308 provided no protection with any adjuvant tested. In one experiment, two inoculations with the porin-R-LPS resulted in a low level of protection, probably owing to priming of the animals for production of O-polysaccharide-specific antibodies. However, one vaccination with rough-strain porin covalently bound to purified O polysaccharide conferred protection equal to that obtained with natural complexes of porin-S-LPS or with living strain 19. A synthetic vaccine containing long chains of O polysaccharide was more effective than one prepared with short chains. Protective vaccines caused the formation of increased concentrations of circulating O-polysaccharide-specific antibodies, although there were individual exceptions to the quantitative association between O-polysaccharide-specific antibodies and protection. Antibodies specific for porin or R-LPS were found in negligible quantities in vaccinated mice. These results provide additional evidence that the O polysaccharide will constitute an essential component of an effective subcellular vaccine against B. abortus and that O-polysaccharide-specific antibodies play an important role in protective immunity in brucellosis.     

Lymphocyte proliferation in response to immunodominant antigens of Brucella abortus 2308 and RB51 in strain 2308-infected cattle.

Lymphocyte proliferation in response to proteins from the Brucella abortus strain 2308 (S2308) and the lipopolysaccharide (LPS) O-antigen-deficient mutant of S2308, strain RB51 (SRB51), was measured in S2308-infected cattle following abortion. Supramammary and superficial cervical lymph node lymphocytes from infected cattle proliferated most when incubated with 27- to 18-kDa proteins of S2308 or SRB51. Proteins of SRB51, which contained no LPS O antigens, induced lymphocyte proliferation similar to that induced by S2308 proteins, which contained LPS O antigens. These results indicate that 27- to 18-kDa proteins, but not LPS O antigens, of S2308 and SRB51 are immunodominant in S2308-infected cattle as assessed by lymphocyte proliferation assays.     

Construction of Cu-Zn superoxide dismutase deletion mutants of Brucella abortus: analysis of survival in vitro in epithelial and phagocytic cells and in vivo in mice.

Cu-Zn superoxide dismutase (SOD) deletion mutants of Brucella abortus S2308, a virulent strain, and S19, a vaccine strain, were generated by gene replacement. A deletion plasmid, pBA delta sodknr, was constructed by excising the Cu-Zn SOD gene (Cu-Zn sod) from a 2.3-kb B. abortus DNA fragment of plasmid pBA20-1527 and inserting a 1.4-kb DNA fragment encoding kanamycin resistance into the Cu-Zn sod excision site. The deletion plasmid was introduced into B. abortus by electroporation, and Southern blot analysis confirmed that the antibiotic resistance fragment had replaced Cu-Zn sod in kanamycin-resistant colonies. The survival and growth of Cu-Zn SOD mutant strains were compared with that of the parental strains in HeLa cells and in the mouse macrophagelike cell line J774. The survival and growth of the Cu-Zn SOD mutant strains were similar to those of their respective parental strains in HeLa and J774 cell lines. The kinetics of infection with these strains were examined in BALB/c mice. The splenic levels of the S19 Cu-Zn SOD mutant recovered from intraperitoneally infected BALB/c mice were approximately 10-fold lower than those of the parental strain through 26 days postinfection. Thereafter, infection sharply declined in both groups, and by 105 days postinfection, no organisms were detected. The splenic levels of the S2308 Cu-Zn SOD mutant were lower than those of wild-type S2308-infected mice. The spleen weights of mice infected with the S2308 Cu-Zn SOD mutant were consistently lower than those of wild-type S2308-infected mice. These results suggest that the antioxidant enzyme Cu-Zn SOD plays a role in the survival and pathogenicity of B. abortus in vivo.     

