Variability in LPS composition, antigenicity and reactogenicity of phase variants of Bordetella pertussis

Comparison of lipopolysaccharides (LPS) from phase variants of different strains of Bordetella phase variants of different strains of Bordetella pertussis has shown a difference in their composition, antigenicity and reactogenicity. Phase I variants of B. pertussis, with the exception of strain 134, contain a preponderance of LPS I whereas the major component of LPS of phase IV variants is LPS II. Sera raised to LPSs of phase I strains, other than 134, cross-react with each other but not with phase IV LPSs; and similarly all sera raised to phase IV LPSs cross-react with each other and with LPS from 134 phase I. The LPSs of all phase I variants, including that of 134, are approximately ten-fold or more reactive in the limulus amoebocyte lysate assay (LAL) than phase IV LPSs. In the human mononuclear cell pyrogen assay phase IV LPSs also stimulated a lower response than phase I LPSs. The B. pertussis phase I LPSs are 10-times more reactive than Escherichia coli standard endotoxin in the LAL assay but 100-times less reactive than E. coli LPS in the monocyte test for pyrogen. The SDS-PAGE profiles of B. pertussis LPSs are quite different from those of B. parapertussis and B. bronchiseptica strains. B. pertussis LPSs produced a typical lipo-oligosaccharide (LOS) pattern. B. bronchiseptica LPS produced a similar pattern but was antigenically distinct from B. pertussis LPSs I and II. B. parapertussis in contrast produced a ladder pattern typical of smooth type LPS.     

Structural analysis of oligosaccharides from lipopolysaccharide (LPS) of Escherichia coli K12 strain W3100 reveals a link between inner and outer core LPS biosynthesis

Lipopolysaccharide (LPS) from Escherichia coli K12 W3100 is known to contain several glycoforms, and the basic structure has been investigated previously by methylation analyses (Holst, O. (1999) in Endotoxin in Health and Disease (Brade, H., Opal, S. M., Vogel, S. N., and Morrison, D., eds) pp. 115-154; Marcel Dekker, Inc., New York). In order to reveal dependences of gene activity and LPS structure, we have now determined the composition of de-O-acylated LPS by electrospray ionization-Fourier transform ion cyclotron-mass spectrometry (ESI-FT-MS) and identified 11 different LPS molecules. We have isolated the major glycoforms after de-O- and de-N-acylation and obtained four oligosaccharides that differed in their carbohydrate structure and phosphate substitution. The main oligosaccharide accounted for approximately 70% of the total and had a molecular mass of 2516 Da according to ESI-FT-MS. The dodecasaccharide structure (glycoform I) as determined by NMR was consistent with MS and compositional analysis. One minor oligosaccharide (5%) of the same carbohydrate structure did not contain the 4'-phosphate of the lipid A. Two oligosaccharides contained the same phosphate substitution but differed in their carbohydrate structure, one (5%) which contained an additional beta-D-GlcN in 1-->7 linkage on a terminal heptose residue (glycoform II) which was N-acetylated in LPS. A minor amount of a molecule lacking the terminal L-alpha-D-Hep in the outer core but otherwise identical to the major oligosaccharide (glycoform III) could only be identified by ESI-FT-MS of the de-O-acylated LPS. The other oligosaccharide (20%) contained an alpha-Kdo-(2-->4)-[alpha-l-Rha-(1-->5)]-alpha-Kdo-(2-->4)-alpha-Kdo branched tetrasaccharide connected to the lipid A (glycoform IV). This novel inner core structure was accompanied by a truncation of the outer core in which the terminal disaccharide L-alpha-D-Hep-(1-->6)-alpha-D-Glc was missing. The latter structure was identified for the first time in LPS and revealed that changes in the inner core structure may be accompanied by structural changes in the outer core.     

