Isolation of a phenol-soluble endotoxin from Leptotrichia buccalis

Neil Goldsworthy and his legacy: Sydney's Institute of Dental Research

Dr. Neil Goldsworthy had a profound effect on the development of dental research in the biological sciences in Australia, including measures for the prevention of dental caries. He played a major role in the establishment of the Institute of Dental Research in Sydney in 1946 and was Director until his sudden death in 1960. Reviewing the Institute's activities provides the opportunity for his achievements to receive due recognition and to show how they influenced its subsequent development. As one who has been associated with the Institute for most of the last 50 years, it is an honor for me to undertake this task.

Ability of bacterial endotoxin to diffuse through human dentin

An in vitro system was developed to determine whether bacterial endotoxin is capable of diffusing through dentin without the use of filtration pressure. Cavities were prepared in five third molar teeth in order to produce a split chamber device consisting of occlusal and pulpal chambers with 0.5 mm of intervening dentin. Endotoxin was introduced into the occlusal chamber and the effluent in the pulpal chamber was sampled every 30 min for 5 h and at 24 h using the limulus lysate assay. In four specimens the initial appearance of endotoxin in the effluent ranged from 15 min to 4 1/2 h. In two specimens the concentration of endotoxin in the effluent leveled off in 4 1/2 and 5 h, respectively, whereas in another two the concentration continued to increase throughout the experiment. In one specimen no endotoxin was detected. The results indicate that endotoxin is capable of passing through 0.5 mm of dentin.

Kinin generation in the gingival inflammatory response to topically applied bacterial lipopolysaccharides

A water-soluble lipopolysaccharide from Salmonella enteritidis and a phenol-soluble lipopolysaccharide from Leptotrichia buccalis were applied topically to the healthy marginal gingivae of beagle dogs. Saline was applied to contralateral areas as an internal control. Increases in vascular permeability were monitored by measurement of gingival fluid, and the collected gingival fluid samples were assayed for kininogenase and kinin activities. Both lipopolysaccharides induced an inflammatory response, as indicated by increased gingival fluid flow. Kininogenase-kinin activities paralleled the increases in gingival fluid flow, with the highest values being associated with peak increases in gingival fluid. The results indicate that both lipopolysaccharides, although different in lipid solubility, penetrate healthy sulcular epithelium and initiate an inflammatory response which is mediated in part by the kallikrein-kinin system. Interrelationships between this system and other inflammatory mediators suggest that kinin generation not only plays a role in the early phases of acute gingival inflammation, but may also contribute to the activation of other mediators appearing later in the response and in chronic inflammatory lesions.

Determination of the presence of root-bound endotoxin using the local Shwartzman phenomenon (LSP)

The purpose of these studies was to determine, by attempting to elicit the local Shwartzman phenomenon (LSP), if the material extracted from periodontally involved root surfaces is actually endotoxin or only an endotoxin-like substance. A total of 719 periodontally involved teeth and 201 unerupted third molars yielded pooled samples of root surface grindings which were treated with the endotoxin extraction technique nonpyrogenic saline or pyrogen-free water. In the first study each of the reconstituted extracts was injected into the right palatal gingiva of New Zealand white rabbits which had demonstrated the ability to elicit an LSP. Commercial Salmonella enteritidis endotoxin was injected into the left (control) palatal gingiva. This was followed 18 hours later by a provoking injection of 200 to 300 microgram S enteritidis endotoxin in a marginal ear vein. In the second study, reconstituted extracts were injected into the right side of the abdomen of New Zealand white rabbits. The left (control) side received varying amounts of S enteritidis endotoxin. Twelve hours later each animal received a provoking injection of 400 microgram of S enteritidis endotoxin in a marginal ear vein. In both studies the animals were examined 6 hours later for visual signs of an LSP. After sacrifice, the tissues were prepared and histologic sections evaluated by two examiners who were unaware of how the specimens had been treated. The results indicated that a heat-stable, phenol-water extractable and highly irritating substance could be obtained from periodontally involved root surfaces. However, it is questionable if the substance extracted was actually endotoxin. If a true endotoxin, it was present only in extremely small amounts.

