UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The adherence properties of Bacteroides gingivalis Singh, Umadatt 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1990_A1 S57.pdf [ 13.17MB ]
JSON: 831-1.0100549.json
JSON-LD: 831-1.0100549-ld.json
RDF/XML (Pretty): 831-1.0100549-rdf.xml
RDF/JSON: 831-1.0100549-rdf.json
Turtle: 831-1.0100549-turtle.txt
N-Triples: 831-1.0100549-rdf-ntriples.txt
Original Record: 831-1.0100549-source.json
Full Text

Full Text

The adherence properties of Bacteroides gingivalis by UMADATT SINGH B . S c , Acadia University, 1980 M.Sc , Acadia University, 1982 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Microbiology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA DECEMBER 1989 © Singh Umadatt, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract A Bacteroides gingivalis adhesin mediating attachment to red blood cells and buccal epithelial cells was isolated, cloned and characterized. The isolation procedure involved gentle stirring of the cells followed by ammonium sulphate precipitation, ion-exchange and gel chromatography. The native molecule had a M r in excess of 10 6 kDa and was made up of subunits with an M r of 43 kDa. Antisera raised to the adhesin and its subunits reacted with antigens on the surface of B. gingivalis cells. No reaction with fimbriae was seen. The IgG fractions from these antisera inhibited the adherence of B. gingivalis to host tissue. Proteolytic enzymes destroyed binding capability of whole cells and of the purified adhesin but the molecular weight of the haemagglutinin was not altered. A genomic library of B. gingivalis DNA was created in £ . coli JM83. 5500 colonies were screened by a colony immunoassay with anti-S. gingivalis serum and by a direct haemagglutinating assay. 337 clones tested positive by the immunoassay and two clones, 1-3,and 1-49 tested positive for haemagglutinating activity. Both haemagglutinating positive recombinants had inserts of 3.2 kb. One clone, 1-49 was chosen for further characterization. E. coli 1-49 expressed a protein of 43 kDa that was not present in E. coli JM 83 control as seen by S D S - P A G E and Western blot analysis. Anti-1-49 serum inhibited the haemagglutinating activity of B. gingivalis and E. coli 1-49. This serum reacted with surface molecules on B. gingivalis and E. coli 1-49 as seen by immunogold electron microscopy and immunofluorescence, and to the purified haemagglutinin by Western blot analysis. Like the haemagglutinin on B. gingivalis, the haemagglutinating activity of E. coli 1-49 was destroyed by heating and proteolytic enzymes but the apparent size of the molecule as determined by SDS-PAGE was not affected. A bacterial coaggregating adhesin from B. gingivalis was isolated and characterized. ii The isolation procedure involved adsorption of the solubilized adhesin on S. mitis followed by elution with glycine buffer. S D S - P A G E of the boiled adhesin revealed a protein with an M r of 46 kDa. Proteolytic digestion destroyed all bacterial aggregating activity and hydrolysed the 46 kd protein. Antisera raised to the 46 kDa protein reacted with surface molecules on all strains of B. gingivalis tested. This antiserum inhibited the coaggregation reaction between B. gingivalis and other bacteria. Vesicles produced by B. gingivalis were found to enhance the binding of S. sanguis to serum coated hydroxy apatite (SeHA). Maximum vesicle mediated binding took place at 37°C and was destroyed by heating. The lipopolysaccharide from several black pigmented bacteroides were isolated and characterized physically, chemically and immunologically. All of the LPS were of the smooth type and contained the sugars rhamnose, glucose, galactose, glucosamine and galactosamine; no KDO or heptose were found. iii TABLE OF CONTENTS Abstract i i Table of contents iv List of tables v i i i List of figures x Abbreviations.Nomenclature and Symbols X J J J Acknowledgement X J V Introduction 1 Specificity of adherence 1 Methods of study 3 Bacterial surface components involved in adherence 5 (1) Firhbrial adhesins 5 (2) Fibrils 7 (3) Non-fimbrial proteinaceous adhesins 8 (4) Carbohydrates 8 Microbial adherence in the oral cavity 9 (a) Ecosystem , 9 (b) Dental plaque 1 1 (c) Adherence 1 2 Bacteroides gingivalis 14 (a) Classification 1 5 (b) Ecology 1 6 (c) Serology 1 7 (d) Pathogenicity 1 8 Virulence factors 21 (1) Adherence 21 (a) Haemagglutination 2 1 (b) Bacterial coaggregation 2 3 iv (c) Adherence to other substrates 2 4 (2) Collagenase 2 5 (3) Other proteolytic enzymes 2 6 (4) Toxic products 2 6 (5) Capsule 2 7 Vesicles 2 7 Lipopolysaccharides 30 Materials and Methods 3 6 Bacteria and cultivation 3 6 Isolation of haemagglutinin 3 7 Haemagglutination assay 3 8 Haemagglutination inhibition assays 3 8 Enzymatic treatment 3 9 Sodium periodate treatment 3 9 B-Mercaptoethanol 3 9 Electrophoretic techniques 4 0 Preparation of B. gingivalis chromosomal DNA 41 Construction of genomic library 4 1 Colony immunoassay 4 2 Immunological procedures 4 4 Binding to epithelial cells... 45 Isolation of bacterial aggregating factor 4 7 Bacterial aggregation assay 4 8 Bacterial aggregating inhibition 4 8 Modification of cell surface 4 9 Vesicles 5 0 Lipopolysaccharides 5 2 Results 5 6 V Isolation of haemagglutinin , 5 6 Characterization of the haemagglutinin 6 0 Dissociation of the haemagglutinin 6 2 Chemical composition 6 2 Inhibition of haemagglutination 63 Immunogold labelling 6 3 Enzymatic treatment 6 7 Cloning of the haemagglutinin 7 3 Genomic library 7 7 Recombinant plasmids 7 7 Restriction analysis 7 7 Southern blot 77 Expression of cloned antigen 8 2 Prevalence of the 43 kDa protein 8 2 Immunogold electronmicroscopy 8 6 Inhibition by antisera 8 6 Binding to buccal epithelial cells 8 9 Characterization of the recombinant-epithelial cell reaction. 9 4 Bacterial coaggregating adhesin 9 8 Purification 9 8 Characterization of the solubilized BA 101 Enzymatic digestion of the BA ; 105 Inhibition of bacterial aggregating activity by antisera.. 105 Immunogold labelling 109 Vesicles 113 Isolation of vesicles 113 Effect of vesicles on attachment of S. sanguis to SeHA 11 3 Kinetics of the binding reaction 11 7 Spectrophotometry adherence assay 1 25 vi Lipopolysaccharides 1 27 Analysis of LPS from Bacteroides 1 27 Chemical analysis of LPS "129 Gas-liquid chromatography of sugars from LPS 131 Fatty acid composition 1 31 Western electroblotting analysis 134 Discussion 1 3 6 Bibliography 1 4 7 vii LIST OF TABLES TABLE PAGE 1 Effect of a number of putative inhibitors on haemagglutinating activity 6 4 2 Effect of various surface modifying substances on the haemagglutinating activity of B. gingivalis 6 5 3 Effect of immune sera on haemagglutinating activity of B. gingivalis 6 6 4 Effect of enzyme treatment on haemagglutinating activity of intact cells and purified haemagglutinin of B. gingivalis.... 71 5 Effect of increasing amounts of proteinase K on haemagglutinating activity of purified B. gingivalis haemagglutinin 7 4 6 Time course of HA inactivation by proteinase K 7 5 7 Effect of PMSF on proteinase K digestion of the purified B. gingivalis haemagglutinin 7 6 8 Effect of immune sera on haemagglutinating activity of E. coli 1-49 91 9 Adherence to buccal epithelial cells 9 2 10 Effect of protease, heat, and putative inhibitors on the attachment to oral epithelial cells 9 6 11 Effect of immune IgG on binding to epithelial cells 9 7 12 Effect of various compounds on aggregation 1 0 3 13 Effact of various substances on the bacterial aggregating activity of B. gingivalis 1 0 4 viii 14 Effect of enzymatic treatment on the bacterial aggregating activity of B. gingivalis and the purified BA 106 15 Effect of immune serum on bacterial aggregating activity 108 16 Effect of anti-46 kDa antiserum on the binding of B. gingivalis to immobilized S. mitis 110 17 Influence of vesicles on the attachment of S. sanguis to SeHA and SHA 117 18 Langmuir isotherm parameters for the adsorption of S. sanguis to SeHA with and without the addition of vesicles 1 21 19 The effect of heat and putative receptor on the attachment of S. sanguis to SeHA beads 1 2 4 20 Visual adherence assay for the attachment of bacteria to SeHA 13 6 21 Chemical content of lipopolysaccharides of the black pigmented oral Bacteroides 13 0 22 Carbohydrate composition of lipopolysaccharides from BPB 132 23 Fatty acid composition of LPS from BPB 1 33 ix LIST OF FIGURES FIGURES PAGE 1 Elution of haemagglutinating activity and protein from a DEAE-Sepharose column 57 2 SDS-PAGE of fractions eluted from a DEAE-Sepharose column.. 5 8 3 Chromatography of HA on Sephacryl S-1000 5 9 4 S D S - P A G E gradient gel analysis of fractions from Sephacryl S-1000 column 61 5 Cells of B. gingivalis reacted with anti-43kDa serum 6 8 6 Immunogold bead labelling of thin sections of B. gingivalis 6 9 7 Immuno gold bead labelling of thin sections of B. gingivalis 7 0 8 Colony immunoblot 7 8 9 Screening of recombinant clones for haemagglutinating activity... 7 9 10 Restriction endonuclease analysis of the recombinant clones 80 11 Linear restriction map of the the cloned insert from recombinant E. coli 1-49 81 12 Southern blot hybridization with the 3.2 kb Pst 1 restriction fragment of E. coli 1-49 8 3 13 Western blot of anti-1-49 antiserum against cell lysate of E. coin -49. 8 4 14 Western blot of the purified haemagglutinin from B. gingivalis and cell lysates of Bacteroides sp. using anti-1-49 serum as the probe 85 15 Immuno gold bead labelling of E. coli 1-49 8 7 x 16 Immuno gold bead labelling of B. gingivalis 8 8 17 Immuno gold bead labelling of E. coli 1-49 90 18 Scanning electron micrograph of bacteria bound to buccal epithelial cells 9 3 19 Electron micrograph of bacteria attached to epithelial cells.. 95 20 Adsorption of the BA by S. mitis 1 00 21 Western-blot analysis of the BA adsorption on S. mitis 1 02 22 S D S - P A G E analysis of the enzymatic digestion of the bacterial aggregating adhesin from B. gingivalis 33277 107 23 Immuno gold bead labelling of B. gingivalis 111 24 Immuno gold bead labelling of thin sections of B. gingivalis. . 112 25 Transmission electron micrograph of a vesicle preparation ... 114 26 B. gingivalis SDS-PAGE profile 11 5 27 Scanning electron micrograph of SeHA incubated with (A) B. gingivalis vesicles, (B) S. sanguis, and (C) B. gingivalis vesicles , 118 28 Adsorption isotherm of S. sanguis bound to SeHA with and without vesicles 11 9 29 Langmuir adsorption isotherm for the adherence of S. sanguis to SeHA, and SeHA pre-incubated with B. gingivalis vesicles.... 1 20 30 Scatchard adsorption isotherm for the adherence of S. sanguis to SeHA pre-incubated with B. gingivalis vesicles 122 31 Adhesion of S. sanguis to SeHA, pre-incubated with increasing quantities of B. gingivalis vesicles 1 23 32 S D S - P A G E analysis of LPS from black pigmented Bacteroides.. 128 xi 33 Reaction of LPS from BPB with antisera to B. gingivalis 33277 .. 135 xii ABBREVIATIONS, NOMENCLATURE AND SYMBOLS Anti-43 kDa Antiserum raised against subunits of the haemagglutinin Anti-HA Antiserum raised against the native haemagglutinin Anti-46 kDa Antiserum raised against the bacterial aggregating adhesin ATCC American Type Culture Collection BA Bacterial agglutinin BPB Black Pigmented Bacteroides DNA Deoxyribonucleic acid EDTA Ethylenediaminetetraacetic acid HA Haemagglutinin HEPES N-2-hydroxyethylpiperazine-N -2-ethanesulphonic acid kb 1000 base pairs kDa 1000 daltons M r Relative molecular mass PAGE Polyacrylamide gel electrophoresis PBS Phosphate buffered saline PBS/T Phosphate buffered saline/ Tween 20 SDS Sodium dodecyl sulphate SHA Saliva-coated hydroxyapatite beads SeHA Serum-coated hydroxyapatite beads SeHA-V Vesicle covered serum-coated hydroxyapatite beads TE Tr is /EDTA xiii ACKNOWLEDGEMENTS I sincerely acknowledge the financial, inspirational and technical support of Dr. B.C. McBride during the course of this project. My sincere thanks to my committee members Drs. R. A. J . Warren, R. E. W. Hancock and R. S. Molday for serving on my committee and their advice throughout the course of this work. I wish to thank Andre Wong for his work with the electron microscope, and Bruce McCaughey for his photographic assistance. Finally, I like to extend a sincere thank you to my wife Patricia for her encouragement and unfailing love during the course of this study. This thesis is dedicated to my wife and my son Michael. xiv INTRODUCTION The introductory section of this thesis will be a general discussion of the principles governing bacterial adherence to host tissue. This will be followed by a description of adherence in the oral cavity and its importance in the pathogenesis of oral organisms. This latter section will include a detailed description of Bacteroides gingivalis including a discussion of potential pathogenic factors. Adhesion or adherence can be defined as the irreversible attachment of an organism to a surface. In the oral cavity this surface can be another bacterium, soft tissue (epithelial cells) or hard tissue (teeth) which may be covered with an organic layer. Any molecule or structure responsible for adhesive activities will be called an adhesin. The entity on surfaces to which adhesins bind will be called the receptor. It has been postulated that when a bacterium approaches a putative receptor it binds via a reversible interaction (147) and other interactions stabilize or contribute towards the irreversible attachment (42). It is to be expected that molecules on the surface of the bacteria would be responsible for this attachment. Specificity of adherence The question arises whether bacteria adhere to host tissue via specific surface molecules or via non-specific interactions. The specificity of the interaction of bacteria with host tissues was first suggested by data from in vivo experiments. Gibbons et al (67) pointed out the apparent preference of particular bacteria for certain tissues over other tissues (tissue tropism). For example Streptococcus salivarius was found in large numbers on the tongue and low numbers in dental plaque and subsequently it was shown to bind to epithelial cells. This idea of the specificity of the interaction between bacteria and host tissue is further supported by the species specificity of certain bacterial infections. For example, gonococcal infections are limited to humans, (as the bacteria responsible for the disease process is capable of attaching to human epithelial cells via their pili) 1 (104), group A streptococcal infection to humans (228) and diarrheagenic E.coli K 88 infection to pigs (106). Cheney et al (21) demonstrated convincingly the high degree of specificity between adherence and disease. They showed that certain diarrheagenic strain of E. coli which were capable of attaching to the gut epithelial cells of rabbit produced diarrhea in these animals, whereas these same strains of E. coli, unable to attach to the epithelial cells of guinea pigs and rats produced no disease in these animals. Susceptibility to certain infections is a genetic trait related to the availability of appropriate receptors. For example it was demonstrated that certain pigs are genetically immune to E. coli K 88 infection. By crossing susceptible and resistant pigs the investigators showed that susceptibility is coded for by chromosomal dominant genes and moreover is mediated by receptors for E. coli K 88 on the brush borders of the intestinal epithelial cell (160, 207). It should be emphasized however, that the colonization of a surface is a complex process which may involve factors such as bacterial motility, chemotaxis and the ability to reproduce in that environment (9). Various non-specific factors also influence bacterial adherence. Healthy surfaces are constantly bathed by secretions laden with antibacterial enzymes and antibodies which are impedements to colonization. Additionally unattached bacteria are swept away by mechanical means such as coughing, sneezing and peristalsis, and the hydrodynamic forces exerted by fluid flowing over the mucosal surfaces, eg saliva and crevicular fluid. Successful colonizers are those capable not only of attaching, but also of multiplying, this is particularly true on surfaces where host cells are continuously desquamated (94). In the oral cavity there are surfaces eg tooth, which circumvent problems of shedding and thus allow a stable flora to develop. Nevertheless, attachment to such surfaces must be of such a nature to resist cleansing mechanisms such as mastication and salivary flow (8, 62). 2 Adherence is a complex process which may depend on several forces interacting at the same time. For instance it has been postulated that binding to oral surfaces involves electrostatic (201), lectin-like (69,162) and hydrophobic interactions (8). Electrostatic and hydrophobic interactions may be less specific in nature but are important in stabilizing the specific lectin-like interaction. Methods of study Several different approaches have been used in an attempt to gain a better insight as to how bacteria bind to a surface. The method chosen has depended on whether the objective was to (a) identify, isolate and characterize the adhesin or receptor, (b) characterize the adhesive interaction, (c) study the expression of genes involved, or (d) study the in vivo activity of the adhesin. A number of adhesins have been isolated by conventional biochemical procedures including ultracentrifugation and column chromatography. One of the best example is E . coli where a variety of unique cell binding adhesins have been purified (46). In oral system adhesins have been purified from a number of organisms including S. salivarius and Actinomycetes viscosus (24,243). The structural characterization of these adhesins has been determined by microscopic examination (55,256,257,260,). An effective method for identifying adhesins has been to study adhesive negative mutants. In these studies spontaneous or chemically induced mutants were isolated following enrichment by adsorption. A comparsion was then made between the cell wall polypeptides of the mutant and wild type. For example, Weiss et al (246) isolated a coaggregating deficient mutant of Bacteroides loeschii which was found to lack a specific 75 kDa cell wall polypeptide. In other cases monospecific antisera was prepared by adsorption of the anti-wild type serum with the mutant, the adhesin was then identified by Western blot analysis. This monospecific antiserum was also used to inhibit adherence of the wild type cells (25,28) or to immunoprecipitate the adhesin (243). 3 A routine method of study for understanding the adherence reaction has been to chemically or enzymatically modify the cell surface of the bacteria or epithelial cell and to determine how adherence is affected. McBride et al (152) showed that Streptococcus sanguis bound to sialic acid suggesting the presence of a sialic acid lectin. Weerkamp and McBride (242) found that treatment of S. salivarius with protease reduced binding to Fusobacterium nucleatum thus implicating a protein as the adhesin. Yamazaki et al (258) showed that trypsin treatment of Eikenella corrodens prevented it from attaching to buccal epithelial cells. They also found that neuraminidase treatment of the epithelial cell enhanced adherence, exposing a galactose residue thus implying that the adherence reaction involved a lectin-like protein on the bacterial surface. The nature of the receptor has been studied by measuring the inhibitory activity of soluble receptor analogues. Well studied examples are the mannose-sensitive haemagglutinin (HA) of enterobacteriaceae (47) and the lactose sensitive A. viscosus-S. sanguis interaction (119). Another example is the adherence of galactose sensitive binding of Eikenella corrodens to epithelial cells (258). Recently, recombinant DNA technology has been used to identify the genes involved in adhesion. The most thoroughly studied examples are the E. coli adhesins which bind the organism to gut epithelial cells. The number of genes involved in the adherence reaction was elucidated and the structural and regulatory sequences have been located (129). Jacobs et al (101) did site specific mutagenesis and localized the amino acids involved in the adhesin-receptor reaction. This will be discussed in detail in subsequent paragraphs. Bacterial surface components involved in adherence Bacteria have a complex cell surface comprised of proteins, carbohydrates and amphiphilic compounds such as LTA and LPS. Each of these components has the potential 4 to act as an adhesin. In some cases the adhesins are organized into polymerized structures such as fimbriae or capsule, which can be visualized by electron microscopy. (1) Fimbrial adhesins (a) Morphology In general two classes of fimbrial adhesin exist, rigid and flexible. An example of the former are the type 1 fimbriae of E. coli which measures between 2 and 10 nm in diameter and may be as long as 4 urn (181). (b) Composition Fimbrial adhesins are proteinaceous polymers whose structural elements are primarily composed of identical subunits (186,197). Since the subunits are normally disassociated by sodium dodecyl sulphate, (220), dilute acid (157) or guanidine hydrochloride (54), it is reasonable to believe that they are held together by hydrogen and hydrophobic bonds. The molecular weight of fimbrial subunits ranges from 8 to 26 kDa (159,220). Most fimbrial adhesins are negatively charged molecules with isoelectric points in the range of 3.7 to 5.6 (183). Genetic and biochemical analysis have shown that there are a number of fimbriae associated proteins which are different from the fimbrial structural subunits and which actually form the adhesin. For example Lindberg et al (129) showed that a distinction existed between the adhesin molecule and the pap pili subunit of E. coli J-96. Clegg et al (17) demonstrated that mutants of Klebsiella pneumoniae can be constructed which possess intact fimbriae but lack the ability to agglutinate erythrocytes. This implies that adhesin responsible for recognition of receptors can be a moiety separate from that of the fimbrial structural element. (c) Role of fimbriae in adherence Fimbriae have a well documented role in adherence. For example in E. coli the mannose sensitive (MR) haemagglutinin was shown to be the typel fimbriae (45). Type 1 and 5 type 2 fimbriae of Actinomyces viscosus have been shown to be involved in binding to saliva coated hydroxyapatite and to S. sanguis respectively (42). The in vivo role of fimbriae in adherence has also been investigated. An example are the enterotoxigenic (ETEC) strains of E. coli which have lost their adherence mediating fimbriae and are no longer pathogenic (20,77). Similarly non-fimbriated gonococcus were found to be non-infective in human studies (104 ). (d) Genetics Recently there have been a number of studies involved in the analysis of fimbrial genes. These studies have shown that there are multiple genes that are responsible for the production, export, assembly and anchorage of the fimbriae. These genes can be located on a plasmid as is the case with the E. coli K 88 fimbriae (180) or on the chromosome as found in E. coli J 96 (172). For example, genes responsible for type 1 fimbriae have been cloned from uropathogenic E. coli J 96 (148). The genes responsible for the formation of this fimbriae are organized into an operon in the order pil E, F, D, C, B, A.hyp. The pil B and pil C genes code for 30,000 and 86,000 M r proteins respectively. The proteins are involved in assembly and anchorage of the fimbriae. The pil E gene codes for a 31 kDa protein, mutagenesis in this gene results in a nonadhesive piliated bacteria. So although type 1 pili are needed for receptor binding sometime their presence is not sufficient for binding. The receptor binding function requires in addition to the polymerized pili the product of the pil E gene (148). The pil F gene codes for an 18.2 kDa protein, mutation in this gene results in pilus of increased length. This led to the conclusion that the pil F gene product act as a competitive inhibitor of pilus polymerization. The pap pili genes which are located on chromosomal DNA were cloned from E. coli J96 (172). This pili is responsible for the MR haemagglutination and uroepithelial attachment of this bacteria. The pap genes are arranged in similar fashion to those of the type 1 fimbriae. By mutant analysis coupled with immunoelectron microscopy it was 6 shown by Lindberg et al (129) that the pap pili are heteropolymers composed of the major pilin protein pap A, the minor pilin pap E and pap F, and the adhesin pap G. These latter three proteins are located at the tip of the pili. It has become clear from studies of type 1 and pap pili that the receptor binding component is different from the fimbrial subunit and that fimbrial length and numbers are regulated by associated genes. A logical question ensues as to why the adhesin is associated only with the pilus tip. An attractive possibility is that the pilus is needed to place the adhesin outside the complete LPS O-chain found on E. coli. This suggestion is based on the finding that smooth strains of E. coli J 96 binds only if the pap A gene is intact, the pap A protein is not required in rough mutants which lack the O-side chain (129) . (2) FibriHae These are structures which were originally found on the surface of gram-positive bacteria where they formed part of the "fuzzy coat". Recently they have been seen on many gram-negative organisms (90). They are different from fimbriae in that they have no definite diameter and are usually long and flexible. Fibrillae may be peritrichous, polar or located in patches on the side of the bacteria. They vary in length from 70 nm to 200 nm. In S. salivarius they have been divided into four distinct groups depending on their adhesive properties, length and organization (243). A fibrillar haemagglutinin has been purified from Campylobacter pylori, an organism responsible for gastritis, peptic and duodenal ulcers in humans (55). Adhesins with fibrillar structures have also been isolated from several other bacteria including E. coli, Yersinia sp. (111), S. sanguis (128) and Bacteroides loeschii (246). (3) Non-fimbrial proteinaceous adhesins Not all proteins on the surface of bacterial cells are incorporated into structural 7 elements such as fimbriae and fibrils; they can also exist as discrete elements on the cell surface. For example E. coli (47) and Enterobacter aerogenes (2) possess non fimbrial non fibrillar surface proteins that are responsible for haemagglutination. Mycoplasma pneumoniae attaches to epithelial cell via a 165 kd protein (7) which binds to sialic acid residues. A 300 kDa protein binds the periodontopathogen Fusobacterium nucleatum to buccal epithelial cell in a galactose sensitive reaction (165). Capnocytophaga ochracea coaggregated with S. sanguis, A. naaslundii and A. israelii via a cell surface protein in a reaction that was inhibited by L-rhamnose and D-fucose (245). Microscopic examination did not reveal any polymerised structures on the cell surface. (4) Carbohydrates Other structures on bacterial cell surface have been known to have adhesive properties; these include capsules and lipopolysaccharides (224,249). One of the most studied example in which surface located carbohydrates is responsible for adherence is that of the extracellular glucan synthesized by S. mutans. This material is involved in mediating the attachment of the organism to teeth and to other S. mutans (87). S. sanguis 34 has carbohydrate receptors which mediate coaggregation with A. viscosus and A, naeslundii lectins. The receptor has been found to consist of repeating subunits of a hexasaccharide containing N-acetylglucosamine, galactose, glucose and rhamnose joined by phosphodiester bonds (52). The plant pathogens Agrobacterium, Corynebacterium, Erwinia, Pseudomonas and Xanthomonas produce extracellular polysaccharides which are believed to bind specifically to the host (204). Agrobacterium tumifaciens is an invasive pathogen which initiates crown gall tumours in a variety of plants. The adherence of virulent Agrobacterium to a specific site on the host cell wall exposed by wounding has been shown by Lippincott and Lippincott (130) to be due to the LPS O-antigen. 8 Microbial adherence in the oral cavity The mouth is a nutrient rich environment ideally suited for the growth of bacteria. Within the mouth there are several types of surfaces which lend themselves to colonization. These include the keratinized epithelium found in such places as the tongue, non-keratinized epithelium which lines the dentogingival sulcus, the pellicle coating the surface of the teeth, and other attached bacteria (66). These surfaces will be modified by the environment providing a multitude of potential receptors. Oral surfaces are subjected to a number of modifying influences including the effect of saliva, crevicular fluid and ingested food. Saliva secreted by major and minor salivary glands continuously rinse oral surfaces to remove non-adherent organisms. Crevicular fluid plays a similar role in the gingival crevice. Both fluids supply nutrients to the indigenous microflora and remove waste products. The food eaten can alter the pH and depending on its composition can favour the colonization of a particular group of organisms (85). For example; fermentation of carbohydrates leads to the production of acids and the selection of an acidogenic flora. The most profound effect is caused by sucrose which in addition to being fermented, is converted to high molecular weight glucans which bind large numbers of the strongly acidogenic S. mutans to the teeth (27). (a) Ecosystem There are several environmental niches in the mouth created by anatomical diversity and a variety of allo-and autogenic factors. The complexity of the oral environment is reflected by diversity of the microbial flora which is estimated to consist of greater than three hundred species (161). These organisms are arranged into unique ecosystems. They modify their environment by consuming oxygen, altering pH, removing nutrients, releasing metabolic end products and elaborating enzymes which may change the nature of the bacterial receptors on the immobilized surfaces. 