Detection of human antibodies binding with smooth and rough LPSs from Proteus mirabilis O3 strains S1959, R110, R45.

Bacteria of the genus Proteus of the family Enterobacteriaceae are facultative human pathogens responsible mainly for urinary tract and wound infections, bacteremia and the development of rheumatoid arthritis (RA). We have analyzed and compared by ELISA the titer of antibodies in plasmas of healthy individuals and in sera of rheumatoid arthritis patients recognizing a potential host cross-reactive epitope (lysine-galacturonic acid epitopes) present in Proteus lipopolysaccharide (LPS). In our experiments LPSs isolated from two mutants of smooth Proteus mirabilis 1959 (O3), i.e. strains R110 and R45, were used. R110 (Ra type mutant) is lacking the O-specific polysaccharide, but possesses a complete core oligosaccharide, while R45 (Re type) has a reduced core oligosaccharide and contains two 3-deoxy-D-manno-oct-2-ulosonic acid residues and one of 4-amino-4-deoxy-L-arabinopyranose residues. Titer of P. mirabilis S1959 LPS-specific-antibodies increased with the age of blood donors. RA and blood donors' sera contained antibodies against S and Ra and Re type of P. mirabilis O3 LPSs. Antibodies recognizing lysine-galacturonic acid epitopes of O3 LPS were detected by ELISA in some plasmas of healthy individuals and sera of rheumatoid arthritis patients. RA patients antibodies reacting with P. mirabilis S1959 S and R LPSs may indicate a potential role of anti-LPS antibodies in molecular mimicry in RA diseases.

A total internal reflection ellipsometry and atomic force microscopy study of interactions between Proteus mirabilis lipopolysaccharides and antibodies.

Specific antigen-antibody interactions play a central role in the human immune system. The objective of this paper is to detect immune complexes using label-free detection techniques, that is, total internal reflection ellipsometry (TIRE) and atomic force microscopy (AFM)-based topography and recognition imaging. Interactions of purified rabbit immunoglobulin G (IgG) antibodies with bacterial endotoxins (Proteus mirabilis S1959 O3 lipopolysaccharides) were studied. Lipopolysaccharide was adsorbed on gold surface for TIRE. In the AFM imaging experiments, LPS was attachment to the PEG linker (AFM tip modification). The mica surface was covered by IgG. In TIRE, the optical parameters Ψ and Δ change when a complex is formed. It was found that even highly structured molecules, such as IgG antibodies (anti-O3 LPS rabbit serum), preserve their specific affinity to their antigens (LPS O3). LPS P. mirabilis O3 response of rabbit serum anti-O3 was also tested by topography and recognition imaging. Both TIRE and AFM techniques were recruited to check for possible detection of antigen-antibody recognition event. The presented data allow for determination of interactions between a variety of biomolecules. In future research, this technique has considerable potential for studying a wide range of antigen-antibody interactions and its use may be extended to other biomacromolecular systems.

Serotyping of Proteus mirabilis clinical strains based on lipopolysaccharide O-polysaccharide and core oligosaccharide structures.

The aim of this work was to serotype Proteus mirabilis urinary tract infection (UTI) strains based on chemically defined O-antigens with the use of two clinical collections from Sweden and Poland consisting of 99 and 24 UTI strains, respectively. A simple two-step serotyping scheme was proposed using enzyme immunoassay with heat-stable surface antigens of Proteus cells and immunoblotting with isolated lipopolysaccharides (LPSs). Using polyclonal anti-P. mirabilis rabbit antisera, 50 Swedish and 8 Polish strains were classified into serogroups O10, O38, O36, O30, O17, O23, O9, O40, O49, O27, O5, O13, O24, O14, and O33. From the Swedish strains, 10 belonged to serogroup O10 and five to each of serogroups O38, O36, and O9. Therefore, none of the O-serogroups was predominant. The majority of the serotyped clinical strains possess acidic O-antigens containing uronic acids and various acidic non-carbohydrate substituents. In immunoblotting, antisera cross-reacted with both O-antigen and core of LPSs. The core region of 19 LPSs bound a single serum, and that of 12 LPSs bound more than two sera. Following bioinformatic analysis of the available sequences, a molecular approach to the prediction of Proteus core oligosaccharide structures was proposed. The identification of the core type of P. mirabilis R110, derived from a serogroup O3 wild strain, using restriction fragments length polymorphism analysis of galacturonic acid transferase is shown as an example. In summary, the most frequent O-serogroups among P. mirabilis UTI stains were identified. The diversity of serological reactions of LPSs is useful for serotyping of P. mirabilis clinical isolates. A possible role of the acidic components of O-antigens in UTI is discussed.