Mutation in the LPS outer core biosynthesis gene, galU, affects LPS interaction with the RTX toxins ApxI and ApxII and cytolytic activity of Actinobacillus pleuropneumoniae serotype 1

Lipopolysaccharides (LPS) and Apx toxins are major virulence factors of Actinobacillus pleuropneumoniae, a pathogen of the respiratory tract of pigs. Here, we evaluated the effect of LPS core truncation in haemolytic and cytotoxic activities of this microorganism. We previously generated a highly attenuated galU mutant of A. pleuropneumoniae serotype 1 that has an LPS molecule lacking the GalNAc-Gal II-Gal I outer core residues. Our results demonstrate that this mutant exhibits wild-type haemolytic activity but is significantly less cytotoxic to porcine alveolar macrophages. However, no differences were found in gene expression and secretion of the haemolytic and cytotoxic toxins ApxI and ApxII, both secreted by A. pleuropneumoniae serotype 1. This suggests that the outer core truncation mediated by the galU mutation affects the toxins in their cytotoxic activities. Using both ELISA and surface plasmon resonance binding assays, we demonstrate a novel interaction between LPS and the ApxI and ApxII toxins via the core oligosaccharide. Our results indicate that the GalNAc-Gal II-Gal I trisaccharide of the outer core is fundamental to mediating LPS/Apx interactions. The present study suggests that a lack of binding between LPS and ApxI/II affects the cytotoxicity and virulence of A. pleuropneumoniae.     

Features of the monosaccharide composition and ultrastructure of lipopolysaccharides of different forms of Shigella sonnei]

The authors present the results of study of chemical monosaccharide composition and the ultrastructure of purified lipopolysaccharides (LPS) of the I and the II phases and the R-form of Sh. sonnei. The amount of lipids in LPS preparations increased with the change from S- into R-form. Galactose content in LPS of the II phase was less than in LPS of the I phase, and it was absent entirely in LPS of R-form. It was demonstrated by negative contrasting that LPS dissociation increased with S leads to R dissociation. A marked similarity was found between macromolecular aggregates of LPS of the II phase and of R-form.     

Plasma lipopolysaccharide (LPS)-binding protein. A key component in macrophage recognition of gram-negative LPS.

LPS-binding protein (LBP) binds with high affinity (Kd approximately equal to 10(-9) M) to lipid A of LPS isolated from rough (R)- or smooth (S)-form Gram-negative bacteria as well as to lipid A partial structures such as precursor IVA. To define the role of LBP in regulating responses to LPS we have examined TNF release in rabbit peritoneal exudate macrophages (M phi) stimulated with LPS or with complete or partial lipid A preparations in the presence or absence of LBP. In the presence of LBP, M phi showed increased sensitivity to S- and R-form LPS as well as synthetic lipid A. Compared with LPS or lipid A, up to 1000-fold greater concentrations of partial lipid A structures were required to induce TNF production. However, consistent with our previous observations that these structures bind to LBP, TNF production was increased in the presence of LBP. In contrast, LBP did not enhance or inhibit TNF production produced by heat-killed Staphylococcus aureus, peptidoglycan isolated from S. aureus cell walls, or PMA. Potentiated M phi responsiveness to LPS was observed with as little as 1 ng LBP/ml. Heat-denatured LBP (which no longer binds LPS), BPI (an homologous LPS-binding protein isolated from neutrophils), or other serum proteins were without effect. LBP-treated M phi also showed a more rapid induction of cytokine mRNA (TNF and IL-1 beta), higher steady-state mRNA levels and increased TNF mRNA stability. These data provide additional evidence that LBP is part of a highly specific recognition system controlling M phi responses to LPS. The effects of LBP are lipid A dependent and importantly, extend to LPS preparations isolated from bacteria of R- and S-form phenotype.     

Chemical and immunological characterization of lipopolysaccharides from phase I and phase II Coxiella burnetii.