Human immune responses to oral microorganisms. II. Serum antibody responses to antigens from Actinobacillus actinomycetemcomitans and the correlation with localized juvenile periodontitis

Human serum antibody responses to antigens from a suspected oral pathogen, Actinobacillus actinomycetemcomitans (Aa), were studied. IgG and IgM isotype antibodies to four antigen preparations, sonicate antigen (SA), leukotoxin (LT), group carbohydrate (LG), and lipopolysaccharide (LPS), were determined using an ELISA. An ELISA inhibition technique was developed to show that human serum antibodies reacting with the LT, LG, or LPS materials were binding to different antigenic moieties in each preparation. Cross-sectional studies of serum IgG antibodies showed that patients with localized juvenile periodontitis (LJP) had a greater frequency of occurrence and a higher level of antibodies to the SA (82%), LT (70%), and LG (62%) antigens compared to all other diseased (11-46%) or normal (4-13%) groups. Serum IgM antibodies to LPS were increased in LJP, generalized juvenile periodontitis, and adult periodontitis patients compared to all other groups. Therefore, while both IgG and IgM antibodies were found against various Aa antigens, the detection of IgG antibodies was most clearly associated with the specific disease classification of LJP. Blocking studies suggested that the human serum responses were specific for the Aa antigens and that the LT, LG, and LPS comprise major antigenic determinants on the organisms to which human serum antibody reacts.

Haemolysis of human erythrocytes by the Fusobacterium nucleatum associated with periodontal disease

Fusobacterium nucleatum and Bacteroides gingivalis are associated with oral disease. They both attach to haemagglutinate and human erythrocytes. Experiments were performed to determine whether haemolysis would occur following attachment of strains of F. nucleatum and Bacteroides species including B. gingivalis. The F. nucleatum strains consistently displayed both haemagglutination and haemolytic activity. The B. gingivalis strains and other Bacteroides species displayed haemagglutination but no measurable haemolytic activity. Varying the concentration of the F. nucleatum whole cells in the standard haemolysis assay suggested a F. nucleatum-erythrocyte binding site interaction. The haemolytic moiety was observed in various cell, cell wall and lipopolysaccharide extracts.

Characterization of the lipopolysaccharide from Vibrio cholerae 395 (Ogawa)

The chemical structure and biological properties of the lipopolysaccharide (LPS) from Vibrio cholerae 395 (Ogawa), isolated by the phenol-water procedure, were studied. Upon acid hydrolysis, the LPS was split into its polysaccharide and lipid A moieties. The polysaccharide contained both neutral (glucose, heptose, fructose) and amino (glucosamine, quinovosamine) sugars. The LPS contained the acid-labile amino sugar, 4-amino-arabinose, which was absent in the Inaba serotype of V. cholerae. The LPS differed from the LPSs of Enterobacteriaceae by the absence of 2-keto-3-deoxyoctonate and the presence of fructose. Analysis of the methylated polysaccharide by gas-liquid chromatography and mass spectrometry showed that it had a branched structure with glucose and heptose residues primarily appearing at the nonreducing-end groups. Interactions with lectins, concanavalin A. and wheat germ agglutinin suggested that terminal glucose residues were alpha linked, whereas terminal glucosamine residues were connected by alpha-1,3 linkages. The major fatty acids of the LPS were C14:0, C16:0, C12h:0, and C14h:0 compounds, of which only the C14h:0 were amide linked, the remainder being ester linked to the backbone. Biological studies showed that the LPS possessed endotoxic properties such as lethality, pyrogenicity, limulus lysate gelation, and ability to induce non-specific resistance to infection. Thus, the LPS from V. cholerae 395 (Ogawa) possessed both common and distinct features as compared with the LPSs from the Enterobacteriaceae.

Chemical composition, ultrastructure and some serological properties of lipopolysaccharides from Leptotrichia buccalis

Lipopolysaccharide (LPS) was extracted with aqueous phenol from Leptotrichia buccalis strains L 11, ATCC 14201 and ATCC 19616. Virtually all the LPS was found in the water phase. LPS could also be extracted with phenol/chloroform/petroleum ether, but not with cold trichloroacetic acid. All LPS preparations contained D-glycero-D-manno-heptose, galactose, glucose, glucosamin, n-dodecanoate, 3-hydroxy-n-tetradecanoate and phosphorus. In addition LPS from L11 contained 2-keto-3-deoxyoctonate and an unidentified aldose. LPS from L11 was serologically distinct from LPS of ATCC 14201 and ATCC 19616. LPS from the two latter strains crossreacted.