Comparative study of electrokinetic potentials and binding affinity of lipopolysaccharides-chitosan complexes.

Electrokinetic properties of complexes of chitosan (Ch) with lipopolysaccharides (LPSs) from Escherichia coli O55:B5, Yersinia pseudotuberculosis 1B 598, and Proteus vulgaris O25 (48/57) and their size distribution were investigated using zeta-potential distribution assay and quasi-elastic light scattering. The interaction of LPS from different microorganisms with chitosan at the same w/w ratio of components (1:1) resulted in the formation of complexes in which the negative charge of LPS was neutralized (LPS from E. coli) or overcompensated (Y. pseudotuberculosis and P. vulgaris). The changing in size of the endotoxin aggregates during binding with chitosan was observed. The binding constants of chitosan with LPSs were determined by a method with using the anionic dye Orange II. The LPS from E. coli possess higher affinity to chitosan in comparison with the two others samples of endotoxin.

Activation of complement by human hemoglobin and by mixtures of hemoglobin and bacterial endotoxin.

Purified human hemoglobin is being developed as an alternative to transfusions of homologous erythrocytes. However, toxicity associated with infusion of hemoglobin has limited the development of this resuscitation fluid. Some observed toxicities, including activation of the complement cascade, have been associated with contamination of hemoglobin solutions by bacterial endotoxin. Recent studies have demonstrated complex formation between hemoglobin and endotoxin, and have documented a resultant increase in the ability of endotoxin to activate coagulation, stimulate tissue factor production by human peripheral blood mononuclear cells, and stimulate tissue factor activity and protein synthesis in cultured human endothelial cells. The process of hemoglobin enhancement of endotoxin toxicity suggests a possible mechanism by which the consequences of endotoxin contamination of hemoglobin solutions, including complement activation, could be magnified. Therefore, we studied the potential of hemoglobin to either fix complement directly, or modify the ability of endotoxin to fix complement. Human crosslinked and native hemoglobins, at concentrations between 0.2 mg/ml and 3 mg/ml, were shown to fix complement. Complement fixation by hemoglobin was identical in normal human serum or in factor B-depleted serum, suggesting that fixation occurred via the classical pathway of complement activation. Complement fixation then was examined with a battery of smooth and rough endotoxins tested in the absence and presence of hemoglobin. Addition of hemoglobin to a solution of a rough Salmonella endotoxin partial structure, from which a single fatty acid had been hydrolyzed from the lipid A portion of the macromolecule, resulted in decreased efficiency of complement fixation. However, addition of hemoglobin had little or no effect on the intrinsic complement fixing abilities of eight other smooth endotoxins, rough endotoxins, or endotoxin partial structures. Our results demonstrated the ability of hemoglobin to fix complement at hemoglobin concentrations which would be achieved during infusion for resuscitation, but failed to demonstrate a reproducible effect of hemoglobin on the activation of complement by endotoxin.

Complement activation by Proteus mirabilis negatively charged lipopolysaccharides.

Proteus mirabilis strains are human pathogens responsible for urinary tract infections and bacteremias and may be involved in rheumatoid arthritis. Lipopolysaccharide (LPS, bacterial endotoxin), the major component of the cell wall, is one of the virulence factors of Proteus. In the presented studies, we have investigated complement activation by LPSs isolated from P. mirabilis O10, O23, O30, and O43 strains, which differ in the number of negative COO- groups on their polysaccharide components. Four P. mirabilis strains studied were sensitive to complement-mediated killing, despite complement binding by their LPSs. The optimal complement binding by LPSs was detected in serum with functional assays for both the classical and alternative pathways. Complement activation in 80% serum by the smooth, uronic acid, and hexosamine containing P. mirabilis LPSs was not critically determined by the structure of their O-chain polysaccharides. One of four LPSs used as a model, P. mirabilis O10 LPS, fragmented C3 in an LPS dose- and time-dependent manner. It was detected by crossed-immunoelectrophoresis and capture ELISA with anti-C3c antibodies. The lower complement activation by 023 LPS correlates with its reduced C3 fragmentation, compared with three other Proteus LPSs studied. Rabbit anti-O antibodies enhanced the complement binding and factor C3 fragmentation by O10, O23, O30, and O43 P. mirabilis LPSs.