Lipopolysaccharides (LPSs) isolated from phase I and phase II Coxiella burnetii (LPS I and LPS II, respectively) were analyzed for chemical compositions, molecular heterogeneity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and immunological properties. The yields of crude phenol-water extracts from phase I cells were roughly three to six times higher than those from phase II cells. Purification of LPSs by ultracentrifugation gave similar yields for both LPS I and LPS II. Purified LPS I and LPS II contained roughly 0.8 and 0.6% protein, respectively. The fatty acid constituents of the LPSs were different in composition and content, with branched-chain fatty acids representing about 15% of the total. beta-Hydroxymyristic acid was not detected in either LPS I or LPS II. A thiobarbituric acid-periodate-positive compound was evident in the LPSs; however, this component was not identified as 3-deoxy-D-mannooctulosonic acid by gas and paper chromatographies. LPS II contained D-mannose, D-glucose, D-glyceromannoheptose, glucosamine, ethanolamine, 3-deoxy-D-mannooctulosonic acid-like material, phosphate, and fatty acids. LPS I contained the unique disaccharide galactosaminuronyl glucosamine and nine unidentified components in addition to the components of LPS II. The hydrophobic, putative lipid A fraction of LPS I and LPS II contained the above constituents, but the hydrophilic fraction was devoid of ethanolamine. The LPS I disaccharide galactosaminuronyl glucosamine was found in both fractions of the acetic acid hydrolysates. Analysis of LPSs by sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by silver staining indicated that LPS II was composed of only one band, whereas LPS I consisted of six or more bands with irregular spacing. Ouchterlony immunodiffusion tests demonstrated that LPS I reacted with phase I but not with phase II whole-cell hyperimmune antibody, and LPS II reacted neither with phase I nor phase II hyperimmune antibody. From these results, it was concluded that the chemical structures of LPSs from C. burnetii were different from those of the LPSs of gram-negative bacteria; however, the LPS structural variation in C. burnetii may be similar to the smooth-to-rough mutational variation of saccharide chain length in gram-negative bacteria.     

Composition of the fractions separated by polyacrylamide gel electrophoresis of the lipopolysaccharide of a marine bacterium.

The sugar composition of lipopolysaccharide (LPS) isolated from whole cells of Alteromonas haloplanktis 214 (previously referred to as marine pseudomonas B-16, ATCC 19855), variant 3, of the lipid A, core, and side-chain fractions derived from it, and of the LPS fractions (LPS I, II, and III) obtained by subjecting it to preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis has been determined. Conditions optimum for the release of constituent monosaccharides by hydrolysis were established. Sugars were quantitated by gas-liquid chromatography of their alditol acetate derivatives. Lipid A was detected by gel electrophoresis and by the spectral shift obtained with a carbocyanin dye. A comparison of the molar ratios of the various fractions suggest that LPS III is an LPS molecule lacking an O-antigenic side chain, whereas LPS I and II are LPS molecules differing in side-chain composition. LPS I may be a mixture of two LPS species. In double immunodiffusion experiments using anti-whole-cell serum, LPS I and II showed a homologous cross-reaction with isolated whole-cell LPS. LPS III as well as lipid A, core, and side-chain fractions failed to give rise to precipitin lines.     

Nature and linkage type of fatty acids present in lipopolysaccharides of phase I and II Coxiella burnetii

The constituent fatty acids of lipopolysaccharides (LPS) of Coxiella burnetii (phase I and II) were qualitatively and quantitatively analysed by combined gas-liquid chromatography/mass spectrometry. The total fatty acid content (per mg LPS) was determined as 90.0 nmol (2.3 wt%) for LPS of phase I cells (LPS I) and 179.1 nmol (4.8 wt%) for LPS of phase II cells (LPS II). Of the 24 different acyl residues characterized (12 to 18 carbon atoms), nine were 3-hydroxy fatty acids (normal, iso- and anteiso-branched) which quantitatively predominated. All 3-hydroxylated fatty acids were found to possess the (R)-configuration, to be exclusively amide-linked and to be acylated at their 3-hydroxyl group. Ester-linked nonhydroxylated fatty acids (normal, iso- and anteiso-branched) were present but ester-bound 3-hydroxy- or 3-acyloxyacyl residues were lacking from C. burnetii LPS I and LPS II. As the major acyl group (R)-3-(12-methyl-tetradecanoyloxy)-12-methyl-tetradecanoic acid was identified. Our results show that the complex fatty acid spectrum of C. burnetii differs considerably from that of LPS of other Gram-negative bacteria. They further suggest an enormous heterogeneity of the lipid A component of C. burnetii LPS I and LPS II.     