9 The most studied ecosystems of the oral cavity include the tongue, buccal surfaces, tooth and gingival crevice. The tongue is an aerobic environment which is thought to be the area of the most rapid cell proliferation (85). S. salivarius is found on the tongue in small numbers on the tooth probably because of its ability to attach to the tongue epithelium. The buccal surfaces are a sparsely populated aerobic ecosystem colonized by a limited number of organisms. The area between the tooth and the gum, known as the gingival crevice is anaerobic in nature and supports the growth of a strict anaerobic population. This niche is relatively isolated from other ecosystems in the mouth (85). Organisms have the possibility of binding to non-keratinized epithelium or to the subgingival regions of the tooth. Food is supplied to the organisms by crevicular fluid which is a nutrient rich serum transudate which seeps into the crevice through the periodontium. The type of bacteria isolated from this habitat reflects the state of health of the gingival tissues. In the healthy state most of the microflora are gram-positive facultative rods and cocci whereas in the diseased state there is a change to gram-negative anaerobic motile rods (221,222). Another unique site in the oral cavity is the tooth. The exposed area of the tooth is enamel which is composed primarily of hydroxyapatite which has phosphate and calcium ions exposed on its surface (120). Since many more phosphate groups than calcium ions are exposed, hydroxyapatite has a net negative charge. However, this surface is rarely exposed directly to the bacteria, since the enamel is covered by a thin layer of organic material termed the acquired pellicle (140). This layer is normally less than a micron in thickness and consists mainly of salivary glycoproteins (4,30). 10 (b) Dental plaque Bacterial accumulations on the tooth surfaces are referred to as dental plaque. Plaque consists of bacteria embedded in an amorphous matrix of salivary, crevicular and bacterial polymers (66,68). Plaque develops rapidly on protected areas of the teeth but will eventually cover all smooth surfaces. The bacterial composition of the dental plaque differs with age of the plaque. When the plaque material is associated with the teeth above the gingival tissue it is designated as supragingival plaque, and when it is below the gingiva it is called subgingival plaque (66). The development of dental plaque involves three types of adherence reactions. The first step is a selective adsorption to the pellicle. Subsequent accumulation of organisms involves bacteria-bacteria binding. By homotypic cell-cell interactions, organisms of one species may accumulate and maintain their position in the growing plaque. Heterotypic cell-cell interactions leading to coaggregation between species are also of importance. Many in vitro studies have shown that oral bacteria can coaggregate when mixed together (51,66,71,73,119,153,155). McBride et al (153) demonstrated that coaggregation supports the colonization of Veillonella in gnotobiotic rats infected with Streptococcus. An unusual example of interbacterial aggregation in plaque are the so called "corn-cob" formations in which coccal forms adhere via surface fibrils to a central filamentous organism (132). The ability of salivary constituents to induce aggregation of a variety of oral streptococci is well documented and has been extensively reviewed by Bowden et al (10) and Gibbons (66). Such interactions may promote homotypic bacterial aggregation in the plaque matrix. Salivary constituents bound to bacterial cell surface may also promote heterotypic bacterial aggregation, thus contributing to the species diversity seen in plaque (52). 11 (c) Adherence Adherence in the oral cavity follows the same principles as that of the adherence of microorganisms to any other surface of the body. The interactions between receptor and adhesin can be ionic, hydrophobic or lectin-like (225). The adhesins can be fimbriae, fibrillae or other proteins, carbohydrates and amphiphilic substances. Adhesion to the tooth surface is studied in an in vitro adherence model which consists of adding bacteria to hydroxyapatite beads treated with saliva to form the in vitro equivalent of the acquired pellicle. These experimental pellicles have similar properties to a naturally acquired pellicle (22,72). Adsorption can be quantified by use of radiolabeled bacteria (128) or by turbidimetric (93) cultural (154) or by visual techniques (179). The saliva coated hydroxyapatite (SHA) model system has been used to study the kinetics of binding. The binding isotherm is obtained by plotting the number of unbound cells (U) at equilibrium against the bound cells (B). Such adherence data can be fitted to a Langmuir adsorption isotherm (3, 72) U/B=K/N +1/NU. A plot of U/B vs U allows an estimation of the dissociation constant K and its reciprocal the affinity constant 1/K from the X intercept. The theoretical number of binding sites (N) on the SHA can be estimated from the slope of the graph. Binding and Langmuir isotherms were used to compare the adherence of different species of oral bacteria to SHA and HA (26). Analysis of such isotherms indicates that the range of binding sites present for different species is much greater on SHA than on the untreated mineral (26). Binding to HA was found to be relatively non-specific whereas binding to SHA was a specific reaction. The salivary pellicle therefore imparts a higher degree of specificity to the adsorption process. In a series of competition experiments Liljemark et al (128) showed that S. sanguis, S. mitis and S. salivarius did not compete with each other for binding sites on SHA. This data provided strong evidence that 12 bacterial adhesins were recognizing specific receptors on the salivary constituents adsorbed to HA and demonstrated that different bacterial species do have unique adhesins recognizing specific SHA receptors. Molecular analysis has revealed that the binding of S. sanguis to SHA involves a number of different adhesins (42, 69, 162) which recognize different receptors on SHA. Some strains have different adhesins for the same receptor while other strains have multiple adhesins for the same receptor. Treatment of SHA with neuraminidase reduced its ability to bind S. sanguis C5, did not affect binding of strain FC-1 (69) and completely eliminated binding of strain 12 (162). Morris and McBride (162) compared the adherence of S. sanguis 12 and S. sanguis 12 na (a variant which did not aggregate) to SHA. Their results showed that strain 12 bound to SHA via two types of salivary receptors. One receptor was sensitive to neuraminidase and the other was sensitive to preincubation at pH 5.0 (37°C) . Strain 12 na has lost the adhesin which binds to the neuraminidase sensitive receptor, but retains the adhesin binding to the pH sensitive receptor. Based on this data a two site binding model was proposed (162). This brief review of the literature clearly shows the complexity of adherence in the oral cavity. Oral organisms have the possibility of binding to host tissue or to other bacteria. The interaction between the microorganism and the surface to which it is attached can be attributed to host synthesized polymers (eg glucans). Regardless of the mechanism, the phenomenon of adherence is characterized by a high degree of specificity. Bacteroides gingivalis The following section will encompass a detailed discussion of Bacteroides gingivalis with emphasis on its adherence properties and other potential virulence factors. 13 B. gingivalis is a gram-negative, non-motile, non-sporulating anaerobic rod. The species is usually defined as short rods but often appears as rods with rounded ends which vary in size and shape from small coccoid forms to long filamentous organisms. When grown on blood agar plates, B. gingivalis and other asaccharolytic Bacteroides produce a black pigment. The chemistry and location of the black pigment gave rise to considerable speculation, but eventually it was shown that it was protohemin with traces of protoporphyrin (177,194,212) Most strains of B. gingivalis require hemin or are stimulated by it and many require vitamin k (menadione) for growth (70,98). The roles of vitamin k and hemin for the growth of B. gingivalis are not fully understood. Gibbons and McDonald (70) have speculated that vitamin k acts as an electron carrier in an electron transport system. This proposal was later supported by Shah et al (210). Vitamin k has also been found to stimulate synthesis of phosphosphingolipids in the cell envelope suggesting a role in membrane permeability (126). It has been suggested that the dark pigment produced by the Bacteroides is a mechanism for storage of hemin. This is consistent with the report of Rizza et al (194) who showed that when B. gingivalis was transferred from high concentrations of hemin to a medium lacking hemin that the cells divided about 8-10 times suggesting that heme had accumulated within the cell. (a) Classification Originally all oral anaerobes isolated from the oral cavity, urine and feces which produced a black pigment when grown on blood agar plates were classified as Bacteroides melaninogenicum (177). In 1947 Schwabacher et al ( 205 ) proposed a new name Fusiformis nigrescens based on pigmentation but this name was rejected and the 7 t h edition of Bergeys Manual (203) retained Bacteroides melaninogenicus. 14 Although Gibbons an coworkers (70) in the 1960>s showed biochemical and immunological heterogeneity among strains of B. melaninogenicus, only one species of black pigmented Bacteroides (BPB) was recognized. An increased understanding of microbial physiology coupled with analysis of metabolic end products, it became evident that B. melaninogenicus could be divided into several subspecies" as a result of their fermentative activities. In 1970 Holdeman and Moore (97) described three different subspecies of B. melaninogenicus on the basis of biochemical activity and quantitative and qualitative differences in the production of volatile fatty acids. The saccharolytic strains were divided into two subgroups: B. melaninogenicus subsp. melaninogenicus which was strongly fermentative and B. melaninogenicus subsp. intermedius which was weakly fermentative. Assacharolytic strains, that is those which did not lower the pH of a glucose-based medium, were grouped in B. melaninogenicus subsp. asaccharolyticus. The studies of Finegold and Barnes (57) showed clearly that the biochemical and genetic characteristics of the saccharolytic and asaccharolytic strains were sufficiently different to justify elevation of the asaccharolytic subspecies to species level. Further studies by Shah et al (209) and van Steenbergen (240) led to the separation of oral and non-oral B. melaninogenicus species based on their genetic heterogenicity particularly among the asaccharolytic strains. This led Coykendall et al (31) to propose the new species B. gingivalis for the asaccharolytic strains isolated from oral sites. B. asaccharolyticus was retained for the non-fermentative Bacteroides sp. from non-oral sites. Results from deoxyribonucleic acid hybridization experiments indicated little similarity between oral and non-oral types of asaccharolytic Bacteroides. The DNA base content of B. gingivalis varied from 46.5 to 48.4 mol % G+C, while that of B. asaccharolyticus varied between 49.2 and 53.6 mol % G+C. Recently, van Steenbergen (240) re-evaluated two strains of B. asaccharolyticus isolated by Sunquist ( G. Sunquist Ph.D thesis, University of Umea. Umea Sweden 1976). 15 Though these strains were classified as B. asaccharolyticus, they had little or no DNA homology with either authentic B. asaccharolyticus or B. gingivalis. A new name B. endodontalis was proposed for these strains. Recently Shah and Collins (211) proposed a new genus Porphyromonas for these closely related species of B. gingivalis, B. endodontalis and B. asaccharolyticus. There are a number of black pigmented bacteroides (BPB) which have been isolated from the mouth. These include Bacteroides denticola, Bacteroides loescheii, Bacteroides buccae and Bacteroides macacae. (b) Ecology The principal habitat of BPB is the oral cavity where they are found in the gingival crevice and in the supragingival dental plaque. There are major differences in the frequency in which different BPB can be isolated from the oral cavity. In the healthy gingival sulcus less than 2 % of the cultivable microflora is made up of BPB (217). Various BPB have been found to be associated with different forms of periodontal d i s e a s e s . B. gingivalis is found in low numbers if present at all, in healthy and gingivitis sites in the oral cavity (44,209,216). In chronic gingivitis the predominant isolate is B. intermedius (216). In adult periodontitis the numbers of B. gingivalis increases significantly (218,250). This species may constitute in excess of 65 % of the flora in the deep periodontal pockets, with the severity of the disease correlating significantly with the percentage of B. gingivalis present (233). (c) Serology Immunological analysis has proven to be an important tool in classifying closely related microorganisms (202). In the BPB early serological work revealed the heterogeneity of antigens in different strains that were classified as B. melaninogenicus (29). This is 16 not unexpected since it was subsequently shown that they were different species. The cross reaction of B. gingivalis with other BPB has been investigated by a number of groups. Mansheim et al (144) demonstrated that antisera to capsular material from B. gingivalis was not serologically cross-reactive with similar material from B. intermedius and B, melaninogenicus. They later showed by immunodiffusion and immunofluoresence that there was no serological cross-reactivity between outer membrane antigens from oral and non-oral strains of the asaccharolytic Bacteroides (146). The antigenic specificity of B. asaccharolyticus and B. gingivalis was demonstrated by Reed et al (192) using immunodiffusion and Immunoelectrophoresis. In their study it was shown that none of the strains of B. asaccharolyticus (obtained from non-oral sites) was antigenically similar to B. gingivalis. Furthermore, no common antigens were observed between the asaccharolytic and saccharolytic BPB strains. These immunogenic differences between B. gingivalis and other oral Bacteroides were further revealed by the use of monoclonal antibodies to B. gingivalis. Naito et al (169) found that five monoclonal antibodies raised to surface antigens of B. gingivalis reacted with sonicated antigen from all strains of B. gingivalis tested. No serological cross-reactivity was seen with antigens from nine other species of BPB tested. Hanazawa et al (88) showed that three of four monoclonal antibodies to B. gingivalis did not react with any other Bacteroides whereas one monoclonal antibody reacted with antigen from B. intermedius. B. gingivalis can be divided into at least two serotypes. Using crossed-immunoelectrophoresis, Parent et al (184) compared the surface antigens of oral B. gingivalis from humans and animals. They identified serotype-specific and cross reacting antigens. The human biotype exhibited 25 surface antigens, 2 of which were specific for the biotype. The animal biotype had 12 surface antigens and 2 were specific 17 for the biotype. Four common antigens were found in all strains of B. gingivalis. Fisher et al (58) demonstrated two serogroups within human strains of B. gingivalis and these groups appeared to be correlated with virulence. Support for the concept of antigenic differences comes from recent work of Bramanti and Holt (14) which indicated that there was a relationship between the pathogenicity of B. gingivalis strains and the electrophoretic patterns of cell envelope proteins. Virulent strains (W 83, W 50) exhibited polypeptide bands at 56 and 49 kd, while avirulent strains showed bands at 72, 53 and 37 kDa. (d) Pathogenicity There is a large body of evidence which show that B. gingivalis is an important oral pathogen. Experimental infections with Bacteroides sp. were first reported in 1937 by Weiss (247) who demonstrated that two clinical isolates of B. melaninogenicus were pathogenic in dogs. Later, direct evidence was obtained which showed that Bacteroides sp. played an important role in the pathogenesis of transmissible mixed anaerobic infections in animal models (41,223,229). Combinations of bacteria isolated from various sources ( plaque, necrotic dental pulp and periodontal pockets) were tested for their capacity to induce abscess formation and transmissible infections when injected subcutaneously into guinea pigs. The results showed a synergistic effect; i.e individual bacterial species or some Bacteroides were not able to induce an infection whereas mixtures of two or more bacteria did produce an infection (83). Most of these experimental infections have another common trait: when the BPB were present in the infections mixture, it was possible to produce a transmissible infection; when the BPB were deleted from the innocula a transmissible infection was not produced, indicating a key role for the Bacteroides species (151). Observations that the BPB can produce animal infections in pure culture have been infrequent. The first such observation was made by McDonald et al (139) who noted that 18 B. melaninogenicus CR2A (B. gingivalis) could produce an infection in guinea pigs when injected in pure culture. This infection was transmissible to other animals. It was also shown that this organism could produce an infection in conventional and germ free mice (224). More recently it has been shown that a number of strains of fl. gingivalis ( W 50, W 83, BH 18/10, 22B4 and RB46D-1) are capable of producing infections when as pure cultures (82,232). Kastelein et al (114) demonstrated S. gingivalis W 83 virulence in guinea pigs and Swiss mice. The guinea pig has been used extensively for the study of infection involving Bacteroides because it was thought that mice were less susceptible to these bacteria (223). However, recently Neiders et al (170) using a mice abscess model showed that eight B. gingivalis isolates from plaque associated with periodontal disease produced secondary lesions when injected intraperitoneally on the abdomen. Histological examination of the tissue showed evidence of invasion of the connective tissue. (e) BPB and periodontal disease The association of any organism with a specific disease process is not easily proven. The first step in determining the etiological agents of different periodontal disease would be to define the microbiota present in sulci of healthy individuals and individuals with different clinical syndromes (36). In the healthy sulci the majority of the organism found are gram-positive rods and cocci. Tanner et al (233) found that for adults with periodontitis the microflora changed to one with a predominance of gram-negative rods. White and Mayrand (250) studied the flora of healthy gingival crevices and periodontal pockets to determine if a relationship exists between the presence of B. gingivalis and the severity of the disease. Samples from patients with severe inflamation showed a greater proportion of S. gingivalis. They found also that 6. gingivalis was not present in the healthy sulci. Zambon et al (262) in a similar type of study found a positive correlation between the proportion of BPB and both the severity of gingival 19 inflammation and periodontal pocket depth suggesting that these organism may contribute to the pathogenesis of certain forms of periodontal diseases. In conjunction with these studies, several researchers did comparative analysis on patients that were under remission from periodontitis after conventional antibiotic and other treatments. Loesche and co-workers (134) treated 5 advanced periodontitis patients with metronidazole. They found that isolates of B. gingivalis comprised from 35-40 % of the subgingival microflora prior to treatment and decreased to 0 % of the flora immediately after treatment. At 6 months after the treatment, the proportion of B. gingivalis had not returned to pretreatment level and neither had the disease. Bacterial infections are frequently accompanied by an immune response that is specific for the pathogenic organism. Several studies have been carried out correlating the presence of serum antibodies to that of BPB especially B. gingivalis. Using an Elisa assay Mouton et al (164) measured levels of antibody specific for B. asaccharolyticus in serum samples obtained from new born infants, children, periodontal normal adults and endentulous adults, patients with adult periodontitis, juvenile periodontitis and acute necrotizing ulcerative gingivitis. They found elevated levels of serum IgG to B. gingivalis in patients with adult periodontitis. Similar results were obtained by Patters and Korman (185) who found that subjects with periodontitis had significantly greater anti-fi. gingivalis IgG and IgA than did periodontal^ healthy subjects. In a more recent study Ebersole et al (50) assessed human systemic antibody levels to B. gingivalis and B. intermedius in subjects with juvenile periodontitis or adult periodontitis and in normal persons. Elevated anti-S. gingivalis serum IgG levels were seen in adult with periodontitis. 20 Virulence factors It can be seen from this brief discussion that there is strong evidence implicating B. gingivalis as an important oral pathogen. The next obvious question was to define the determinants which make this organism a pathogen. Specific adhesins, proteases, endotoxins and capsules have all been looked at as potential virulence factors (142). (1) Adherence B. gingivalis grows in the sulcus where it is exposed to the cleansing action of crevicular fluid and thus it must be able to adhere (9). The organism has been shown to possess a number of adhesive properties including the ability to agglutinate erythrocytes, bind to epithelial cells and to coaggregate with a number of oral plaque organisms (214). (a) HaemannltJtination The haemagglutinating activity of the BPB was first studied by Okuda and Takazoe (175) who showed that the reaction was limited to B. gingivalis. They found that saliva and serum inhibited haemagglutinating activity of B . gingivalis. Confusion still exists as to the nature of the adhesin responsible for the haemagglutinating activity of this bacteria. Okuda et al (175) suggested that the haemagglutinating activity was due to a fimbriae. They showed that fimbriated strains haemagglutinated whereas non-fimbriated strains did not and therefore they concluded that the adhesin was fimbrial. Slots and Gibbons (214) showed that partially purified pili (fimbriae) from 6 haemagglutinin (HA) positive strains of B. asaccharolyticus possessed haemagglutinating activity and concluded that fimbriae were responsible for haemagglutination. However, electron microscopic examination of the preparations revealed the presence of aggregated pili mixed with globular material. Attempts to eliminate the globular material were unsuccessful and thus it was questionable whether 21 the fimbriae were responsible for the haemagglutinating activity. LPS and surface polysaccharides did not exhibit haemagglutination nor did they inhibit the reaction (174). Boyd and McBride (11) fractionated the outer membrane of B. gingivalis and found that the haemagglutinating activity was associated with a fraction containing low molecular weight LPS, protein and loosely bound lipid. They also demonstrated that there was no correlation between the presence of fimbriae and haemagglutination. This result was further reinforced by Suzuki et al (231) who screened 53 isolates of B. gingivalis for fimbriae by negative staining and immunological methods.They found that over 50 % of the isolates had no fimbriae but had haemagglutinating activity, thus casting further doubt on the hypothesis that fimbriae were the HA of B. gingivalis. Recently fimbriae from B. gingivalis have been isolated and their morphological, biological, immunological and chemical properties characterized (260,261). These structures are heat stable, curly filaments approximately 5 nm in width which are devoid of haemagglutinating activity. The fimbrilin subunit which can be dissociated by boiling in SDS has an apparent molecular weight of 43 kDa. Dickinson et al (40) cloned the gene encoding the fimbrial subunits into E. coli. They found that the gene was present in a single copy on the bacterial chromosome and that the predicted size of the mature protein was 36 kDa. The protein sequence had no marked similarity to other fimbrial sequences and no homologous sequences could be found in other BPB spp. suggesting that the fimbrillin represents a class of fimbrial subunit protein of limited distribution. There have been several claims to the purification of the HA from B. gingivalis. Recently Inoshita et al (100) isolated an exohaemagglutinin from the culture medium of B. gingivalis by ultracentrifugation followed by gel filtration and affinity chromatography. Biochemical analysis and S D S - P A G E revealed that the isolated HA contained 3 major protein bands and no detectable LPS. Haemagglutination inhibition 22 studies showed that the HA was inhibited by L-arginine and arginine containing peptides suggesting that the arginine residues may be part of the ligand-binding sites on erythrocytes. Okuda et al (176) also isolated a HA from culture supernatant, in this case ammonium sulphate precipitation was followed by column chromatography with a hydrophobic column, DEAE-Sephadex and Sephadex G-100 gel filtration columns. When the purified HA was analysed on S D S - P A G E several protein bands were present. In electron microphotographs of the preparation, the HA appeared to have a vesicle or tube like structure. Haemagglutinating activity was destroyed by heating and inhibited by L-arginine and L-lysine. They suggested that the guanido group on arginine may act as a contact residue between the adhesin and the receptor on erythrocytes. Monoclonal antibody against the HA bound to cell surface and inhibited haemagglutinating activity of both the cells and the purified HA. Recently, Mouton et al (163) identified a cell bound HA of B. gingivalis by use of crossedimmunoaffinity electrophoresis. A polyclonal antiserum with specificity restricted to the HA was produced and on immunoblotting experiments two antigens with apparent molecular weights of 33 and 38 kDa were detected. The antiserum inhibited haemagglutinating activity. (b) Bacterial coaggregation As indicated, B. gingivalis attaches to other bacteria associated with dental plaque. Slots and Gibbons (214) showed that B. gingivalis coaggregated with several species of common oral isolates including S. sanguis, S. mitis, S. mutans, S. salivarius, A.viscosus, A. israelii and A. naeslundii. This interaction unlike haemagglutination was not inhibited by either saliva or serum suggesting that the reaction was critical for the survival of B. gingivalis in the oral cavity. The inhibition data indicated that there was two types of adhesin on the surface of B. gingivalis. Boyd and McBride (11) demonstrated that the outer membrane contained the coaggregation activity . They were 23 able to separate the HA and the bacterial agglutinin (BA) adhesins. The coaggregating preparation contained a number of protein, carbohydrates and high molecular weight LPS but no fimbriae like structure. Schwarz et al (206) using A. viscosus attached to salivary coated hydroxyapatite (actinobeads) and ^H-labelled B. gingivalis carried out a quantitative evaluation of the cohesion of B. gingivalis and A. viscosus. They found that B. gingivalis adhered tenaciously to the actinobeads , thus showing that cohesion with bacteria in preformed plaque could fosters establishment of B. gingivalis on teeth. With the exception of Boyd and McBride (11) no attempt has been made to identify or isolate the adhesin responsible for coaggregation of B. gingivalis. (c) Adherence to other substrates In addition to haemagglutination and bacterial coaggregation, B. gingivalis has been shown to bind to a number of other substrates found in the oral cavity. Lantz and coworkers (123) reported that strains of B. gingivalis specifically bind and degrade fibrinogen. Binding was rapid, reversible and saturable. They speculate that interaction with fibrinogen may mediate colonization and establishment of B. gingivalis in the periodontal microbiota. In another study, Winkler et al (253) demonstrated the attachment of B. gingivalis to an in vitro basement-membrane like matrix and to selected individual macromolecular constituents of this matrix. Preincubating the bacteria with fibronectin decreased the binding of B. gingivalis. Similar results were obtained when the matrices were preinubated with anti-fibronectin antibodies. Recently Naito and Gibbons (168) studied the attachment of B. gingivalis to hydroxyapatite beads treated with either human type 1 or type 1V collagen or to 24 particles of bovine collagen. They found that collagen on hydroxyapatite surfaces promoted attachment . B. gingivalis also bound to particles of collagen. Heating of bacteria abolished their adhesive properties. Human serum, fibronectin and protease inhibitors were inhibitory to the reaction. Cimasoni et al (23) showed that B. gingivalis can attach to serum, saliva and crevicular fluid coated hydroxyapatite beads. (2) Collaoenase Some types of periodontal diseases are characterized by the loss of gingival connective tissue. Collagen is the major constituent of the gingival connective tissue and while this protein is resistant to a wide variety of proteolytic enzymes, it is degraded by bacterial and tissue collagenases (133). For this reason the presence of collagenase is an important factor in the pathogenesis of any oral pathogen. Gibbons and McDonald (70) were the first to demonstrate the collagenolytic activity of B. gingivalis. The enzyme was found to be cell associated and to be stimulated by reducing agents such as dithiothreitol and cysteine (149, 198). Lantz and her associates ( J . Dent. Res. 289, Abstr. 1076, 1983) indicated that the enzyme may be within the periplasmic space. Maximum activity was seen when the cells were grown in peptide deficient medium or when the culture reached stationary growth phase (150). Roeterink et al (200) showed that among the BPB, B. gingivalis produced the most damage to collagen tissue of the palate of mice. Mayrand and Grenier (150) showed that B. gingivalis is the only BPB to produce a true collagenase. As opposed to eukaryotic collagenases which cleave undenatured collagen at a single site, the collagenase from B. gingivalis hydrolysis collagen into small peptides (149, 198). B. gingivalis also produces a factor which induces the production and activation of mammalian procollagenase (H. Birkedal-Hansen, J . Periodontal Res. 62: 101 Abstr. 