Structures of the O-antigens of Proteus bacilli belonging to OX group (serogroups O1-O3) used in Weil-Felix test.

Structures of the O-specific polysaccharide chains of lipopolysaccharides from Proteus group OX strains (serogroups O1-O3) used as antigens in Weil-Felix test for diagnosis of rickettsiosis, were established. From them, the acid-labile polysaccharide of Proteus vulgaris OX19 (O1) is built up of the following branched pentasaccharide repeating units connected via a phosphate group: [structure in text] where QuiNAc stands for 2-acetamido-2,6-dideoxyglucose (N-acetylquinovosamine). The basis of serospecificity of the Proteus group OX antigens and their cross-reactivity with human anti-rickettsial antibodies is discussed.

Structural and serological studies of the O-specific polysaccharide of the bacterium Proteus mirabilis O10 containing L-altruronic acid, a new component of O-antigens.

An acidic O-specific polysaccharide from the lipopolysaccharide of Proteus mirabilis O10 contains 2-acetamido-2-deoxy-D-glucose, 2-acetamido-2-deoxy-D-galactose, D-galacturonic acid, and L-altruronic acid, the last-named sugar having not been found hitherto in O-antigens. Structure of a branched tetrasaccharide repeating unit of the polysaccharide was established by 1H and 13C NMR spectroscopy, including two-dimensional COSY and rotating-frame NOE spectroscopy. The lateral L-altruronic acid residue plays the immunodominant role in manifestation of the O10 specificity of Proteus, whereas a disaccharide fragment of the main chain in common with the O-specific polysaccharide of P. mirabilis O43 provides the one-way serological cross-reactivity between anti-O10 serum and O43-antigen.

Structures of new acidic O-specific polysaccharides of the bacterium Proteus mirabilis serogroups O26 and O30.

The polysaccharide chains of the lipopolysaccharides of the Proteus mirabilis serogroups O26 and O30 were studied using sugar and methylation analysis and 1H and 13C NMR spectroscopy, including two-dimensional correlation spectroscopy and rotating-frame NOE spectroscopy. The polysaccharides were found to be acidic due to the presence of D-galacturonic acid and its amide with L-lysine in serogroup O26 or D-glucuronic acid in serogroup O30, and the structures of their tetrasaccharide repeating units were established. The O26-specific polysaccharide is structurally and serologically related to the O-specific polysaccharide of P. mirabilis O28, which includes amides of D-GalA with L-lysine and L-serine [Radziejewska-Lebrecht, J. et al. (1995) Eur. J. Biochem. 230, 705-712].

Structures of the O-specific polysaccharides and a serological cross-reactivity of the lipopolysaccharides of Proteus mirabilis O24 and O29.

Strains of Proteus mirabilis belonging to serogroups O24 and O29 are frequent in clinical specimens. Anti-P. mirabilis O24 serum cross-reacted with the lipopolysaccharide (LPS) of P. mirabilis O29 and vice versa. The structures of the O-specific polysaccharides (OPSs, O-antigens) of both LPSs were established using sugar analysis and one- and two-dimensional 1H- and 13C-NMR spectroscopy and found to be different. SDS-PAGE and Western immunoblotting suggested that the serological cross-reactivity of the LPSs is due to a common epitope(s) on the core-lipid A moiety, rather than on the OPS. Therefore, the epitope specificity and the structures of the O-antigens studied are unique among Proteus serogroups.

The generation of superoxide anion in blood platelets in response to different forms of Proteus mirabilis lipopolysaccharide: effects of staurosporin, wortmannin, and indomethacin.