Characterization of LPS mutants of peanut specific Bradyrhizobium (GN17)

Two mutants of Bradyrhizobium sp. (Arachis) strain GN17 having altered lipopolysaccharide (LPS) composition were isolated upon random Tn5 mutagenesis to study their binding with peanut root lectin (PRA II). These mutant strains designated as GN17M1 and GN17M2 produced rough colonies and showed autoagglutination. Flow cytometric analyses indicated that strain GN17M1 bind to PRA II with highest efficiency. Both the mutants synthesized only high molecular weight lipopolysaccharides as observed by silver staining of polyacrylamide gel. The LPSs from both the mutants cross-reacted with anti-GN17 LPS, however, GN17M1 LPS showed 3 times higher cross-reactivity as detected by ELISA. Carbohydrate analysis by high performance anion exchange chromatography (HPAEC) showed that glucose was the major constituent of the purified LPS from the parent strain whereas mannose appeared as major component in the GN17M2 LPS. Equivalent amount of glucose and galactosamine with significant amount of mannose and galactose was the characteristics of the GN17M1 LPS. Purified LPS from GN17M1 and GN17M2 were respectively 17 and 10 times more potent inhibitors of PRA II activity than that of parent strain GN17. Similar binding efficiencies of the mutant LPS towards PRA II was also observed by ELISA. The results of this study indicate that the composition and the arrangement of the LPS are crucial for lectin binding.     

Effect of anti-dementia drugs on LPS induced neuroinflammation in mice

Inflammation has been recently implicated in pathogenesis of dementia disorders. Effect of anti-dementia (Acetylcholinesterase inhibitor) drugs tacrine, rivastigmine and donepezil were studied on neuroinflammation induced by intraperitoneal administration of lipopolysaccharide (LPS) in mice. Interleukin-2 (IL-2) and isoforms of acetylcholinesterase (AChE) were estimated in different brain areas as marker for neuroinflammation and cholinergic activity respectively. LPS significantly increased the level of IL-2 in all the brain areas while enhancement of AChE activity varied in brain areas. It was found that administration of tacrine, rivastigmine and donepezil in mice significantly attenuated the LPS induced increased levels of IL-2 along with the significant reduction of AChE activity predominantly in salt soluble (SS) fraction as compared to the detergent soluble (DS) fraction in a dose dependent manner. In vitro effect of LPS was also studied in different brain areas. LPS significantly increased the AChE activity in SS fractions but the significant increase was not found in DS fractions. The present study indicate that cholinesterase inhibitor anti-dementia drugs are effective against LPS induced neuroinflammation that may be linked to enhanced cholinergic activity.     

Lipopolysaccharides in a traditional pertussis vaccine

Analysis of the lipopolysaccharide (LPS, endotoxin) in cell sonicates of four Danish vaccine strains of Bordetella pertussis (3803, 3825, 3843 and 3860) and of purified strain 3803 LPS in sodium dodecyl sulphate-polyacrylamide gel electrophoresis by silver staining, showed identical profiles. The LPS profile revealed a dominant, brownish LPS II band and a minor, faster-migrating, black-stained LPS I band. However, the ratio of LPS I to LPS II in the preparation of purified LPS differed slightly from the cell sonicates. Using marker LPS, the molecular weights of LPS I and LPS II were estimated at 5.4 and 6.0 kD, respectively. Seven different lots of whole cell pertussis vaccine were assayed for LPS in the Limulus Amoebocyte Lysate test and were found to contain 0.9-2.8 micrograms LPS/ml. No significant difference in the content of LPS in similar dilutions of the individual strains was observed. In addition, the distribution of free and cell-bound LPS in four pertussis vaccines was investigated. Most of the LPS was found to exist as free LPS. During several months, the course of both LPS and pertussis toxin (Pt) release in freshly killed B. pertussis preparations was followed. In the first few weeks, 35-50% of the LPS was released and after 5-6 months of storage 60-80% had been released. In contrast, less than 10% of the biologically active pertussis toxin was released during the experimental period. The possibility of producing a safer whole cell pertussis vaccine by reducing the amount of free LPS without reducing the protective value correspondingly is discussed.     

Immunologic properties of bacterial lipopolysaccharide (LPS). IV. Cellular basis of the unresponsiveness of C3H/HeJ mouse spleen cells to LPS-induced mitogenesis.

