Brucella are highly infectious bacterial pathogens responsible for brucellosis, a frequent worldwide zoonosis. The Brucella genus has recently expanded from 6 to 11 species, all of which were associated with mammals; The natural host range recently expanded to amphibians after some reports of atypical strains from frogs. Here we describe the first in depth phenotypic and genetic characterization of a Brucella strains isolated from a frog. Strain B13-0095 was isolated from a Pac-Man frog (Ceratophyrus ornate) at a veterinary hospital in Texas and was initially misidentified as Ochrobactrum anthropi. We found that B13-0095 belongs to a group of early-diverging brucellae that includes Brucella inopinata strain BO1 and the B. inopinata-like strain BO2, with traits that depart significantly from those of the “classical” Brucella spp. Analysis of B13-0095 genome sequence revealed several specific features that suggest that this isolate represents an intermediate between a soil associated ancestor and the host adapted “classical” species. Like strain BO2, B13-0095 does not possess the genes required to produce the perosamine based LPS found in classical Brucella, but has a set of genes that could encode a rhamnose based O-antigen. Despite this, B13-0095 has a very fast intracellular replication rate in both epithelial cells and macrophages. Finally, another major finding in this study is the bacterial motility observed for strains B13-0095, BO1, and BO2, which is remarkable for this bacterial genus. This study thus highlights several novel characteristics in strains belonging to an emerging group within the Brucella genus. Accurate identification tools for such atypical Brucella isolates and careful evaluation of their zoonotic potential, are urgently required.

Haemophilus parasuis is a porcine respiratory pathogen, well known as the aetiological agent of Glä sser's disease. H. parasuis comprises strains of different virulence, but the virulence factors of this bacterium are not well defined. A neuraminidase activity has been previously detected in H. parasuis, but the role of sialylation in the virulence of this bacterium has not been studied. To explore the relationship between sialic acid (Neu5Ac) and virulence, we assessed the distribution of genes involved in sialic acid metabolism in 21 H. parasuis strains from different clinical origins (including nasal and systemic isolates). The neuraminidase gene nanH, together with CMPNeu5Ac synthetase and sialyltransferase genes neuA, siaB and lsgB, were included in the study. Neuraminidase activity was found to be common in H. parasuis isolates, and the nanH gene from 12 isolates was expressed in Escherichia coli and further characterized. Sequence analysis showed that the NanH predicted protein contained the motifs characteristic of the catalytic site of sialidases. While an association between the presence of nanH and the different origins of the strains was not detected, the lsgB gene was predominantly present in the systemic isolates, and was not amplified from any of the nasal isolates tested. Analysis of the lipooligosaccharide (LOS) from reference strains Nagasaki (virulent, lsgB + ) and SW114 (non-virulent, lsgB " ) showed the presence of sialic acid in the LOS from the Nagasaki strain, supporting the role of sialylation in the virulence of this bacterial pathogen. Further studies are needed to clarify the role of sialic acid in the pathogenicity of H. parasuis.

Brucella spp. are Gram-negative bacteria that behave as facultative intracellular parasites of a variety of mammals. This genus includes smooth (S) and rough (R) species that carry S and R lipopolysaccharides (LPS), respectively. S-LPS is a virulence factor, and mutants affected in the S-LPS O-polysaccharide (R mutants), core oligosaccharide or both show attenuation. However, B. ovis is naturally R and is virulent in sheep. We studied the role of B. ovis LPS in virulence by mutating the orthologues of wadA, wadB and wadC, three genes known to encode LPS core glycosyltransferases in S brucellae. When mapped with antibodies to outer membrane proteins (Omps) and R-LPS, wadB and wadC mutants displayed defects in LPS structure and outer membrane topology but inactivation of wadA had little or no effect. Consistent with these observations, the wadB and wadC but not the wadA mutants were attenuated in mice. When tested as vaccines, the wadB and wadC mutants protected mice against B. ovis challenge. The results demonstrate that the LPS core is a structure essential for survival in vivo not only of S brucellae but also of a naturally R Brucella pathogenic species, and they confirm our previous hypothesis that the Brucella LPS core is a target for vaccine development. Since vaccine B. melitensis Rev 1 is S and thus interferes in serological testing for S brucellae, wadB mutant represents a candidate vaccine to be evaluated against B. ovis infection of sheep suitable for areas free of B. melitensis.