551, 1987).These results suggests that B. gingivalis has the potential to destroy periodontal connective tissue directly or indirectly. 25 (3) Other proteolytic activity There are a number of reports of proteolytic enzymes found in B. gingivalis. Various amidopeptidases (1) thiol (178) and trypsin-like (230) proteases have been purified from this organism. Grenier et al (78) using bovine serum albumin conjugated to polyacrylamide demonstrated the presence of 8 proteases with different M f . These enzymes have been shown to be active against fibrinogen (123) fibronectin (125) and other glycoproteins found on epithelial cell surfaces and in the connective tissue matrix. One of these was purified and proved to be a pro-gly specific enzyme of molecular weight 29,000 (78). Some clinically important bacteria produce enzymes which hydrolyse serum proteins involved in host defense against microbial infections (110). Bacteroides gingivalis is capable of degrading immunoglobulins (237 ) as well as the C 3 and C 5 polypeptide of the complement system in vivo and in vitro (230). This organism is also capable of degrading human plasma proteinase inhibitors alpha-1-antitrypsin and alpha-2-macroglobulin (18). By degrading complement proteins and host proteinase inhibitors B. gingivalis can evade opsonization and phagocytosis by host cells and permits greater destruction of host tissue (227). (4) Tissue toxicity Culture supernatant or whole cells of B. gingivalis are cytotoxic for monkey kidney cells and squamous epithelial cells (79). Butyrate and propionate, two by products of B. gingivalis metabolism are potent inhibitors of several human cell lines (239). Other potentially toxic factors include LPS , ammonia, indole, volatile sulphur compounds ( hydrogen sulphide and methylmercaptan) (235). The mercaptans have been shown to increase cell permeability. 26 (5) Capsule Electron microscopic observation revealed a layer of ruthenium red material 15 nm thick covering the outer surface of several strains of B. gingivalis (90). Okuda and Takazoe (173) demonstrated that encapsulated strains of B. melaninogenicus were resistant to phagocytosis whereas non capsulated strains were not. These observations were later confirmed and expanded by van Steenbergen et al (238) who showed by chemiluminescence that virulent strains of B. gingivalis (W 83, W 50) were more resistant to phagocytosis than non-virulent strains such as 376 and HG185. The virulent strains had a thicker capsule and a more hydrophilic surface than the avirulent strains. This observation indicates that a major difference between virulent and non-virulent strains of B. gingivalis is the thickness of their capsular layer. Vesicles This section will include a general description of vesicles with emphasis on the vesicles produced by Bacteroides gingivalis and their role as a potential pathogenic factor. (1) Formation All gram-negative bacteria have the potential to produce vesicles. Vesicles are formed by a budding off of the outer membrane and they can remain attached to the cell or be released. The growth conditions appear to influence the quantity of vesicles formed. McKee et al (156) compared growth in hemin excess and hemin limitation and found that B. gingivalis released five times more vesicles when grown under hemin limiting conditions. Knox et al (115) showed that B. coli, produced large quantities of vesicles with lysine as the limiting factor. In the oral cavity several genera including H ae m o p h i I u s (39) along with several spec ies of Bacteroides (19,122,131,189,209,255) release these vesicular structures. These vesicles are similar to structures previously found in dental plaque (61,86). 27 (2) Composition Vesicles are derived from the outer membrane, and thus it is reasonable to assume that the chemical constituents of these two structures will be similar. William and Holt (251) compared outer membrane and vesicles polypeptide profiles of several Bacteroides and found that the major protein bands were similar but the vesicular preparation had more minor polypeptide bands. The LPS and carbohydrate composition of the outer membrane and vesicles have not been compared. (3) Biological activities The surface of many bacteria plays a crucial role in pathogenicity. Several determinants of virulence are surface components and therefore vesicles by virtue of their mode of formation would be expected to retain these attributes. Vesicles have been shown to possess a number of biological activities, the following sections describe some of the biological functions of vesicles. (a) Adherence It is not unreasonable to assume that vesicles should possess the same adherence properties as that of the outer membrane and a number of studies have shown that this is true. Scanning electron microscopy revealed that like the whole cells of B. gingivalis vesicles can attach to hydroxyapatite coated with saliva, serum and crevicular fluid (23). Purified B. gingivalis vesicles also binds to erythrocytes causing them to agglutinate (80). Similarly Celesk and London (19) showed that Cytophaga DR 2002 attaches to SHA beads. B. gingivalis vesicles can act as a bridge to permit the coaggregation of two bacteria which did not normally aggregate when mixed as is the case with Eubacterium saburreum and Capnocytophaga ochracea (80). Treatment of the vesicles with heat or proteolytic enzymes destroyed their ability to act as a bridge whereas similar treatment of the Eubacterium or Capnocytophaga did not affect attachment. This would indicate unimodal interaction of a vesicular protein or glycoprotein with a non-proteinaceous component on the bacterial cells. 28 The aggregative properties associated with vesicles suggests that they may play an important role in the colonization of the periodontal pocket (118). The accumulation of diverse sulcular flora could be mediated by intergeneric binding mediated by B. gingivalis vesicles. Vesicles would induce coaggregation resulting in the formation of microcolonies which could be resistant to phagocytosis by macrophage. The examples cited above relate to vesicles with properties similar to that of outer membrane. There are also reports of vesicles with different adherence properties than that of the parent cell from which they were derived. For example Kagermeir et al (109) studied the aggregation of Cytophaga with A. israelii. They found that vesicles derived from Cytophaga did not bind to A. israelii but whole cells did. They suggested two possible explanation (1) the receptors on the outer membrane are located in areas which do not give rise to vesicles, (2) the receptors are anchored to the peptidoglycan or cytoplasmic membrane of the bacteria. (b) Enzymatic activity Enzymes are another biological activity associated with both outer membranes and vesicles. Grenier and Mayrand (80) compared the proteolytic activities of vesicles and whole cells of B. gingivalis and found them to be similar. In a recent study, Grenier et al (78) characterized the proteolytic activities of outer-membranes and vesicles of B. gingivalis by S D S - P A G E using as a substrate bovine serum albumin covalently conjugated to the polyacrylamide. A total of eight major proteolytic activities were present in both outer membranes and vesicles. Pseudomonas tragi vesicles have been shown to have proteolytic activity associated with them (48). Using immunological procedures, Thompson et al (234) showed that a neutral protease was present on the cell surface and in the cell-bound vesicles. A wide variety of other enzymes have been found to be present on vesicles. For example 29 the periodontal pathogen Capnocytophaga show an increase in the secretion of acid and alkaline phosphatases during the growth cycle (189). It has been estimated that over 50 % of these enzymes were released with the vesicles. The vesicles elaborated by Actinibacillus actinomycetecomitans contain a leucotoxin. Lai et al (122) observed that leucotoxic strains of this organism produce numerous vesicles whereas non-leucotoxic strains released no vesicles. It has been postulated that vesicles, due to their small size, may reach sites not available to the whole cell. An interesting example of this is B. succinogenes a common rumen bacteria which produces large quantities of vesicles which contain more than 50% of the extracellular carboxymethyl cellulase activity (84). These vesicles are located between the adhering cells and cellulose fibres. The secreted vesicles are thought to facilitate hydrolysis of the substrate by reaching sites not available to the whole cells. It may be that vesicles by virtue of their small size, adherence properties and enzymatic activities may play a role in the pathogenesis of bacteria. Vesicles may have been responsible for the immunofluorescence seen by Pekovic and Fillery (187) at the surface as well as in the interstitial spaces between epithelial cells in the oral cavity. Lipopolysaccharides Lipopolysaccharides are a heterogenous collection of molecules unique to the outer cell envelope of gram-negative bacteria. This polymeric bacterial amphiphile is made up of three major constituents which are linked covalently (56). The lipid A portion consisting of five or six fatty acids is attached to diglucosamine phosphate (254). Covalently attached to the lipid A is the rough core region of LPS consisting of 11 to 14 saccharides including unique octoses, heptoses and a variety of hexoses substituted with phosphate and ethanolamine. This rough core is often but not always capped with a repeating structure of saccharides of variable length which bears the name O-antigen and can constitute the major antigenic structure of gram-negative cells (59,188). 30 Lipopolysaccharides may be found in non-covalent association with envelope proteins (99,182). When LPS is disassociated from its surrounding molecules there is disruption of the outer envelope which leads to a breakdown of the cell permeability barrier as measured by the penetration of antibiotics and other bactericidal agents (53). Thus LPS is critical for the structural integrity of the cell and for changes in the function of some of the LPS associated envelope proteins (38). Lipid A structure The lipid A is usually covalently attached to the octose-2-keto-3-deoxyoctonate which forms the proximal part of the rough core. The large majority of lipid A's studied contain a 6 (1-6) linked-D-glucosamine disaccharide that contains two phosphoryl group, one bound as an ester and the other to the glycosidic hydroxyl group of the reducing glucosaminyl residue (6). Free lipid A can be prepared from LPS by mild acid hydrolysis. The lipid A of Salmonella has been subjected to intense analysis and its structure has been elucidated. It contains straight chain and hydroxy fatty acids joined by amide linkage to C-3 of glucosamine (254). There is a high degree of chemical similarity between the lipid A's of bacteria belonging to the family Enteriobacteriaceae (13). This is further emphasized by the fact that antibody to E. coli lipid A cross reacts with the lipid A of other bacteria in this family (182). Mutharia et al (166) showed that monoclonal antibody to the lipid A from Pseudomonas aeruginosa reacted with a variety of gram-negative species tested, showing the conservation of the lipid A moiety. However, there are lipid A which show no serological cross reactivity. For example antiserum to lipid A of Salmonella does not react with lipid A of R. palustrus and R vividus (135). Biological effect of lipid A Lipid A is the endotoxic principle of gram-negative bacterial lipopolysaccharide, the polysaccharide portion serves as a solubilizing carrier for the insoluble lipid A. It has 31 been shown that the endotoxic activity of the soluble free lipid is comparable to that of intact LPS (64). LPS and lipid A exhibit a high affinity for cell membranes, for other lipids and for proteins. The lipid reacts with several different organs. For example in the liver, lipid A leads to the depletion of carbohydrate reserves, inhibition of enzyme induction and inhibition of glucogenesis (76). Cell types sensitive to LPS include macrophages, polymorphonuclear leukocytes, platelets, B-lymphocytes and fibroblasts (13). Lymphocytes are induced to undergo mitogenesis by LPS and lipid A. In addition LPS and lipid A activate a clotting enzyme on the standard lipid A assay, the Limulus amoebocyte assay (13). Lipid A is also pyogenic and induces the Schwartzman phenomenon and hemodynamic changes leading to shock and death (137,138). However, due to different chemical structures there is also a difference in levels of biological activity, for example Brucella sp. lipid A has been found to exhibit very low toxicity in experimental animals (120) whereas lipid A from Salmonella sp. displayed high toxicity in animals (65,105,193, 248). Core The core of LPS usually contains D-glucosamine, D-glucose and D-galactose and an inner lipid A proximal region consisting of an oligosaccharide of core specific sugars, L-glycero-D-manno-heptose (L-D-hep) and 2-keto-3-deoxy-D-manno-octonate (KDO) each forming a branched trisaccharide (91). Evaluation of the core structure of LPS was greatly facilitated when so-called rough (R) mutants of Salmonella were found. Depending on the defect, these mutants synthesized complete or incomplete core structures linked to lipid A but were devoid of O-chain (137). The R mutants of Salmonella were found to be non-virulent, were phagocytized without opsonization by macrophage and were sensitive to toxic agents (195). 32 O-side chain The O-specific chains of LPS are made up of repeating units of identical oligosaccharides (196) containing a number of different sugars. However, in some cases an oligomer of a single sugar type is found as in the case of E. coli 09 (190, 137). The number of repeating units usually varies and this gives rise to the typical ladder like profile of LPS when analysed on sodium dodecyl sulphate polyacrylamide gel electrophoresis (136,213). Antigenicity The O-chain contains the immunodominant structures against which anti-LPS antibodies are directed during infection or on immunization (63,64,136). It has been found that active vaccination with E. coli J-5 LPS and passive immunization with antiserum to J -5 protects against bacteremia from a variety of gram-negative species (15). It was found that immunization with LPS and the polysaccharide side chain, promoted clearance of P. aeruginosa (32,89). Virulence The role of LPS in the virulence of P. aeruginosa was studied in mice (33). The virulence of several strains was found to be related to the state of the O-polysaccharide side chain. Strains with complete side chains were far more virulent than those with abbreviated structures. Strains without O-side chains were avirulent and were cleared rapidly. These strains were also serum sensitive indicating that LPS O-chain polysaccharide acts to mask the complement binding site that initiates cell lysis (33). B. gingivalis LPS (1) Composition The LPS from BPB have not been subjected to the same extensive analysis as has the LPS from Salmonella and Escherichia coli. It has been reported that the LPS from BPB lacks 2-keto-3-deoxyoctonate (KDO), heptose and B-hydroxy myristic acid which are 33 common to enterobacterial LPS (95,113). However, recently using mass spectroscopic and gas-liquid-chromotographic analysis (103) revealed the presence of small quantities of these compounds in LPS from certain strains of B. gingivalis and B. intermedius. Chemical analysis of the O-side chain showed the presence of both neutral and amino sugars with rhamanose, mannose, galactose and glucose the most predominant neutral sugars and glucosamine the predominant amino sugar (102,117,145). The lipid A portion was shown to contain over 20 fatty acids (103) with a predominance of palmitic acid, hexadecanoic acid and stearic acids (117,145). Some investigators could not detect 6-OH-myristic acid in some preparations (140) while others reported the presence of hydroxy fatty acids (167). (2) Mitooenicitv The mitogenic effects of LPS from BPB were studied and compared to that of E. coli (117). It was found that the two LPS'S affected the LPS-nonresponsive C3H/HeJ spleen cells equally. This response was specific to B cells . This may have biological significance since it has been reported that established lesions of chronic inflammatory periodontal disease in humans are primarily infiltrated with B-cell (49,141,208). (3) Bone resorption The correlation between the incidence of B. gingivalis and alveolar bone loss is well documented (215,218). Millar et al (158), using a low and high molecular weight fraction of LPS in a bone resorption assay found that both fractions stimulated the release of 4 5 C a from prelabelled fetal rat bones. More importantly they found that collagen formation was reduced by as much as 40% . These two results suggest that LPS may play an important role in alveolar bone loss that is associated with periodontal disease. Koga et al (117) found that B. gingivalis LPS stimulated interleukin-1 production from macrophage to activate thymocytes. IL-1 is known to have many activities including enhancement of immune responses, stimulation of thymocyte 34 proliferation, activation of B-cells and stimulation of bone resorption. These findings show that one of the mediators of alveolar bone loss in human periodontitis could be IL-1 stimulated by B. gingivalis LPS. (4) Endotoxic effects The biological potency of LPS from B. gingivalis and other BPB is low in comparison to LPS from aerobic gram-negative organisms when assessed by pyrogenicity, chick embryo lethality and the Schwartzmann test (167,124). Mansheim et al (145) found that LPS preparations gave no positive skin reactions in rabbit in doses up to 1mg, whereas 12.5 u.g oiSalmonella typhi endotoxin gave a positive reaction. Limulus lysate gelation and chick embryo lethality occurred at doses 30-fold or more greater than that in Salmonella endotoxin controls. (5) Serology Dahlen and Baltzer (34) found that antibody to lipid A of B. gingivalis did not cross react with lipid A from any of the oral anaerobes they studied. Anti-lipid A from Enterobacteriaceae did not cross react with lipid A from any of the BPB. These results confirm that Bacteroides species have a unique lipopolysaccharide structure (96). The studies to be described in this thesis were undertaken with the objective of identifying, isolating and characterizing the adhesins used by B. gingivalis in its attachment to host tissue and to other bacteria in dental plaque. In addition the LPS of a number of the BPB was studied. 35 Material and methods Bacteria and cultivation Bacteroides gingivalis 33277, B. levii 29147, B. denticola 33185, B. asaccharolyticus 25260 were obtained from the American Type Culture Collection (Rockville, Md.). B. gingivalis W 12, W 50 and W 83 were obtained from G . Bowden, University of Winnipeg, Manitoba, Canada. B. intermedius BMA and 2332H and B. gingivalis 2D were isolated and characterized in our laboratory. Escherichia coli JM 83 [ara A (lac-pro AB) rspL (=strA) 0 80 lacZAM15] and JM 101 [supE thi A (lac-proAB) {F'traD36 proABIaclq lacZ AM15}] were used as recipients (241) for the plasmid vector pUC 18 [Apr] (259). Streptococcus mitis, S. sanguis, S mutans and S. salivarius were human oral isolates. All bacterial cultures were stored in 10 % glycerol at - 7 0 ° C . Each experiment was started from a frozen stock culture. Bacteroides strains were grown in Brain-Heart Infusion broth (BHI) (BBL Microbiology systems, Cockeyville, MD.) supplemented with hemin (5 ug/ml) and menadione (0.5 ug/ml) at 3 7 ° C in an anaerobic chamber (Coy Manufacturing, Ann Arbor, Michigan) containing an atmosphere of N 2 : H 2 : C 0 2 (85: 10: 5) for 48 h. When solid medium was needed, agar (1.5 %) was added to BHI broth together with whole blood at a final concentration of 5 %. S. mitis and other streptococci were grown at 3 7 ° C in trypticase soy broth (TSB, BBL Microbiology systems, Cockeysville, MD.) supplemented with yeast extract (3 g/liter). E. coli JM 83 and JM 101 were grown in Luria-Bertani broth (10 g tryptose, 5 g NaCl, 5 g yeast extract per liter, pH 7.5) medium at 3 7 ° C . When solid medium was needed, agar (1.5 %) was added to LB broth. Ampicillin, 100 ug/ml, (Sigma) was added to the medium when required. For adherence assays, S. sanguis 12 was grown overnight in presence of 1.5 u.Ci of [3H]-methyl-thymidine per ml of culture medium. The cells were harvested by centrifugation, washed three times in 50 mM HEPES buffer (pH 7.2), sonicated for 5 sec (30 % duty cycle, output 5, Sonifier Cell Disrupter 350), washed twice more in 36 HEPES buffer and finally resuspended to an absorbance of 4.0, as measured at 580 nm. Isolation of the haemagglutinin B. gingivalis was grown for 48 h after which the cells were harvested by centrifugation at 10,000 x g for 10 min. Cells were washed twice in 0.05 M phosphate buffer (pH 7.4) and resuspended to 1/10 the original culture volume in 20 mM Tris-hydrochloride (pH 7.4) containing 0.15 M NaCl and 10 mM MgCI 2- The suspension was stirred for 15 min at room temperature on a magnetic stirrer. The cells were pelleted by centrifugation and the proteins in the supernatant were precipitated by the addition of ammonium sulphate to 40 % . After stirring at 4 ° C overnight, the precipitated proteins were collected by centrifugation and resuspended in a small volume of 20 mM Tris-HCl buffer (pH 8.0). This material was dialysed extensively against 3 mM Tris-HCl (pH 8.0). The dialysate was clarified by centrifugation at 10,000 x g for 20 min. The dialysate was then subjected to ion-exchange chromatography, on a column (16x1.5 cm) of DEAE-Sepharose which was equilibrated with 20 mM Tris-HCl (pH 8.0). After loading the sample, the column was washed with 100 ml of 20 mM Tris-HCl (pH 8.0), eluted with a linear gradient of 0-0.3 M NaCl in 20mM Tris-HCl, (pH 8.0). Fractions of 2 ml were collected and screened individually for haemagglutinating activity and analysed on S D S - P A G E . Fractions which exhibited haemagglutinating activity and showed a single silver staining band at 43 kDa on 12 % S D S - P A G E were pooled and lyophilized. This lyophilized sample was resuspended in 20 mM Tris-HCl (pH 8.0) and applied to a Sephacryl S-1000 (Pharmacia Fine Chemicals Uppsala, Sweden) column (80 x 1.5 cm) equilibrated with 20 mM Tris-HCl buffer (pH 8.0). The column was eluted with 20 mM Tris-HCl (pH 8.0) containing 50 mM NaCl. Fractions of 3 ml were collected and were screened individually for haemagglutinating activity and analysed on SDS-PAGE. 37 Haemagglutination assay Haemagglutinating activity was assessed in U-bottom polystyrene microtitre plates (Dynatech Corp.) with formalized sheep or human erythrocytes resuspended in phosphate buffered saline (PBS) pH 7.2 at a concentration of 2 % (v/v). B. gingivalis and recombinant clones were washed twice in PBS and resuspended to an A 6 6 0 of 1 and 3 respectively. Equal volumes (100 u.l) of erythrocytes and bacterial cells or fractionated samples were mixed together and the plates shaken for 20 min after which they were allowed to sit for 1h. Results were scored on the basis of the settling pattern of the erythrocytes. To establish aggregation titres, the bacterial cells or fractionated samples were serially diluted in PBS (pH 7.2) and the last well containing visibly detectable aggregated erythrocytes was taken as the end point of the reaction. Haemagglutination inhibition assays (a) A number of substances were tested for their ability to block the haemagglutinating activity of B. gingivalis. The materials to be tested were dissolved in PBS (pH 7.2) to the concentration indicated, serially diluted in 50 u.l of the same buffer and mixed with an equal volume of bacterial cells or purified haemagglutinin (HA). The concentration of HA or the number of bacterial cells used was twice the highest dilution which gave clear haemagglutination in a preliminary titration with erythrocytes. The mixtures were incubated at room temperature for 30 min with shaking and then 100 ul of formalized red blood cells was added. The mixtures were incubated for 20 min and the titre determined. (b) Antisera against whole cells, the prepared haemagglutinin and the 43 kDa protein were evaluated for their ability to inhibit haemagglutination. The antisera were serially diluted in two fold steps in PBS after they were standardized for protein (100 u.g/ml). The concentration of purified HA or the number of bacterial cells used was twice the highest dilution which gave clear haemagglutination in a preliminary titration with 38 erythrocytes. The mixtures were gently shaken for 30 min after which 100 uJ of 2 % erythrocytes was added to each well. After incubation for a further 20 min the plates were allowed to sit for 1h before the inhibitory titre of the antisera were recorded. Enzymatic treatment Intact cells of B. gingivalis and the recombinant E. coli were resuspended ( A Q 6 0 3.0) in H E P E S (N-2-hydroxy-ethylpiperazine-N'-2-ethanesulfonic acid) buffer ( pH 7.2) and incubated with an equal volume of enzyme at 3 7 ° C for 60 min. The enzymes and appropriate buffers were pronase, proteinase k, mixed glycosidases, papain, lysozyme, mutanolysin, chymotrypsin, phospholipase C (HEPES buffer, pH 7.2) trypsin, (20 mM phosphate buffer, pH 8.0) and papain (0.05 M acetate buffer, pH 5.8). All enzymes were used at a concentration of 1 mg per ml. After incubation, the reaction mixtures were chilled and the cells washed three times with ice-cold PBS (pH 7.2). The cells were resuspended in PBS to their original volume and then assayed for haemagglutinating activity at 4 ° C . The purified haemagglutinin was treated in similar fashion and its haemagglutinating activity was evaluated at 4 ° C without removal of the enzymes. Controls without bacterial cells or purified HA were included for each enzyme. Sodium periodate treatment B. gingivalis cells and the purified HA were suspended in 0.12 M acetate buffer (pH 5.0) containing 0.05 M sodium periodate and incubated at 4^C for 2 h. Subsequently the cells were washed three times and suspended in PBS buffer to an A 6 6 0 n m of 3.0 and assayed for haemagglutinating activity. The purified adhesin was also treated in similar manner and then assayed for activity. Controls containing only sodium periodate were incubated similarly and were assayed for haemagglutinating activity. B-Mercaptoethanol Washed cells and purified adhesin were suspended in PBS (pH 7.2) containing 0.05 M 6-mercaptoethanol and incubated for 2 h at room temperature. The cells were washed 39 three times and resuspended in PBS buffer (pH 7.2) and assayed for HA activity. The purified haemagglutinin was also assayed for activity but without the removal of the reducing agent. Appropriate controls containing reducing agent but no bacterial cells or HA were included. Heat B. gingivalis cells or purified adhesin were suspended in PBS buffer (pH 7.2) and heated in a water bath at 7 0 ° C , 8 5 ° C and 100°C for 1, 5, 10 and 30 min after which they were assayed for haemagglutinating activity. Electrophoretic techniques Sodium dodecyl sulphate polyacrymlamide gel electrophoresis was carried out with polyacrylamide concentrations of 12 %, 15 % or a gradient of 5 to 20 % using the buffer system of Laemmli (121). Non-denaturizing PAGE was carried out in the absence of SDS in the gel and buffer systems. For polypeptide analysis samples were solubilised in 0.125 M Tris-HCl buffer containing 4 % S D S , 20 % glycerol, 10% 8-mercaptoethanol and 0.01 % bromophenol blue (pH 6.8). For lipopolysaccharides (LPS), 20 % sucrose was added. The mixtures were boiled for 5 min unless stated otherwise. Molecular weight standards were: myosin (200,000), phosphorylase B (97,400), bovine serum albumin (68,000), ovalbumin (45,000), 3-chymotrypsinogen (25,700), 6-lactoglobulin (18,400) and cytochrome C (12,300). Gels were stained for protein with Coomassie Brilliant blue or silver nitrate and for LPS by the method of Tsai and Frasch (236). For Western blotting (17), material was transferred electrophoretically onto nitrocellulose paper in 25 mM Tris-HCI-192 mM glycine and 20 % methanol buffer (pH 8.3). Transfer was carried out at 25 V for 18 h in a Bio-Rad Trans-Blot Cell after which the voltage was increased to 60 V for 2 h. After blocking the unreactive sites with bovine serum albumin the transblotted material was reacted with specific antiserum and 40 visualized by the procedure described in the Bio-Rad Immuno-Blot (GAR-HRP) assay kit. Prestained markers were used to calibrate molecular weights in Western blots. Preparation of B. ainaivalis chromosomal DNA Chromosomal DNA from B. gingivalis was prepared by established methods. Briefly, cells from 500 ml of a log phase culture B. gingivalis cells were pelleted by centrifugation and washed twice in T E S buffer (20 mM Tris-HCl, 5 mM EDTA, 100 mM NaCl, pH 7.0). Cells were resuspended in T E S buffer, SDS and RNase A were added to a final concentration of 1.6% and 50 ug/ml respectively and the mixture was incubated at 3 7 ° C until cell lysis was complete (30-60 min). Proteinase k was then added to a final concentration of 25 ug/ml and incubation was continued for 1h. The lysate was extracted twice with equal volumes of phenol and once with chloroform before the DNA was precipitated with 95 % ethanol. The DNA was further purified by cesium chloride density gradient centrifugation. Plasmid DNA was isolated by an alkaline lysis procedure (143) and separated from RNA and chromosomal DNA by cesium chloride density gradient centrifugation in the presence of ethidium bromide. Cesium chloride and ethidium bromide were removed by dialysis and butanol extraction respectively. The DNA was then extracted with phenol-chloroform, ethanol precipitated and stored in T E buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) at 4 ° C . Construction of genomic library Partial digestion of B. gingivalis DNA for cloning was carried out as described by Darzins et al (37) with some modifications. The optimal time for digestion of 50 ug of DNA with 5 units of Pst 1 restriction enzyme at 3 7 ° C was worked out by removing aliquots from the reaction mixture at different times and analysing on a 0.7 % agarose gel. Chromosomal DNA was partially digested with Pst 1 enzyme and separated 41 electrophoretically on a 0.7 % agarose gel. Fragments of 3-9 kb were electroeluted and ligated to alkaline phosphatase treated pUC 18 vector which had been linearised with Pst 1 enzyme. T4 DNA ligase (.1u.g), ATP and dithiothreitol were added to a final concentration of 0.1 and 1mM respectively. Ligation was carried out at 1 2 ° C in a final reaction volume of 20 u.l. The ligated DNA was used to transform E. coli JM 83 or E. coli JM 101 which had been made competent by calcium chloride treatment (143). After incubation at 3 7 ° C for 1 h in 1 ml of LB broth, the transformants were plated on LB Agar containing ampicillin at 100 p.g per ml, X-gal (5-bromo-4 chloro-3-indolyl-B-D-galactopyranoside)( 50 u.l of a 2 % solution) and IPTG (isopropyl-B-D-thiogalactopyranoside) ( 20 u.1 of a 2 % solution). White colonies were screened for haemagglutinating activity as follows: Colonies were individually picked and macerated in 100 U.I of PBS. 100 u.l of erythrocytes were added to each well, the plates were shaken for 20 min at room temperature and allowed to sit for 1 h before being scored for•>••-haemagglutination. Colony immunoblot assay White colonies carrying recombinant plasmids were replica plated on LB agar supplemented with ampicillin and X-gal and incubated at 3 7 ° C overnight. The cells were lysed by exposure to chloroform vapour for 15 min. Nitrocellulose filters (BA85; Schleicher &Schuell, Inc., Keene, N.H.) were blotted onto the agar plates for 1 h and air dried after which the filters were placed in a blocking solution of 3 % bovine serum albumin in PBS. Blots were then reacted with antibody raised to whole cells of B. gingivalis. The antisera had been extensively adsorbed with E. coli 83 (pUC 18). Following removal of the first antibody the blots were treated as described in the Bio-Rad Immuno-Blot (GAR-HRP) assay kit as described previously. Restriction analysis of recombinant plasmids Plasmids were isolated from recombinant clones by the alkaline lysis method as described by Maniatis et al (143). Restriction endonuclease digestions using single and 42 double digestions were performed under the conditions described by the manufacturer (Bethesda Research Laboratories, Gaithersburg, MD.). The DNA fragments were resolved by electrophoresis in a 1.0 % agarose gel using 50 mM Tris-HCl, 50 mM borate, 0.002 M N a 2 EDTA buffer containing 0.5 ug P© r m ' ° f ethidium bromide. The size of DNA fragments was estimated by comparing the distance of migration to a logarithmic plot of the migration of standard restricted lambda DNA run on the same gel. Southern blot analysis The recombinant plasmid was digested to completion with Pst 1 enzyme and the fragments separated in a 0.7 % agarose gel. The insert DNA was electroeluted (Bio-Rad Electroeluter) and then nick translated using biotinylated A T P nucleotide, DNA Polymerase 1 and DNA 1/DNase 1 as described by the manufacturer (Nick Translation System, Bethesda Research Laboratories, Gaithersburg, MD.). B. gingivalis, S. sanguis and the recombinant plasmid DNA were digested with Pst 1 enzyme and the DNA fragments were separated by electophoresis on 1 % agarose gel. The DNA was transferred to a nitrocellulose membrane by Southern transfer (226) and the filter was baked at 8 0 ° C for 2 h. After prehybridization for 4 h at 4 2 ° C in 50 % formamide and 5x Denhart's solution (1 % Ficoll, I % Polyvinylpyrrolidone, 1 % BSA) in the presence of 250 ug/ml denatured sperm DNA, the bound DNA was hybridized to the labelled probe at 4 2 ° C in 45 % formamide for 18 h. The membrane was washed twice at room temperature in 2 % S S C buffer (3 M sodium acetate, 0.3 M sodium citrate, pH 7.0) containing 0.1 % SDS for 3 min each followed by washing twice in 0.16 % SSC-0.1 % SDS at 5 0 ° C for 15 min each. The membrane was incubated for 1h at 6 5 ° C in 3 % BSA in 0.1 M Tris-HCl , 0.15 M NaCl (pH 7.5) and then reacted with streptavidin-alkaline phosphatase (SAAP) conjugate for 20 min. After washing in 0.1 M Tris-HCl buffer (pH 7.5) for 30 min the filter was incubated in nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolylphosphate (BCIP). The incubation was carried out in the dark until reactive bands appeared (10-40 min). To terminate the reaction, the filter was washed for 5 min in 20 mM Tris-0.5 mM N a 2 EDTA pH 7.5. 