Lipopolysaccharide (LPS), the major component of the outer membrane of Gram-negative bacteria may activate blood platelets. The aim of our study was to evaluate the effects of different forms of Proteus mirabilis LPS and isolated lipid A and polysaccharide part on the production of superoxide radicals in blood platelets and to estimate the role staurosporin, wortmannin and indomethacin on this process. We compared the generation of superoxide radicals in platelets treated with LPS after preincubation with inhibitors of the signal transduction pathways, namely staurosporin (inhibitor of protein kinase C), wortmannin (inhibitor of phosphoinositide 3-kinase), and indomethacin (inhibitor of cycloxygenase). Our results demonstrate that all LPS molecules and their fragments caused a stimulation of O2- generation in platelets (P<.5). LPSS1959 had the strongest stimulatory effect. Straurosporin and wortmannin, but not indomethacin inhibited O2- production in LPS-stimulated platelets. Staurosporin (8 nM) and wortmannin (50 nM) caused about 50% inhibition of thrombin-induced O2- generation in platelets, while indomethacin (10 microM) had only a slight inhibitory effect on this process. Our results provide support that in LPS- and thrombin-activated platelets, at least part of O2- generation in platelets, while indomethacin (10 microM) had only a slight inhibitory effect on this process. Our results provide support that in LPS- and thrombin-activated platelets, at least part of O2- is generated due to the activation of the enzymes (protein kinase C and phosphoinositide 3-kinase) involved in signal transduction pathway. Cycloxygenase seems to be not involved in this process.

Lipopolysaccharide phage receptor of Proteus mirabilis 1959 strain. Site of phage-mediated hydrolysis in O-specific polysaccharide.

The oligosaccharides isolated as products of phage "otto" provoked degradation of LPS provided by the host strain P. mirabilis 1959 were investigated. The site of phage mediated hydrolysis was shown to be the linkage 1,4(2) between galactosamine and glucuronic acid in the main chain of O-specific polysaccharide. The pentasaccharide (with one lysine residue) represents one phage cleaved repeating unit in the O-specific polysaccharide P. mirabilis 1959.

Inhibition of mouse liver cytochrome P-450 by gram-negative bacteria lipopolysaccharides.

The ability of bacterial endotoxins, of different origin, to modify the level of mouse liver cytochrome P-450 was investigated. Endotoxins, (lipopolysaccharides, LPSs) were isolated from Proteus, Escherichia, Salmonella, Bacteroides and Coxiella strains. The most potent inhibitor of cytochrome P-450 activity was S. typhi 0101 LPS, which at a dose of 1 microgram/mouse reduced the cytochrome P-450 activity to 59%. E. coli O55:B5, S. typhimurium, P. mirabilis O3, and C. burnetii LPSs, at dose 10 micrograms/mouse, decrease cytochrome P-450 level from 56 to 69%. B. ovatus LPS significantly suppressed the expression of cytochrome P-450 only at the highest dose used-100 micrograms/mouse. The comparison of inhibitory activity of P. mirabilis complete, S and R types of LPSs indicate that lipid A portion of LPSs are sufficient to decrease the cytochrome P-450 level. However, the core oligosaccharide of LPS significantly enhance that inhibition. The isolated O-specific polysaccharide part of P. mirabilis O3 LPS did not decrease cytochrome P-450 level. The comparison of biological activity of Proteus LPSs, tested by chromogenic Limulus amebocyte lysate (LAL) test, demonstrated the enhancement effect of O-polysaccharide part of tested LPSs.

Structure of the 0-specific polysaccharide of Proteus mirabilis S 1959.

The complete structure of the 0-specific polysaccharide of the strain Proteus mirabilis S 1959, as analyzed by 13C NMR, is presented. Some data demonstrating the significant heterogeneity of the 0-specific chain in the investigated lipopolysaccharide are also described.

Comparison of serological reactions of Rickettsiae-infected patients and rabbit anti-Proteus OX antibodies with Proteus OX2, OX19 and OXK lipopolysaccharides.

The reactivity of anti-Rickettsiae human antibodies with Proteus OX cells is used as a presumptive rickettsial diseases diagnostic test Weil-Felix reaction. In presented studies we compare the reactivity of human anti-Rickettsiae and rabbit anti-Proteus antibodies with series of Proteus OX2, OX19 and OXK lipopolysaccharides (LPS). Polyclonal rabbit anti-OX2, anti-OX19 and anti-OXK sera reacted only with the homologous LPS--OX2, OX19 and OXK, respectively. The antibodies of Japanese spotted fever patients were less specific and reacted with OX2 as well as OX19 LPS. The antibodies of scrub typhus patients recognized Proteus OXK LPS, only. The serological reactions of O-antigen of P. mirabilis S1959 indicated that this, previously serologically not classified, strain may belong to the OXK group. Bacteria, used in the studies, came from the American, Japanese and European strain collections. The series of OX2, OX19 and OXK LPS, isolated from these Proteus strains, presented pattern of electrophoretic mobility and serological reactivities specific for each of OX-group types.

