The pathogenesis of Coxiella burnetii infections was studied in rabbits and in vitro. We studied (a) the pathogenesis of experimental endocarditis, (b) the effect of human serum complement on lipopolysaccharide (LPS) phase variants and (c) the role of LPS in phase variation. A. Endocarditis. C. burnetii strains from acute Q fever and from endocarditis patients differ in LPS structure and in their ability to induce fever in animals. Studies using a rabbit model of endocarditis showed no differences between the abilities of the Nine Mile strain (typical of acute Q fever strains) and the Priscilla strain (typical of endocarditis strains) to colonize catheter-induced cardiac vegetations. An in vitro assay showed no differences between these strains in their abilities to adhere to fibrin-platelet clots. B. Serum complement-susceptibility. The rough LPS phase variant was serum sensitive, whereas the smooth and intermediate LPS phase variants were resistant. The smooth and intermediate phase variants, however, differed in their interactions with the complement system, suggesting that the mechanisms by which they resisted complement-killing were different. C. Role of LPS in phase variation. The smooth LPS on native phase I variants sterically blocked the binding of antibodies to surface proteins shared with phase II cells. Although any of these surface-exposed proteins which are accessible to antibodies on phase II cells may therefore confer apparent phase II serospecificity, these studies suggest that the unique phase determinant is in the phase II LPS. The significance of this project lies in defining host defence mechanisms controlling C. burnetii infections, in determining whether certain strains have a predilection to cause endocarditis, and in developing effective antimicrobial therapy for C. burnetii endocarditis.
