Different Bactericidal and Inflammatory Activities of Human and Mouse Blood.

We performed side-by-side experiments to compare the behavior of four strains of Escherichia coli and one strain of Pseudomonas aeruginosa in fresh human and mouse blood. Bacteria were multiplied in mouse whole blood and plasma but were killed in human whole blood and plasma. The percentage of granulocytes associated with fluorescence-labeled heat-killed E coli relative to total leukocytes counted was higher in human compared to mouse blood as assessed by flow cytometry analysis. Concentrations of proinflammatory cytokines were high in human blood, but undetectable in mouse blood despite high concentrations of bacteria. We conclude that bacterial killing, phagocytosis, and cytokine induction in blood during human bacteremia with these organisms are probably not mimicked in mouse models of bacterial challenge. Understanding the mechanisms for low cytokine induction with high bacterial loads in mouse blood may be helpful to interpret murine models of bacteremia and develop new approaches for treating sepsis in humans.

Purified and Recombinant Hemopexin: Protease Activity and Effect on Neutrophil Chemotaxis.

Infusion of the heme-binding protein hemopexin has been proposed as a novel approach to decrease heme-induced inflammation in settings of red blood cell breakdown, but questions have been raised as to possible side effects related to protease activity and inhibition of chemotaxis. We evaluated protease activity and effects on chemotaxis of purified plasma hemopexin obtained from multiple sources as well as a novel recombinant fusion protein Fc-hemopexin. Amidolytic assay was performed to measure the protease activity of several plasma-derived hemopexin and recombinant Fc-hemopexin. Hemopexin was added to the human monocyte culture in the presence of lipopolysaccharides (LPS), and also injected into mice intravenously (i.v.) 30 min before inducing neutrophil migration via intraperitoneal (i.p.) injection of thioglycolate. Control groups received the same amount of albumin. Protease activity varied widely between hemopexins. Recombinant Fc-hemopexin bound heme, inhibited the synergy of heme with LPS on tumor necrosis factor (TNF) production from monocytes, and had minor but detectable protease activity. There was no effect of any hemopexin preparation on chemotaxis, and purified hemopexin did not alter the migration of neutrophils into the peritoneal cavity of mice. Heme and LPS synergistically induced the release of LTB4 from human monocytes, and hemopexin blocked this release, as well as chemotaxis of neutrophils in response to activated monocyte supernatants. These results suggest that hemopexin does not directly affect chemotaxis through protease activity, but may decrease heme-driven chemotaxis and secondary inflammation by attenuating the induction of chemoattractants from monocytes. This property could be beneficial in some settings to control potentially damaging inflammation induced by heme.

Hemopexin in severe inflammation and infection: mouse models and human diseases.

INTRODUCTION: Cell-free plasma hemoglobin is associated with poor outcome in patients with sepsis. Extracellular hemoglobin and secondarily released heme amplify inflammation in the presence of microbial TLR ligands and/or endogenous mediators. Hemopexin, a plasma protein that binds heme with extraordinary affinity, blocks these effects and has been proposed as a possible treatment approach to decrease inflammation in critically ill patients. METHODS: We studied mouse models of endotoxemia, burn wound infections and peritonitis in order to assess if a repletion strategy for hemopexin might be reasonable. We also measured hemopexin in small numbers of three patient populations that might be logical groups for hemopexin therapy: patients with sepsis and ARDS, patients with severe burns, and premature infants. RESULTS: Despite severe disease, mean plasma hemopexin levels were increased above baseline in each murine model. However, plasma hemopexin levels were decreased or markedly decreased in many patients in each of the three patient populations. CONCLUSIONS: Potentially different behavior of hemopexin in mice and humans may be important to consider when utilizing murine models to represent acute human inflammatory diseases in which heme plays a role. The findings raise the possibility that decreased hemopexin could result in insufficiently neutralized or cleared heme in some patients with ARDS, burns, or in premature infants who might be candidates to benefit from hemopexin administration.