Search

Capnocytophaga canimorsus is a Gram-negative commensal of dog's mouth causing severe human infections. A strain isolated from a human fatal infection was recently shown to have a sialidase, to inhibit the bactericidal activity of macrophages and to block the release of nitric oxide by LPS-stimulated macrophages. The present study aimed at determining the prevalence of C. canimorsus in dogs and the occurrence of these hypothetical virulence factors. C. canimorsus could be retrieved from the saliva of 61 dogs out of 106 sampled. Like in clinical isolates, all dog strains had a sialidase and 60% blocked the killing of phagocytosed Escherichia coli by macrophages. In contrast, only 6.5% of dog strains blocked the release of nitric oxide by LPS-challenged macrophages, suggesting that this property might contribute to virulence. The comparative analysis of 69 16S rDNA sequences revealed the existence of C. canimorsus strains that could be misdiagnosed.

The inner-membrane protein YscU has an important role during the assembly of the Yersinia enterocolitica type III secretion injectisome. Its cytoplasmic domain (YscU(C)) recognizes translocators as individual substrates in the export hierarchy. Activation of YscU entails autocleavage at a conserved NPTH motif. Modification of this motif markedly changes the properties of YscU, including translocator export cessation and production of longer injectisome needles. We determined the crystal structures of the uncleaved variants N263A and N263D of YscU(C) at 2.05 A and 1.55 A resolution, respectively. The globular domain is found to consist of a central, mixed beta-sheet surrounded by alpha-helices. The NPTH motif forms a type II beta-turn connecting two beta-strands. NMR analysis of cleaved and uncleaved YscU(C) indicates that the global structure of the protein is retained in cleaved YscU(C). The structure of YscU(C) variant N263D reveals that wild type YscU(C) is poised for cleavage due to an optimal reaction geometry for nucleophilic attack of the scissile bond by the side chain of Asn263. In vivo analysis of N263Q and H266A/R314A YscU variants showed a phenotype that combines the absence of translocator secretion with normal needle-length control. Comparing the structure of YscU to those of related proteins reveals that the linker domain between the N-terminal transmembrane domain and the autocleavage domain can switch from an extended to a largely alpha-helical conformation, allowing for optimal positioning of the autocleavage domain during injectisome assembly.

Many Gram-negative bacteria use a type III secretion (T3S) system to directly inject effector molecules into eucaryotic cells in order to establish a symbiotic or pathogenic relationship with their host. The translocation of many T3S proteins requires specialized chaperones from the bacterial cytosol. SycD belongs to a class of T3S chaperones that assists the secretion of pore-forming translocators and, specifically chaperones the translocators YopB and YopD from enteropathogenic Yersinia enterocolitica. In addition, SycD is involved in the regulation of virulence factor biosynthesis and secretion. In this study, we present two crystal structures of Y. enterocolitica SycD at 1.95 and 2.6 A resolution, the first experimental structures of a T3S class II chaperone specific for translocators. The fold of SycD is entirely alpha-helical and reveals three tetratricopeptide repeat-like motifs that had been predicted from amino acid sequence. In both structures, SycD forms dimers utilizing residues from the first tetratricopeptide repeat motif. Using site-directed mutagenesis and size exclusion chromatography, we verified that SycD forms head-to-head homodimers in solution. Although in both structures, dimerization largely depends on the same residues, the two assemblies represent alternative dimers that exhibit different monomer orientations and overall shape. In these two distinct head-to-head dimers, both the concave and the convex surface of each monomer are accessible for interactions with the SycD binding partners YopB and YopD. A SycD variant carrying two point mutations in the dimerization interface is properly folded but defective in dimerization. Expression of this stable SycD monomer in Yersinia does not rescue the phenotype of a sycD null mutant, suggesting a physiological relevance of the dimerization interface.

Capnocytophaga canimorsus is a commensal bacterium from the canine oral flora, which can cause septicemia or meningitis in humans upon bite wound infections. C. canimorsus 5 (Cc5), a strain isolated from a patient with fatal septicemia, was used to investigate the interaction between C. canimorsus and J774.1 mouse macrophages. J774.1 cells infected at high multiplicity with Cc5 did not phagocytose nor kill Cc5 within 120 min of infection, unless the bacteria were opsonized with specific antibodies. Opsonization with complement, however, did not increase phagocytosis. Moreover, infection of J774.1 cells with live Cc5 led to the release of a soluble factor, which interfered with the ability of macrophages to kill other phagocytosed bacteria. These results provide an example of how C. canimorsus neutralizes the innate immune system.

The complete nucleotide sequence and organization of the Yersinia enterocolitica strain A127/90 serotype O:8 biotype 1B virulence plasmid (pYVa127/90) were determined. In contrast to other biotype 1B strains, usually isolated in the USA, strain A127/90 was isolated in Japan. The organization of the genes carried by the 66591-bp pYVa127/90 plasmid is globally similar to that of 67720-bp pYVe8081, the virulence plasmid from 8081, an American isolate of serotype O:8 biotype 1B. Differences are noticeable in the amino acid sequences of the SycH chaperone, the YopM cytotoxin and the plague-protective antigen LcrV. Interestingly, pYVa127/90 contains the entire coding sequence of the YlpA lipoprotein, which is missing from pYVe8081. In addition, we identified five potential new genes. Several vestigial insertion sequence elements appeared to be different between the two plasmids. These observations suggest that both pYV plasmids are from a common ancestor but may have evolved independently following the conquest of two distant ecological niches.