Search

Several rough strains of Escherichia coli, Salmonella enterica and Proteus mirabilis were cultivated in the presence of (14C)acetate, which incorporated into their phospholipids and lipopolysaccharide. Phospholipids were removed from the cells with ethanol extraction. However, as shown by thin layer chromatography, methanol additionally extracted remarkable quantities of lipopolysaccharide from deep rough strains, but not from bacteria containing complete or nearly complete core.

Adjuvant activities of isogenic Salmonella enterica, serovar Typhimurium, O-6,7 and O-4,5,12 lipopolysaccharide (LPS), lipid A and Bordetella pertussis LPS were compared by immunizing groups of mice subcutaneously with diphtheria and tetanus toxoid vaccine alone or mixed with one of the LPS derivatives. Five weeks later the mice were bled and the tetanus and diphtheria antibodies in the sera were measured. All the LPS derivatives efficiently increased the antibody responses when compared to the vaccine alone, but the mannose-rich O-6,7 LPS and lipid A were significantly more potent than O-4,5,12 LPS and B. pertussis LPS. We conclude that the quality of the O antigen influences the adjuvant activity of LPS.

The group-specific antigen of Chlamydia trachomatis serotype L2 was chemically analyzed. It is composed of typical lipopolysaccharide (LPS) components, i.e., D-glucosamine, long-chain 3-hydroxy fatty acids, 2-keto-3-deoxyoctonic acid, and phosphate in a molar ratio of approximately 2:5:3:2.6, respectively, resembling enterobacterial LPS of the Re chemotype. For the first time, 3-hydroxydocosanoic acid (3-OH C22:0) was found as an LPS constituent.

We used crossed immunoelectrophoresis to study detergent-solubilized components of the outer membrane of Salmonella typhimurium under nondenaturing conditions. The antisera used were raised against nondenatured outer membrane preparations. Lipopolysaccharide and lipoprotein were identified easily as discrete precipitates when they were solubilized with Triton X-100. However, solubilization of the porins with Triton X-100 resulted in a complex precipitate pattern, indicating incomplete dissociation of protein-protein interactions. A clear-cut pattern was obtained when the porins were first solubilized and denatured with hot sodium dodecyl sulfate, followed by removal of the sodium dodecyl sulfate and renaturation in the presence of Triton X-100. Our findings suggested that crossed immunoelectrophoresis can be used to study the antigenicity of nondenatured porins and the antibody responses to them.

The antigenic properties of the lipopolysaccharide (LPS) of Chlamydia trachomatis L2 were investigated. By means of passive hemolysis, passive hemolysis inhibition, and absorption experiments, it was shown that antiserum raised against chlamydial elementary bodies contained at least two different antibody specificities which reacted with different antigenic determinants of chlamydial LPS. One of these antibodies cross-reacted with enterobacterial Re LPS, recognizing a structure which is shared by both LPSs, whereas the reactivity of the second antibody was restricted to chlamydial LPS. The former antibody could be absorbed with Salmonella minnesota Re LPS, whereas the latter was not affected by this absorption. Therefore, chlamydial LPS possesses two distinct antigenic determinants, one of which is C. trachomatis specific, the other of which is responsible for the cross-reactivity with enterobacterial Re-type LPS. Both antigenic determinants were destroyed during mild acid-catalyzed hydrolysis. It was further shown that free chlamydial lipid A exhibits antigenicity that cross-reacts with free enterobacterial lipid A. This antigenicity, however, as in enterobacterial LPS, is present in a cryptic form, i.e., it is unmasked only after acid hydrolysis of LPS.

Peripheral blood lymphocytes were incubated with glutaraldehyde-fixed Salmonella bacteria. This resulted in rapid activation of nonspecific cytotoxic potential of the lymphocytes. Both originally noncytotoxic, high-density Percoll-fractionated cells, and cytotoxic natural killer (NK) cell-enriched low-density cells were activated. The induction of originally noncytotoxic cells into activated killer (AK) cells was apparently independent of interferon (IFN), whereas the activation of the NK cell-enriched fractions also involved IFN production. Neither the AK nor NK activity were associated with significant bactericidal activity. The IFN-independent induction of AK activity was not dependent on the O-antigenic polysaccharide part of the lipopolysaccharide (LPS) on the bacterial cell surface, because both smooth (S) strains with differing O-antigenic structures (S-4,12 and S-6,7) and a rough (Re) strain without O-antigen were effective inducers. Isolated LPS, and especially alkali-hydrolyzed (O-deacylated, detoxified) LPS (ALPS) interfered with the induction of cytotoxicity. At concentrations of 10 to 30 micrograms/ml, ALPS totally inhibited the induction of AK activity without affecting the endogenous NK activity. Thus contact with bacteria can lead to the emergence of AK cells, and a bacterial product can effectively block this activation. These phenomena stress the complexity of interactions with host defenses that can take place during bacterial infection.

The lipopolysaccharides (LPS) of Chlamydia trachomatis, Acinetobacter calcoaceticus var. anitratus, and Re mutants of Salmonella sp. were shown to share related immunodeterminants , as demonstrated by double immunodiffusion and immunoblotting from sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels. The cross-reactive material in the extracellular slime of A. calcoaceticus var. anitratus was shown to be released LPS. The Acinetobacter LPS was found to separate in sodium dodecyl sulfate-polyacrylamide gel electrophoresis into three fractions. The cross-reactive component was the fraction migrating fastest, at a rate identical to Re-type LPS of Salmonella sp. The Acinetobacter LPS could be used as antigen in complement fixation assays performed on paired sera of patients with chlamydial pneumonia; it gave results identical to those of the chlamydial complement fixation glycolipid antigen conventionally used in such assays in 9 of 10 patients.

Exposure of human peripheral blood mononuclear (MN) cells to deesterified (alkali-treated) lipopolysaccharide (LPS-OH) and then to 51Cr rendered the cells susceptible to 51Cr release in the presence of specific antibody and complement. The assay was optimized by using rough (Rb2 or Re) LPS. 51Cr release did not occur from cells preexposed to untreated or electrodialyzed LPS. Studies of isolated monocytes and lymphocytes revealed that the majority of the 51Cr released was derived from monocytes. The optimum concentration of LPS-OH was 10 micrograms/ml. Antiyersinia agglutinin-positive serum, but not a negative serum, obtained from patients with reactive yersinia arthritis caused 51Cr release from MN cells pretreated with yersinia LPS-OH. This implies that during yersinia infection antibodies are generated that can attack the cell membrane--LPS-OH complex. We conclude that the method provides a tool to demonstrate binding of LPS to MN cells in a manner that leads to cell injury in an immune host.

A high degree of non-specific resistance to a lethal systemic Escherichia coli infection was induced in mice by pretreatment with a small dose (less than 5 micrograms/mouse) of the homologous lipopolysaccharide (LPS) or with heterologous rough-type LPS from E. coli K-12. The route of LPS administration, intraperitoneally or subcutaneously, did not influence the development of resistance, suggesting that a systemic cell activation was responsible for the effect. The enhanced elimination of bacteria was similar to that in mice recovering from a sublethal E. coli infection. In the LPS-treated mice, elimination of the challenge bacteria from the peritoneal cavity and the blood started 3-4 h after challenge whereas, in controls, the bacterial numbers continued to increase until the mice died. The detoxified LPS derivative, monophosphoryl lipid A (MPL), also increased the survival of mice infected with E. coli O18:K1. However, the dose of MPL required for optimal infection resistance was 100-fold greater than that of native, E. coli K-12 LPS, corresponding to the 100-fold reduced toxicity of MPL for mice and rabbits in lethality and pyrogenicity assays.

A strong immunological cross-reaction between a major glycolipid antigen of Chlamydia and the innermost (Re) core of the lipopolysaccharide of enteric bacteria was demonstrated with the aid of mutants in which the Re structure is exposed. The chlamydial glycolipid resembled the Re lipopolysaccharide in molecular size, solubility, and endotoxic properties and may thus be functionally equivalent to lipopolysaccharide, an essential and characteristic component of the outer membrane of Gram-negative bacteria.

A crude major outer membrane (porin) preparation, obtained from a rough strain of Salmonella typhimurium earlier shown to be protective both in active and passive immunization of mice against challenge with smooth S. typhimurium, was further purified. Removal of the main impurities, lipopolysaccharide (LPS) and lipoprotein, was accompanied by loss of protective capacity in passive immunization experiments. Reconstitution with rough LPS restored the protective capacity. Protection was, however, concluded not to be due to anti-LPS, because a large fraction of the anti-LPS antibodies could be removed from the protective rabbit antiserum with an LPS immunosorbent without loss of protection.

The B lymphocyte mitogens, bacterial lipopolysaccharide (LPS), pokeweed lectin, and tuberculin, induced proliferation in density-inhibited monolayer cultures of chick embryo fibroblasts. The stimulation was seen both as an early increase in sugar uptake and cell volume and later as an increase in thymidine incorporation and cell number. The concentration of LPS maximally stimulating fibroblasts was remarkably low, about 0.1-1 ng/ml. LPS preparation with very different sugar chains gave quite similar results indicating that the architecture of the hydrophilic carbohydrate part is not critical for the mitogenic effect.