Opsonized virulent Brucella abortus replicates within nonacidic, endoplasmic reticulum-negative, LAMP-1-positive phagosomes in human monocytes

Cells in the Brucella spp. are intracellular pathogens that survive and replicate within host monocytes. Brucella maintains persistent infections in animals despite the production of high levels of anti-Brucella-specific antibodies. To determine the effect of antibody opsonization on the ability of Brucella to establish itself within monocytes, the intracellular trafficking of virulent Brucella abortus 2308 and attenuated hfq and bacA mutants was followed in the human monocytic cell line THP-1. Early trafficking events of B. abortus 2308-containing phagosomes (BCP) were indistinguishable from those seen for control particles (heat-killed B. abortus 2308, live Escherichia coli HB101, or latex beads). All phagosomes transiently communicated the early-endosomal compartment and rapidly matured into LAMP-1(+), cathepsin D(+), and acidic phagosomes. By 2 h postinfection, however, the number of cathepsin D(+) BCP was significantly lower for live B. abortus 2308-infected cells than for either Brucella mutant strains or control particles. B. abortus 2308 persisted within these cathepsin D(-), LAMP-1(+), and acidic vesicles; however, at the onset of intracellular replication, the numbers of acidic B. abortus 2308 BCP decreased while remaining cathepsin D(-) and LAMP-1(+). In contrast to B. abortus 2308, the isogenic hfq and bacA mutants remained in acidic, LAMP-1(+) phagosomes and failed to initiate intracellular replication. Notably, markers specific for the host endoplasmic reticulum were absent from the BCPs throughout the course of the infection. Thus, opsonized B. abortus in human monocytes survives within phagosomes that remain in the endosomal pathway and replication of virulent B. abortus 2308 within these vesicles corresponds with an increase in intraphagosomal pH.     

Killing of Brucella abortus by bovine serum

Studies of the serum bactericidal system in bovine brucellosis were undertaken to investigate the role of the humoral immune response in protection of cattle against the facultative intracellular parasite Brucella abortus. Fresh sera from normal control cattle, infected cattle, and cattle immunized with B. abortus cell envelopes were collected before treatment and during the course of immunization or infection. Normal fresh bovine serum or fresh agammaglobulinemic serum from colostrum-deprived calves was effective in killing smooth virulent B. abortus 2308, but rough strains RB51 (a rough mutant of strain 2308) and 45/20 were much more sensitive to serum. The difference in susceptibility to serum was shown to be correlated with differences in lipopolysaccharide chemotype, with the more resistant strain 2308 having O polysaccharide and the more susceptible strains 45/20 and RB51 lacking O side chains. By treatment of fresh serum with MgCl2 and EGTA [ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid] killing was shown to occur via the classical pathway of complement activation. When antibody to B. abortus was present, killing of strain RB51 increased but killing of smooth strain 2308 decreased. The earliest antibody response in serum from infected animals did not interfere with killing. When affinity-purified bovine immunoglobulins specific for B. abortus smooth lipopolysaccharide were added to fresh normal bovine serum, immunoglobulin G1 (IgG1) and IgG2 isotypes blocked killing but IgM and IgA isotypes did not. Thus, it appears that serum from previously unexposed animals or animals early during infection can kill smooth B. abortus, an appropriate defense mechanism before the organism becomes intracellular. At later stages of infection, blocking antibodies predominate.     

Validation of the abbreviated Brucella AMOS PCR as a rapid screening method for differentiation of Brucella abortus field strain isolates and the vaccine strains, 19 and RB51

The Brucella AMOS PCR assay was previously developed to identify and differentiate specific Brucella species. In this study, an abbreviated Brucella AMOS PCR test was evaluated to determine its accuracy in differentiating Brucella abortus into three categories: field strains, vaccine strain 19 (S19), and vaccine strain RB51/parent strain 2308 (S2308). Two hundred thirty-one isolates were identified and tested by the conventional biochemical tests and Brucella AMOS PCR. This included 120 isolates identified as B. abortus S19, 9 identified as B. abortus strain RB51, 57 identified as B. abortus biovar 1, 15 identified as B. abortus bv. 2, 1 identified as B. abortus bv. 2 (M antigen dominant), 7 identified as B. abortus bv. 4, and 22 identified as B. abortus S2308 and isolated from experimentally infected cattle. The Brucella AMOS PCR correctly identified each isolate as RB51/S2308, S19, or a field strain of Brucella.     