43 Immunological procedures (a) Serology B. gingivalis 33277 and recombinant E. coli cells in the late exponential phase of growth were pelleted, washed twice in PBS buffer (pH 7.4) and resuspended in the same buffer to a concentration of 1.0 x 10 9 cells/ml. The cell suspensions were mixed with an equal volume of Freunds complete adjuvant and injected intramuscularly into New Zealand white rabbits. Seven days later the rabbit was injected with antigen suspended in Freunds incomplete adjuvant. This was followed by three weekly injections of cells in PBS, serum was collected 5 days after the last injection. Antisera against the purified adhesin was prepared in similar fashion using 50 ug of protein per injection. To obtain antisera to the 43 kDa and 46 kDa proteins, rabbits were immunized with purified protein eluted from SDS-PAGE. Gels were stained with Coomassie Brilliant blue and the appropriate band was excised from the gel. The gel strips were destained in 25 % isopropanol-10 % acetic acid and the protein was eluted using the Bio-Rad electroeluter apparatus. The immunization protocol was similar to that described above. IgG was obtained by passage of the antisera over a column of Protein-A Sepharose C L 4B (Sigma St. Louis Mo.) followed by elution with 0.5 M glycine-HCI ( pH 2.3), All antisera were stored in small aliquots at - 2 0 ° C . (b) Immunogold labelling of whole cells B. gingivalis cells were washed twice in PBS and resuspended in the same buffer to give an A 6 g 0 of 1.0. 200 u.l of the cell suspension was incubated with an equal volume of the appropriate IgG for 1.5 h at room temperature. Unbound IgG was removed by washing twice in PBS, the cells were then resuspended in 200 uJ of PBS and 15 u,l of gold beads (5 nm) conjugated to goat anti-rabbit IgG (EM GAR G5; Janssen Life Sciences Products, Beerse Belgium) were added and incubated at 4 ° G overnight. Unbound secondary antiserum was removed by washing twice in PBS. Cells were resuspended in 100 u.l of PBS and 25 u.l was placed on a copper grid. After drying, the cells were negatively stained with either 5 % uranyl acetate or 1 % phosphotungsic acid. 44 Observations were made with a Philips EM 300 transmission electron microscope. For immunofluoresence, 200 u.1 of cells resuspended to an AggQ of 1.0 were incubated with the appropriate antisera for 1 h at room temperature. The cells were washed free of unbound primary antiserum and incubated with FITC conjugated goat anti-rabbit IgG for 1 h. Unbound antibody was removed by washing in PBS and the cells were visualized by epifluoresence microscopy. (c) Thin sections Thin section immunoelectron microscopy,was done as follows. B. gingivalis cells (AggQ 0 f 1.0) were resuspended in 2.5 % gluteraldehyde and incubated for 1 h at 4 ° C . The cells were then pelleted by centrifugation, washed twice in PBS, (each washing was for 5 min) resuspended in PBS and left at 4 ° C overnight. Cells were fixed in 1 % O s 0 4 for 30 min at 4 ° C , washed twice for 5 min in water and then dehydrated progressively for 5 min each in 30, 50, 70, 95 and 100 % ethanol. They were then embedded in epon-adaldite resin for 72 h at 5 7 ° C before sectioning. Thin sections were picked up on copper grids. The grids were immersed in 3 % hydrogen peroxide for 10 min and then gently washed in filtered 20 mM Tris-HCl buffer (pH 7.2) for 10 min. Grids were immersed in 10 % BSA for 10 min, transferred to primary antisera for 2 h, washed twice in 20 mM Tris-HCl buffer and placed in goat anti-rabbit IgG conjugated to gold beads for 2 h. The grids were washed in Tris buffer followed by distilled water and then stained with 5 % uranyl acetate. Observations were made with a Philips EM 300 electron microscope. Binding to epithelial cells Human epithelial cells were collected by scraping oral and mucosal surfaces with a wooden applicator stick. Loosely bound bacteria were removed by filtration on a 8.0 um pore size filter (Nucleopore Corporation, Plessanton, CA) as described by Gibbons et al (74). The cells were suspended in saline solution to a final concentration of 1.0 x 10 5 45 cells per ml as determined by direct microscopic count. 1 ml of bacterial cells (1.0 x 1 0 8 per ml ) and 1 ml of epithelial cell suspension were incubated at room temperature for 1 h after which the unattached and loosely bound cells were removed by the membrane filtration system. The epithelial cells were examined under phase contrast microscope and the number of bacteria attached per cell was determined. At least 50 epithelial cells were examined for each reaction mixture. Scanning electron microscopy After the incubation of bacteria with epithelial cells, the unbound and loosely bound bacteria were removed by washing in the filtration system as described above and the epithelial cells processed for scanning electron microscopy by the method of Cimasoni et al (23). Briefly, the epithelial cells were fixed with 2.5 % glutaraldehyde in PBS for 30 min, post fixed with 1 % osmium tetroxide for 30 min and washed in distilled water for 5 min. The cells were then tanned in 2 % tannic acid for 20 min, washed in water for 5 min and fixed again in 1 % osmium tetroxide for 30 min. After rinsing in distilled water, the cells were dehydrated in an ethanol series (30, 50, 70 %) for 5 min at 4 ° C . Final dehydration in 95 and 100 % ethanol was carried out at room temperature. After dehydration, the samples were transferred to Teflon membrane filters (0.45 um, Millipore Corp. Bedford Mass.) and dried with a Ladd critical point dryer. They were then coated with palladium gold using a Hummer V1 sputter coater and examined in a model S-100 Cambridge scanning electron microscope at 25 KV. Dissociation of bacterial HA Attempts were made to dissociate the purified haemagglutinin into its subunits by a variety of treatments. Lyophilised samples were treated with 8 M urea (pH 7.0), 6 M guanidine hydrochloride (pH 7.0), 1 % SDS, 3 M LiCI, 3 M sodium thiocyanate (pH 7.0), 0.1 N HCI, 0.1 N NaOH and 1 % octyl phenoxy polyethoxyethanol (Triton X-100). Treatments were carried out at 37^C for 18 h. Reagents were removed by dialysis against water before analysis by SDS-PAGE. 46 Enzvmes and reagents All restriction enzymes, polynucleotide kinase, T 4 DNA ligase and lambda DNA were from Bethesda Research Laboratories, (Gaithersburg, MD). Calf intestinal phosphatase was from Boehringer-Mannheim (Indianapolis, IN.). IPTG, X-gal, naphthol AS-MX phosphate, salmon sperm DNA, DNase, RNase, lysozyme, lipase, mixed glucosidase, proteinase k, trypsin, chymotrypsin, pepsin and papain were from Sigma Chemical Co., (St. Louis, Mo). Isolation of the bacterial aggregating factor B. gingivalis cells were pelleted by centrifugation and washed twice in PBS (pH 7.3). The cells were resuspended in a tenth of the original culture volume in 0.05 M phosphate buffer (pH 7.4) and agitated at room temperature for 15 min, using a magnetic stirrer. The cells were removed by centrifugation and ammonium sulphate was added to the supernatant to give a final concentration of 40% (w/v). The mixture was incubated at 4^C overnight with stirring. The precipitated proteins were removed by centrifugation and resuspended in 20 mM Tris-HCl buffer (pH 8.0). The suspension was dialysed extensively against 10 mM Tris-HCl pH 8.0 to remove the salt. Any material which precipitated during dialysis was removed by centrifugation. The supernatant was used as the starting material for the purification of the bacterial binding adhesin. The coaggregating partner S. mitis was pelleted by centrifugation, washed twice in 50 mM HEPES buffer (pH 7.2) containing 10 mM CaCI 2 - The cells were resuspended in the same buffer to an AggQ of 5.0. 5 ml of S. mitis suspension was incubated with 3 ml of the 40 % ammonium sulphate precipitate. The mixture was incubated at room temperature with shaking for 20 min. The aggregates were pelleted by centrifugation at 2000 x g for 5 min. The pellet was saved and 5 ml of S. mitis suspension was added to the supernatant which was shaken at room temperature for a further 20 min. Again the aggregates were pelleted by centrifugation and more S. mitis added to the supernatant. 47 This process was repeated until no more aggregates were formed. The aggregates were pooled and washed once in HEPES buffer. The adsorbed material was then eluted by incubating the aggregates in 10 ml of 0.15 M NaCI-0.1 M glycine pH 2.3 with shaking. After 10 min, the pH was bought to neutrality by the addition of 0.15 NaCI-0.1 M N a 2 H P 0 4 (pH 7.5). The S. mitis was removed by centrifugation and the supernatant was dialysed extensively against water and lyophilized. This material was resuspended in 3 ml of H E P E S buffer and its bacterial aggregating activity was quantitated. Bacterial aggregation assay Bacterial aggregation was measured as described by McBride et al (152). Reaction mixtures contained 100 ul of S. mitis resuspended in 50 mM H E P E S buffer (pH 7.2) ( A g 6 0 of 3.0 ) and 100 of of aggregating substrate. B. gingivalis used as aggregating substrate was standardized to A ggQ of 2.0. Reaction mixtures were shaken for 30 min at room temperature and observed periodically. To establish an aggregation titre, B. gingivalis or any aggregating substrate was serially diluted and the last tube containing detectable aggregation was taken as the end point of the reaction. Since aggregates formed with the cell free substrate were smaller than those formed with whole B. gingivalis cells, evaluation of the degree of aggregation was made with a Leitz steromicroscope at 20 X magnification. Bacterial aggregation inhibition A number of compounds including carbohydrates, amino acids and other substances were tested for their ability to inhibit the coaggregation reaction of B. gingivalis with S. mitis. Briefly, the putative inhibitor was serially diluted and 50 u.l of B. gingivalis or the solubilised adhesin with a bacterial aggregating titre 2 times that required for visible aggregation was added to each tube. After shaking for 30 min at room 48 temperature, 50 ul of S. mitis was added and the mixture was shaken for a further 30 min before the results were determined. A similar assay was used to test the inhibitory activity of anti-adhesin serum. Modification of cell surfaces (a) Enzymes B. gingivalis and S. mitis cell suspensions (A 6 6 0 of 3.0) were prepared in H E P E S buffer at pH 7.2 for pronase and subtilisin treatment; phosphate buffer pH 8.0 for trypsin treatment; and 0.05M acetate buffer at pH 5.8 for pepsin treatment. The mixtures were incubated with the enzymes at 37^C for varying times. Cells were washed 3 times with H E P E S buffer and then checked for their ability to coaggregate. Controls without enzymes were treated in similar fashion. The purified bacterial aggregating adhesin was also incubated with the different enzymes after which BA activity was evaluated at 4 ° C without the removal of the enzymes. (b) B-mercaptoethanol Washed cells were suspended in H E P E S buffer pH 7.0 containing 0.05M 8-mercaptoethanol and incubated for 2 h at room temperature. The cells were washed 3 times and resuspended in H E P E S buffer and their BA activity was measured. Control without reducing agent were included. (c) Sodium periodate treatment Cells suspension of B. gingivalis was treated for 60 min with 10 mg/ml sodium periodate in 0.25 M potassium phosphate-buffer (pH 7.4) at room temperature. The cells were washed in HEPES buffer prior to assaying. 49 (d) Heat treatment: Cell suspensions of B. gingivalis and S. mitis were placed in water baths at 50, 70, 85 and 100°C for 5, 10 or 30 min then evaluated for coaggregation . Elisa-assav for bacterial coaggregation inhibition Briefly, S. mitis was resuspended to an A 6 6 0 n m of 1.0 and 0.1 ml was added to wells of a microtitre plate (Dynatech) and incubated for 24 h at 4 ° C . The plate was washed three times with phosphate buffer saline containing 0.05% tween (PBS/T) (pH 7.5). 0.1 ml of 3 % BSA in PBS was added to each well and incubated for 1h at room temperature with shaking. The wells were washed 3 x in PBS/T and 100 u.l of a suspension of B. gingivalis prepared as follows: ( an equal volume of B. gingivalis cell suspension with 4 times the number of cells required for aggregation was added to serially diluted anti-46 kDa serum followed by incubation for 30 min at room temperature) was added to each well and incubated for 30 min with shaking. The unbound bacteria were removed by washing the wells with PBS/T. 0.1 ml of anti-S. gingivalis or anti-46 kDa was added to the wells and incubated with shaking for 1h. The wells were rinsed, washed three times with PBS/T prior to adding 0.1 ml of a 1:1000 dilution of goat anti-rabbit IgG alkaline phosphatase conjugate. Following incubation for 1h at room temperature.the wells were rinsed 3 times with PBS/T and then 0.1 ml of substrate was added to the wells and the plate incubated at 3 7 ° C . Adsorbance was measured at 405 nm on an Elisa plate reader. The percent inhibition was determined by relating the absorption readings to those obtained for non-immune serum. Vesicles Preparation of vesicles Fifteen litres of a three-day old culture of B. gingivalis were concentrated to 250 ml by passage through an ultrafiltration system (Millipore Corporation, Bedford, MA.) with a membrane molecular weight cut off of 100,000. Cells were removed by two centrifugations at 10,000 x g for 30 min. The supernatant containing the vesicles was 50 dialysed overnight against 10 litres of 50 mM Tris-HCl , 1 mM dithiothreitol, (pH 9.5). The vesicles were collected by centrifugation at 90,000 x g for 2 h and then lyophilized. An outer membrane fraction from B. gingivalis was obtained by the method of Boyd and McBride (11). Adherence assay The assay was based on the method of Clark et al ( 26) with some modifications. 40 mg of washed and dried spheroidal hydroxyapatite (HA) beads (BDH, Poole, England) were added to glass scintillation vials together with 0.5 ml of pre-immune rabbit serum and then shaken for 2 h at room temperature and left at 4 ° C overnight. The excess serum was aspirated and the beads were washed twice with distilled water (10 ml) and a third time with phosphate buffer (1mM K H 2 P O 4 , 5 0 mM NaCl, 1 mM CaCI 2 , 0.01 mM MgCI 2 pH 6.5). Two-hundred u.g (dry weight) of vesicles from B. gingivalis 33277 were added in a volume of 200 ul of phosphate buffer, and the vials were shaken at 4 ° C for 1 h. The unattached vesicles were removed by aspiration, the beads washed twice in phosphate buffer and resuspended in 0.5 ml of the same buffer. Radiolabeled bacteria (0.5 ml) (specific activity 3.6 x 10 5 dpm/10 8 cells) were added and the mixture was incubated at 4 ° C for 1 h with constant agitation. Unattached bacteria were removed with three washes of phosphate buffer. On the third wash, the beads were transferred to plastic scintillation vials, excess buffer was aspirated and the beads were dried at 6 0 ° C for 1 h. Water (100 u,l) a n d tissue solubilizer (0.5 ml) (Amersham Corp. Arlington Heights, IL.) were added and the vials were left at 6 0 ° C overnight. Scintillation fluid (10 ml) was added to each vial and the radioactivity monitored in a Tracor Analytic Scintillation counter to determine the number of attached bacteria. All assays were performed in triplicate. For saliva-coated HA assays, 0.5 ml of clarified, heat treated saliva replaced the serum. When indicated, the inhibitors were pre-incubated for 1 h at 4 ° C with the serum-coated HA beads before the addition of 51 the bacteria. Spectrophotometric assay In order to screen several species of oral streptococci for their ability to bind to vesicle-adsorbed SeHA beads, a modification of the adherence was used. The bacteria to be tested were resuspended in HEPES buffer to an absorbance of 0.6 at 580 nm. The cells were added (1 ml) to SeHA beads (100 mg) which had been pre-incubated with vesicles (500 u,g) or buffer in a volume of 500 u,l. The mixture was shaken for 1 h at 4 ^ C . After allowing the beads to settle, (5 min) the supernatant was removed and the absorbance measured at 580 nm. Lipopolvsaccharide Purification of LPS Lipopolysaccharide from the nine different strains and species of BPB was prepared by the method of Darveau and Hancock (35) with minor variations. Briefly, cells from 2 litres of culture were pelleted and washed twice in phosphate buffered saline (PBS) pH 7.2. Ten grams of cells, wet weight, were resuspended in 15 ml of 10 mM Tris-HCl buffer (pH 8.0) containing 2 mM MgCI 2 , 100 u\g of pancreatic DNase (DN 100 Sigma Chemical Co.) per ml and 25 ug of pancreatic RNase (R-4821 Sigma Chemical Co. St. Louis Mo.) per ml. The cells were lysed by passage through a French pressure cell (Amicon Corp) at 1500 lb per inch 2 three times and then sonicated (Sonifer 350 Branson Sonic Power C O . ) for 1 min at 70 % output to ensure complete breakage of cells. RNase and DNase were added to a final concentration of 100 and 200 ug/ml respectively and the suspension was incubated at 3 7 ° C for 2 h. After incubation, 5 ml of 0.5 M tetrasodium EDTA, 2.5 ml of 20 % SDS and 2.5 ml of 10 mM Tris-HCl (pH 8.0) were added, and the sample vortexed to ensure solubilization of the components. Pronase (Sigma Chemical Co.) was added to give a final concentration of 200 ug/ml and the sample was incubated at 3 7 ° C overnight. Two volumes of 0.375 M MgCI 2 in 95 % 52 ethanol were added and the mixture cooled to 0 U C to precipitate the LPS. After centrifugation the pellet was resuspended in 25 ml of 2 % SDS, 0.1 M N a 4 E D T A in 10 mM Tris-HCl (pH 8.0) and briefly sonicated. After heating at 8 5 ° C for 10 min the pH was raised to 9.5 and pronase added (25 ug/ml) and the sample incubated overnight at 3 7 ° C . The LPS was precipitated with two volumes of 0.375 M MgCI 2 in ethanol and more ethanol was added to a final concentration of 80 %. The suspension was centrifuged, the pellet resuspended in water and dialysed extensively against water and lyophilized. Characterization of the LPS Chemical analysis Protein was determined by the method of Bradford et al (12) using bovine serum albumin as a standard. Total carbohydrates were measured by the phenol-sulphuric acid method (43) with glucose as the standard. Hexoses were quantified by the anthrone reaction (199) using glucose as the standard. Hexosamines were determined by the method of Elson and Morgan (5). The method of Dische and Schettles (108) was used to determine methylpentoses with rhamnose as the standard. Heptose and KDO were quantitated by the colorimetric methods of Wright and Rebers (257) and Karkhanis et al (112) respectively. Gas-liquid-chromatographic analysis of neutral and amino sugars The methods of Niedermeir (171) and Henry et al (92) were used to convert the sugars of the LPS to alditol acetates. Briefly, 10 mg of LPS was hydrolysed with 4 N HCI at 100^C for 1, 2 and 4 h in a sealed glass tube under nitrogen. The lipid A portion was removed by centrifugation and inositol and mannosamine were added as internal standards. 1.5 ml of distilled water was added to the vial and the mixture was mixed thoroughly. The solution was transferred to a 50 ml beaker and the pH adjusted to between 6.8 and 7.2 with Dowex 1 HCO3" resin. The resin was removed by filtration through a Buchner funnel and the reaction mixture was lyophilized. The dried sample was then dissolved in 0.5 ml water and cooled to 0 ° C . 1 ml of 0.25 M sodium 53 borohydride was added and the reduction was allowed to proceed overnight at 4 ° C . One drop of 6 N HCI was added to decompose the sodium borohydride. After the solution was lyophilized, approximately 1 ml of methanol was added to the dry residue in order to convert the boric acid to the volatile trimethyl borate. The methanol was removed under reduced pressure at room temperature. The methanol treatment was repeated 6 times for each sample. 0.2 ml of a 1:1 solution of pyridine and acetic anhydride was added to the residue . Acetylation was allowed to proceed for 30 min in a 1 0 0 ° C water bath. The dried alditol acetates were dissolved in 0.1 ml of chloroform ( G . L C . grade) and analysed on a GP 3% SP-2340 on 100/200 Supelcoport (Supelco Chromatography Supplies) glass column ( 2000 x 1 mm) in a Perkin-Elmer Sigma 3B Dual FID Chromatograph. Reference alditol acetate derivatives of pure sugars were prepared in similar fashion. Resolution of the alditol acetate derivatives was achieved by use of the following programme: injector/detector temperature 2 5 0 ° C , initial temperature 1 8 0 ° C , final temperature 2 4 0 ° C , ramp rate 2°C/min, helium flow rate 30 ml/min. Fatty acids analysis The method of Flesher et al (60) was used to analyse fatty acids. LPS (3 mg) was hydrolysed in 2 ml of 4 N NaOH for 5h at 100^C in glass vials sealed under nitrogen gas. After adding the internal standard (methyl-tetra decanoic acid) the hydrolysate was neutralized with 4 N HCI and extracted three times with 2 ml of chloroform. The organic phase was dissolved in 1.5 ml of 0.5 N NaOH in 95% methanol and incubated at room temperature for 30 min with shaking in a teflon capped test tube. Boron trichloride (Sigma Chemical Co., St. Louis, MO.) was added and the mixture was heated to 5 0 ° C for 30 min, cooled and mixed with 1 ml water. The solution was shaken with 1 ml of hexane for 1 min and the hexane layer was removed and concentrated. 5 ul was injected into a Perkin-Elmer Sigma 3B Dual FID Chromatograph with a 10% diethyeneglycol succinate column on Chromosorb W 80/100 mesh (Supelco, Belleforte, Pa.). The carrier gas was 54 helium with a flow rate of 20 ml/min. The temperature of the injector was 200 U C, the column temperature 180°C and the flame ionization temperature was 2 5 0 ° C . Individual fatty acids standards were treated in similar fashion and analysed under the same conditions. The NHI fatty acid methyl ether mixture (Supelco) was used as a standard. 55 Results Isolation of the haemagglutinin Haemagglutinin was removed from B. gingivalis by mixing the cells with a magnetic stirrer at room temperature. Following removal of the cells, the HA was recovered by precipitation with ammonium sulphate . This material was applied to a DEAE-Sepharose column and eluted with a gradient of 0-0.3 M NaCl. Figure 1 shows a single peak of protein and haemagglutinating activity. S D S - P A G E analysis of the fractions with haemagglutinating activity revealed that fractions eluted before 0.15 M NaCl ( fractions 28-31) had several protein bands (Fig. 2 lanes B and C) whereas fractions eluted between 0.18-0.2 M NaCl exhibited a single protein staining band (Fig. 2 lanes D and E ) at 43 kDa. Those fractions containing a single protein band and strong haemagglutinating activity were pooled, dialysed against water and lyophilized. This material was chromatographed on a Sephacryl S-1000 column ( 80 x 1.5 cm). Two protein peaks were obtained as seen in Figure 3. The first peak which eluted in the void volume contained strong haemagglutinating activity, the second peak contained significant quantities of protein but little haemagglutinating activity. All fractions with haemagglutinating activity were analysed by 12 % SDS-PAGE. During the process of staining it was possible to identify two closely spaced bands which tended to merge into a single band if staining was allowed to continue for too long. To try and separate these bands, the fractions were analysed on either a 15 % S D S - P A G E with electrophoresis continued for 4 h after the dye front emerged from the gel, or on a 5 to 20 % gradient gel. Both methods resolved the two proteins with the gradient gel giving the best results. On staining, two distinct bands one at 43 kDa and the other at 41 kDa were seen. Fractions 20-25 contained mostly the 43 kDa band and good haemagglutinating activity, fractions 25-35 had a mixture of the 43 56 Figure 1. Elution of haemagglutinating activity ( • ) and protein ( • ) from the 40 % ammonium sulphate precipitate using a DEAE-Sepharose column. The column (16x1.5 cm) was eluted with a linear gradient of 0.0-0.3 M NaCl in 20 mM Tris-HCl (pH 8.0). Fractions were collected and assayed for haemagglutinating activity after dialysis against water. Fraction # 57 Figure 2. SDS-PAGE of fractions eluted from a DEAE-Sepharose column. Lanes: A, 40 % ammonium sulphate precipitate; B and C, fractions 29 and 31 eluted before 0.15 M NaCl; D,E and F, fractions 32, 33 and 35 eluted between 0.18 and 0.2 M NaCl. 10 ug of protein was loaded in lanes A and B, 8 u.g in lane C and 5 ug in lanes D,E and F. 58 Figure 3. Chromatography of HA on Sephacryl S-1000. Pooled and concentrated HA containing fractions from DEAE-Sepharose chromatography were applied to a Sephacryl S-1000 column (80 x1.5 cm) and eluted with 20 mM Tris-HCl (pH 8.0) containing 50 mM NaCl. Each fraction (3 ml) was assayed for haemagglutinating activity ( • ) and absorbance at 280 nm ( 0 ) . 