Lymphoid cells obtained from the C3H/HeJ mouse strain respond abnormally to LPS in vitro, as shown by the fact that they are unable to make a mitogenic response to some LPS preparations and make only a low mitogenic response to other LPS preparations. In contrast, cells from a closely related C3H substrain, the C3H/St, are highly responsive to both types of LPS preparations. Experiments were carried out to determine the cellular basis of these genetically determined LPS response differences. This question was approached by studying the mitogenic response to LPS in cultures containing mixtures of various combinations of B cells, T cells, and macrophages from C3H/HeJ and C3H/St mice. Experiments utilizing an LPS preparation to which the C3H/HeJ is totally unresponsive (negative LPS) revealed, first, that either spleen cells, or partially purified T cells and/or macrophages obtained from C3H/St, could not restore the ability of C3H/HeJ spleen cells to respond to LPS, indicating that the C3H/HeJ is not deficient in an LPS-specific helper cell population which may be required for mitogenesis. Secondly, the addition of either spleen cells or partially purified T cells or macrophages from the C3H/HeJ to spleen cells from the C3H/St did not inhibit the mitogenic response to LPS, suggesting that the presence of suppressor cell activity is also not involved. Experiments analogous to those described, except utilizing another LPS preparation to which the C3H/HeJ is partially responsive (positive LPS), also failed to demonstrate reconstitutive or suppressive effects when C3H/HeJ and C3H/St spleen cells were admixed. The results obtained indicate that the defect in the C3H/HeJ mouse strain that limits its responsiveness to positive LPS and which renders it totally unresponsive to negative LPS appears to be an intrinsic defect in the capacity of B cells to react to the mitogenic stimulus of LPS.     

Restoration of LPS responsiveness of C3H/HeJ mouse lymphocytes by microinjection of cytoplasmic factor(s) from LPS-stimulated normal lymphocytes

Cytosol fractions of lymphocytes from LPS-responder mice (C3H/HeN) were prepared and injected into B lymphocytes of LPS-nonresponder mice (C3H/HeJ) by a microinjection technique utilizing polyethyleneglycol-mediated cell fusion. The B lymphocytes of C3H/HeJ mice microinjected with cytosol prepared from LPS-stimulated C3H/HeN cells became normally responsive to LPS. Microinjection of cytosol itself did not stimulate C3H/HeJ cells to proliferate or differentiate into immunoglobulin-producing cells, and the cells injected with cytosol had to be restimulated with LPS in order to proliferate and differentiate. These data suggested that C3H/HeJ B cells acquired LPS responsiveness by microinjection of cytoplasmic factor(s) from LPS-stimulated C3H/HeN cells and that these factor(s) may be one of the components involved in normal signal transmission from cell surface to nucleus in the early stages of the LPS response. The apparent m.w. of the cytoplasmic factor(s) is 100,000 by gel filtration. Chromatofocusing analysis suggested that these factors may consist of two components with the same m.w.     

Quantitative estimates of the distribution of homoserine dehydrogenase isozymes in maize tissues

The low molecular weight threonine-resistant (class I) and the higher molecular weight threonine-sensitive (class II/III) isozymes of homoserine dehydrogenase (EC 1.1.1.3) isolated from Zea mays L. were shown to differ in stability during incubations in the presence of urea. Class II/III was inactivated by urea in a time- and concentration-dependent manner, with complete inactivation occurring within 24 hours at 5 degrees C in 4.0 m urea. Under identical conditions, neither the activity nor the properties of class I were affected. Therefore, it was possible to estimate the amounts and properties of both maize isozymes in crude mixtures by measurements of enzyme activity before and after treatment with urea.The relative amounts of the two isozymes proved to be tissue-specific. When shoots of etiolated seedlings were extracted under optimum conditions, the resultant preparations contained about 16% class I and 84% class II/III. This distribution of isozymes, as well as the regulatory properties of class II/III, were constant during growth of the seedlings between 4 and 13 days. Enzyme preparations isolated from shoots of light-grown plants contained higher proportions of class I. The two isozymes were not uniformly distributed within leaves, as the basal meristematic region contained high levels of II/III and small amounts of I. During leaf maturation, the amount of II/III declined while the level of I remained constant or increased slightly. As a result, nearly half of the enzyme extracted from leaf tips was class I. The synthesis of specific members of the aspartate family of amino acids might be expected to differ when the ratio of threonine-sensitive to threonine-resistant homoserine dehydrogenase is altered. However, additional information on the subcellular localization and the catalytic characteristics of the two enzymes is required for evaluation of this possibility.     