Characterization of the lipopolysaccharide from Actinobacillus actinomycetemcomitans Y4 and N27

The lipopolysaccharide (LPS) from Actinobacillus actinomycetemcomitans strains Y4 and N27 was isolated by the phenol-water procedure. Morphologically, the molecule consisted of ribbon and branched filaments which comprised 3% of the cellular dry weight. Chemical analysis of the isolated and purified LPSs of both strains showed them to consist of carbohydrate, lipid, 2-keto-3-deoxyoctonate, heptose, hexosamine, and phosphate. The major fatty acids of the lipid A moiety were saturated C14 and beta-OH C14 compounds. Rhamnose, fucose, galactose, glucose, heptose, glucosamine, and galactosamine comprised the monosaccharide portion of the LPS. Biological activity studies revealed both LPS molecules to be active in the Schwartzman reaction and in in vitro 45Ca bone resorption, as well as in macrophage activation and lethality and in platelet aggregation.

Bacterial endotoxin: a role in chronic inflammatory periodontal disease?

A review of current information was undertaken examining suggested roles for bacterial endotoxin in periodontal disease. Evidence for the presence of endotoxin in plaque, gingival crevicular fluid and periodontally involved cementum is discussed. The biological effects of endotoxin are numerous, and its abilities to induce tissue inflammation, resorb rat foetal bone, activate the complement pathway, stimulate macrophages and activate immune responses are examined in relation to the suggestion that endotoxin is an important factor in the pathogenesis of periodontal disease.

Detection of a fibroblast proliferation inhibitory factor from Capnocytophaga sputigena

The addition of a sonic extract of Capnocytophaga sputigena to the culture fluid to human fibroblasts resulted in an inhibition of cell proliferation. The inhibition was dose-related (200 micrograms/ml caused a 90% inhibition, and 1,000 micrograms/ml caused a complete cessation of growth). The growth inhibition was not due to alterations in culture medium, pH or ionic strength, or to the effects of the C. sputigena lipopolysaccharide.

Biological and chemical characterization of endotoxin from Capnocytophaga sputigena

An endotoxin was isolated from Capnocytophaga sputigena strain 4 by a modification of the hot phenol-water method. The extraction procedure yielded a lipopolysaccharide which accounted for approximately 1.5% of the dry weight of the cells. The material was composed of 18.6% lipid (as C(15) fatty acid), 46.5% neutral sugar including 9.6% hexose, 18.3% 6-deoxy sugar, 1.0% 2-keto-3-deoxy sugar, and 4.8% heptose. Hexosamine, protein, and phosphorus were found in quantities amounting to 9.0, 2.9, and 2.0% of the dry weight, respectively. No pentose or nucleic acid was detected. Acid hydrolysis resulted in the release of the constituent sugars and the formation of an insoluble precipitate. The lipopolysaccharide was tested for numerous biological activities characteristic of endotoxins. The pyrogenicity was relatively low; the fever index 40 was 17 mug, and 10 mug was required to give the characteristic biphasic fever response. The toxicity of the extract was very low, with a 50% chicken embryo lethal dose of 15.6 mug and a 50% mouse embryo lethal dose of greater than 8 mg. Similarly, the C. sputigena endotoxin had modest effects on leukocytes when compared with endotoxin standards from other organisms. The extract exhibited little or no mitogenicity when tested on mouse spleen lymphocytes. It was not toxic to human peripheral polymorphonuclear leukocytes and caused the release of only a small (13%) portion of lysosomal enzymes. Although the C. sputigena lipopolysaccharide caused significant activation of mouse peritoneal macrophages, the dose required was twice that of an Escherichia coli endotoxic standard. However, the Limulus amoebocyte lysate clotting activity of the lipopolysaccharide was comparable to that of an Serratia marcescens lipopolysaccharide standard, and passive hemagglutination tests revealed that 1 mug of the lipopolysaccharide was capable of sensitizing 1 ml of a 2% sheep erythrocyte suspension for agglutination with an antiserum prepared against C. sputigena whole cells.

The alteration of osteoclast morphology by diphosphonates in bone organ culture

Two diphosphonates alter the morphology of the osteoclast, as they inhibit the calcium45 release from bones stimulated to resorb by lipopolysaccharide. Disodium dichloromethylene diphosphonate was more potent than disodium ethane-1-hydroxy-1, 1-diphosphonate in both inhibiting 45calium release and altering osteoclast morphology. Alteration in the morphology of osteoclasts is associated with little or no change in the morphology of the surrounding non-osteoclast cells. These results indicate a specific morphological effect of diphosphonates on osteoclasts.