Comparison of the chemical structure and biological activities of the glycolipids of Salmonella minnesota R595 and Salmonella typhimurium SL1102

Review: Evidence against the hypothesis that antibodies to the inner core of lipopolysaccharides in antisera raised by immunization with enterobacterial deep-rough mutants confer broad-spectrum protection during Gram-negative bacterial sepsis

Antisera to rough enterobacterial mutants of chemotypes Ra, Rc, and Re have been reported to confer broad-spectrum protection against wild-type smooth strains. It has been hypothesized that binding and neutralization of lipopolysaccharides (LPS) by antibodies to common core epitopes underlies such protection. This review summarizes experiments by our laboratory and others that do not confirm this concept and proposes reasons for the divergent results. Studies indicating broad-spectrum protection by rough-mutant antisera often had defects in experimental design or methodology. These include the failure: (i) to use matched pre- and postimmune sera from the same donors to control for variable protective activity of normal sera; (ii) to exclude the role of natural and polyclonally stimulated antibodies with proven protective activity against the infecting bacterial strain (e.g. O-specific, capsular, Pseudomonas exotoxin A); (iii) to exclude protective effects of acute-phase serum factors; (iv) to exclude protective effects of endotoxin contamination after adsorption or fractionation of antibody preparations; (v) to use non-boiled bacteria and LPS not subjected to acid-hydrolysis or gel-fractionation, and to exclude nonspecific adsorption, to demonstrate physiologically meaningful binding of rough-mutant antibodies to smooth enterobacteria and their LPS.

Column liquid chromatography of endotoxins

A new, fast and highly reproducible column liquid chromatographic method was elaborated for the analysis and small-scale preparative isolation of endotoxin from Serratia marcescens Bizio (ATCC No. 264). This procedure detects contaminants of such preparations with high sensitivity and it is capable of separating them from endotoxic components. Extensive heterogeneity of both 5% trichloroacetic acid and phenol-water-extracted endotoxin preparations was recorded. Heterogeneity among the endotoxic components of purified preparations could also be detected by this method. Measurements of biological activities, such as Limulus amoebocyte lysate activation, lymphoblastogenesis (mitogenicity) and tumor necrosis factor (TNF) liberation were carried out on the chromatographically separated fractions. During these studies, non-toxic but in vitro TNF-generating components of crude endotoxin extracts were also detected.

Purification of specific precipitinogen and extraction of endotoxin from Haemophilus influenzae

After purifying a Haemophilus influenzae precipitinogen from endotoxic activity by means of ultracentrifugation, column chromatography (Sepharose 6B) and ion exchange chromatography (DEAE Sephadex A25) a fraction was obtained which still contained a specific precipitinogen that was virtually free of endotoxin. Furthermore, during the chromatographic procedures fractions with a high and a low molecular weight endotoxic activity were found. The limulus lysate test was more sensitive in the high molecular weight fractions and the LD50 in mice in the low molecular weight fractions with endotoxic activity.

Myeloperoxidase-Cl--H2O2 bactericidal system: effect of bacterial membrane structure and growth conditions

The human-myeloperoxidase-Cl--H2O2 bactericidal system killed (i) smooth Enterobacteria spp. greater than or equal to rough and (ii) static-grown, stationary-phase bacteria greater than or equal to aerated-grown, log-phase. This is in contrast to human neutrophil granule extracts (involved in nonoxidative bactericidal mechanisms) that kill rough and aerated log-phase Enterobacteria spp. much more efficiently.

Susceptibility of lipopolysaccharide mutants to the bactericidal action of human neutrophil lysosomal fractions

Acetate extracts of purified human neutrophil granules (a mixed population containing specific and azurophil granules) were dialyzed against phosphate-buffered saline (pH 7.0) and tested for bactericidal activity against smooth parent and rough mutant, gram-negative bacteria. Rough (Re) mutants of Escherichia coli, Salmonella typhimurium, and Salmonella minnesota were exquisitely more sensitive to extracts of human polymorphonuclear leukocyte granules than were their smooth (S) parents. The mean lethal dose (LD(50)) for the parent strains was 25 to 50 mug of granule protein per ml. As much as 500 mug of extract per ml failed to kill 100% of the S parents. The LD(50) for the rough mutants was 1.5 to 2.0 mug of the same granule extract per ml; 100% killing occurred with 5 to 10 mug of lysosomal protein per ml. Conditions affecting the growth of the bacteria greatly affected their sensitivity to the granule extracts. Granule extract killed bacteria grown with aeration to log phase 10 to 15 times more efficiently than the same bacteria grown to stationary phase under static conditions. The bactericidal incubation mixture also influenced results, in that greater killing occurred with tryptone than with phosphate or N-2-hydroxyethyl piperazine-N'-2-ethanesulfonic acid-buffered saline. Bactericidal activity depended on lysosomal protein concentration, time, and temperature. Boiled lysosomal fractions failed to kill the S parents but retained 20 to 50% of their ability to kill the Re mutants. Parents (smooth) were killed more efficiently at pH 5 to 6, whereas their Re mutants were killed more efficiently at pH 7 to 8.

Relationship of structure to function in bacterial endotoxins. IX. Differences in the lipid moiety of endotoxic glycolipids

Chemical, immunochemical, chromatographic, and endotoxic properties of five chromatographically pure glycolipids were compared. The preparations were extracted by chloroform-methanol from three Escherichia coli, one Salmonella minnesota, and one S. typhimurium Re heptoseless mutant strains. The local Shwartzman skin assay, the nonspecific resistance-enhancing effect, and the Limulus assays could not distinguish among the five glycolipids, all five being active in all three assays. Significant differences could be seen when the tumor resistance-enhancing effect of the glycolipids in mice was compared with the nonspecific TA3-Ha murine mammary adenocarcinoma growing in ascites form. Even greater variation was observed in the capacity of the preparations to enhance the nonspecific resistance of mice to virulent S. typhi 0901 infections. The data show that the five glycolipids are quite dissimilar in their biological effects. Similarly, thin-layer chromatography and molecular ratio determinations showed that differences exist in the chemical structure of the glycolipids. Accordingly, we claim that not only the polysaccharide but the lipid moiety as well may vary in various gram-negative endotoxin preparations.