0 1 0 2 0 3 0 4 0 5 0 6 0 Fraction # 59 and 41 kDa bands and little haemagglutinating activity (Fig. 4). Fractions 40-43 had little haemagglutinating activity and small quantity of the 43 kDa protein. These results show that there is a correlation between the presence of the 43 kDa protein and haemagglutinating activity. Additional evidence that the 43 kDa protein was the HA was obtained by adsorbing fractions from peaks 1 and 2 with erythrocytes. In all cases, haemagglutinating activity and most of the 43 kDa protein were removed by the red blood cells, the 41 kDa protein was not adsorbed. It was difficult to elute the adsorbed protein from the erythrocytes (0.15 M glycine-0.15 M NaCl, pH 2.3) but the small percentage of the material that eluted displayed haemagglutinating activity and . There was a question as to whether the adsorbed HA was inactivated during elution or simply did not elute. The latter seems to be more likely as it was shown that antisera to the 43 kDa protein reacted positively with the surface of red blood cells that had been subjected to the elution procedure following a adsorption of the haemagglutinating activity. Characterization of the haemannlutinin SDS-PAGE (12 %) of the purified adhesin of B. gingivalis showed a single band at 43 kDa when the sample was boiled for 5 min in 4 % SDS- 2 % 3-mercaptoethanol. In the absence of boiling no material entered the running gel but a band stained at the junction of the stacking and running gel. Heat treatment of the sample in the presence of SDS at 3 7 ° C , 7 0 ° C and 8 0 ° C for 30 min did not disassociate the molecule into a form which could migrate into the gel. 60 Figure 4. SDS-PAGE gradient gel analysis of fractions from Sephacryl S-1000 column. Lanes: A, molecular weight standards; B, fraction 24 showing the pure 43 kDa band; C, fraction 28 showing 43 and 41 kDa bands; D, fraction 45 showing the 41 kDa protein band. All fractions were boiled for 5 min in 4 % SDS. 5 u.g of protein was loaded in lanes B and C and 2 ng in lane D. 61 Dissociation of the haemagglutinin The inability of the HA to migrate into the gel without boiling in SDS suggested that it existed in a multimeric form of high molecular weight. The fact that the HA eluted in the void volume of the Sephacryl column provided additional evidence for this. As heating destroyed all haemagglutinating activity, other dissociating procedures were attempted in order to obtain non-denatured subunits and to gain some understanding of the bonds that held the subunits together. Urea (8M), guanidine hydrochloride (6M), thiocyanate (3M), lithium chloride (1M), deoxycholate (0.5 %), Triton X-100 (1 %) and HCI (0.1 N) and various combinations of these compounds were found to be ineffective in dissociating the HA. Reducing compounds such as 8-mercaptoethanol also had no effect on dissociating the HA. Electrophoresis of an unboiled sample of the HA on a 7 % polyacrylamide gel under non-denaturing condition resulted in a smear running the length of the gel (data not shown). This suggests a mixture of varying size and may explain the presence of the minor quantities of the 43 kDa protein in peak 2 from the Sephacryl S-1000 column. When the non-denaturing gel was cut into 1 cm strips and the protein eluted and electrophoresed after boiling in SDS, the protein from each segment showed only a single 43 kDa band (results not shown). Because of the confusion surrounding the structural properties of B. gingivalis HA, it was decided to look for fimbriae or other structural elements in the purified 43 kDa fraction. The purified HA was negatively stained with 5 % uranyl acetate and analysed by transmission electron microscopy. No fimbriae or any other polymerized structure was seen. Chemical composition of the HA Purified preparations of the HA were found to contain approximately 90 % protein and 10 % carbohydrate based on a dry weight basis. Staining boiled and unboiled samples for LPS proved to be negative. 62 Heat treatment Heating of the purified HA at 1 0 0 ° C for 1 min destroyed all haemagglutinating activity. The same result was obtained with whole cells of B. gingivalis. Inhibition of haemagglutination A number of putative receptor analogues were tested for their ability to inhibit erythrocyte agglutination by B. gingivalis cells and by the purified HA. With the exception of fetuin at 100 ug per ml none of the compounds had any effect on haemagglutinating activity (Table 1). Sodium periodate, EDTA and CaCI 2 had no effect on haemagglutinating activity but the reducing compound 6-mercaptoethanol was found to double the activity (Table 2). Similar results were obtained for both whole cells and the purified HA of B. gingivalis. This is not thought to be significant since it was within the realm of experimental error. Antisera raised against whole cells of B. gingivalis, the purified haemagglutinin and the 43 kDa subunit were tested for their ability to inhibit haemagglutination of B. gingivalis. In the assay, the IgG fractions of these antisera were resuspended to 100 ug per ml and serially diluted in PBS prior to adding the bacteria. Following incubation for 30 min, erythrocytes were added. As seen in Table 3, anti-43 antiserum had an inhibitory titre of 128, the same as that for B. gingivalis and for the purified HA antisera. Preimmune serum had a slight effect with an inhibitory titre of 4. Similar results were seen by Slots et al (214) with non-immune serum. Immunoaold labelling The location and distribution of the HA on B. gingivalis was determined with antisera to the purified 43 kDa as a probe. After incubation of the whole cells with the primary 63 Table 1. Effect of a number of putative inhibitors on haemagglutinating activity. Substance effect on haemagglutinatin activity3 D-mannose 0 D-glucose 0 D-galactose 0 D-ribose 0 D-fucose 0 D-fructose 0 Lactose 0 Rhamnose 0 D-glucosamine 0 D-galactosamine 0 N-Acetylglucosamine 0 N-Acetylgalactosamine 0 L-lysine 0 L-Arginine 0 L- Leucine 0 Fetuin (100u.g/ml) -aEffect on activity designated as follows: enhancement (+), inhibition (-), no effect (0). ^All the test substances were at an initial concentration of 100mM unless otherwise indicated. 64 Table 2. Effect of various surface modifying substances on the haemagglutinating activity of B. gingivalis. Substance effect on haemagglutinating activity3 EDTA (1 mM) 0 C a C I 2 (10 mM) Periodic acid (0.01 M) B-Mercaptoethanol (.2 %) c 0 0 + aEffect on activity designated as follows: enhancement (+), inhibition (-), no effect (0). '-'All the test substances were at an initial concentration of 100mM unless otherwise indicated. increased HA activity two fold. 65 Table 3. Effect of immune sera 3 on haemagglutinating activity of B. gingivalis. IgG Inhibitory titre Anti-S. gingivalis 1 28 Anti-43 kDa 128 Anti-HA 128 Preimmune serum 4 a l g G fractions were prepared by passage of the serum over a Protein A Sepharose column. The IgG'S were resuspended to a final concentration of 100 u\g per ml for use in the inhibition assay. ^minimum antibody titre required to inhibit haemagglutination. 66 antiserum, the samples were reacted with goat-anti-rabbit IgG conjugated to gold beads. As seen in Figure 5 there was an even distribution of the adhesin over the cell surface as indicated by the gold bead labelling pattern. The labelling does not seem to be indicative of any structural entity such as fimbriae or pili. Nonimmune serum did not react with the surface of the B. gingivalis cells as no gold beads were seen when these preparations were viewed under the electron microscope. Thin sections of B. gingivalis cells were prepared and reacted in similar manner to whole cells with primary and secondary antisera before being examined under the transmission electron microscope. As seen in Figure 6, most of the labelling is concentrated in the periplasm and on the outer membrane, while the cytoplasm is basically free of any labelling. It is of interest to note that there appears to be clumping of gold beads on the outside of the cells. These clumps may represent HA which was displaced during preparation of the specimen. A control using pre-immune serum as the primary antiserum was carried out and the result is seen in Figure 7; no labelling is visible. Enzvmatic treatment The effect of a variety of enzymes on the haemagglutinating activity of B. gingivalis was evaluated. B. gingivalis cells and the purified adhesin were incubated with the enzymes for 1 h after which samples were removed from the reaction mixtures and assayed for haemagglutinating activity. The result is summarized in Table 4. Pronase and proteinase K were the most effective in destroying the haemagglutinating activity of the whole cells and purified HA. The haemagglutinating activity titre dropped from 512 to 0 for the purified HA and from 256 to 32 for whole cells, a decrease of over 80 %. Of the other proteolytic enzymes, trypsin and subtilisin were the most active in destroying haemagglutinating activity reducing it by 75 % for both whole cells and purified HA. Papain, pepsin and chymotrypsin had no effect on the haemagglutinating activity of B.gingivalis cells but reduced the activity of the purified HA by 50 %. 67 Figure 5. Cells of B. gingivalis reacted with anti-43kd serum and then incubated with gold beads conjugated to goat anti-rabbit IgG. Cells are negatively stained with uranyl acetate. Bar 0.5 am. Figure 6. Immunogold bead labelling of thin sections of B. gingivalis. The sectioned cells were reacted with anti-43 kDa serum and then incubated with gold beads conjugated to goat anti-rabbit IgG. Thin sections were negatively stained with phosphotungstic acid. Arrow indicate clumps of gold beads on the outer membrane. Bar, 0.5 u.m. 69 Figure 7. Immunogold bead labelling of thin sections of B. gingivalis. Sections were reacted with preimmune serum and then incubated with gold beads conjugated to goat anti-rabbit IgG. Thin sections were negatively stained with phosphotungstic acid. Bar, 0.5 um. 7 0 Table 4. Effect of enzyme treatment on the haemagglutinating activity of intact cells and the purified haemagglutinin of B. gingivalis. Haemagglutinating Titre Treatment3- Intact cells Purified Haemagglutinin Positive control 256 512 Pronase 32 0 Proteinase K 32 0 Papain 256 256 Pepsin 256 256 Trypsin 64 1 28 Chymotrypsin 256 256 Subtilisin 64 256 Neuraminidase 256 512 Lysozyme 1 28 256 Mixed glycosidases 256 512 Phospholipase C 64 1 28 Mutanolysin 256 512 Enzyme concentration was 1 mg per ml. 71 With the exception of phospholipase C and lysozyme which reduced the haemagglutinating titre from 256 to 64 and 256 to 128 for both cells and HA respectively, none of the other enzymes tested had any inhibitory effect on the haemagglutinating activity on B. gingivalis. On S D S - P A G E , preparations of these latter two enzymes displayed several protein bands indicating that they were contaminated. In addition to assaying for haemagglutinating activity, samples from the protein digestions were mixed with SDS-6ME, boiled for 5 min and electrophoresed on 12 % S D S - P A G E . The 43 kDa band was still present in samples digested with pronase and proteinase K. This was surprising in light of the fact that the haemagglutinating activity was destroyed in these latter two preparations. The meaning of this observation is not clear but it is possible that: (1) the proteolytic enzymes released a few amino or carboxy terminal amino acids but not enough for a difference in molecular weight to manifest itself; (2) the enzymes prevented haemagglutinating activity by occupying the binding site of the adhesin and not by hydrolysing the molecule; (3) the HA is a minor contaminant which coincidentally migrates with a similar sized non-haemagglutinin protein. The first possibility was investigated by subjecting treated and untreated samples to isoelectric focusing. In all cases the samples behaved in an identical manner, ie, they had the same pi of 6.4. The other possibility that the HA is a minor protein at 43 kDa is doubtful for two reasons; first as explained previously in the absence of boiling no bands were seen on S D S - P A G E , and secondly in the adsorption experiments using red blood cells, all of the 43 kDa band was adsorbed as seen by SDS-PAGE. To preclude the possibility that the enzyme simply blocked the active site of the HA the following set of experiments were carried out: (1) the HA was incubated with varying amounts of proteinase K for 60 min after which the reaction mixtures were assayed for haemagglutinating activity, (2) the HA was incubated with 100 u.g of proteinase K and at 72 different time intervals samples were removed and assayed for haemagglutinationg activity. As can be seen in Table 5, after 60 min, 10 ng of proteinase K reduced the HA activity from a titre of 128 to 32 and 50 u.g totally destroyed all activity. Table 6 , shows the time course of proteinase K inactivation of haemagglutination. There was no effect for 10 min, and then a gradual decline, 50 % of activity was lost at 20 min 88 % at 30 min and complete loss by 60 min. When these reactions were carried out with heat inactivated proteinase K or at 4^C there was no loss of haemagglutinating activity. The possibility that the enzyme caused steric hinderance and thereby inactivate the HA was not ruled out by these experiments. To address this problem protease inhibitors were used to inactivate proteinase K and the enzyme was then checked to see its effect on haemagglutinating activity. In preliminary experiments it was shown that hydrolysis of azocasein by 100 u.g of proteinase K was completely inhibited by 20 jxg of PMSF. The effect of PMSF treated proteinase K is shown in Table 7. In the absence of inhibitor, all HA activity was destroyed, but as the quantity of PMSF increased, so did the HA activity. PMSF by itself had no effect on HA activity. This indicates that proteinase K had to be active to destroy haemagglutinating activity. Cloning of the haemagglutinin The surprising results observed when the HA was digested with proteolytic enzymes and analysed on SDS-PAGE introduced some confusion as to the exact identity of the HA and as such to elucidate this problem it was decided to clone the gene(s) encoding the HA from B. gingivalis into E. coli. 73 Table 5. Effect of increasing amounts of proteinase K on haemagglutinating activity of purified B. gingivalis haemagglutinin ug proteinase K added 3 HA titre 0 1 28 5 128 1 0 32 20 8 40 2 50 0 1 00 0 200 0 100 (boiled) 1 28 3 Incubated for 60 min. 74 Table 6. Time course of HA inactivation by proteinase K. Time of incubation 3 (min) HA titre. 0 5 1 0 20 30 45 60 1 20 1 28 1 28 1 28 64 1 6 2 0 0 3 Incubated with 100 ug of enzyme 75 Table 7. Effect of P M S F a on proteinase K digestion of the purified B. gingivalis haemagglutinin. ug of PMSF added 0 HA titre 0 0 5 0 1 0 4 1 5 32 2 0 128 3 0 128 50 128 PMSF control 0 128 Boiled proteinase K + PMSF 128 aPhenylmethyl sulfonylfluoride. D Amount of PMSF added to 100 ug of proteinase k. c No enzyme added, 20 ug of PMSF. 76 Genomic library Using Pst 1 enzyme to restrict B. gingivalis chromosomal DNA and the Pst 1 site of the vector pUC 18, a genomic library was constructed in E. coli JM 83. A total of 5,500 transformants were tested for the expression of antigens by the filter binding enzyme immunoassay and for haemagglutinating activity by a biological assay. Three hundred and thirty seven clones were positive by the immuno assay ( Fig. 8), and two clones (1-3 and 1-49) tested strongly positive for haemagglutination in the biological assay while a third clone gave weak activity ( Fig. 9). Recombinant plasmids The size of the DNA inserts was determined by digesting the recombinant plasmids with Pst 1 enzyme followed by electrophoresis on 1 % agarose gel . Figure 10 shows that both recombinant clones had 3.2 kb DNA inserts. Clone 1-49 was chosen for further characterization. Restriction analysis The 3.2 kb insert DNA was treated with a variety of restriction enzymes in single and double digestions. The physical map of restriction sites is shown in Figure 11. As can be seen Bam H1, EcoR 1, Pst 1 and several other common restriction enzymes did not restrict the insert DNA. Southern blot analysis Southern blot analysis was performed to confirm that the DNA insert was derived from the B. gingivalis DNA. As can be seen in Figure 12, the 3.2 kb probe generated by Pst 1 digestion of the recombinant plasmid hybridized to identical size fragments in DNA from both B. gingivalis and the recombinant plasmid. The control DNA from S. sanguis did not hybridize with the probe (Figure 12, lane 3). 77 Figure 8. Colony immunoblot. White colonies, obtained by transforming E. coli JM 83 with B. gingivalis DNA, were replica plated and lysed in situ. Lysed colonies were blotted onto nitrocellulose paper and reacted with anti-B. gingivalis ATCC 33277 serum. A, B. gingivalis cells; B, E. coli JM 83 (pUC18); C, E. coli 1-49; 78 Figure 9. Screening of recombinant clones obtained by transforming E. coli JM 83 with B. gingivalis DNA for haemagglutinating activity. Recombinant colonies were macerated in 0.1 ml of PBS and mixed with an equal volume of 2 % v/v erythrocytes. The arrows indicate the positive clones. i 2 3 1 4 5 5 y 8 9 M 0 11 A A A A B ©©©©©0©0©©0@ ©©©©©©000® i>O©©©O0OO0O ©©©©©©0@@ ©©©©©©@©o©@@ 79 Figure 10. Restriction endonuclease analysis of the haemagglutinin recombinant clones. 0.7 % agarose gel electrophoresis. Lanes : A, molecular weight markers, Hind 111 digest of lambda DNA; B, pUC 18 digested with Pst 1, C, E. coli 1-49 DNA; E, E. coli 1-49 digested with Pst 1, the arrow indicates the insert of 3.2 kb; F, E. coli 1-3 digested with Pst 1. 80 Figure 11. Linear restriction map of the the cloned insert from recombinant E. coli 1-49. Abbreviations: P, Pst 1; K, Kpn 1; S, Sma 1; B, Bgl 1; X, Xho 1; A, Sph 1; H, Hinc 11; C, Cla 1. 81 Expression of cloned antigen The polypeptide encoded by the insert DNA in recombinant 1 -49 was identified by SDS-P A G E and Western immunoblotting. On 5 to 20 % SDS-PAGE a faint band at 43 kDa can be seen sometimes in the recombinant lysate which is not present in cell lysate of control E. coli JM 83 (pUC 18). Polyclonal antisera was raised to 1-49 and exhaustively adsorbed with control E. coli JM 83 (pUC 18). The adsorbed antisera reacted with a 43 kDa band from the cell lysate of 1-49 (Fig. 13) but not with any similar band in E. coli JM 83 (pUC 18) lysate. There are several bands that reacted with the serum although the serum was adsorbed extensively with E. coli JM 83 (pUC 18). These bands presumably represent E. coli antigens which are not expressed on the cell surface and would not bind antibody when whole cells were used to adsorb the antisera. There appears to be a band at about 18 kDa that is present in the recombinant lysate but not in the control lysate, however, in the original blot this was not evident. To determine where the majority of the 43 kDa protein was localized, outer membrane, periplasmic and cytoplasmic preparations of 1-49 were probed with antisera to the recombinant clone on Western blot. Most of the 43 kDa protein was found on the outer membrane and a small amount in the periplasm. Antisera to 1-49 was used to probe B. gingivalis cell lysate and the purified haemagglutinin in a Western blot. As can be seen in Figure 14, this antisera only reacted with a band at 43 kDa in both preparations. This demonstrates that immunologically similar proteins are found in 1-49 and in B. gingivalis. Prevalence of the 43 kDa protein in Bacteroides Several species and strains of Bacteroides were checked by Western blot for the presence of an immunologically similar 43 kDa protein using anti 1-49 antiserum and antiserum to the purified haemagglutinin. Both antisera reacted with a 43 kDa band in all strains of B. gingivalis but not to any band in the other species of Bacteroides tested (Fig. 14). 82 Figure 12. Southern blot hybridization with the 3.2 kb Pst 1 restriction fragment of E. coli 1-49. Panel A, DNA fragments separated by agarose gel electrophoresis. Panel B, Hybridization with the 3.2 kb biotinylated probe. Lanes: 1, molecular weight markers, Hind 111 digest of lambda DNA; 2, Pst 1 digested B. gingivalis DNA; 3, Pst 1 digested S. sanguis DNA; 4, Pst 1 digested E. coli 1-49 DNA. B 83 Figure 13. Western blot of anti-1-49 antiserum against cell lysate of E. coli 1-49 using the GAR-HRP labelling system. Lanes: A, molecular weight markers (kDa); B, and D, cell lysate of E. coli 1-49 10 ug and 20 ug of protein respectively; C, and E, cell lysate of E. coli JM 83 (pUC18) 10 and 20 ug of protein respectively. The arrow indicates the 43 kDa band. A B C D E 200 Figure 14. Western blot of BPB cell lysates and the the purified HA using anti-1-49 serum as the probe and using the GAR-HRP system for detection. Lanes: A, cell lysate of B. gingivalis 33277; B, cell lysate of B. intermedius; C, cell lysate of B. asaccharolyticus; D, purified haemagglutinin of B. gingivalis 33277; E,F,G and H, cell lysates of B. gingivalis W 12, W 50, W 83, and 2D respectively. A B C D E F G H 9 7 . 4 4 6 2 5 . 7 1 8 . 4 85 Immunogold electronmicroscopv To localize the cloned antigen and to provide additional evidence for its surface location in E. coli, antisera to B. gingivalis, the haemagglutinin and to 1-49 were examined for their ability to bind to the cell surface in immunofluoresence and immunogold labelling assays. Cells were incubated with one of the antisera, washed and then incubated with either gold beads conjugated to goat anti-rabbit IgG or FITC labelled goat anti-rabbit IgG. E. coli 1-49 reacted positively with all of the antisera in immunofluorescence labelling. There was uniform labelling of the cell surface, although the intensity of the labelling was low (results not shown). Immunogold labelling of the cell surface of 1-49 with anti-S. gingivalis serum is seen in Figure 15. The recombinant protein appeared to be distributed in a relatively even manner over the cell surface. The concentration of the protein appeared low, this is consistent with the result obtained by immunofluoresence, SDS-PAGE and Western blot analysis. B. gingivalis was reacted with serum to 1-49 followed by gold beads conjugated to goat anti-rabbit IgG. As can be seen in Figure 16, the gold labelling appeared to be uniform over the cell surface. Another point of interest is the total absence of labelling of any structural material such as pili . Anti-S. sanguis serum did not label B. gingivalis and E. coli 1-49 cells. None of the antisera reacted with E. coli JM 83 (pUC 18) control cells. Similarly antisera to the 43 kDa subunit of the haemagglutinin reacted with surface antigen on E. coli 1-49 as seen in Figure 17. Again the labelling is sparse but uniform over the entire surface of the cell. Inhibition by antisera Antisera to the purified HA, anti-43, anti-1-49 and anti-S. gingivalis were tested for their inhibitory effect on the haemagglutinating activity of recombinant 1-49. The result is shown in Table 8. Anti-S. gingivalis anti 43 kDa and anti-HA serum all 86 Figure 15. Immuno gold bead labelling of E. coli 1-49. Cells were reacted with anti-S. gingivalis serum followed by reaction with gold beads conjugated to goat anti-rabbit IgG and then negatively stained with uranyl acetate. Bar, 0.5 u.m. 87 Figure 16. Immuno gold bead labelling of B. gingivalis. Cells were reacted with anti-E. coli 1-49 serum then incubated with gold beads conjugated to goat anti-rabbit IgG and negatively stained with 5 % uranyl acetate. Bar, 0.5 am. 88 had inhibitory titres of 256. Anti-1-49 serum had an inhibitory titre of 64 whereas pre-immune serum had an inhibitory titre of 4. Binding to buccal epithelial cells It is well established that B. gingivalis binds to epithelial cells. Thus it was of interest to know whether the HA is an epithelial cell binding adhesin. This question was answered by mixing bacteria with freshly prepared human epithelial cells. Microscopic analysis revealed that washed epithelial cells had an indigenous population of approximately 5 bacteria per epithelial cell (Table 9 ). E. coli JM 83 (pUC 18) did not possess adhesins and therefore did not bind to the epithelial cells whereas E. coli 1 -49 which carried the B. gingivalis HA bound in large numbers to the epithelial cells. Binding was strong as bacterial cells were not removed by vigorous washing. It was noted that there appeared to be two populations of epithelial cells, one which possessed receptors available to the adhesin and bound 1-49 and another population which either lacked the receptor or in which the receptor was masked. The reaction of E. coli 1-49 with buccal epithelial cells was examined by scanning electron microscopy. As seen in Figure 18, E. coli 1-49 binds in large numbers to the cells whereas control E. coli JM 83 (pUC 18) did not bind at all. In another control it can be seen that B. gingivalis bound to the epithelial cells. Epithelial cells with attached E. coli 1-49 and B. gingivalis cells were sectioned and viewed under transmission microscope. As seen in Figure 19 panel A, B. gingivalis cells bound to several villi-like structures on the epithelial cell surface. In panel B, E. coli 1-49 bound to a number of points on the epithelial cell with plicate-like structures. Examination of the preparations under the electron microscope revealed the presence of two distinct populations of epithelial cells; one supporting binding with the plicate 89 Figure 17. Immuno gold bead labelling of E. coli 1-49. The recombinant E. coli 1-49 was reacted with anti-43 kDa serum and then incubated with gold beads conjugated to goat anti-rabbit IgG. The cells were negatively stained with uranyl acetate. Bar, 0.5 urn. 90 Table 8. Effect of immune sera 3 on haemagglutinating activity of E. coli 1-49 Antisera Inhibitory titre Anti-S. gingivalis 256 Anti-43 kDa 256 Ant i -E . CO//-1-49 64 Anti-HA 256 Preimmune serum 4 3 IgG fractions were prepared by passage of the serum over a Protein A Sepharose column. The IgG-s were resuspended to a final concentration of 100 u.g per ml for use in the inhibition assay. D minimun antibody titre required to inhibit haemagglutination 91 Table 9. Adherence to buccal epithelial cells. Organism added Avo. no. of bacteria/epithelial cell3-. None 5 b ± 2 fl. gingivalis 33277 1 3 7 ± 3 7 E. coli 1-49 7 0 ± 2 6 E. coli JM 83 (pUC 18) 5 ± 2 3 number of epithelial cells counted was 50. D represents the number of bacteria bound to buccal epithelial cells after washing and prior to addition of the test bacteria. 92 Figure 18. Scanning electron micrograph of bacteria bound to human buccal epithelial cells. Bacteria and epithelial cells were incubated, washed and then coated with palladium gold. Panels : A, B. gingivalis 33277; B, E. coli 1 -49; C, E. coli JM 83 (pUC 18). Bar, 0.5 urn. 93 like structures and those not supporting binding with smooth surfaces. The surface of the epithelial cell under went a morphological change when bacteria attached. Characterization of the recombinant-epithelial cell adherence reaction The nature of the the reaction of E. coli with buccal epithelial cells was investigated in a series of experiments described in Table 10. In all cases the recombinant was compared to B. gingivalis. Proteinase K treatment of either B. gingivalis or E. coli 1 -49 destroyed their ability to attach to epithelial cells, similar treatment of the epithelial cells did not affect the number of bacteria that bound. This would indicate that a protein on the bacterial surface is responsible for the attachment. Heating the bacteria and/or epithelial cells destroyed all binding indicating a protein-protein lectin like interaction. Several investigators have shown that arginine and lysine were capable of inhibiting the binding of B. gingivalis to erythrocytes. These two compounds were assayed for their ability to inhibit the binding of B. gingivalis and E. coli 1-49 to epithelial cells. The result is seen in Table 10. Neither of them proved to be effective in inhibiting the attachment of bacteria to epithelial cell. Inhibition bv antisera Anti-fi. gingivalis, anti-43 kd, and anti-1-49 IgG were assayed for their ability to inhibit the binding of B. gingivalis and E. coli 1-49 to buccal epithelial cells. The result is shown in Table 11. The attachment of both bacteria was reduced by more than 70 % when incubated with 50 ug of each of the immunoglobulin. Pre-immune IgG had no effect on on the binding reaction. Bacterial coaaareaatina adhesin A number of investigators have shown that B. gingivalis will bind to other species of bacteria. With the exception of Boyd and McBride (11) very little has been done to 94 Figure 19. Electron micrograph of bacteria attached to human buccal epithelial cells. The cells were embedded in epon-adaldite, sectioned and then stained with uranyl acetate. Panels: A, B. gingivalis; B, E. coli 1-49. Bar, 0.5 um. 9 5 Table 10. Effect of protease, heat, and putative inhibitors on the attachment to oral epithelial cells. Organism/treatment Avg. no of bacteria bound per epithelial cell. B. gingivalis (none) 137 E. coli 1 -49 (none) 7 0 E. coli JM 83 (pUC18) (none) 5 B. gingivalis (proteinase K 10 ug) 14 E. coli 1-49 (proteinase K 10 ug) 5 B. gingivalis (70°'C/10 min) 16 E. coli 1-49 ( 7 0 ° C /10 min) 5 B. gingivalis (lysine 100 mM) 117 E. coli 1-49 (lysine 100 mM) 5 7 B. gingivalis (arginine 100 mM) 120 E. coli 1-49 (arginine 100 mM) 5 6 96 Table 11. Effect of immune l g G a on binding to epithelial cells. Bacteria Antisera Avg. no. of bacteria/epithelial cell. B. gingivalis none 138 Anti-S. gingivalis 3 0 Anti-43 kDa 3 7 Ant i -E. coli 1-49 4 5 Preimmune 124 E. coli 1-49 none 6 8 Anti-S. gingivalis 1 8 Anti-43 kDa 2 0 Ant i -E. coli 1-49 21 Preimmune 6 4 a l g G at 50 u.g per ml 97 identify the adhesin responsible. This study was an attempt to isolate the bacterial coaggregating adhesin from e. gingivalis. Prior to the isolation of the adhesin, coaggregating tests were carried out with a number of oral bacteria. S. mitis was found to be the best coaggregating partner for B. gingivalis A T C C 33277. Purification Chromatography The precipitate obtained after mixing B. gingivalis cells with a magnetic stirrer and precipitating the proteins from the supernatant by ammonium sulphate was used as the starting material for the purification of the bacterial coaggregating adhesin. This material was fractionated on Sephadex G-100, G-200 and on Sephacryl S-400 columns in the presence and absence of sodium deoxycholate. None of the fractionation procedures resulted in the purification of the adhesin. In the absence of sodium deoxycholate in the eluting buffer the coaggregating activity eluted in the void volume. On SDS-PAGE each fraction displayed the same protein profile as the starting material. In the presence of sodium deoxycholate the material eluted in several peaks with bacterial aggregating activity distributed throughout the peaks. Bacterial coaggregating activity could not be recovered following chromatography on DEAE-Sepharose. Adsorption assay The inability to effect a purification by standard chromatographic procedures led to the decision to try selective adsorption to S. mitis. Coaggregates form rapidly when S . mitis is mixed with B. gingivalis and for this reason it was thought that it would have good potential as a substrate for binding the solubilized adhesin. The dialyzed and concentrated 40 % ammonium sulphate precipitate was incubated with S. mitis, and the mixture agitated for 10 min. Aggregates formed by interaction of the adhesin with S. mitis were removed by centrifugation and more S. mitis cells were added to the supernatant. This process was repeated until no more aggregates formed. The 98 protein profiles of the adsorbed supematants are shown in Figure 20. As can be seen in lane A, the starting material contained several protein bands. After the first adsorption with S. mitis there was a marked decrease in two bands, one at approximately 46 kDa and the other at 53 kDA. After subsequent adsorptions, there was a further reduction in the concentration of these bands and after the fourth adsorption the 46 kDa band disappeared altogether. Other protein bands including the band at 43 kDa show no reduction . The aggregated S. mitis cells were washed in H E P E S buffer and the bacterial coaggregating adhesin was eluted. Several different eluehts were tested including 1 M NaCl, 1 % SDS and 0.15 M glycine-HCI-0.15 M NaCl pH 2.5. Incubation in the acidic glycine buffer for 10 min gave the best results. Following neutralization, the eluate was dialysed and lyophilized. It was found to have had strong bacterial aggregating activity. Only about 5 % of the bacterial aggregating activity was recovered in the glycine eluate. Attempts were made to remove more of the adhesin but it was found that although this was possible by prolonged exposure to 0.1 M glycine-HCI-0.15 M NaCl (pH 2.3), the activity of the eluted material was markedly reduced probably due to exposure to the low pH for protracted periods. It seemed likely that the majority of the adhesin remained bound to the S. mitis. This possibility was investigated by immunofluoresence analysis. S. mitis cells recovered following glycine elution were incubated with antibody raised against B. gingivalis or the 46 kDa protein and then with FITC labelled goat anti-rabbit IgG. The cells fluorescesed strongly with either antibody indicating that significant quantities of the bacterial aggregating adhesin were still attached to the cells. The eluate was analysed by SDS-PAGE and Western blot. On SDS-PAGE only two 99 Figure 20. Adsorption of the BA by S. mitis. S D S - P A G E analysis of supernatant recovered after removal of aggregates formed by incubating material containing solubilized BA with S. mitis., Lanes: A, starting material; B, supernatant after the first adsorption; C, supernatant after the second adsorption; D, supernatant after the third adsorption; E, supernatant after the fourth adsorption. 