Succinic acid production by Bacteroides fragilis. A potential bacterial virulence factor.

We previously demonstrated that Bacteroides species elaborate a factor that impairs polymorphonuclear leukocyte (PMN) migration. This factor, active at a pH of 5.5 but not at 7.4, had similar inhibitory characteristics as the Bacteroides metabolite succinic acid. We examined the kinetics of production of this inhibitory factor and correlated its presence with succinic acid concentrations in the culture filtrate. In the four to six hours after Bacteroides fragilis 9032 reached the stationary phase of growth, there was a large increase in succinic acid production with an accompanying reduction in culture pH. This correlated well with the generation of a leukocyte-inhibitory factor. Pure succinic acid in sterile culture medium at concentrations similar to those detected in the 18-hour culture filtrate (20 mmol/L) produced comparable reduction in PMN migration. Following gel filtration chromatography of B fragilis culture filtrate, the fractions containing succinic acid labeled with carbon 14 inhibited PMN migration.

A Bacteroides by-product inhibits human polymorphonuclear leukocyte function.

We have previously demonstrated that Bacteroides fragilis enhanced Escherichia coli-induced lethality in the rat fibrin-clot peritonitis model. As a possible mechanism for this phenomenon, it was hypothesized that B fragilis inhibited host defense mechanisms, allowing the E coli to flourish and kill the animal. Culture filtrates of three Bacteroides species were tested in vitro for their effect on human polymorphonuclear leukocyte (PMN) chemotaxis and random migration. Two of these, B fragilis and Bacteroides distasonis, impaired PMN migration. The other, Bacteroides thetaiotaomicron, caused variable inhibition of neutrophil chemotaxis. The ability of the culture filtrates to inhibit neutrophil function appeared to depend on two factors: (1) adequate growth of the Bacteroides culture, permitting production of the leukotoxic factor, and (2) reduction of the culture pH to a level at which the putative toxin could exert its effect. Further studies revealed that the factor was heat stable, had a molecular weight less than 500, and that its effect on PMNs was only partially reversed by multiple washings. This supports the concept that Bacteroides species may contribute to the pathogenicity of mixed infections by producing a factor that inhibits host neutrophil function.

Mechanism of the adjuvant effect of hemoglobin in experimental peritonitis. IX: The infection-potentiating effect of hemoglobin in Escherichia coli peritonitis is strain specific.

Hemoglobin solutions have been said to consistently increase the lethality of otherwise nonlethal bacterial inocula in experimental models of Escherichia coli peritonitis. We tested the capacity of stroma-free hemoglobin to potentiate the lethality of each of 26 separate clinical isolates of E. coli. The LD50 of each strain with and without stroma-free hemoglobin was then correlated with the ability of that strain to express putative "virulence characteristics": the expression of 0 (lipopolysaccharide) and K (capsular) antigens, the ability to produce colicin V, the capacity to hemagglutinate mammalian red cells in the presence of 1% mannose, and the ability to secrete alpha-hemolysin. No perfect correlations were found. The LD50 of only four of the 26 strains of E. coli was affected by hemoglobin. Each of these four strains could hemagglutinate red cells and secreted alpha-hemolysin. Many other strains whose lethality was not increased by hemoglobin also had these virulence properties. We must conclude that the infection-potentiating effect of hemoglobin cannot be shown for most clinical isolates of E. coli and that the mechanism cannot be correlated with the usual "virulence characteristics" of E. coli.

Mechanism of the adjuvant effect of hemoglobin in experimental peritonitis: VIII. A leukotoxin is produced by Escherichia coli metabolism in hemoglobin.

Hemoglobin, but not albumin, has long been recognized as an infection potentiating factor in experimental Escherichia coli peritonitis, but the mechanism has defied definition. We have shown previously that stroma-free hemoglobin is not toxic to polymorphonuclear neutrophils. To test the hypothesis that hemoglobin provides a nutritional boost to the growth of E. coli in vivo, we inoculated E. coli into dialysis bags containing equivalent amounts of stroma-free hemoglobin or albumin. These bags were implanted into the peritoneal cavity of rats and at intervals the fluid was removed and the bacteria enumerated. This technique allows for intraperitoneal bacterial growth but eliminates the variables of lymphatic clearance and phagocytic ingestion. The growth rate of E. coli was the same irrespective of the nutritional supplement in the bag. Thus there is no experimental support for the notion that hemoglobin directly accelerates E. coli proliferation under in vivo conditions. To test the hypothesis that a leukocyte toxin may result from E. coli growth in hemoglobin, we exposed normal human neutrophils to the sterilized contents of the peritoneal dialysis bags. In vitro function of the neutrophils (viability, random migration, chemotaxis, phagocytosis, bacterial killing, and chemiluminescence) was significantly depressed by prior exposure to hemoglobin supernatants that had supported E. coli proliferation in vivo. Stroma-free hemoglobin had minimal adverse effects. Albumin supernatants that had supported E. coli proliferation in vivo had significantly less effect on neutrophil function even though the endotoxin levels were identical to the hemoglobin E. coli solutions. We must conclude that leukotoxins result from E. coli growth in solutions of pure hemoglobin. The data support the idea that the infection potentiating effect of hemoglobin in vivo is due to such leukotoxins.