Expression of lipocortins in human bronchial epithelial cells: effects of IL-1beta , TNF-alpha, LPS and dexamethasone

In this study, we investigated the expression of lipocortin I and II (annexin I and I in the human bronchial epithelium, both in vivo and in vitro. A clear expression of lipocortin I and II protein was found in the epithelium in sections of bronchial tissue. In cultured human bronchial epithelial cells we demonstrated the expression of lipocortin I and II mRNA and protein using Northern blotting, FACScan analysis and ELISA. No induction of lipocortin I or II mRNA or protein was observed after incubation with dexamethasone. Stimulation of bronchial epithelial cells with IL-1beta, TNF-alpha or LPS for 24 h did not affect the lipocortin I or II mRNA or protein expression, although PGE(2) and 6-keto-PGF(1alpha) production was significantly increased. This IL-1beta- and LPS-mediated increase in eicosanoids could be reduced by dexamethasone, but was not accompanied by an increase in lipocortin I or II expression. In human bronchial epithelial cells this particular glucocorticoid action is not mediated through lipocortin I or II induction.     

Highly-purified Helicobacter pylori LPS preparations induce weak inflammatory reactions and utilize Toll-like receptor 2 complex but not Toll-like receptor 4 complex

Helicobacter pylori is recognized as an etiologic agent of gastroduodenal diseases. Among toxic substances produced by H. pylori, LPS exhibits extremely low endotoxic activity as compared to the typical LPSs, such as that produced by Escherichia coli. We found that the LPS-low-responder stomach cancer cell line MKN28, which expresses Toll-like receptor 4 (TLR4) at extremely low levels, showed similar levels of interleukin-8 (IL-8) induction by H. pylori or E. coli LPS preparations. Weak IL-8 induction by H. pylori LPS preparations was suppressed by expression of a dominant negative mutant of TLR2 but not of TLR4. Data from luciferase reporter analysis indicated that cotransfection of TLR2-TLR1 or TLR2-TLR6 was required for the activation induced by H. pylori LPS preparations. In conclusion, the H. pylori LPS preparations significantly induce an inflammatory reaction via the receptor complex containing TLR2-TLR1 or TLR2-TLR6 but not that containing TLR4. The TLR2-TLR1 complex was preferentially recognized by the H. pylori LPS preparations over the TLR2-TLR6 complex. Whereas the magnitude of response to H. pylori LPS preparation was markedly less than that to E. coli LPS preparation in LPS-high-responder cells strongly expressing TLR4, it was comparable to that of E. coli LPS in low-responder cells expressing negligible amount of TLR4.     

Neutrophil chemotactic factor produced by lipopolysaccharide (LPS)-stimulated human blood mononuclear leukocytes: partial characterization and separation from interleukin 1 (IL 1)

LPS stimulated human blood mononuclear leukocytes to produce a chemotactic factor for human neutrophils. The effect of LPS was dose-dependent; 10 micrograms/ml was optimal for production of chemotactic factor. Chemotactic activity was detected 3 hr after LPS stimulation, and reached its peak at 12 hr. No activity was detected in culture supernatants of unstimulated cells, provided LPS-free media were selected. Isoelectric point of the factor, determined by chromatofocusing, was approximately 8 to 8.5. Molecular weight was approximately 10 kilodaltons by Sephacryl S-200 gel filtration or by HPLC gel filtration on TSK-2000 and -3000 columns in succession. The gel filtration fractions were also assayed for IL 1 activity. The elution position of IL 1 activity corresponded to a m.w. of 18. There was no chemotactic activity in the IL 1 activity peak. Furthermore, highly purified natural Il 1 alpha and -beta and recombinant Il 1 alpha and -beta did not exhibit chemotactic activity for neutrophils in our assay. Among mononuclear leukocytes, the monocyte was the principal producer of neutrophil chemotactic factor. These results suggest that a chemotactic factor for neutrophils, different from IL 1, is produced by LPS-stimulated blood monocytes.