10 u.g of protein was loaded in each lane except lane A which was loaded with 30 ug of protein. A B C D E 100 bands were present, one at 43 kDa and another at 46 kDa. Similar results were seen when samples were immunoblotted against anti-B. gingivalis serum (Figure 21). In the absence of boiling no silver staining material entered the gel. The inability of the BA to migrate into the gel without boiling in SDS suggested that it existed in a multimeric form. As heating destroyed all bacterial aggregating activity, other solubilization procedures were tried in order to obtain nondenatured subunits. Urea (8 M), guanidine-HCI (6 M), LiCI (3 M), SCN" (3 M) and combinations of these compounds did not dissociate the BA molecule. Characterization of the solubilized BA To try and gain an insight into the nature of the adhesin-receptor reaction a number of potential receptor analogues were tested to see if they were capable of blocking the reaction of the solubilized BA with S. mitis. As can be seen in Table 12 none of the amino acids tested at a concentration of 100mM were effective in inhibiting the BA activity of B. gingivalis. Of the sugars tested, only galactose and galactose containing compounds inhibited bacterial aggregation, but the high concentrations required ( 500 mM) suggests that they were acting in a non-specific manner. Several surface modifying substances were tested for their effect on the bacterial aggregating activity (Table 13). EDTA, C a C I 2 and the reducing agents 8-mercaptoethanol and dithiothreitol had no effect on the bacterial aggregating activity whereas periodic acid decreased activity by 50 %. This is in direct contrast to the HA whose activity was enhanced by B-mercaptoethanol and unaffected by periodic acid. Heating of B. gingivalis cells or the BA at 8 5 ° C for 10 min or at 1 0 0 ° C for 1 min destroyed all bacterial aggregating activity. 101 Figure 21. Western-blot analysis of the supernatant remaining after adsorption of the BA with S. mitis. Samples were electrophoresed on 12 % S D S - P A G E . The blot was probed with anti-S. gingivalis serum. Lanes: A, starting material; B, supernatant after the second adsorption; C, supernatant after the fourth adsorption; D, material eluted from the aggregates with 0.15 M glycine-0.15 M NaCl, pH 2.3. 9 7-4 • 6 8 • 4 5 • A B C D 2 5-7 • 1 8-4 • 102 Table 12. Effect of various compounds on aggregation. Compound" Effect on bacterial aggregating activity3 D- mannose 0 0 D-glucose 0 D-galactose _c D-ribose 0 D-fucose 0 D-fructose 0 Lactose 0 Rhamnose 0 D-galactosamine _c D-glucosamine 0 N-acetylglucosamine 0 N-acetylgalactosamine _c L-Lysine 0 L-Arginine 0 L-leucine 0 aEffect on activity designated as follows: enhancement (+), inhibition (-), no effect (0). bAII compounds were tested at a concentration of 100 mM unless indicated otherwise. c Concentration: 500mM, inhibition 50 %. 1 0 3 Table 13. Effect of various substances on the bacterial aggregating activity of B. gingivalis. Substances Effect on bacterial aggregating activity3 EDTA (1 mM) 0 C a C I 2 (10 mM) 0 Periodic acid (0.01 M) - b B-mercaptoethanol (.2 %) 0 Dithiothreitol 50 mM 0 aEffect on activity desugnated as follows: enhancement (+), inhibition (-), no effect (0). D periodic acid inhibited the bacterial aggregating activity by 50 %. 104 Enzvmatic digestion of the adhesin The effect of proteolytic enzymes on the bacterial aggregating activity of B. gingivalis whole cells and the purified BA was evaluated. The enzymes were incubated with the bacterial cells or adhesin at 3 7 ° C for 1 h after which samples were assayed for bacterial aggregating activity. For B. gingivalis, the enzymes were removed by washing the cells three times with H E P E S buffer. The enzymes were not removed from the reaction mixtures containing the purified adhesin. In these cases bacterial aggregating activity was measured at 4 ° C . As can be seen in Table 14 all of the proteases tested reduced activity. Pronase and proteinase K eliminated all the bacterial aggregating activity of the purified adhesin. Of the other proteolytic enzymes trypsin was the most active, reducing the bacterial aggregating titre by over 80 % for both the whole cells and the purified adhesin. Neither lysozyme, mixed glycosidases nor phospholipase C had any effect on the activity. Treatment of S. mitis with the same enzymes had no effect on the bacterial aggregating activity. SDS-PAGE analysis of the proteinase K and pronase digestion mixtures is seen in Figure 22. The 46 kDa band was hydrolysed completely leaving the 43 kDa band. This hydrolysis was accompanied by a loss in bacterial aggregating activity, indicating that the 46 kDa protein is associated with the bacterial binding activity of B. gingivalis. Inhibition of bacterial aggregating activity bv antisera To provide more evidence that the 46 kDa protein is the BA, specific antiserum raised against the 46 kDa protein which had been eluted from an S D S - P A G E was assayed for its ability to block bacterial aggregation. The antisera was serially diluted and preincubated with B. gingivalis cells prior to the addition of S. mitis. As can be seen in Table 15, the inhibitory titre of the anti-46 kDa serum and anti-6. gingivalis serum was 256. In comparison, pre-immune serum and anti-43 kDa serum had no inhibitory effect on the bacterial aggregating activity. 105 Table 14. Effect of enzymatic treatment on the bacterial aggregating activity of B. gingivalis and the purified BA . Bacterial Aggregating Titre. Enzyme. 3 Intact cells. Purified adhesin. None 256 256 Pronase 16 0 Proteinase k 8 0 Papain 64 32 Pepsin 1 28 64 Trypsin 32 32 Chymotrypsin 1 28 1 28 Lysozyme 256 256 Mixed glycosidases 256 256 Phospholipase C 256 256 3 Enzymes concentration was 100 ug per ml. 106 Figure 22. SDS-PAGE analysis of the enzymatic digestion of the bacterial aggregating adhesin from B. gingivalis 33277. 10 ug of enzyme was used in each of the digestion mixture. Lanes: A, pronase digestion; B, proteinase K digestion; C, trypsin digestion; D, mixed glycosidase digestion; E, phospholipase C digestion; F, lysozyme digestion; G, pronase control; H, proteinase k control; I, trypsin control; J, phospholipase C control; L, undigested control of bacterial aggregating adhesin. The arrow indicates the 46 kDa protein. 107 Table 15. Effect of immune serum on bacterial aggregating activity. Treatment Inhibitory Titre. Anti- B. gingivalis 128 Anti- 43 kDa 2 Anti- 46 kDa 25 6 Preimmune serum 0 a l g G fractions contained 100 ug per ml of protein. 108 An Elisa type assay was developed to provide a more quantitative estimate of the inhibitory effect of the antisera. Both antisera were resuspended to 100 ug per ml of IgG and serially diluted in HEPES buffer. B. gingivalis was added and the mixtures were incubated for 1 h at room temperature with shaking. 100 u,l of the mixture was added to wells of a microtitre plate coated with S. mitis cells and incubated for 1 h for binding to take place. The unbound cells were removed by washing and primary antiserum (anti-46 kDa) was added. After incubation, unbound antisera was washed away and goat anti-rabbit IgG conjugated to alkaline phosphatase was added. As seen in Table 16 preincubation of the B. gingivalis with anti-46 kDa serum eliminated its ability to bind to the immobilized S. mitis. Preimmune serum had no effect. Immunogold labelling To localize the BA and determine if it was associated with any particular structure on the cell surface, immunogold labelling was carried out on whole cells and thin sections of B. gingivalis. Figure 23 shows whole cells of B. gingivalis reacted with anti-46 kDa primary sera followed by gold beads conjugated to goat anti-rabbit IgG. In this electron microphotograph it can be seen that gold beads are distributed evenly over the cell surface. Numerous fields were examined but there was no evidence for the labelling of fimbriae or pili. When thin section of B. gingivalis were reacted with antisera to the 46 kDa protein (Fig. 24) most of the gold beads were located on the outer membrane or the periplasmic space. Clumps of gold beads can be seen in areas around the thin sections suggesting that the adhesin may have been dislocated during cell processing. 109 110 Table 16. Effect of anti-46 kDa antiserum on the binding of B. gingivalis to immobilized S. mitis. — 405 nm Dilution of antisera 3 Antisera Anti-46 kDa Preimmune 2 0.08 1.0 4 0.20 0.95 8 0.35 0.90 1 6 0.66 0.85 32 0.80 0.90 64 0.83 0.90 1 28 0.96 0.95 256 1.0 1.0 512 1.0 1.0 3 Initial concentration of antisera was 100 ug protein per ml. 110 Figure 23. Immuno gold bead labelling of B. gingivalis. Cells were reacted with anti- 46 kDa serum and then incubated with gold beads conjugated to goat anti-rabbit IgG. Cells were negatively stained with uranyl acetate. Bar, 0.5 urn. 111 Figure 24. Immuno gold bead labelling of thin sections of B. gingivalis. The sections were reacted with anti-46 kd serum, incubated with goat anti-rabbit IgG conjugated to gold beads and then negatively stained with 5 % uranyl acetate. The arrow indicate the accumulation of the gold beads in the outermembrane. Bar, 0.5 urn. 112 Vesicles The possibility that vesicles released by B. gingivalis retained the adhesive properties of the whole cell was investigated. It was decided to assess the role of vesicles in mediating attachment of S. sanguis to model tooth surfaces. The model chosen was hydroxyapatite (HA) to which had been adsorbed a pellicle derived from serum (SeHA). Serum was used to coat HA because it is similar in composition to crevicular fluid which bathes surfaces in the periodontal pocket. S. sanguis was chosen because it is an early colonizer and is found in significant numbers in subgingival plaque Isolation of vesicles Vesicles of B. gingivalis 33277 were isolated from 3 day old culture supernatants by ultrafiltration followed by ultracentrifugation. The yield was found to be 25 mg (dry weight) of vesicles per litre of culture. Electron microscopic examination of the vesicular preparation revealed that it was homogeneous in regard to morphology (Fig. 25) and free of fimbriae. The average size of the vesicles was approximately 25 nm. The protein and LPS profiles of the outer membrane and vesicles of B. gingivalis are shown in Figure 26. They are similar supporting the contention that the vesicles are derived from the outer membrane. Vesicles also have haemagglutinating and bacterial aggregating activity. Effect of vesicles on attachment of S. sanguis 12 to SeHA Figure 27 panel A shows the attachment of vesicles to SeHA . The result of adding vesicles to SeHA (SeHA-V) on the binding of S. sanguis is shown in Table 17. In the 8 presence of vesicles the number of S. sanguis attached was 4.8 x 10 cells. In the absence of vesicles the number attached decreased at least 10 fold to 4.6 x 10 7 cells. SHA beads had 8.0 x 10 7 bacteria attached. When vesicles were included this number 113 Figure 25. Transmission electron micrograph of a vesicle preparation of B. gingivalis prepared by ammonium sulphate precipitation of 5 day old culture supernatant. Bar 50 nm. 1 if 1 i fifes . ' •- v -A-114 Figure 26. B. gingivalis SDS-PAGE profile of vesicles and outer membrane.. Gel 1, stained for protein with silver nitrate. Gel 2, stained with silver for lipopolysaccharide. Lane A, vesicles; Lane B, outer membranes. 10 and 25 ug of material was loaded in lanes A and B respectively. A B A B was 2.8 x 10° cells, an increase of 3.5 times. The electron micrographs seen in Figure 27 illustrates the differences described in Table 17. In the absence of vesicles very few bacteria attached to SeHA (Fig. 27 B). However, in the presence of vesicles the number of S. sanguis increased significantly (Fig. 27 C). Kinetics of the bindino reaction A typical binding isotherm for the adherence of S. sanguis to SeHA with and without vesicles is seen in Figure 28. The number of S. sanguis which adsorbed to SeHA-V beads increased as the number of bacteria added increased. Saturation of the beads did not occur until the number of S. sanguis added exceeded 2.5 x 10 8 cells. In the absence of vesicles a similar pattern of binding was seen but the total number of S. sanguis bound was 10 fold less. When the data for the adherence of S. sanguis to SeHA with and without vesicles were analysed by the Langmuir equation, the curve indicated independent and identical binding sites (Figure 29). Reaction parameters calculated from the Langmuir isotherm (Table 18 ) show that the number of binding sites for the vesicle treated SeHA was about four times higher than that of untreated SeHA. The dissociation constant for vesicle treated SeHA was more than twice that of untreated beads indicating stronger attachment of S. sanguis to the vesicle covered SeHA. Analysis of the adherence data in the form of a Scatchard plot (Figure 30 ) revealed a curve with a negative slope indicating that the sites of interaction were homogenous and that there was no positive cooperativity between the sites. The effect of increasing the quantity of vesicles on the attachment of S. sanguis to the SeHA was evaluated and the result is shown in Figure 31. As can be seen the number of bacteria bound increased as the quantity of vesicles added increased. The curve began to plateau after 100 u\g of vesicles was added and the reaction was 116 Table 17. Influence of vesicles on the binding of S. sanguis to SeHA and SHA Treatment SeHA SeHA-V SHA SHA-V # of bacteria attached 4.6 x 10 7 4.8 x 10 8 8.0 x 10 7 2.8 x 10 8 117 Figure 27. Scanning electron micrograph of SeHA incubated with (A) B. gingivalis vesicles, (B) S. sanguis, and (C) B. gingivalis vesicles followed by S. sanguis. Hydroxyapatite beads were incubated with serum followed by vesicles and bacteria then coated with palladium gold.Bar, 1 um. Figure 28. Adsorption isotherm of S. sanguis bound to SeHA with and without vesicles. Hydroxyapatite beads were incubated with serum followed by vesicles and then bacteria. B, bound cells ( x 10 7 ) ; U, unbound cells (x 10 8 ) . ( B ) SeHA with vesicles; ( • ) SeHA without vesicles. 30 -\ 119 Figure 29. Langmuir adsorption isotherm for the adherence of S. sanguis to SeHA, and SeHA pre-incubated with B. gingivalis vesicles. Hydroxyapatite beads were incubated with serum followed by vesicles and bacteria. B, bound cells x 10 7; U, unbound cells x 10 8 . ( • ) SeHA with vesicles; ( • ) SeHA without vesicles. o H > 1 • 1 • 1 ' 1 0 10 20 30 40 U 120 Table 18. Langmuir isotherm parameters for the adsorption of S. sanguis to SeHA and SeHA-V. N K r SeHA (without vesicles) 8.4 x 10 7 1.2 x 10 0 9 2 SeHA (with vesicles) 3.5 x 10 8 2.8 x 10" 9 0.98 N, Number of binding sites; K, dissociation constant; r, correlation coefficient. 121 Figure 30. Scatchard adsorption isotherm for the adherence of S. sanguis to SeHA pre-incubated with B. gingivalis vesicles. Hydroxyapatite beads were incubated with serum followed by vesicles and bacteria. B, bound cells x 10 7 , U, unbound cells x 10 8 . 0.20-1 0.15-0.10-0.05-o.oo-I • i i 1 i 1— 0.0 ' 1.0 2.0 3.0 B 122 Figure 31. Adhesion of S. sanguis to S e H A pre-incubated with increasing quantities of B. gingivalis vesicles. Hydroxyapatite beads were incubated with serum followed by vesicles and radiolabeled bacteria. 30 - i CO o mg of uesicles added. 123 Table 19. The effect of heat and putative inhibitors on the attachment of S. sanguis to SeHA-V beads. Treatment or inhibitor # of cells attached ( x 1 0 7 ) none (without vesicles) 4.1 none (with vesicles) 25.7 heated bacteria (70°C/10 min) 16.3 heated vesicles (70°C/10 min) 4.1 sialic acid (50 u,g/ml) 21.5 lactose (100 mM) 15.8 arginine (100 mM) 13.3 124 eventually saturated after the addition of 400 u.g of vesicles to the serum coated hydroxyapatite beads. Experiments were carried out to find the optimum time of incubation for adsorption of vesicles to SeHA and the optimum time for attachment of S. sanguis to SeHA-V. Previous to this all incubations were carried out for 2 h. The optimum time for adsorption of vesicles to SeHA and for the attachment of S. sanguis to SeHA-V was found to be 1 h. 3 7 ° C was found to be the optimum temperature for binding. Pretreatment of the vesicles at 7 0 ° C for 10 min completely eliminated their ability to promote the attachment of S. sanguis to SeHA (Table 19). The heat treatment did not affect the ability of vesicles to adsorb to SeHA as evaluated by scanning electron microscopy. Heating S. sanguis at 7 0 ° C for 10 min reduced its ability to bind to vesicle treated SeHA by 35 %. Several substances were tested for their ability to inhibit the adherence of S. sanguis to SeHA-V. Of these, arginine and lysine at 100 mM concentration reduced the binding by 48 and 38 % respectively (Table 19). Sialic acid reduced binding by 16 %. Spectrophotometric adherence assay A number of species and strains of oral streptococci were assayed to measure the effect of vesicles on their binding to SeHA. As can be seen in Table 20, the attachment of two of the three strains of S. salivarius were markedly enhanced by the presence of vesicles. Of the of S. mutans tested only M2B binding was enhanced. In the case of S. sanguis, only strain 12 showed enhanced binding. This concludes the study on the adherence reactions of B. gingivalis. The HA has been isolated and cloned, the bacterial aggregating adhesin has been identified and purified and the role of vesicles in mediating attachment of oral streptococci to SeHA has been studied. 125 Table 20. Spectrophotometric assay of the adherence of oral streptococci to vesicles immobilized on SeHA. Number of bacteria attached (x 10 8) Organism without vesicles with vesicles S. sanguis 12 11.0 14.0 NY101 9.83 10.7 E 15.5 16.7 J11 8.73 9.4 S. salivarius K8 15.5 17.8 K3 12.0 16.7 HB 10.5 14.4 S. mutans 19645 11.8 12.4 B13 9.4 9.4 M2B 7.5 11.4 Lipopolysaccharides One of the original plan of this study was to determine whether LPS of B. gingivalis was directly involved in haemagglutination and bacterial aggregation. For this reason LPS were isolated and preliminary chemical characterization was completed. In terms of their biological properties no evidence could be found for a role in adherence. Analysis of LPS from Bacteroides LPS from the oral BPB was prepared by the method of Darveau and Hancock (35). This material was found to constitute approximately 4 % of the bacterial dry weight. Spectra analysis of the LPS (250 u.g/m!) indicated that they were free of nucleic acid. The size range of LPS from BPB was investigated by sodium-dodecyl sulphate polyacrylamide gradient gel electrophoresis (5 to 20 %). As can be seen in Figure 32, all of the LPS preparations displayed a prominent silver staining band which ran close to the dye front. This represent the lipid A portion plus the core region of the LPS. All of the B. gingivalis strains (lanes 1-5) display the typical ladder like profile of smooth type LPS. The molecular weight range between core material and the 200,000 molecular weight marker with a concentration of LPS material between the 18-25 kDa and 60-70 kDa range. Strain 2D appears to have more bands in the high molecular weight area than strains W 12 and W 50, this result was found to be the same for all LPS preparations of these two organisms. The two strains of B. intermedius ( lanes 6,7) have very similar profiles on SDS-PAGE, but they have fewer number of high molecular weight bands when compared to the LPS of the B. gingivalis strains. B. asaccharolyticus (lane 8) has LPS of three molecular weight ranges, between 10 and 12 kDa, 20 and 25 kDa and 60 and 70 kDa. 127 Figure 32. S D S - P A G E analysis of LPS from black pigmented Bacteroides. Lanes 1 through 5, B. gingivalis W 12, W 50, W 83, 2D and 33277 respectively; Lanes 6 and 7, B. intermedius BMA and 2332H respectively; Lane 8, B. asaccharolyticus; Lane 9, B. denticola; Lane 10, B. levii; Lane 11, B. melaninogenicus; Lane 12, E. coli. 12 11 10 9 8 7 6 5 4 3 2 1 200^ 4 3*-18-4 1 2 -3--I 128 B. denticola has a large amount of core and very little of the higher molecular weight material. The same result was found for the LPS of B. levii. This would indicate that the ratio of carbohydrates to lipid is low in the LPS of these two species when compared to the other BPB . Chemical analysis of LPS The LPS were characterized with regard to their carbohydrate and lipid composition. All analysis were done in duplicate and on two different samples. The quantity of protein present was less than 0.2 % of the total dry weight. Table 21 gives the results obtained when the nine LPS-s were assayed for carbohydrates and lipids. Total carbohydrates was measured by the phenol sulphuric method (43). The total percentage of carbohydrates for the four strains of B. gingivalis were similar varying between 60 % for W 83 to 67 % for W 12. The total carbohydrates of the other BPB averaged from 35 % for B. denticola to a high of 55 % for B. intermedius BMA. The methylpentoses varied between a high of 25 % for B. denticola to a low of 4 % for B. intermedius 2332H. The lipid portion of B. gingivalis displayed more variability when compared to the total carbohydrates with a high of 30 % for W 50 and a low of 19 % for 2D. All other BPB had higher percentage of lipid A with a high of 45 % for B. denticola to a low of 27 % for t B. intermedius BMA. No 2-keto-3-deoxyoctonate (KDO) or heptose were found by chemical analysis in any of the BPB lipopolysaccharides analysed, whereas LPS from E. coli contained these two sugars. 129 130 Table 21. Chemical composition ( in percentages ) of LPS from oral BPB. Organism Composition (%) Pro 3 . ( C H 2 0 ) n b Me. Pent 0. FA d . B. gingivalis W 12 0.1 67 1 3 21 W50 0.05 63 1 1 30 W 83 0.1 60 1 3 22 2D 0.1 60 17 19 B. asaccharolyticus 0.1 4 8 9 3 0 B. intermedius BMA 0.1 5 5 8 2 7 2332H 0.1 4 8 4 3 3 B. denticola 0.05 35 25 45 B. levii 0.1 4 0 1 3 4 0 3 protein, D total carbohydrates, , c methylpentoses, fatty acid. 130 Gas liquid chromatography of sugars from LPS of Bacteroides The LPS'S were hydrolysed in 4 N HCI under nitrogen in glass vials, derivatised to their alditol acetates and analysed by gas liquid chromatography. The result is shown in Table 22. The sugars present in the core and O-antigenic side chains of LPS's were rhamnose, mannose, galactose, glucose, galactosamine and glucosamine. B. denticola and B. levii were the exceptions in that they were devoid of mannose. In the LPS of B. gingivalis an unknown peak was found with retention time similar to that of the internal standard myoinositol, but its identification was not known for certain so it was classified as an unknown. Fatty acid composition The fatty acids from the LPS of 9 Bacteroides were hydrolysed and methylated before being analysed by gas liquid chromatography. As can be seen in Table 23, the major fatty acids were myristic ( C 1 4 : 0 ) palmitic ( C 1 6 : 0 ) , heptadeconoic ( C 1 7 I S 0 ) and stearic ( C - ^ . Q ) . For the B. gingivalis strain there was an unknown with retention time close to that of myristic acid which had a retention time of about 10 min. Only B. gingivalis showed the presence of pentadeconic and heptadeconic acids although they were present in amounts less than 5 % of the total fatty acid content.There was considerable variation within the species in the quantity of any particular fatty acid. Myristic acid was present at a concentration of 32 % for B. levii and 25 % for B. intermedius BMA, all of the other BPB had concentrations between 11 and 20 %. Palmitic acid was present in the highest concentration in all of the BPB tested. B. gingivalis W 50 had the highest concentration at 40 % followed by B. gingivalis W 12 at 35 % and B. levii at 30 % . Stearic acid was present in B. gingivalis W 83 at a concentration of 43 % folloiwed by B. denticola at 40 %. B. intermedius showed the presence of pentadeconoic acid which was not present in any other species. 131 Table 22. Carbohydrate compositiorTof lipopolysaccharides from BPB. Bacteria Rha a Man 0 Gal° Glu d h NH 2 g lu e h NH 2 gal f Unl B.gingivalis W 12 6.5 10.3 12 29 + + 25 W50 13.7 2.1 8.5 41 + + 27 W83 4.5 4 28.5 12 + + 30 W2d 5 4.4 13.3 60 + + 13 B. asaccharolyticus 4 9 18 32 + + -B. intermedius 2332 2 3 4.3 86 + + -BMA 7.5 10.8 21 54 + + -B. denticola 6 - 18.4 75 + - -B. levii 23 3 55 + _ As a percentage of the total carbohydrate present. a rhamnose, D mannose, c galactose, d glucose, e galactosamine, *-glucoasmine, 9 unidentified sugar with identical retention time. +, present. -, not present. h, glucosamine and galactosamine were not quantitated. 132 Table 23. Fatty acid composition of LPS from BPB. Bacteria % of total fatty acid C 14 a C15:1 D C 1 6 c C17:1 d C l 7 e C18 f C 1 8 : 1 a U 1 h 112' B. gingivalis W 12 1 3 5 35 3 7 1 7 - 1 7 -W50 1 1 2 40 2 4 21 - 1 1 -W83 1 5 7 1 2 2 6 43 - 1 0 -2D 1 4 3 1 9 4 1 1 30 - 1 3 -B. asaccharolyticus 1 7 - 27 - - 23 - 7 20 B. intermedius BMA 25 - 1 7 - - 20 1 3 - 1 7 2332H 1 9 - 1 8 - 1 5 22 - - 20 B. denticola 20 - 27 - - 40 - - -B. levii 32 _ 30 1 7 1 1 _ 9 -, not present. 3 myristic acid, D pentadecanoic acid, c palmititic acid, d methyl heptadecanoic acid, e heptadecanoic acid, * stearic acid, a oleic acid, n unidentified fatty acid, 1 unidentified fatty acid. 133 Western electroblottina analysis LPS separated by S D S - P A G E and transferred from the electrophoretogram to nitrocellulose paper by the Western blot technique (17) was reacted with polyclonal antisera to whole cells of different strains and species of Bacteroides. Antisera to B. gingivalis strains reacted only with B. gingivalis LPS despite the similar profiles obtained on S D S - P A G E when compared to other species such as B. asaccharolyticus (Fig. 33). Antisera to 33277 reacted with LPS from all B. gingivalis strains but not with LPS from other strains. 134 Figure 33. Reaction of LPS from BPB with antisera to B. gingivalis 33277. Lanes 1 through 4, B. gingivalis W 50, W 83, 2D and 33277 respectively; Lane 5, B. asaccharolyticus; Lanes 6 and 7, B. intermedius BMA and 2332H respectively; Lane 8, B. denticola; Lane 9, B. levii; Lane 10, E. coli. 135 Discussion Haemaqalutination Haemagglutination is a distinctive characteristic of B. gingivalis which differentiates it from other asaccharolytic BPB species (175). It is generally accepted that bacteria that haemagglutinate possess specific cell surface adhesins and that the presence of these adhesins can be an essential virulence determinant of many infectious diseases. Since the pathogenicity associated with B. gingivalis stems from adhesive interactions with eukaryotic and procaryotic cells (214) the adhesins of this organism are of particular interest. Recently, there have been three reports on the purification of the B. gingivalis H A (100,163,176), but analysis of the purified material by S D S - P A G E revealed that several protein bands were present and it was not possible to link a specific band to the HA. In the study reported here B. gingivalis HA was purified and shown to be a protein with an apparent molecular weight of 43 kDa. The size of the native HA was found to be in excess of 10 6 daltons as was indicated by its elution in the void volume of a Sephacryl S-1000 column and by its inability to enter a 5 % polyacrylamide gel if it had not been boiled in SDS. Molecules of this size are indicative of structures such as pili or fimbriae and indeed previous reports had suggested that HA was associated with pili (175). However, no evidence was found to corroborate this, on the contrary evidence obtained from this study suggests that the HA is not on a pili. For example, transmission electron microscopic examination of the purified HA and cells of B. gingivalis 33277 did not revealed any fimbriae. Immunogold labelling of B. gingivalis with either anti-43 kDa or anti-HA sera did not show the presence of these structures. Fimbriae were not detected in the purified preparations of the HA. 136 Attempts were made to obtain active subunits of the HA by methods which usually dissociate fimbriae of gram negative organisms. Fimbrial proteins from E. coli are insoluble in their native form but may be disaggregated by guanidine hydrochloride and HCI or by treatment with detergents and other chaotropic agents (16). However, none of the treatments tried had any effect on dissociating the B. gingivalis HA complex with the exception of boiling in SDS. Prolong exposure to this detergent at 37 ° C had no effect in dissociating the HA. This indicates that the bonds holding the subunits of the HA together are very strong and in that sense are similar to adhesins of A. viscosus and S . salivarius (242). The HA has all of the properties of a porin protein in terms of its molecular weight and abundance on the cell surface, however there is not enough evidence to conclude that it is a porin. Pronase and proteinase K destroyed the haemagglutinating activity of the purified HA, yet on SDS-PAGE there was no perceivable change in the migration or quantity of the 43 kDa protein. In E. coli a similar phenomenon was seen for a non-fimbrial haemagglutinin with a molecular weight in excess of 10 6 daltons. Protease treatment destroyed the haemagglutinating activity of the molecule but the molecule was not hydrolysed as seen by S D S - P A G E (75). An explanation for this reaction could be (1) the'enzymes only removed a few amino acids from either the amino or carboxy end of the protein and this is not enough for a difference in molecular weight to manifest itself on S D S - P A G E , (2) there was a minor protein at 43 kDa which was masked by the 43 kDa protein, (3) the proteolytic enzymes undergo self degradation and the fragments acted as inhibitors by occupying the active site of the adhesin, or (4) the enzymes do not hydrolyse the HA but in some manner block the active site. The possibility that a few amino acids were removed by the enzymes was not thought to be the answer since both digested and undigested HA had the same pi although it must be noted that if the amino acids removed by the enzymes were not charged then there would be no change in the pi. The chance that a minor protein of 43 kDa is present is remote since the red blood cell were shown to adsorb all of the 43 kDa protein. The third and fourth possibilities seems unlikely as proteases preincubated before being added to the HA showed no inhibitory activity. 