No abstract

We have developed an ELISA for IgM and IgG antibodies to the core glycolipid (CGL) of the Re mutant Salmonella minnesota R 595, and to lipid A. Anti-CGL antibodies have been detected in sera from 37% of healthy blood donors, whereas anti-lipid A activities were found in 13% of individuals only. The anti-CGL and anti-lipid A activities were examined in patients in a surgical intensive care unit, selected on the basis of a definite risk of infectious complications due to Gram-negative bacteria. Of the patients who developed such infections, the rate of favourable outcome was significantly higher in patients with either stable positive or increasing anti-CGL activities than in patients found to be negative. Our results provide clear evidence that anti-CGL antibodies contribute to host defence against various Gram-negative bacteria.

Mice were passively immunized with sera from blood donors active for rough lipopolysaccharides (LPS), the J5 (Rc chemotype) mutant of Escherichia coli O111:B4, and the Re595 (Re chemotype) mutant of Salmonella minnesota. All protected the mice against lethal challenge with smooth E. coli WF96 LPS, E. coli and Salmonella rough mutant LPS, or free lipid A. Epitopes recognized by monoclonal antibodies (MAbs) reacting with the LPS of S. minnesota Re595 or lipid A were localized in the 2-keto-3-deoxy-D-manno-octulosonic acid (KDO) region and on lipid A. Core-reactive MAbs reacted with their homologous Re LPS and with free lipid A. One, GL11, cross-reacted with the KDO alone. MAbs GL6, GL11, L.4, L.6, and L.8 protected the actinomycin D-sensitized mice against the lethal effects of LPS from E. coli WF96, Salmonella enteritidis, E. coli J5, S. minnesota Re595, and free lipid A. The GL11 antibody was also protective when injected after LPS challenge. These results indicate that antibodies directed against the core glycolipid of S. minnesota Re595 LPS may be useful as an additive form of therapy that may enable decreased mortality during gram-negative bacterial sepsis.

Free endotoxin was assayed in filtered samples of E. coli suspensions submitted to the bactericidal and bacteriolytic action of 10% human serum. The Limulus amoebocyte lysate test, a passive hemolysis inhibition assay based on O antigenic specificity and the determination of 3-OH-myristic acid by mass spectrometry were used as assay methods differing from one another with regard to the part of the endotoxin macromolecule involved in the reaction. The biological activity of endotoxin was assessed in a mouse lethality test. The bactericidal and bacteriolytic action of human serum on sensitive strains of E. coli released quantities of endotoxic lipopolysaccharide (LPS) amounting to 3,000-12,000 ng/ml, for an inoculum of 1--3 x 10(8) colony-forming units. The material thus appearing in the medium was shown to react with the Limulus amoebocyte lysate, to be lethal for actinomycin D-sensitized mice and to bear O antigen, as well as 3-OH-myristic acid, a marker of lipid A. Samples of serum depleted of lysozyme by adsorption onto bentonite, and displaying a strictly bactericidal effect, released approximately 80% of the quantity of LPS appearing in the medium in a control experiment performed with untreated serum. The LPS release is therefore mainly linked to the bactericidal effect of antibody and complement. The amount of LPS released depended on the concentration of divalent cations in the medium, being reduced by an increase in the concentration of calcium and magnesium beyond the values optimal for complement activity. This effect was already observed for an increase in the concentration of divalent cations too low to alter the bactericidal or bacteriolytic effects. The significance of the release of endotoxin by complement dependent bactericidal reactions occurring in vivo is discussed.