The role of integrase/recombinase xerD and monofunctional biosynthesis peptidoglycan transglycosylase genes in the pathogenicity of Brucella abortus infection in vitro and in vivo

Brucella abortus clones identified previously using a green fluorescence protein reporter system after 4h macrophage infection provided insight regarding possible genes involved in early host-pathogen interaction. Among identified genes were an integrase/recombinase (xerD) gene involved in cell division, and a monofunctional biosynthesis peptidoglycan transglycosylase (mtgA) gene that catalyzes the final stages of the peptidoglycan membrane synthesis. Here, we evaluate the in vitro and in vivo survival of B. abortus xerD and mtgA insertional mutants. B. abortus xerD::kan and B. abortus mtgA::kan demonstrated no significant growth defects in broth culture when compared to the parental strain, S2308. Also, neither gene was required for B. abortus S2308 replication in RAW 264.7 macrophages. However, experimental evidence using interferon regulatory factor 1 knockout mice, a mouse strain highly susceptible to virulent Brucella, revealed that mice infected with B. abortus xerD::kan or B. abortus mtgA::kan survived longer than mice infected with S2308. Additionally, in immunocompetent BALB/c mice, B. abortus xerD::kan had a significantly lower level of bacterial survival when compared to S2308. Together, these results suggest that B. abortus xerD and mtgA genes play a role during the initial phase of infection in mice.     

The Role of B-Cell Memory in Secondary IgG and IgM Responses

Mice were primed with the hapten 3-nitro-4-hydroxy-5 iodophenacetic acid (NIP) conjugated to chicken globulin (cg) and were boosted 2, 6, or 12 months later with CG conjugates of the related haptens 3,5-diiodo-4-hydroxyphenacetic acid (DIP) or 3-nitro-4-hydroxphenacetic acid (NP). Accelerated secondary responses were demonstrated both in the 7S and 19S class. Fine-specificities of secondary-response antibodies were studied by the hapten inhibition method of haptenated bacteriophage inactivation. 7S antibodies were found to have the fine-specificity of anti-NIP antibodies regardless of whether DIP or NP was the booster hapten ('original antigenic sin'). 19S antibodies had the fine-specificity of anti-DIP when DIP was the booster hapten. NP as the booster hapten resulted in 19S antibodies whose fine-specificity was intermediate between anti-NIP and anti-NP. A strong B-cell memory could thus be demonstrated in the 7S antibody response and a weak B-cell memory in the 19S antibody response.     

Brucella abortus RB51 induces protection in mice orally infected with the virulent strain B. abortus 2308

Brucellae are gram-negative, facultative intracellular bacteria which are one of the most common causes of abortion in animals. In addition, they are the source of a severe zoonosis. In this trial, we evaluated the effect of oral inoculation of Brucella abortus RB51 in mice against a challenge infection with B. abortus 2308. First, we showed that a gastric acid neutralization prior to the oral inoculation contributed to a more homogeneous and consistent infection with both vaccine strain B. abortus RB51 and virulent strain B. abortus 2308. Successively, we assessed the clearance and the immune response following an oral infection with B. abortus RB51. Oral inoculation gave a mild infection which was cleared 42 days after infection, and it induced a delayed humoral and cell-mediated immune response. Finally, we immunized mice by oral inoculation with B. abortus RB51, and we challenged them with the virulent strain B. abortus 2308 by an oral or intraperitoneal route 42 days after vaccination. Oral inoculation of B. abortus RB51 was able to give protection to mice infected with the virulent strain B. abortus 2308 by the oral route but not to mice infected intraperitoneally. Our results indicate that oral inoculation of mice with B. abortus RB51 is able to give a protective immunity against an oral infection with virulent strains, and this protection seems to rely on an immune response at the mucosal level.     