137 The strongest evidence implicating the 43 kDa protein as the HA was provided by cloning the HA gene into E. coli. Two clones with haemagglutinating activity were identified and one, 1-49 was characterized. This recombinant produced a B. gingivalis protein with a molecular weight of 43 kDa as seen by SDS-PAGE and Western blot analysis. The 43 kDa protein was immunologically similar to the 43 kDa subunit of the HA of B. gingivalis as shown by western blot. Immunogold and immunofluoresence staining with anti-43 kDa or anti-1-49 sera showed that this molecule was present on B. gingivalis and the recombinant 1-49. Anti 1-49 serum inhibited the haemagglutinating activity of B. gingivalis and likewise anti-43 kDa serum inhibited the haemagglutinating activity of 1-49. The two proteins, the 43 kDa from B. gingivalis and the 43 kDa protein from 1-49 behaved similarly when exposed to protease. They both lost their ability to agglutinate erythrocytes but their mobility in S D S - P A G E was not affected. Both were sensitive to heating. The surface location of the HA in E. coli 1-49 suggest that B. gingivalis surface antigens can be processed as well as expressed in E. coli. The insert DNA probably codes for a leader sequence which was removed after which the protein was transported to the surface of the outer membrane of the recombinant. B. gingivalis colonizes the sulcus a region which is continuously flushed by the serum-like crevicular fluid. Previous studies by Slots and Gibbons (214) had shown that serum was inhibitory to haemagglutinating activity of B. gingivalis and therefore cast doubt upon the significance of HA reaction in mediating binding in vivo. IgG prepared from nonimmune serum when compared to IgG prepared from serum to the HA, the 43 kDa, E. coli 1-49 and to B. gingivalis displayed only limited inhibitory activity and confirms the observation of Boyd and Mc Bride that serum has only a small effect on binding (11). 138 The haemagglutinating activity of a number of bacteria is inhibited by sugars such as mannose, lactose and galactose (47,118 ) however, in this study none of the putative receptor analogues tested were found to be inhibitory to the activity of the HA of B. gingivalis. Arginine and lysine were reported to be receptor analogues for the erythrocytes (100,176) but the high concentration required suggests the reaction is non specific. The immunogold staining pattern seen when B. gingivalis was reacted with either anti-43 kDa or anti-1-49 sera suggested that the HA was evenly distributed over the surface of the cell. Thin section analysis of the cell revealed that the majority of the gold beads were localized on the outer membrane and in the periplasm. This is consistent with the work of others who also reported that the adhesin was distributed in an even manner over the cell surface (163,176). Immunogold bead labelling with anti-43 kDa serum showed that the protein was present on the surface of all strains of B. gingivalis but not on other BPB. This is consistent with the inability of these cells to haemagglutinate and with the lack of immunological cross reactivity between species of BPB. There was variation in the intensity of fluoresence suggesting that the different strains of B. gingivalis express different levels of the 43,000 Mr protein or that there is a difference in the accessibility of the protein to the antibody. In the oral cavity, the ability of B. gingivalis to attach to epithelial cells may be important in its colonization of the sulcus. B. gingivalis has been shown to be capable of attaching to epithelial cells giving rise to the question of whether the HA and epithelial cell binding adhesin are the same. The results obtained from this study show that the HA and the epithelial cell binding factor are one and the same. Evidence that supports this conclusion are: (1) anti-43 kDa, anti-HA and anti-1-49 IgG inhibited both haemagglutinating and binding to epithelial cells; (2) heat destroyed both activity 139 simultaneously (3) protease destroyed both activity simultaneously (4) the recombinant E. coli 1-49 gained both haemagglutinating and epithelial cell binding capability. The binding of E. coli 1-49 to buccal epithelial cells was examined by phase contrast, scanning electron and transmission electron microscopy. There appeared to be two population of cells, one which supported binding and the other (10 %) which did not . Similar results were found by Gibbons et al (69) when they looked at the binding of oral bacteria to epithelial cells. They found that fibronectin was a potent inhibitor of attachment. Mild trypsin treatment of the epithelial cells enhanced binding by removing and unmasking the bacterial receptors. Transmission electron microscopy of thin section of bacteria attached to epithelial cells showed that bacteria attached to plicate like extensions on the epithelial cell from their surface. Epithelial cells which did not support binding did not have these projections, they had smooth surfaces. Similar results have been reported by Williams et al (252) studying the attachment of enteropathogenic E. coli attached to gut epithelial cells. The explanation for this phenomenon is not known but it appears as if on recognition of the adhesin by the receptor, the cytoskeleton of the epithelial cells are transformed to produce these extensions. Bacterial agglutinin In the oral cavity bacteria accumulate by binding to host tissue or to other bacteria (214). This latter is extremely important as most of the interaction which lead to accumulation of bacteria involve bacteria-bacteria interaction. B. gingivalis is an organism which colonize in areas where microflora has been established, thus it can be imagined that the ability to bind to other plaque organism is important in the ecology of the organism. In the studies reported here it was shown that bacterial aggregating activity can be attributed to a surface protein (M r 46 kDa). The purified protein is able to induce aggregation of several species of oral bacteria. Purification of this adhesin 140 proved to be relatively difficult since none of the conventional biochemical procedures tested were effective. However, specific adsorption of BA with S. mitis yielded a pure and active fraction. There was a problem in that only 5 % of the material could be recovered from S. mitis. Immunofluoresence of S. mitis with anti-46 kDa serum revealed that much of the BA remained associated with the cells even after extreme attempts to remove it. Whether a covalent bond is formed between the adhesin and the receptor is not known but it is conceivable that a peptide bond can be formed between an N H 2 group on the adhesin and a -COOH group on the receptor such a system was recently described for the binding of C. albicans to epithelial cells. The molecular weight of the adhesin molecule on SDS-PAGE was approximately 46 kDa and it always copurified with a protein of approximately 43 kDa. Since the material with BA activity had HA activity, this suggests that both haemagglutinating and bacterial aggregating adhesins are part of a complex. An explanation for this association could be that the bacteria placed more than one adhesin in a complex with the majority of proteins in the complex responsible for haemagglutination, since this protein is in much greater abundance than the 46 kDa protein. Immunogold electron microscopy with anti-46 kDa serum showed that the bacterial coaggregating adhesin was evenly distributed located on the surface of the cell and did not appear to be associated with fimbria.When B. gingivalis was sectioned and labelled with anti-46 antiserum, most of the labelling was concentrated in the periplasm and outer membrane. There were no fimbriae present in purified preparations of the bacterial coaggregating adhesin as seen by electron microscopy. Anti-46 antiserum inhibited BA activity of B. gingivalis to the same extent as that of anti-S. gingivalis antiserum, whereas preimmune serum had no inhibitory effect. Anti-43 antiserum had no effect on BA activity, thus proving that BA and HA activities are found on two immunologically distinct molecules. On Western blot analysis there was 141 no cross-reactivity between the 46 kDa BA and antiserum to E. coli 1-49. Several reports have identified bacteria-bacteria adhesin-receptor reactions which are inhibited by sugars. Kolenbrander et al (118 ) has identified a number of coaggregation reaction of oral bacteria which can be inhibited by galactose and lactose. In this study several sugars were tested as putative inhibitors for their ability to inhibit bacterial aggregation between B. gingivalis and S. mitis. None of the sugars tested proved to be inhibitory with the exception of galactose and galactose containing sugars, but the high concentration needed indicate a non-specific effect. Many gram negative bacteria coaggregate with gram positive bacteria. This includes the B P B B. loeschii which causes the aggregation of S. sanguis and A. israelii (245). Using anti-46 kDa serum several BPB were tested for the presence of immunologically similar molecules on their cell surface. All strains of B. gingivalis tested expressed an immunologically similar 46 kDa protein as determined by Western blot analysis. None of the other BPB tested reacted with the antiserum indicating the uniqueness of the B. gingivalis coaggregation reaction. The purified bacterial aggregating adhesin caused the aggregation of several species of bacteria tested including S. sanguis and S. salivarius. This would indicate that B. gingivalis has only a single adhesin for coaggregation of a number of organisms rather than an adhesin for each coaggregating partner. This is not the case with some organisms which possess a number of adhesins, each capable of reacting with a different bacteria (245 ) . Vesicles Bacterial binding to organic matrices adsorbed to hydroxyapatite is a complex process. Most studies have involved looking at saliva-coated hydroxyapatite but more recently the 142 role of serum and other organisms has been evaluated (131,214). In trying to assess the role of a receptor source such as saliva or serum it is important to note that other constituents may play a role, eg dextran, glucosyltransferase (242). These supplementary factors will increase the diversity of organisms able to bind by increasing the range of potential receptors. Presumably they will also eliminate some potential binding sites by occupying receptors. One such modifier of the binding activity of saliva or serum derived pellicle could be the vesicles released by the Gram negative plaque microflora. These structures resemble the outer membrane in terms of their chemistry and it is expected that they would possess those binding properties associated with the whole cell outer membrane. These vesicles would represent a concentrated array of immobilized putative receptors if they were bound to the pellicle surface. The B. gingivalis vesicles reported here were found to bind avidly to serum- or saliva-coated HA. The bound vesicles then served as a source of receptors for other oral organisms. The vesicles were tightly bound to the insoluble substratum as they resisted removal by washing even when they were acting as a bridge between HA and the streptococci. The experiments reported here show that vesicles can act as a supplementary form of receptor with the potential to mediate the attachment of a heterogeneous group of organisms to plaque and would therefore contribute to the microbial complexity of plaque. The vesicle mediated binding reaction would not take place at all subgingival sites but rather would only occur at specific sites and at those times when an appropriate flora was present. At those times the vesicles could have a significant effect on colonization. The electrophoretic analysis of the vesicles supports the contention that vesicles are similar to outer membranes. Williams and Holt (251) showed that vesicle protein profiles were somewhat more complex than outer membranes but contained the same major proteins. It is possible that the minor bands seen in vesicles are partial degradation products which arise from the proteolytic activity which is known to be 143 present in vesicles (80,81). The increase in binding of S. sanguis, S. mutans and S. salivarius to vesicle treated SeHA correlated with the ability of these cells to coaggregate with B. gingivalis whole cells. Those strains which did not coaggregate were not affected by the presence of vesicles adsorbed to SeHA whereas those that coaggregate very strongly e.g. S . salivarius, showed a marked increase in binding. While the assay demonstrates the potential of the vesicles to mediate binding there are obviously other factors that will determine whether an organism can utilize this binding mechanism in vivo. S. sanguis can be isolated from the sulcus in significant numbers and therefore would be in a position to avail itself of this receptor. On the other hand S. salivarius is not commonly found in the sulcular flora and presumably would not bind via this mechanism. Vesicles also induce aggregation of S. sanguis and thus it can be envisaged that accumulation of S. sanguis within the plaque matrix would be mediated by vesicles derived from B. gingivalis. Adsorption of vesicle by organisms such as S. sanguis would have the added beneficial effect of removing a potentially harmful factor. The vesicles would be prevented from coming into contact with the gingival tissues where they could exert this pathogenic effect. Langmuir and Scatchard analysis of S. sanguis binding to the SeHA immobilized vesicles shows a single type of interaction. This suggests that the binding as measured in these assays is to the vesicle and not to serum or saliva derived receptors implying that the vesicles block the pellicle receptors. The electron micrographs show that vesicles cover a significant portion of the SeHA surface and that binding of S. sanguis to the vesicles would exclude a high percentage of potential pellicle binding sites. The coating of HA beads with serum was done to more closely mimic the subgingival hard tissue surfaces which are bathed in crevicular fluid. Bacterial binding to serum or crevicular fluid coated beads differs from binding to SHA. Cimasoni et al. (23) and Liljemark et al. (127,128) found that in contrast to saliva, human crevicular fluid and 144 serum inhibited binding of S. sanguis to HA. The vesicles would thus provide a mechanism for S. sanguis to adhere. Slots and Gibbons (214) reported that serum interfered with the binding of B. gingivalis to SHA, by extrapolation it would be expected that the vesicles would be affected in a similar way. However it was shown in this study that vesicles readily attach to SeHA. This confirms earlier work by Cimasoni et al (23). which showed by scanning electron microscopy that B. gingivalis cells bound to SeHA. The enhanced binding to immobilized vesicles demonstrates the potential importance of bacterial vesicles as mediators of adherence to hard tissues and indicates that the pellicle can be an extremely complex structure containing a diverse array of host and bacterially derived receptors. This type of interaction could also be important in promoting the succession of microorganisms within the sulcus. Accumulation of B. gingivalis within the sulcus would be accompanied by the release of large numbers of vesicles which would have the opportunity to attach to the serum coated enamel or cementum surfaces. This in turn would create improved opportunities for colonization by other organisms. In the right circumstances, organisms such as S. sanguis or Actinomycoces viscosus (R.G. Ellen and D.A. Grove, J . Dent. Res. 67:268) would colonize and might eventually displace B. gingivalis thus contributing to the cyclical nature of periodontal infections. Lipipolvsaccharides Lipopolysaccharides are known to have a broad spectrum of biochemical and immunological activities. There are conflicting reports about the LPS of the BPB-s with regard to their chemical constituents and biological activities.lt has been reported that the composition of LPS from BPB are unique; they lack KDO, heptose and B-hydroxy myristic acid and their biological potency is considerably lower than that of enterobacterial LPS (102,103,113,117,145,167). In this study, the LPS from 145 several strains of BPB were analysed physically, chemically and immunologically. Chemical analysis of LPS of the BPB revealed that they have much in common but dissimilarities are present as well. Similar to previous reports (102,117,34,95) the majority of the sugars present were hexoses, followed by hexosamine and pentoses. No KDO nor heptoses were found by chemical analysis which is in agreement with most of the earlier reports in this area. However, in a recent publication (103) it was reported that KDO was found in minute quantities in some strains of Bacteroides. One way of explaining this discrepancy is that the KDO may be masked by phosphate groups and not liberated by acid hydrolysis, thus it is not recognized by the thiobarbituric acid assay of Karkhani et al (112). The study described here underscores the complexity and specificity of the reaction involved in the attachment of B. gingivalis to host tissue and to other bacteria. The epithelial cell binding adhesin have been purified and cloned and the bacterial coaggregating adhesin purified and characterized. The role of vesicles in the adherence of B. gingivalis in the sulcus was elucidated and the LPS from several BPB were characterized. Further studies should be directed to (1) production of isogenic mutants for adherence to host tissue so that the role of this adhesin in the pathogenesis of B. gingivalis can be evaluated, (2) cloning of the bacterial aggregating adhesin and the production of mutants so that the role of this adhesin can be evaluated in the disease process. 146 Bibliography 1. Abiko, Y., M. Hayakawa, S. Murai, and H. Takiguchi. 1985. Glycylprolyl dipeptidylaminopeptidase from Bacteroides gingivalis. J . Dent. Res. 64: 106-111. 2. Adegboles, R.A., and D.C. Old. 1984. Fimbrial and non-fimbrial haemagglutinin in Enterobacter aerogenes. J . Med. Microbiol. 19: 35-43. 3. Appelbaum, B., E. Golub, S.C. Holt, and N. Rosan. 1979. In vitro studies of dental plaque formation, adsorption of oral streptococci to hydroxyapatite. Infect. Immun. 25: 717-728. 4. Armstrong, W . C , and A.F. Hayward. 1968. Organic integuments of human enamel. Caries Res. 2: 294-321. 5. Ashwell, G. 1957. Colourimetric analysis of sugars. In. S.P. Colowick and N.O. Kaplan (eds.). Methods in enzymology. Vol. 3. p.95. Academic Press , New York. 6. Baltzer, I.M., P. Gemski, and C.R. Alving. 1984. Heterogeneity of lipid A. Rev. Inf. Dis. 6: 444-448. 7. Baseman, J.B., R.M. Cole, D.C. Krause, and D.K. Leith. 1982. Molecular basis for cytoadsorption of Mycoplasma pneumoniae. J . Bacteriol. 151: 1514-1522. 8. Beachey, E.H. 1981. Bacterial adherence: Adhesin-receptor interactions mediating the attachmnet of bacteria to mucosal surfaces. J . Inf. Dis. 143: 325-343. 9. Beachey, E.H., B.l. Einenstein, and I. Ofek. 1982. Bacterial adherence and infectious diseases. In Current Concepts. The Upjohn Co. Kalamazoo. Mich. p.1-52. 10. Bowden, G.H.W., D.C. Ellwood, and I.R. Hamilton. 1979. Microbial ecology of the oral cavity. Adv. Microb. Ecol. 3: 135-217. 11. Boyd, J . , and B.C. McBride. 1984. Fractionation of haemagglutinating and bacterial binding adhesins of Bacteroides gingivalis. Infect. Immun. 45: 403-409. 12. Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of proteins utiling the principle of protein-dye binding. Anal. Biochem. 72: 248-254. 13. Bradley, S .G. 1979. Cellular and molecular mechanisms of action of bacterial endotoxins. Ann. Rev. Microbiol. 33: 67-94. 14. Bramanti, T .E . , and S.C. Holt. 1987. Characterization of the major membrane proteins of selected Bacteroides gingivalis. J . Dent. Res. 66: 223. 15. Brande, A.I., E.J. Ziegler, and J.A. McCutchan. 1978. Antiserum treatment of Gram-negative bacteremia. Schireiz. Med. Wochenschur. 108: 1872-1876. 147 Gram-negative bacteremia. Schireiz. Med. Wochenschur. 108: 1872-1876.148 16. Braun, V., and K. Hantke. 1981. Bacterial cell surface receptors. In "Organization of Procaryotic Cell Membranes". B.K. Ghosh (ed.), Vol. 2, pp.1-73. C R C Press. Boca Raton, Florida. 17. Burnette, W.N. 1981. Western Blotting: Electrophoretic transfer of proteins from sodium-dodecyl sulphate polyacrylamide gels for radiographic detection with antibody and radio iodinated Protein A. Anal. Biochem. 112: 195-203. 18. Carlsson, J . , B.F. Herrmann, J.F. Hoffing, and G.K. Sundquist. 1984. Degradation of human proteinase inhibitors alpha-1 antitrypsin and alpha-2-macroglobulin by Bacteroides_gingivalis. Infect. Immun. 43: 644-648. 19. Celesk, R.A., and J . London. 1980. Attachment of oral Cytophaga species to hydroxyapatite containing surfaces. Infect. Immun. 29: 768-777. 20. Chan, R.C. Lian, J.W. Costerton and S.D. Aeries. 1982. The use of specific antibody to demonstrate the glycocalyx and spatial relationship of K 99, F 41-, enterotoxigenic strain of E. coli colonizing the ileum of colostrum deprived calves. C.J . Comp. Med. 47: 150-156. 21. Cheney, C P . , P.A. Schad, S.B. Formal, and E.C. Boedeker. 1980. Species specificity of in vitro Escherichia coli adherence to host intestinal cell membranes and its correlation with in vivo colonization and infectivity. Infect. Immun. 28: 101 9 -1027 . 22. Ciardi, J .E. , G.F.A. McCray, P.E.Kolenbrander, and A. Lau. 1987. Cell to cell interaction of Streptococcus sanguis and Propionibacterium acnes on saliva coated hydroxyapatite. Infect. Immun. 55: 1441-1446. 23. Cimasoni, G., M. Song, and B.C. McBride. 1987. Effect of crevicular fluid and lysosomal enzymes on the adherence of Streptococci and Bacteroides to hydroxyapatite. Infect. Immun. 55: 1484-1489. 24. Cisar, J.O., S.H. Curl, P.E. Kolenbrander, and A.E. Vatten. 1983. Specific absence of type 2 fimbriae on a coagglutination-defective mutant on Actinomyces viscosus T14V. Infect. Immun. 40: 759-765. 25. Cisar, J.O., A.L. Sandberg, and S.E. Mergenhagen. 1984. The function and distribution of different fimbriae on strains of Actinomyces viscosus and Actinomyces naeslundii. J . Dent. Res. 63: 393-396. 26. Clark W.B., L.L. Bammann, and R.J. Gibbons. 1978. Comparative estimates of bacterial affinities and adsorption sites on hydroxyapatite surfaces. Infect. Immun. 19: 846-853. 27. Clark, W.B., and R.J. Gibbons. 1977. Influence of salivary components and extracellular polysaccharide synthesis from sucrose on the attachment of 148 Streptococcus mutans 6715 to hydroxyapatite surfaces. Infect. Immun. 18: 514-523. 28. Clark, W.B., T.T. Wheeler, and J.O. Cisar. 1984. Specific inhibition of adsorption of Actinomyces viscosus T14V to saliva-treated hydroxyapatite by antibody against type 1 fimbriae. Infect. Immun. 43: 497-501. 29. Courant, P.R., and R.J. Gibbons. 1967. Biochemical and and immunological heterogeneity of Bacteroides melaninogenicus. Arch. Oral Biol. 12: 1605-1613. 30. Cowman, R.A., R.J. Fitzgerald, and S.J. Schaefer. 1986. Role of salivary factors in the nitrogen metabolism of plaque-forming oral streptococci. In. H.M. Stiles. W.J. Loesche and T .C . O'Brien (eds.). Microbial aspects of dental caries. Information Retrieval Inc. Washington D.C. and London, p. 465-475. 31. Coykendall, A.L., F.S. Kaezmarek, and J . Slot. 1980. Genetic heterogeneity in Bacteroides asaccharolyticus (Holdeman and Moore, 1970). Finegold and Barnes 1977. (Approved Lists. 1980) and proposal of Bacteroides gingivalis sp. nov. and Bacteroides macacae (Slots and Genco) comb. nov. Int. J . Syst. Bacterid.; 30: 5 5 0 - 5 6 4 . 32. Cryz, O., S .J .E . Furer, and R. Germanier. 1984. Protection against fatal Pseudomonas aeruginosa burn wound sepsis by immunization with lipopolysaccharide and high molecular weight polysaccharide. Infect. Immun. 43: 795-799. 33. Cryz, S.J . , T.L. Pitt, E.Furer, and R. Germanier. 1984. Role of lipopolysaccharide in virulence of Pseudomonas aeruginosa. Infect. Immun. 44: 5 0 8 - 5 1 3 . 34. Dahlen, G. , and I.M. Baltzer. 1983. Lipid A in anaerobic bacteria. Infect. Immun. 39: 466-468. 35. Darveau, R.P., and R.E.W. Hancock. 1983. Procedure for isolation of bacterial lipopolysaccharides from both smooth and rough Pseudomonas aeruginosa and Salmonellajyphimurium strains. J . Bacteriol. 155: 831-838. 36. Darwish, S., T. Hyppa, and S.S. Socransky. 1978. Studies of the predominant cultivable microbiota of early periodontitis. J . Periodont. Res. 13: 1-16. 37. Darzins, A., and A.M. Chakrabarty. 1984. Cloning of genes controlling alginate biosynthesis from a mucoid cystic fibrosis isolate of Pseudomonas aeruginosa. J . Bacteriol. 159: 9-18. 38. Day, D.F., and J.M. Ingram. 1975. In vitro studies of an alkaline phosphatase cell wall complex from Pseudomonas aeruginosa. Can. J . Microbiol. 21: 9-16. 39. Deich, R.A., and L.C. Hoyer. 1982. Generation and release of DNA-binding vesicles by Haemophilius influenza during induction and loss of competence. J . 149 Bacteriol. 152: 855-864. 40. Dickerson, P.D., M.A. Kubiniee, F. Yoshimura, and R.J. Genco. 1988. Molecular cloning and sequencing of the gene encoding the fimbrial subunit protein of Bacteroides gingivalis. J . Bacteriol. 170: 1658-1665. 41. Dowell, V.R. Jr., and G . L Lombard. 1981. Pathogenic members of the genus Bacteroides. In. M.P. Starr, H. Stolp, H.G. Truper, A. Balows and H.G. Schlegel (eds.). The prokaryotes. A handbook on habits, isolation and identification of bacteria. Springer-Varlag. Berlin. 2: p.1425-1449. 42. Dolye, R.J., W.E. Nesbitt, and K.G. Taylor. 1982. On the mechanism of adherence of Streptococcus sanguis to hydroxylapatite. FEMS Microbiol. Lett. 15: 1-5. 43. Dubois, M., K.A. Gilles, J.K. Hamilton, P.A. Rebers, and F. Smith. 1956. Colorimetric method for determination of sugars and related substances. Anal. Biochem. 28: 350-356. 44. Duerden, B.l. 1980. The isolation and identification of Bacteroides spp. from normal human gingival flora. J . Med. Microbiol. 13: 89-101. 45. Duguid, J.P., E.S. Anderson, and I. Campbell. 1966. Fimbriae and adhesive properties in Salmonellae. J . Path. Bact. 92: 107-119. 46. Duguid, J.P., W. Smith, G. Dempster, and P.J. Edmonds. 1955. Non-flagellar filamentous appendages ("fimbriae") and haemagglutinating activity in Bacterium coli. J . Pathol. Bacteriol. 70: 335-348. 47. Duguid, J.P. , S. Clegg and M.J. Wilson. 1979. The fimbrial and non-fimbrial haemagglutinin of Escherichia coli. J . Med. Microbiol. 12: 213-223. 48. Dutson, T.R., A.M. Pearson, J.F. Price, G.C. Spink, and P.J.V. Tarvant. 1971. Observations by electron microscopy on pig muscle innoculated and incubated with Pseudomonas tragi. Appl. Microbiol. 22: 1152-1158. 49. Ebersole, J.L., M.J. Steffer, and S.C. Holt. 1988. Lipopolysaccharide interactions with human gingival fibroblast. I.A.D.R. Mont. #1015: 50. Ebersole, J.L., M.A. Taubman, D.J. Smith, and D.E. Frey. 1986. Human responses to oral microorganisms. Patterns of systemic antibody levels to Bacteroides species. Infect. Immun. 51: 507-513. 51. Ellen, R.P., and I.B. Balcenzak-Roczkowski. 1973. Interbacterial aggregation of Actinomycetes naeslundii and dental plaque streptococci. J . Periodont. Res. 12: 11-20. 52. Ellen, R.P. 1985. Specificity of attachment as a tissue tropic influence on oral bacteria. In S .E . Mergenhagen and B. Rosan (eds.), Molecular basis of oral microbial adhesion. American Society for Microbiology, Washington, D.C. p.33-39. 150 53. Ennis, H.L., and M.I. Bloomstein. 1974. Antibiotic sensitive mutants of Escherichia coli possess altered outer membranes. Annals of the New York Academy of Sciences. 235: 593-600. 54. Eshdat, Y., F.J. Silverblatt, and N. Sharon. 1981. Dissociation and reassembly of Escherichia coli type 1 pili. J . Bacteriol. 148: 308-313. 55. Evans, D.G., D.J. Evans, J .J . Moulds, and D.Y. Graham. 1988. N-Acetyl-neuraminyllactose-binding fibrillar hemagglutinin of Campylobacter pylori. A putative colonization factor antigen. Infect. Immun. 56: 2896-2906. 56. Ewing, W.A., and W.J. Martin. 1974. Enterobacteriaceae. In. E.H. Lennette, E.H. Spaulding and J.P. Truant (eds.) Manual of Clinical Microbiology, 2nd ed. American Society for Microbiology, Washington, D.C. p.187-221. 57. Finegold, S.M., and .M. Barnes. 1977. Report of the ICSB Taxonomic Subcommittee on Gram-negative anaerobic rods. Proposal that the saccharolytic and asaccharolytic strains at present classified in the species Bacteroides melaninogenicus be classified on two species as Bacteroides melaninogenicus and Bacteroides asaccharolyticus. Int. J . Syst. Bacteriol. 27: 388-391. 58. Fisher, J .G. , J .J . Zambon, P. Chen, and R.J. Genco.1987 Bacteroides gingivalis serogroups and correlation with virulence. J . Dent. Res. 65: 816. 817. 