Investigation of influence of Proteus mirabilis LPS polysaccharide part on the murine immune cells activation.

The biological activities were investigated of Proteus mirabilis lipopolysaccharides (LPSs) and their fragments, namely, the induction of nitric oxide (NO) and interleukin 1 (IL-1) synthesis by murine macrophages and the proliferation of murine spleen cells. The O-specific polysaccharide (F1), core oligosaccharide (F2) and lipid A were as effective as intact LPS in stimulating murine macrophages to produce NO. IL-1 synthesis was also induced by all studied types of endotoxins (S, Ra, Re) and partial structures, however F1, F2 and lipid A fractions required the presence of serum. In contrast to LPS, the O-specific polysaccharide, core oligosaccharide and lipid A were not able to induce the blast response of murine non-adherent splenocytes.

Galacturonic acid as the terminal constituent in the R core polysaccharide of proteus R110 (Ra) mutant.

Terminal non-reducing position of GalUA residue in the R core region of P. mirabilis R110 (Ra)mutant was established by GLC/MS analysis of methylated degraded polysaccharide. This position of GalUA is exceptional as compared with the terminal constituents present in the known structures of R core polysaccharides in Enterobacteriaceae. As it was shown in serological study, GalUA does not play a role of immunodominant in the examined Proteus R core region.

Structure and serological characterization of an Nepsilon-[(R)-1-carboxyethyl]-L-lysine-containing O-chain of the lipopolysaccharide of Proteus mirabilis O13.

In this paper we present the structure and describe serological properties of the O-specific polysaccharide of Proteus mirabilis O13 lipopolysaccharide, which contains a unique component: an amide of D-galacturonic acid (D-GalA) with an unusual amino acid, Nepsilon-[(R)-1-carboxyethyl]-L-lysine (alaninolysine, AlaLys). Selective chemical degradations of either GalA or AlaLys resulted in the loss of the serological reactivity of the polysaccharide with anti-O serum against P. mirabilis O13. Neither synthetic stereoisomers of AlaLys nor the isolated amide of GalA with AlaLys inhibited the reaction of the O-antiserum with the homologous lipopolysaccharide. The O-antiserum did not cross-react with the lipopolysaccharide of Providencia alcalifaciens O23 containing an amide of D-glucuronic acid with AlaLys. These data showed that both uronic acid and amino acid components of the amide play an important role in manifesting the P. mirabilis O13-specificity, but the full specific epitope also includes another O-specific polysaccharide component(s). A cross-reactivity of anti-O13 serum with some other P. mirabilis strains was observed and attributed to a common heat-stable antigen(s) different from the lipopolysaccharide.

Serum IgG antibodies in children and adults reacting with Helicobacter pylori lipopolysaccharides.

The presence of IgG antibodies reacting with Helicobacter pylori lipopolysaccharides (LPSs) in sera from children and adults diagnosed as H. pylori-infected, as well as healthy persons, was tested. There was no correlation between the production of antibodies reacting with H. pylori surface proteins and LPSs. Also no correlation between reactivity of tested sera with H. pylori antigens and deep rough mutant (Re types) enterobacterial LPSs was revealed. The prevalence of anti-LPS IgG in randomly selected children was relatively high.

The effect of removal of D-fructose on the antigenicity of the lipopolysaccharide from a rough mutant of Vibrio cholerae Ogawa.

The lipopolysaccharide (LPS) of a rough mutant (95R) of Vibrio cholerae Ogawa has been investigated chemically and serologically. D-Fructose was released from LPS under conditions (10mM trifluoroacetic acid, 60 degrees, 1 h) that liberated no other sugar constituent of the LPS (2-amino-2-deoxy-D-glucose, D-glucose, L-glycero-D-manno-heptose). Upon periodate oxidation, D-fructose and D-glucose were oxidised quantitatively, whereas approximately 50% of heptose was periodate-resistant. The data indicate that D-fructose does not link the polysaccharide and lipid A portion as proposed earlier, and suggest that D-fructose is present as a branch. By passive hemolysis inhibition, it was shown that the release of D-fructose paralleled the exposure of an antigenic determinant cryptic in LPS.