Relation of structure to function in bacterial O-antigens--VII. Endotoxicity of 'lipid A'

Mutations of Francisella novicida that alter the mechanism of its phagocytosis by murine macrophages

Infection with the bacterial pathogen Francisella tularensis tularensis (F. tularensis) causes tularemia, a serious and debilitating disease. Francisella tularensis novicida strain U112 (abbreviated F. novicida), which is closely related to F. tularensis, is pathogenic for mice but not for man, making it an ideal model system for tularemia. Intracellular pathogens like Francisella inhibit the innate immune response, thereby avoiding immune recognition and death of the infected cell. Because activation of inflammatory pathways may lead to cell death, we reasoned that we could identify bacterial genes involved in inhibiting inflammation by isolating mutants that killed infected cells faster than the wild-type parent. We screened a comprehensive transposon library of F. novicida for mutant strains that increased the rate of cell death following infection in J774 macrophage-like cells, as compared to wild-type F. novicida. Mutations in 28 genes were identified as being hypercytotoxic to both J774 and primary macrophages of which 12 were less virulent in a mouse infection model. Surprisingly, we found that F. novicida with mutations in four genes (lpcC, manB, manC and kdtA) were taken up by and killed macrophages at a much higher rate than the parent strain, even upon treatment with cytochalasin D (cytD), a classic inhibitor of macrophage phagocytosis. At least 10-fold more mutant bacteria were internalized by macrophages as compared to the parent strain if the bacteria were first fixed with formaldehyde, suggesting a surface structure is required for the high phagocytosis rate. However, bacteria were required to be viable for macrophage toxicity. The four mutant strains do not make a complete LPS but instead have an exposed lipid A. Interestingly, other mutations that result in an exposed LPS core were not taken up at increased frequency nor did they kill host cells more than the parent. These results suggest an alternative, more efficient macrophage uptake mechanism for Francisella that requires exposure of a specific bacterial surface structure(s) but results in increased cell death following internalization of live bacteria.

Application of a novel apparatus, the quartz chemical analyzer, to the determination of endotoxin in blood

A novel apparatus called a quartz chemical analyzer (QCA) has been developed using a quartz crystal resonator. This apparatus measures sample viscosity changes based on resonant frequency changes of the quartz crystal. The apparatus was used to determine bacterial endotoxin concentrations by monitoring the gelation reaction of Limulus amebocyte lysate. The QCA determined endotoxin concentrations with good accuracy and reproducibility in the range of 0.001-3 EU/ml for endotoxin standard (JP XII). For endotoxin determination in human whole blood and plasma samples, the inhibitory reaction was eliminated by pretreatment of a fourfold dilution at 60 degrees C and incubation for 30 min. There are many advantages of the QCA method compared with the turbidimetric and chromogenic methods. For example, QCA can measure sample viscosity changes with high sensitivity and accuracy because QCA detects minor resonant frequency changes and the frequency data give a numerical value for easy quantitation. QCA can examine turbid samples, and the required quantities of samples and reagents are small, since the quartz crystal detects sample viscosity changes directly. The endotoxin determination time may be shortened by raising the reaction temperature, and QCA can detect other types of coagulation reactions.

Studies on the topography of the catalytic site of acetylcholinesterase using polyclonal and monoclonal antibodies

Polyclonal and monoclonal antibodies were generated against a synthetic peptide (25 amino acid residues) corresponding to the amino acid sequence surrounding the active site serine of Torpedo californica acetylcholinesterase (AChE). Prior to immunization, the peptide was either coupled to bovine serum albumin or encapsulated into liposomes containing lipid A as an adjuvant. To determine whether this region of AChE is located on the surface of the enzyme and thus accessible for binding to antibodies, or located in a pocket and thus not accessible to antibodies, the immunoreactivity of the antibodies was determined using enzyme-linked immunosorbent assay (ELISA), immunoprecipitation, Western blots, and competition ELISA. The polyclonal antibody and several of the monoclonal antibodies failed to react with either Torpedo or fetal bovine serum AChE in their native conformations, but showed significant cross-reactivity with the denatured enzymes. Human serum butyrylcholinesterase, which has a high degree of amino acid sequence homology with these AChEs, failed to react with the same antibodies in either native form or denatured form. Chymotrypsin also failed to react with the monoclonal antibodies in either form. Eighteen octapeptides spanning the entire sequence of this region were synthesized on polyethylene pins, and epitopes of representative monoclonal antibodies were determined by ELISA. The reactivity of peptides suggest that a portion of the 25 mer peptide in AChE containing the active site serine is the primary epitope. It is not exposed on the surface of the enzyme and is most likely sequestered in a pocket-like conformation in the native enzyme.