59. Fisher, M.W. 1977. A polyvalent human gamma-globulin immune to Pseudomonas aeruginosa: passive protection of mice against lethal infection. J . Infect. Dis. 136 (Suppl.): 181-185. 60. Flesher, A.R., and R.A. Insel. 1978. Characterization of lipopolysaccharide of Haemophilus influenzae. J . Infect. Dis. 138: 719-729. 61. Frank, R.M., and G . Hoover. 1970. An ultrastructural study of human supragingival dental plaque formation. In. W.D. McHugh (ed.) Dental Plaque. S. Livingstone Ltd. London, p.85-105. 62. Freter, R. 1980. Prospects for preventing the association of harmful bacteria with host mucosal surfaces. In. E.H. Beachey (ed.). Bacterial adherence. Chapman and Hall, London, p.439-458. 63. Galanos, C , M.A. Freudenberg, S. Hase, F. Jay, and E. Ruschmann. 1977. Biological activities and immunological properties of lipid A. In. Schlessenger D. (ed.) Microbiology. Washington, D.C. American Society for Microbiology, p.267-276. 64. Galanos, C , and O. Luderitz. 1971. Electrodialysis of lipopolysaccharides and their conversion to uniform salt forms. Eur. J . Biochem. 54: 603-610. 65. Galanos, C , and O. Luderitz, and O. Westphal. 1971. Preparation and properties 151 of antisera against lipid A component of bacterial lipopolysaccharides. Eur. J . Biochem. 24: 116-122. 66. Gibbons, R.J., and J . van Houte. 1975. Bacterial adherence in oral micobial ecology. Ann. Rev. Micobiol. 29: 19-44. 67. Gibbons, R.J., and J . van Houte. 1980. Bacterial adherence and the formation of dental plaques. In. E.H. Beachey (ed.) Bacterial adherence. Chapman and Hall, LOndon. p .60-104. 68. Gibbons, R.J., and J . van Houte. 1980. Bacterial adherence and the formation of dental plaques. In. Bacterial adherence. E.H. Beachey (ed.) Receptors and recognition, Series B. volume 6, Chapman and Hall, London, p. 62-104. 69. Gibbons, R.J., and I. Etherden. 1982. Enzymatic modification of bacteria receptors on saliva-treated hydroxyapatite surfaces. Infect. Immun. 36: 52-58. 70. Gibbons, R.J., and J.B. McDonald. 1960. Hemin and vitamin k compounds are required factors for the cultivation of certain strains of Bacteroides melaninogenicus. J . Bacteriol. 80: 164-170. 71. Gibbons, R.J., and M. Nygaard. 1970. Interbacterial aggregation of plaque bacteria. Arch. Oral Biol. 15: 1397-1401. 72. Gibbons, R.J., and E.C. Moreno, and D.M. Spinell. 1976. Model delinating the effects of a salivary pellicle on the adsorption of Streptococcus mitis to hydroxyapatite. Infect. Immun. 14: 1109-1112. 73. Gibbons, R.J., and J . van Houte. 1973. On the formation of dental plaque. J . Period. 44: 347-360. 74. Gibbons, R.J., and J. van Houte. 1979. Selective bacterial adherence to oral epithelial surfaces and its role as an ecological determinant. Infect. Immun. 3: 567-573. 75. Goldhar, J . , R. Perry, J.R. Golechi, H. Hoschutzky, B. Jann.and K. Jann. 1986. Non-fimbrial mannose-resistant adhesins from uropathogenic Escherichia coli 083:K1:H4 and 014:K?:H11. Infect. Immun. 55: 1837-1842. 76. Goodrum, K.J., R.N. Moore, and L.J. Berry. 1978. In: Toxins: Animals, Plant and Microbiol. P. Rosenberg (ed), New York: Pergamon. p. 899-905. 77. Gorbach, S.L., B.H. Kean, S.G. Evans, D.J. Evans, and D. Bessudo. 1975. Travellers diarrhea and toxigenic Escherichia coli. N. Engl. J . Med. 292: 933-936. 78. Grenier, D., G. Choa, and B.C. McBride. 1988. Characterization of SDS-stable Bacteroides gingivalis proteases by polyacrylamide gel electrophoresis. Infect. Immun. 65. 125-132. 152 79. Grenier, D., and D. Mayrand. 1985. Cytotoxic effects of culture supernatants of oral bacteria and various organic acids on Vero cells. Can. J . Microbiol. 31: 302-304. 80. Grenier, D., and D. Mayrand. 1987. Functional characterization of extracellular vesicles produced by Bacteroides gingivalis. Infect. Immun. 55: 111-117. 801 Grenier, D., and B.C. McBride. 1987. Isolation of a membrane associated Bacteroides gingivalis glycyl-prolyl protease. Infect. Immun. 55: 3131-3136. 82. Grenier, D., and D. Mayrand. 1987. Selected characteristics of pathogenic and non-pathogenic strains of Bacteroides gingivalis. J . Clin. Microbiol. 25: 738-740. 83. Grigsby, W.R., and C.B. Sabiston. 1975. The periodontal disease process. J . Oral Path. 5: 175-188. 84. Grolean, D., and C.W. Forsberg. 1981. Cellulytic activity of the rumen bacterium Bacteroides succinogenes. Can. J . Microbiol. 27: 517-530. 85. Gustafsson, B.E., C.E. Quesnel, L S . Lanke, C. Lundquist, H. Grahnen, B.E. Bonow, and B. Krasse. 1954. The vipeholm dental caries study: the effect of different levels of carbohydrate intake on caries activity in 436 individuals observed for five years. Acta. Odontol. Scand. 11: 232-364. 86. Halhoul, N., and J.R. Colvin. 1980. The ultrastructure of bacteria plaque attached to the gingiva of man. Arch. Oral Biol. 20:115-118. 87. Hamada, S., and H.D. Slade. 1980. Biology, immunology and cariogenicity of Streptococcus mutans. Microbiol. Rev. 44: 331-384. 88. Hanazawa, S., K. Saito, Y. Ohmori, H. Nishihara, S. Fujiwara, and S. Kitano. 1984. Production of monoclonal antibody that recognize specific and cross-reactive antigens of Bacteroides gingivalis. Infect Immun. 46: 285-287. 89. Hancock, R.E.W., L M . Mutharia, L. Chan, R.P. Darveau, D.P. Spert, and G.B. Pier. 1983. Pseudomonas aeruginosa isolates from patients with cystic fibrosis: a class of serum sensitive nontypable strains deficient in lipopolysaccharide O-side chains. Infect Immun. 42: 170-177. 90. Handley, P.S., and L S . Tipler. 1986. An electron microscope survey of the surface structures and hydrophobicity of oral and non-oral species of the bacterial genus Bacteroides. Arch. Oral Biol. 31: 325-335. 91. Hellerquist, C .G. , B. Lindberg, and J . Lonngren. 1971. Fragmentation analysis of the Salmonella newportlipopolysaccharide. Acta. Chem. Scand. 25: 601-606. 92. Henry, R.J., A.B. Blakeney, P.J. Harris, and B.A. Stone. 1983. Detection of neutral and amino sugars from glycoproteins and polysaccharides as their alditol 153 acetates. J . Chrom. 256: 419-427. 93. Hillman, J.D., J . van Houte, and R.J. Gibbons. 1970. Sorption of bacteria to human enamel powder. Arch. Oral Biol. 15: 899-903. 94. Hoffman, H., and M.E. Frank. 1966. Microbial burden of mucosal squamous epithelial cells. Acta. Cytology. 10: 272-285. 95. Hofstad, T. 1968. Chemical characteristics of Bacteroides melaninogenicus endotoxin. Arch. Oral Biol. 13: 1149-155. 96. Hofstad, T. 1968. Serological properties of lipopolysaccharides from oral strains of Bacteroides melaninogenicus. J . Bacteriol. 97: 1078-1082. 97. Holdeman, L.V., and W.E.C. Moore. 1970. Outline of chemical methods in anaerobic bacteriology. 2nd ed. Virginia Polytechnic Institute and State University. Anaerobic Laboratory, Blacksburg. 98. Hunt, D.E., J.V. Jones, and V.R. Dowell, Jr. 1986. Selective medium for the isolation of Bacteroides gingivalis. J . Clin. Microbiol. 23: 441-445. 99. Ingram, J.M., K.J. Cheng, and J.W. Costerton. 1973. Alkaline phosphatase of Pseudomonas aeruginosa. The mechanism of the secretion and release of the enzyme from whole cells. Can. J . Microbiol. 19: 1407-1415. 100. Inoshita, E., A. Awano, T. Haniska, H. Tamagawa, S. Shizukuishi, and A. Tsunemitsu. Isolation and some properties of exohaemagglutinin from the culture medium of Bacteroides gingivalis 381. Infect Immun. 52: 421-427. 101. Jacobs, A.C. , B.A. Simons and F.K. deGraff. 1987. The role of lysine 132 and arginine 136 in the receptor binding domain of the K 99 fibrillar subunit. The Embo Journal. 6: 1805-1808. 102. Johne, B., and K. Bryn. Chemical composition and biological properties of a lipopolysaccharide from Bacteroides intermedius. Acta. Path. Microbial. Scand. Sect. 94: 262-271. 103. Johne, B., I. Olsen, and K. Bryan. 1988. Fatty acids and sugars in lipopolysaccharides from Bacteroides intermedius, B. gingivalis and B. loeschii. Oral Microbiol. Immun. 6. 112-120. 104. Johnson, A.P., D. Taylor-Robinson, and Z.A. McGee. 1977. Species specificity of attachment and damage to oviduct mucosa by Neisseria gonorrhoeae. Infect. Immun. 18: 833-839. 105. Johnson, K.J., P.A. Ward, S. Govalnick, and M.J. Osborn. 1977. Isolation from human serum of an inactivator of bacterial lipopolysaccharide. Am. J . Pathol. 88: 5 5 9 - 5 7 4 . 154 106. Jones, G.W., and J.M. Rutter. 1974. Contribution of the K 88 antigen of E. coli to enteropathogenicity protection against disease by neutralizing the adhesive properties of K 88 antigen. Am. J . Clin. Nutr. 27: 1441-1449. 107. Jones, S .J . 1972. A special relationship between spherical and filamentous microorganisms in mature human dental plaque. Arch. Oral Biol. 17: 613-616. 108. Kabat, E.A., and M.M. Mayer. Experimental Immunochemistry. Second edition. Charles C. Thomas, Springfield, Illinois, p.178. 109. Kagermeier, A., J . London, and P.E. Kolenbrander. 1984. Evidence on the participation of N-acetylated amino sugars in the coaggregation between Cytophaga species strain DR 2001 and Actinomyces israelii PK16. Infect. Immun. 44: 299-305. 110. Kamata, R., K. Matsumoto, R. Okamura, T. Yamamoto, and H. Maeda. 1985. Serratial 56 k protease as a major pathogenic factor in Serratial keratitis. Ophthalmology. 92: 1452-1459. 111. Kappernd, G., E. Namork and H.J. Skanpeid. 1985. Temperature-inducible surface fibrillae associated with the virulence plasmid of Yersinia enterocolitica and Yersinia pseudotuberculosis. Infect. Immun. 47: 561-566. 112. Karkhanis, Y.D., J.Y. Zelter, J .J . Jackson, and D.J. Carlo. 1978. A new and improved microassay to determine 2-keto-3 deoxyoctonate in lipopolysaccharide of Gram-negative bacteria. Anal. Biochem. 85: 595-601. 113. Kasper, D.L., A.A. Lindberg, A. Weinstaub, A.B. Onderdonk and J . Lonngren. 1984. Capsular polysaccharide and lipopolysaccharides from two strains of Bacteroides fragilis. Rev. Infect. Dis. 6: 529-539. 114. Kastelein, P., P. van Steenbergen, J.M. Bras, and J . De Graff. 1981. An experimental induced phelegmonous abscess by a strain of Bacteroides gingivalis in guinea pigs and mice. Antoine Van Loewenhoch. 47: 1-9. 115. Knox, K.W., M. Vesk, and E. Work. 1966. Relationship between excreted lipopolysaccharide complexes and surface structures of a lysine-limited culture of Escherichia coli. J . Bacteriol. 92: 1206-1217. 116. Knutton, S. P.H. Williams, D.R. Lloyd, D.C.A. Candy, and A.S. McNeish. 1984. Ultrastructural study of adherence to and penetration of cultured cells by two invasive Escherichia coli strains isolated from infants with enteritis. Infect. Immun. 44: 5 9 9 - 6 0 8 . 117. Koga, T., T. Nishihara, T. Fujiwara, T. Nisizawa, N. Okahashi, T. Noguchi, and S. Hamanda. 1985. Biochemical and immunobiological properties of lipopolysaccharide (LPS) from Bacteroides gingivalis and comparison with LPS from Escherichia coli. Infect. Immun. 47: 638-647. 1 5 5 118. Kolenbrander, P.E., R.N. Anderson, and L V . Holdermann. 1985. Coaggregation of oral Bacteroides species with other bacteria: central role in coaggregation bridges and competitions. Infect Immun. 48: 741-746. 119. Kolenbrander, P.E., and R.N. Anderson. 1985. Use of coaggregation-defective mutants to study the relationship of cell to cell interactions and oral micrbial ecology. In S .E . Mergenhagen and B. Rosan (eds.). Molecular basis of oral microbial adhesion. American Society for Microbiology, Washington, D.C. p.164-171. 120. Kreutzer, D.R., C S . Buller, and D.C. Robertson. 1979. Chemical characterization and biological properties of lipopolysaccharides isolated from smooth and rough strains of Brucella abortus. Infect. Immun. 23: 811-818. 121. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 27: 680-685. 122. Lai, C.H., M.A. Listgarten, and B.F. Hammond. 1981. Comparative ultrastructure of leukotoxic and non-leukotoxic strains of Actinobacillus actinomycetemcomitans. J . Periodont. Res. 16: 379-389. 123. Lantz, M., R.W. Rowland, L.M. Switalski, and M. Hook. 1986. Interaction of Bacteroides gingivalis with fibrinogen. Infect. Immun. 55: 201-205. 124. Larjava, M., V-K. Uitto, E. Eerola, and M. Haapasalo. 1987. Inhibition of gingival fibroblast growth by Bacteroides gingivalis. Infect. Immun. 54: 201-208. 125. Larjava, H., V-J. Uitto, M. Haapasalo, J . Heino, and M. Vuento. 1987. Fibronectin fragmentation induced by dental plaque and Bacteroides gingivalis . Scand. J . Dent. Res. 95: 308-314. 126. Lev, M., and A.F. Milford. 1971. Vitamin K stimulation of sphingolipid synthesis. Biochem. Biophys. Res. Commun. 45: 358-362. 127. Liljemark, W.F., and S.V. Schaner. 1977. Competitive binding among oral streptococci to hydroxyapatite. J . Dent. Res. 56: 157-165. 128. Liljemark, W.F., and S.V. Schaner. 1975. Studies on the bacterial components which bind Streptococcus sanguis and Streptococcus mutans to hydroxyapatite. Arch. Oral. Biol. 20: 609-615. 129. Lindberg, F., B. Lund, L. Johansson, and S. Novmark. 1987. Localization of the receptor binding protein adhesin at the tip of the bacterial pilus. Nature. 328: 8 4 - 8 7 . 130. Lippincott, B.B., and J.A. Lippincott. 1969. Bacterial attachment to a specific wound site as an essential stage in tumour initiation by Agrobacterium tumifaciens. J . Bacteriol. 97: 620-628. 156 131. Listgarten, M.A., and C H . Lai. 1979. Comparative ultrastructure of Bacteroides melaninogenicus subspecies. J . Periodontal Res. 14: 332-340. 132. Listgarten, M.A., H. Mayo, and M. Amsterdam. 1973. Ultrastructure of the attachment device between coccal and filamentous microorganisms in corn-cob formation of dental plaque. Arch. Oral Biol. 18: 651-656. 133. Loesche, W.J., K.V. Paunio, M.P. Woolfolk, and R.N. Hochett. 1978. Collagenolytic activity of dental plaque associated with periodontal pathology. Infect Immun. 9: 329-336. 134. Loesche, W.J., S.A. Syed, E.C. Morrison, B. Langhon, and N.S. Grossman. 1981. Treatment of periodontal infections due to anaerobic bacteria with short term treatment with metronidazole. J . Clin. Periodontal. 8:29-44. 135. Luderitz, O., C. Galanos, V. Lehmann, H. Mayer, E.T. Rietschel, and J . Weckesser. 1978. Chemical structure and biological activities of lipid A's from various bacterial families. Natrwissenschafer. 65: 575-585. 136. Luderitz, O., O. Westphal, A.M. Staub, and H. Nikaido. 1971. Isolation and chemical and immunological characterization of bacterial lipopolysaccharides. In "Microbial Toxins" G. Weinbaum, S. Kadis and S.J . Ajl, (eds.) Academic Press, New York. Vol.4: p.145-233. 137. Luderitz, O., M.A. Freudenberg, C. Galanos, V. Lehmann, E.T. Rietschel, and D.H. Shaw. 1982. Lipopolysaccharides of Gram-negative bacteria. Current Top. Memb. Transp. 17: 79-151. 138. Luderitz, O., K. Tanamoto, C. Galanos, C P . McKenzie, H. Baude, U. Zahringer, E. Rietschel, S. Kusomoto, and T. Shiba. 1984. Lipopolysaccharides: Strucural principles and biological activities. Rev. Infect. Dis. 6: 428-431. 139. Mac Donald, J.B., S.S. Socransky, and R.J. Gibbons. 1963. Aspects of the pathogenesis of mixed anaerobic infections of mucous membranes. J . Dent. Res. 42: 5 2 9 - 5 4 4 . 140. Mac Gaughey, C , and E.C. Stowell. 1971. Adsorption of salivary proteins by hydroxyapatite: Relations between the effect of calcium ions, hydrogen ions, temperature, and exposure time. J . Dent. Res. 50: 542-550. 141. Mackler, B.F., T . C Waldrop, P. Schur, P.B. Robertson, and B.M. Levy. 1978. IgG subclasses in human periodontal disease. 1. Distribution and incidence of IgG subclasses bearing lymphocytes and plasma cells. J . Periodont. Res. 13: 109-119. 142. Macy, J.M. 1981. Nonpathogenic members of the genus Bacteroides. In. M.P. Starr, H. H. Stolp, H.G. Truper, A. Balows, and H.G. Scvhlegel (eds.). The prokaryotes. A handbook on the habitats, isolation and identification of bacteria. Springer-Vorlag. Berlin. 2: p. 1450-1463. 157 143. Maniatis, T., E.F. Fritsch, J . Sambrook. 1982. J . Molecular Cloning, a laboratory manual. Cold Spring Harbour Laboratory, Cold Spring Harbour, N.Y. 144. Mansheim, B.J., C.A. Solstad, and D.L. Kasper. 1978. Identification of a subspecies specific capsular antigen from Bacteroides melaninogenicus subspecies asaccharolyticus by immunofluorescence and electron microscopy. J . Infect Dis. 138: 736-741. 145. Manshiem, B.J., A.B. Onderdonk, and D.L. Kasper. 1978. Immunochemical and biological studies of the lipopolysaccharide of Bacteroides melaninogenicus subspecies asaccharolyticus. J . Immunol. 120: 72-78. 146. Manshiem, B.J., and S.E. Coleman. 1980. Immunochemical differences between oral and nonoral strains of Bacteroides asaccharolyticus. Infect, and Immun. 27: 5 8 9 - 5 9 6 . 147. Marshal, K.C., R. Stout, and R. Mitchell. 1971. Mechanism of the initial events in the sorption of marine bacteria to surfaces. J . Gen. Microbiol. 68: 337-348. 148. Maurer, L., and P.E. Orndorff. 1987. Identification and characterization of genes determining receptor binding and pilus length of Escherichia coli type 1 pili. J. Bacteriol. 169: 640-645. 149. Mayrand, D., B.C. McBride, T. Edwards, and S. Jensen. 1980. Characterization of Bacteroides asaccharolyticus and Bacteroides melaninogenicus oral isolates. Can. J . Microbiol. 26: 1178-1183. 150. Mayrand, D., and D. Grenier. 1985. Detection of collagenase activity in oral bacteria. Can. J . Microbiol. 31: 134-138. 151. Mayrand, D., and B.C. McBride. 1980. Ecological relationships of bacteria involved in a simple, mixed anaerobic infection. Infect. Immun. 27: 44-50. 152. McBride, B.C., and M.T. Gisslow. 1977. Role of sialic acid in saliva-induced aggregation of Streptococcus sanguis. Infect. Immun. 18: 35-40. 153. McBride, B.C., and J.S. van der Hoeven. 1981. Role of interbacterial adherence in colonization of the oral cavities of gnotobiotic rats infected with Streptococcus mutans and Veillonella alcalescens. Infect. Immun. 33: 467-472. 154. McGauhey, C , and E. Stowell. 1967. The adsorption of human salivary proteins and porcine submaxillary mucin to hydroxyapatite. Arch. Oral Biol. 12: 815-828. 155. Mcintire, F.C. 1985. Specific surface components and microbial coaggregation. In. S.E. Mergenhagen and B. Rosan (ed.), Molecular basis of oral microbial adhesion. American Society for Microbiology, Washington, D.C. p.152-158. 156. McKee, A.S., A.S. McDermund, A. Baskerville, A.B. Dowsett, D.C. Ellwood, and 158 P.D. Marsh. 1986. Effect of hemin on the physiology and virulence of Bacteroides gingivalis. Infect. Immun. 52: 349-355. 157. McMichael, J .C . , and J.T. Ou. 1979. Structure of common pili from Escherichia coli. J . Bacteriol. 138: 969-973. 158. Millar, J.S. , E.G. Goldstein, M.J. Levine, and E. Hausmann. 1986. Modulation of bone metabolism by two chemically distinct lipopolysaccharide fractions from Bacteroides gingivalis. Infect. Immun. 51: 302-306. 159. Mooi, F.R., and K.F. de Graaf. 1979. Isolation and characterization of K 88 antigens. FEMS Microbiol. Lett. 5: 17-23. 160. Moon, H.W., E.M. Kohler, R.A. Schneider, and S.C. Whipp. 1980. Prevalence of pilus antigen enterotoxin types and enterotoxigenic Escherichia coli from neonatal pigs. Infect. Immun. 27: 222-230. 161. Moore, W.E.C. , R.R. Ranney, and L.V. Holdeman. 1982. Subgingival microflora in periodontal disease: cultural studies. In R.J. Genco and S .E . Mergenhagen (ed.), Host-parasite interactions in periodontal diseases. American Society for Microbiology, Washington, D.C. p.13-26. 162. Morris, E.J., and B.C. McBride. 1984. Adherence of Streptococcus sanguis to salivary coated hydroxyapatite, evidence for two binding sites. Infect. Immun. 43: 6 5 6 - 6 6 3 . 163. Mouton, C , D. Bouchard, M. Deslauriers, and L. Lamonde. 1989. Immunochemical identification and preliminary characterization of a non-fimbrial haemagglutinin adhesin of Bacteroides gingivalis. Infect. Immun. 57: 566-573. 164. Mouton, C , P.C. Hammond, J . Slots, and R.J. Genco. 1981. Serum antibodies to oral Bacteroides asaccharolyticus (Bacteroides gingivalis). Relationship to age and periodontal disease. Infect. Immun. 31: 182-192. 165. Murray, P.A., D.G. Kern, and J.D. Winkler. 1988. Identification of a galactose-binding lectin on Fusobacterium nucleatum FN-2. Infect. Immun. 56: 1314-1319. 166. Mutharia, L.M., G. Crockford, W.C. Bogard, and R.E.W. Hancock. 1984. Monoclonal antibodies specific for Escherichia coli J.5 lipopolysaccharide: Cross reaction with other Gram-negative bacterial species. Infect. Immun. 45: 631-636. 167. Nair, B.C., W.R. Mayberry, R. Dziak, P.B. Chen, M.J. Levine, and E. Hausmann. 1983. Biological effects of a purified lipopolysaccharide from Bacteroides gingivalis. J . Periodontal Res. 18: 40-49. 168. Naito, Y., and R.J. Gibbons. 1988. Attachment of Bacteroides gingivalis to collagenous substrates. J . Dent. Res. 67: 1075-1080. 169. Naito, Y., K. Okuda, T. Kato, and I. Takazoe. 1985. Monoclonal antibodies against 159 surface antigens of Bacteroides gingivalis. infect. Immun. 50: 231-235. 170. Neiders, M.E., P.B.Chen, H. Suido, H.S. Reynolds, J.J.Zambon, M. Schlossman.and R.J. Genco. 1988. Herterogeneity of virulence among strains of Bacteroides gingivalis.J. Periodont. Res. 24:192-198. 171. Niedermeier, W. 1971. Gas chromatography of neutral and amino sugars in glycoproteins. Anal, biochem. 40: 465-475. 172. Normark, S., D. Lark, R. Hull, M. Norgen, M. Bagn, P. O'Hanley, G. Schoolnik, and S. Falkow. 1983. Genetics of digalactose-binding adhesin from a uropathogenic Escherichia coli. Infect. Immun. 41: 942-949. 173. Okuda, K., and I. Takazoe. 1973. Antiphagocytic effects of the capsular structure of a pathogenic strain of Bacteroides melaninogenicus. Bull. Tokyo Dent. Coll. 14: 99-104. 174. Okuda, K., J . Slots, and R.J. Genco. 1981. Bacteroides gingivalis, Bacteroides asaccharolyticus and Bacteroides melaninogenicus subspecies: cell surface morphology and adherence to erythrocytes and human buccal epithelial cells. Curr. Microbiol. 6: 7-12. 175. Okuda, K., and I. Takazoe. 1974. Haemagglutinating activity of Bacteroides melaninogenicus. Arch. Oral Biol. 19: 415-416. 176. Okuda, K., A. Yamamoto, Y. Naito, I. Takazoe, J . Slots, and R.J. Genco. 1986. Purification and properties of haemagglutinin from culture supernatant of Bacteroides gingivalis. Infect. Immun. 54: 659-665. 177. Oliver, W.W., and W.B. Wharry.1921. Notes on some bacterial parasites of the human mucous membranes. J . Infect. Dis. 28: 341-345. 178. Ono, M., K. Okuda, and I. Takazoe. 1987. Purification and characterization of a thiol-protease from Bacteroides gingivalis strain 381. Oral Microbiol. Immunol. 2: 77-81. 179. Ostavik, D., F.W. Kraus, and L.C. Henshaw. 1974.In vivo attachment of streptococci to the tooth surface. Infect. Immun. 9: 794-800. 180. Ott, M., T. Schmoll, W. Goebel, I. Van Die, and J . Hacker. 1987. Comparison of the genetic determinant coding for the S-fimbrial adhesin (sfa) of Escherichia coli to other chromosomally encoded fimbrial determinants. Infect. Immun. 55: 1 940-1 943 . 181. Ottow, J .C .G . 1975. Ecology, physiology and genetics of fimbriae and pili. Ann. Rev. Microbiol. 29: 79. 182. Papahadjopoulos, D., and N. Miller. 1967. Phospholipid model membranes. 1. 160 Structural characteristics of hydrated liquid crystals. Biochemica. et. Biophysica. Acta. 135: 624-638. 183. Parachych, W., P.A. Sastry, L.S. Frost, M. Carpenter, G.D. Armstrong, and H.T. Watts. 1979. Biochemical studies on pili isolated from Pseudomonas aeruginosa strain PAO. Can. J . Microbiol. 25: 1175-1179. 184. Parent, R., C. Mouton, L. Lamonde, and D. Bouchard. 1986. Human and animal serotypes of Bacteroides gingivalis defined by cross-immunoelectrophoresis. Infect. Immun. 51: 908-918. 185. Patters, M.R., and K.S. Korman. 1982. Serum antibodies to Bacteroides species in human periodontitis. J . Periodont. Res. 17: 474-477. 186. Pearce, W.A., and T.M. Buchanan. 1980. Structure and cell membrane-binding properties of bacterial fimbriae In. Bacterial Adherence. E.H. Beachey (ed.). Chapman and Hall, London, p. 289. 187. Pekovic, D.D., and E.D. Filler/. 1984. Identification of bacteria in immunopathological mechanisms of human periodontal disease. J . Periodont. Res. 19: 329-351. 188. Pier, G.B., H.F. Sidbury, and J .C. Sadoff. 1978. Protective immunity reduced in mice by immunization with high molecular weight polysaccharide from Pseudomonas aeruginosa. Jnfect. Immun. 22: 919-925. 189. Poirier, P.T., and S.C. Holt. 1983. Acid and alkaline phosphatases of Capnocytophaga species-1. Production and cytological localization of the enzymes. Can. J . Microbiol. 29: 1350-1360. 190. Prehm, P., B. Jann, and K. Jann. 1976. The 09 antigen of Escherichia coli. Structure of the polysaccharide chains. Eur. J . Biochem. 67: 53-56. 191. Pulverer, G . H.L. Ko, Z. Wegrznowicz, and J . Jeljasze-Wiez. 1977. Clotting and fibrinolytic activities of Bacteroides melaninogenicus. Zbl. Bakt. Hyg. 239: 5 1 0 - 5 1 3 . 192. Reed, M.J., J . Slots, C. Mouton, and R.J. Genco. 1980. Antigenic studies of oral and non oral Black Pigmented Bacteroides strains. Infect. Immun. 29: 564-574. 193. Rietschel, E.T., C. Galanos, O. Luderitz, and O. Westphal. 1982. The chemistry and biology of lipopolysaccharide and their lipid A component. In. Webb D.R. (ed.) Immunopharmacology and regulation of leucocyte function. New York. Vol. 2: p. 1 8 3 - 2 2 9 . 194. Rizza, V., P.R. Sinclair, D.C. White and P.R. Courant. 1968. Electron transport system the of protoheme requiring anaerobe Bacteroides melaninogenicus. J . Bacteriol. 96: 665-671. 161 195. Roantree, R.J. 1971. The relationship of lipopolysaccharide structure to bacterial virulence. In. "Microbial Toxins" S. Kadis, G. Weinbaum and S.J . Ajl ( eds.), Vol. 5, pp. 1-32. Academic Press, New York. 196. Robbins, P.W., and T. Uchida. 1962. Studies on the classical basis of the phage conversion of O-antigens in the E- group Salmonella. Biochemistry. 1: 323-335. 197. Robertson, J.N., P. Vincent, and M.E. Ward. 1977. The preparation and properties of gonococcal pili. J . Gen. Microbiol. 102: 169-174. 198. Robertson, P.B., M. Lantz, P.T. Marucha, K.S. Kornman, C . L Trummel, and S.C . Holt. 1982. Collagen activity associated with Bacteroides species and Actinobacillus actinomycetencomitans. J . Periodontal Res. 17: 275-283. 199. Roe, J.H. 1955. The determination of sugar in blood and spinal fluid with the anthrone reagent. J . Biol. Chem. 212: 335-343. 200. Roerterink, C.H., T.J.M. van Steenbergen, W.F.B. deJong, and J. de Graff. 1984. Histopathological effects in the palate of the rat induced with different black-pigmented Bacteroides strains. J . Periodontal Res. 19: 292-301. 201. Rolla, G. 1976. Inhibition of absorption-general considerations In: Microbial aspects of dental caries. Voh 1. p.309-324. 202. Rosan, B. 1976. Relationship of the cell wall composition of group H streptococci and Streptococcus sanguis to their serological properties. Infect Immun. 13: 1144-1153. 203. Roy, T.E., and C D . Kelley. 1939. Bacteroides. Castellani and Chalmers. In. D.H. Bergey, R.S. Breed, E.G.D. Murray and A.P. Hitchans (eds.). Bergey's manual of determinative bacteriology. Fifth edition. The Williams and Williams Co. Baltimore. MD. 204. Schuster, M.L., and D.P. Coyne. 1974. Survival mechanisms of phytopathogenic bacteria. Ann. Rev. Phytopath. 12: 199-217. 205. Schwabacher, H., D.R. Lucas, and C. Rimington. 1947. Bacterium melaninogenicum- a misnomer. J . Gen. Microbiol. 1: 109-120. 206. Schwarz, S., R.P. Ellen, and D.A.Grove. 1987. Bacteroides gingivalis-Actinomycetes viscosus cohesive interactions as measured by a quantitative assay. Infect. Immun. 55: 2391-2397. 207. Sellwood, R., R.A. Gibbons, G.W. Jones, and J.M. Rutter. 1975. Adhesion of enteropathogenic Escherichia coli to pig intestinal brush borders: the existence of two pig phenotypes. J . Med. Microbiol. 8: 405-411. 208. Seymour, G.J. , and J.S. Greenspan. 1979. The phenotypic characterization of 162 lymphocyte sub-populations in established human periodontal disease. J . Periodont. Res. 14: 39-46. 209. Shah, H.N., R.A.D. Williams, G.D. Bowden, and M.J. Hardie. 1976. Comparison of the biochemical properties of Bacteroides melaninogenicus from human dental plaque and other sites. J . Appl. Bacteriol. 41: 473-492. 210. Shah, H.N., and M.D. Collins. 1980. Fatty acid and isoprenoid quinone composition in the classification of Bacteroides melaninogenicus and related taxa. J . Appl. Bacteriol. 48: 75-87. 211. Shah, H.N., and M.D. Collins. 1988. Proposal for reclassification of Bacteroides asaccharolyticus, Bacteroides gingivalis and Bacteroides endodontalis in a new genus Porphyromonas. Int. J . Syst. Bacteriol. 38: 128-131. 212. Shah, H.N., R. Bonnet, B. Matteen, and R.A. Williams. 1979. The porphyrin pigmentation of subspecies of Bacteroides melaninogenicus. Biochem. J . 180: 45-50. 213. Shibaer, U.N., T.N. Druzhinina, A.N. Popova, S.S. Rozhnova, and V.A. Kilesso. 1979. Mechanisms of O-specific polysaccharide biosynthesis in Salmonella serogrougs C2 and C3. Eur. J . Biochem. 101: 309-316. 214. Slots, J . , and R.J. Gibbons. 1978. Attachment of Bacteroides melaninogenicus subsp. asaccharolyticus to oral surfaces and its possible role in colonization of the mouth and of periodontal pockets. Infect. Immun. 19: 254-264. 215. Slots, J . , and E. Hausmann. 1979. Longitudinal study of experimentally induced periodontal disease in Macaca arctoides: relationship between microflora and alveolar bone loss. Infect. Immun. 23: 260-269. 216. Slots, J . , D. Moenbo, J . Langeback, and A. Frandsen. 1978. Microbiotic of gingivitis in man. Scand. J . Dent. Res. 86: 174-181. 217. Slots, J . 1977. Microflora in the healthy gingival sulcus in man. Scand. J . Dent. Res. 85: 247-254. 218. Slots, J . 1977. The predominant cultivable microflora of advanced periodontitis. Scand. J . Dent. Res. 85: 114-121. 219. Smalley, J.W., and A.J . Birss. 1987. Trypsin-like activity of the extracellular membrane vesicles of Bacteroides gingivalis W 50. J . Gen. Microbiol. 133: 2883-2894. 220. Smit, J . , M. Hermodson, and N. Agabian. Caulobacter crescentus pilin. J . Biol. Chem. 256: 3092-3097. 221. Socransky, S.S., J.L. Dzink, and C M . Smith. 1985. Chemically defined medium for oral microorganisms. J . Clin. Microbiol. 22: 303-305. 163 222. Socransky, S.S. 1970. Relationship of bacteria to the etiology of periodontal disease. J . Dent. Res. 49: 203-222. 223. Socransky, S.S., and R.J. Gibbons. 1965. Required role of Bacteroides melaninogenicus in mixed anaerobic infections. J . Infect. Dis. 115: 247-253. 224. Socransky, S.S., and R.J. Gibbons. 1965. Required role of Bacteroides melaninogenicus in mixed anaerobic infection. J . Infect. Dis. 115: 247-253. 225. Sonju, Y., T.B. Christensen, L. Kornstad, and G. Rolla. 1974. Electron microscopy, carbohydrate analyses and biological activities of the proteins adsorbed in two hours to tooth surfaces in vivo. Caries. Res. 8: 113-122. 226. