Bacterial injectisomes deliver effector proteins straight into the cytosol of eukaryotic cells (type III secretion, T3S). Many effectors are associated with a specific chaperone that remains inside the bacterium when the effector is delivered. The structure of such chaperones and the way they interact with their substrate is well characterized but their main function remains elusive. Here, we describe and characterize SycO, a new chaperone for the Yersinia effector kinase YopO. The chaperone-binding domain (CBD) within YopO coincides with the membrane localization domain (MLD) targeting YopO to the host cell membrane. The CBD/MLD causes intrabacterial YopO insolubility and the binding of SycO prevents this insolubility but not folding and activity of the kinase. Similarly, SycE masks the MLD of YopE and SycT covers an aggregation-prone domain of YopT, presumably corresponding to its MLD. Thus, SycO, SycE and most likely SycT mask, inside the bacterium, a domain needed for proper localization of their cognate effector in the host cell. We propose that covering an MLD might be an essential function of T3S effector chaperones.

Capnocytophaga canimorsus is a bacterium of the canine oral flora known since 1976 to cause rare but severe septicemia and peripheral gangrene in patients that have been in contact with a dog. It was recently shown that these bacteria do not elicit an inflammatory response (H. Shin, M. Mally, M. Kuhn, C. Paroz, and G. R. Cornelis, J. Infect. Dis. 195:375-386, 2007). Here, we analyze their sensitivity to the innate immune system. Bacteria from the archetype strain Cc5 were highly resistant to killing by complement. There was little membrane attack complex (MAC) deposition in spite of C3b deposition. Cc5 bacteria were as resistant to phagocytosis by human polymorphonuclear leukocytes (PMNs) as Yersinia enterocolitica MRS40, endowed with an antiphagocytic type III secretion system. We isolated Y1C12, a transposon mutant that is hypersensitive to killing by complement via the antibody-dependent classical pathway. The mutation inactivated a putative glycosyltransferase gene, suggesting that the Y1C12 mutant was affected at the level of a capsular polysaccharide or lipopolysaccharide (LPS) structure. Cc5 appeared to have several polysaccharidic structures, one being altered in Y1C12. The structure missing in Y1C12 could be purified by classical LPS purification procedures and labeled by tritiated palmitate, indicating that it is more likely to be an LPS structure than a capsule. Y1C12 bacteria were also more sensitive to phagocytosis by PMNs than wild-type bacteria. In conclusion, a polysaccharide structure, likely an LPS, protects C. canimorsus from deposition of the complement MAC and from efficient phagocytosis by PMNs.

No abstract

No abstract