Detergent-accelerated hydrolysis of bacterial endotoxins and determination of the anomeric configuration of the glycosyl phosphate present in the "isolated lipid A" fragment of the Bordetella pertussis endotoxin

Due to the formation of micelles, severance of the hydrophilic (poly- or oligosaccharide) and hydrophobic ("Lipid A") domains of bacterial lipopolysaccharides at pH 3.4 or 4.5 and 100 degrees is slow and sometimes does not proceed at all; partially degraded fragments are usually formed. At pH 3.4 (100 degrees) in aqueous 1% sodium dodecylsulphate (SDS), both lipopolysaccharides of the Bordetella pertussis endotoxin are cleaved within 20-30 min, but 80% of the glycosidically bound phosphate present in the hydrophobic domain is lost. Other endotoxins behave similarly. At pH 4.5 (100 degrees) and in the absence of detergent, hydrolysis of the glycosidic bonds of 3-deoxy-D-manno-2-octulosonic acid residues of the B. pertussis endotoxin is negligible but, in aqueous 1% SDS, severance of the two regions of LPS 1 is complete within 1 h (that of LPS-2 requires 3-4 h), and the glycosidically bound phosphate of the isolated hydrophobic region is preserved. Comparison of the rate of acid-catalysed hydrolysis of the glycosidically bound phosphate present in this "isolated Lipid A" preparation with that of 2-deoxy-2-[(3R)-3-hydroxytetradecanamido]-alpha- and -beta-D-glucopyranose 1-phosphates established that the former 1-phosphate was the alpha anomer.

Thin-layer chromatography of endotoxins, their derivatives and contaminants

Thin-layer chromatographic (TLC) separation techniques were used to analyze the heterogeneity of various preparations which included smooth and rough endotoxins (ET), Lipid A precipitates and synthetic Lipid A samples and a novel cytotoxic bacterial lipid. Furthermore, carbohydrate-rich split products (PS) of ET were also separated on commercial silica-coated plates. Satisfactory results were obtained by two-dimensional TLC or by the combination of chromatography followed by high-voltage electrophoresis in the separation of PS of ET cleaved by mild acetic hydrolysis. Several spray reagents were found which were eminently suitable to detect carbohydrate containing compounds. Less specific but generally useful spray reagents were also developed which gave strong color reactions with lipids, proteinaceous and carbohydrate containing split products of the ET preparations. Improved chromatographic resolution has also revealed substantial heterogeneity in both rough and smooth ET samples. Three biological activities of the separated components could be determined. These were antigenicity detected by reactivity with monoclonal antibodies on the TLC plates, endotoxicity, determined by the Limulus amoebocyte lysate (LAL) test and direct cytotoxicity of P815 cells in vitro. Considerable amounts of non-endotoxic and non-antigenic contaminants could be detected in all preparations tested. Significant amounts of free Lipid A were also found in smooth ETs. Thus a new level of complexity is recognized by TLC within these preparations.

Chromogenic assay of endotoxin in platelet poor or rich plasma

In order to determine which sample preparation, platelet rich plasma (PRP) or platelet poor plasma (PPP), is more suitable for clinical endotoxin assay, we investigated the binding of endotoxins to platelets by comparing the amount of endotoxin in PRP with that in PPP, using a newly developed colorimetric assay with chromogenic substrate (Boc-Leu-Gly-Arg-pNA). When purified endotoxins were added to human whole blood, the amount of endotoxin recovered in PPP was significantly lower than that in PRP for all endotoxins tested except that from E. coli 0111:B4 and their ability to bind to platelets was varied depending on the species of bacteria from which they were purified. However, the amount of endotoxin in PRP obtained from surgical patients (n = 50) was almost same as that in PPP with a correlation coefficient r = 0.95, indicating that natural endotoxins circulating in human blood may not bind to platelets and that PPP can be used for endotoxin assay as well as PRP.

Antigenic heterogeneity of lipid A of Haemophilus influenzae

The chemical structure and biologic function of the lipid A portion of lipopolysaccharide are not identical among gram-negative bacteria. This study indicates that antigenically heterogeneous lipid A exists among strains of Haemophilus influenzae. An immunoglobulin G3 murine monoclonal antibody, 3D2, produced against a nontypable H. influenzae strain 3524 has specificity for a site on the lipid A portion of the H. influenzae lipopolysaccharide. With the Western blot and immunodot assay, 3D2 recognized this lipid A determinant on 14 of 24 (58%) of strains of nontypable H. influenzae and in 51 of 95 (54%) strains of H. influenzae type b. This lipid A epitope has a high degree of specificity for H. influenzae, since it is not present on the lipid A of 39 gram-negative strains from 14 non-Haemophilus species. In addition, studies of 36 strains of six Haemophilus species other than H. influenzae and 8 strains of 4 species of Actinobacillus did not contain the 3D2 epitope. Enzyme-linked immunosorbent assay analysis with a kinetic assay and enzyme-linked immunosorbent assay inhibition confirmed the antigenic heterogeneity of H. influenzae lipid A. Thin-layer chromatography demonstrated that the 3D2 epitope is associated with a chloroform-soluble lipid moiety in the lipid A. Fluorescent antibody analysis of H. influenzae indicated that the epitope is on the cell surface. The monoclonal antibody was not bactericidal for strain 3524, and it did not inhibit the bactericidal action of normal human serum against the same strain. These studies demonstrate that the lipid As of H. influenzae are antigenically heterogeneous.

Heterogeneity of lipid A: comparison of lipid A types from different gram-negative bacteria

Chloroform-soluble purified lipid A preparations from 10 sources, including five Escherichia coli strains (EH100, K-12, O127, O111, RCDC), two Salmonella strains (Salmonella typhimurium, Salmonella minnesota R595), Shigella sonnei II, and a hybrid of Shigella flexneri and E. coli K-12, were compared with lipid A from S. flexneri. Purified lipid A from S. flexneri was earlier found to be composed of eight fractions. The various lipid A preparations were assayed by thin-layer chromatography. Chromatograms were stained for phosphate or carbohydrate by molybdenum blue or orcinol, respectively. The number of major bands found for each lipid A preparation varied between 7 and 10, depending on the source. Comparable bands, based on Rf, were found among all of the different lipid A preparations, but the quantity of each band varied between the sources of lipid A. Four bands (designated 2, 3, 7, and 8) were abundant in every preparation. Variations of conditions used for preparing lipid A, such as changing of hydrolysis time, did not affect the appearance of lipid A on thin-layer chromatography. Change in the type of acid used for hydrolysis also did not affect the band pattern, but it did change the quantitative amounts of the various bands to some degree.

Lipid A fractions analyzed by a technique involving thin-layer chromatography and enzyme-linked immunosorbent assay

A modified enzyme-linked immunosorbent assay (ELISA), with alkaline phosphatase as enzyme, was used for the study of antigenicity of lipid A fractions directly on thin-layer chromatographic (TLC) plates. For visualization a gel slab containing the enzyme substrate was placed on the plate containing enzyme-conjugated antibodies. The plate was read by a thin-layer chromatogram spectrophotometer. The immunoassay was both highly specific and quite sensitive. Sensitivity was superior to levels obtained by staining the plate with molybdenum blue (for phosphate) or orcinol (for carbohydrate). Fractions of lipid A from Escherichia coli 0111, Shigella flexneri or Salmonella minnesota R595, after being separated by thin-layer chromatography, were analyzed using rabbit anti-(lipid A) serum. Patterns obtained by scanning the same plates for phosphate staining and for the TLC-ELISA corresponded well. For comparison with TLC-ELISA, an inhibition assay was run using a tube ELISA. The tube ELISA, run in aqueous medium, showed that fractions 6-8 (those having the highest RF values) had the least activities. In contrast, TLC-ELISA did not detect large differences between fractions 2-7. This discrepancy probably reflected limited aqueous solubility of fractions 6 and 7. We conclude that TLC-ELISA might reveal antigenic activities of lipids that could be missed by other methods. The data suggested that all fractions, except for fraction 8, were similar in their antigenicity by TLC-ELISA. Differences in antigenicity between the fractions occurred when the fractions were tested in free form in an aqueous environment and these differences possibly could have been due to different solubilities of individual fractions.

Molecular mechanisms in endotoxin fever

Two important concepts are presented in this review. First, endotoxin fever, like all fevers, is mediated by a host product, leukocytic pyrogen (LP). The mechanism by which LP production is initiated by endotoxin is discussed and evidence is provided which clearly distinguishes the biological and physical differences between LP and endotoxins. The second concept is that many of the molecular and neurochemical mechanisms by which LP causes fever by its action on the hypothalamic thermoregulatory center are also observed when endotoxins are introduced into the central nervous system. Thus, there may be experimental and clinical situations in which endotoxins can directly affect the hypothalamus and initiate fever. Although this bi-modal effect of endotoxin on the production of fever can occur, the importance of LP in mediating endotoxin and other fevers cannot be overstated.

Studies on the optimal conditions of CSF generation by endotoxic LPS and its PS derivative in mice

Not only the endotoxic LPS preparations, but a non-toxic, lipid-free, non-mitogenic hydrolytic breakdown product of it (called PS) is also capable of inducing colony stimulating factor (CSF) release (1). Due to difficulties to reproduce above findings it became necessary to study the optimal conditions to obtain CSF active PS preparations. It was found that the CSF generating component of the highly heterogeneous PS mixture is sensitive to acidic hydrolyses, but it is less sensitive than the toxic site in the lipid moiety of the LPS. Carefully controlled optimal hydrolytic conditions give PS preparations which have less than one percent residual endotoxicity but maintained 40 to 80% of the original CSF generating capacity. Prolonged hydrolysis will destroy this activity too. Optimal dose of LPS and PS for CSF induction in mice differed widely. For LPS the optimal dose is 25 micrograms, injecting more gave a much reduced or non-detectable CSF level. Optimal dose for PS was 160 micrograms, and this generated a significantly higher CSF level than 25 micrograms LPS. At concentrations below 25 micrograms, LPS was clearly more active than PS. The CSF level reached its peak at 3-4 hours after other LPS or PS injection. Intravenous route was sometimes but not always more effective than intraperitoneal.

Biological activities of chemically modified endotoxins from Vibrio cholerae

Lipopolysaccharides from Vibrio cholerae, NIH 41, Ogawa 5321 and Inaba 66/64 were treated with succinic anhydride, phthalic anhydride and dinitrophenyl ethylene diamine, and the resultant derivatives, sodium succinyl lipopolysaccharide, sodium phthalyl lipopolysaccharide, and dinitrohpenyl lipopolysaccharide obtained respectively were investigated for various biological activities. The succinylation and phthalylation of lipopolysaccharide decreased the 3-hydroxy lauric acid, a major ester-linked fatty acid of these bacteria, and as a result of which these modified products exhibited lower toxic activities in chick embryos, mouse and in generalized local Shwartzmann reaction in rabbits than their parent lipopolysaccharides. The dinitrophenylation of lipopolysaccharide increased its toxicity in chick embryos and mice, but dinitrophenyl lipopolysaccharide was completely inactive in Shwartzmann reaction in rabbits. However, despite the loss of these biological activities, these modified derivatives of lipopolysaccharide retained and increased the activities in pyrogenecity and in various immunological properties.

Biological activities of fragments derived from Bordetella pertussis endotoxin: isolation of a nontoxic, Shwartzman-negative lipid A possessing high adjuvant properties

Endotoxin from fresly sedimented Bordetella pertussis cells, isolated by the phenol/water procedure when submitted to kinetically controlled, mild acidic hydrolysis released a polysaccharide (polysaccharide 1), a complex lipid (lipid X), and a glycolipid. When treated with somewhat stronger acid, the glycolipid yielded a second polysaccharide (polysaccharide 2) and another complex lipid (lipid A). The intact pertussis endotoxin had all the usual properties of endotoxins extracted from enteric bacteria. Lipid X and the intermediary glycolipid retained all the endotoxic properties of the unfractionated endotoxin. In lipid A, pyrogenicity was reduced to a very low level and toxicity and Shwartzman reactivity were absent; however, this fraction retained most of the endotoxin's antiviral activity, and its adjuvant power was considerably higher than that of the intact endotoxin. Lipid A elicited nonspecific resistance against challenge with certain bacteria, but not against others.

Studies on the lipid components of endotoxin. II. Chemical analyses and biological properties of neutral, polar-I and polar-II subfractions

Lipid components obtained from Salmonella typhosa O-901 endotoxin by acid hydrolysis were separated into neutral, polar-I and polar-II lipid fractions by silica gel column chromatography. These lipids were further separated by silica gel column and/or thin-layer chromatography. The subfractions were analyzed by thin-layer chromatography, gas chromatography and infrared spectrophotometry. Seven subfractions obtained from the neutral lipid fraction contained lauric, myristic, palmitic, 3-OH-myristic acid, artificial products of 3-OH-myristic acid, or a small amount of two unidentified fatty acids. These fatty acids and glucosamine were commonly detected in six subfractions obtained from the polar-I lipid fraction. Fatty acids, glucosamine, and O-phosphorylethanolamine were detected in all of the 13 subfractions obtained from the polar-II lipid fraction. Chick embryo lethal activity, rabbit pyrogenicity and in vitro interferon inducing activity were found in three polar-I lipid subfractions and five polar-II lipid subfractions, but not in neutral lipids. The activities were highest in a polar-II lipid subfraction which contained smaller amounts of O-phosphorylethanolamine and glucosamine than the other subfractions. However, no particular chemical constituent(s) related to the biological activities could be found. Prolonged acid hydrolysis of the polar-II lipids gave rise to neutral and polar-I lipids. Chemical and biological aspects of the lipid constituents of endotoxin are discussed.

Immunologic properties of bacterial lipopolysaccharide (LPS). II. The unresponsiveness of C3H/HeJ Mouse spleen cells to LPS-induced mitogenesis is dependent on the method used to extract LPS

The C3H/HeJ mouse strain, previously shown to be a nonresponder to bacterial lipopolysaccharide (LPS)-induced mitogenesis in vitro, was demonstrated by the present studies to be competent to respond mitogenically to LPS, but only to LPS preparations obtained by selected extraction methods. These preparations appear to be confined to LPS isolated by mild extraction techniques, such as TCA or butanol. In contrast, those obtained by techniques utilizing phenol were only weakly stimulatory or completely nonstimulatory for spleen cells from the C3H/HeJ. All LPS preparations tested, on the other hand, were highly stimulatory for cells from another mouse strain, namely the C3H/St. The critical importance of the method of extraction of LPS on its mitogenic activity for C3H/HeJ cells was stressed by experiments in which LPS was prepared from Escherichia coli K235 using either of two procedures. In these experiments, phenol-extracted LPS, although mitogenic in the C3H/St, was completely nonstimulatory in the C3H/HeJ; whereas, butanol-extracted LPS was highly stimulatory in both strains of mice. This striking difference was attributed to a destructive effect of phenol on LPS, as demonstrated by the fact that treatment of butanol LPS with phenol resulted in a total loss of its mitogenic activity in the C3H/HeJ, but in only a partial loss in the C3H/St. In general, the mitogenic response observed with selected LPS preparations in the C3H/HeJ was quantitatively lower and more transient than that seen with the C3H/St, although qualitatively these responses appeared to be similar. This was evidenced by the observation that in both mouse strains LPS was a specific mitogen for B cells, a property which was also attributed in both strains to the same distinct structural region of the LPS molecule, that is lipid A. A preparation of LPS that failed to stimulate B cells from the C3H/HeJ nonetheless had the capacity to block activation of these B cells by a stimulatory preparation of LPS. These results strongly suggest that mitogenic stimulation of B cells by LPS is a function of the structural integrity of both the LPS molecule and putative B-cell receptors for LPS.