Generation of chemotactic factor for granulocytes and monocytes from serum by fractions of Brucella abortus.

Several fractions isolated from Brucella abortus were examined for their ability to generate chemotactic factor from normal serum. Phenol phase lipopolysaccharides exhibited activity equivalent to that obtained with E. Coli lipopolysaccharide. A carbohydrate-rich aqueous methanol fraction was inhibitory at high concentrations, but a non-dialysable component of this fraction contained a potent stimulator of chemotactic activity. Protein-rich fractions from both strain 19 and strain 2308 were inactive. Preheating the serum at 56 degrees for 30 min prevented generation of chemotactic activity by the various fractions, suggesting a role for serum complement. No chemotactic activity was produced by Brucella fractions in C5-deficient DBA/2J mouse serum.     

Role in virulence of a Brucella abortus protein exhibiting lectin-like activity

Brucella abortus is a facultative, intracellular zoonotic pathogen which can cause undulant fever in humans and abortions in cattle. A 14-kDa protein of B. abortus was previously identified to be immunogenic in animals infected with Brucella spp. In this study, we discovered that the 14-kDa protein possessed immunoglobulin binding and hemagglutination properties that appeared to be based on the protein's lectin-like properties. Hemagglutination inhibition experiments suggested that the 14-kDa protein has affinity towards mannose. Disruption of the gene encoding the 14-kDa protein in virulent B. abortus strain 2308 induced a rough-like phenotype with an altered smooth lipopolysaccharide (LPS) immunoblot profile and a significant reduction in the bacterium's ability to replicate in mouse spleens. However, the mutant strain was stably maintained in mouse spleens at 2.0 to 2.6 log(10) CFU/spleen from day 1 to week 6 after intraperitoneal inoculation with 4.65 log(10) CFU. In contrast to the case for the smooth virulent strain 2308, in the rough attenuated strain RB51 disruption of the 14-kDa protein's gene had no effect on the mouse clearance pattern. These findings indicate that the 14-kDa protein of B. abortus possesses lectin-like properties and is essential for the virulence of the species, probably because of its direct or indirect role in the synthesis of smooth LPS.     

Genetic Characterization of a Tn5-Disrupted Glycosyltransferase Gene Homolog in Brucella abortus and Its Effect on Lipopolysaccharide Composition and Virulence

We constructed a rough mutant of Brucella abortus 2308 by transposon (Tn5) mutagenesis. Neither whole cells nor extracted lipopolysaccharide (LPS) from this mutant, designated RA1, reacted with a Brucella O-side-chain-specific monoclonal antibody (MAb), Bru-38, indicating the absence of O-side-chain synthesis. Compositional analyses of LPS from strain RA1 showed reduced levels of quinovosamine and mannose relative to the levels in the parental, wild-type strain, 2308. We isolated DNA flanking the Tn5 insertion in strain RA1 by cloning a 25-kb XbaI genomic fragment into pGEM-3Z to create plasmid pJM6. Allelic exchange of genomic DNA in B. abortus 2308 mediated by electroporation of pJM6 produced kanamycin-resistant clones that were not reactive with MAb Bru-38. Southern blot analysis of genomic DNA from these rough clones revealed Tn5 in a 25-kb XbaI genomic fragment. A homology search with the deduced amino acid sequence of the open reading frame disrupted by Tn5 revealed limited homology with various glycosyltransferases. This B. abortus gene has been named wboA. Transformation of strain RA1 with a broad-host-range plasmid bearing the wild-type B. abortus wboA gene resulted in the restoration of O-side-chain synthesis and the smooth phenotype. B. abortus RA1 was attenuated for survival in mice. However, strain RA1 persisted in mice spleens for a longer time than the B. abortus vaccine strain RB51, but as expected, neither strain induced antibodies specific for the O side chain.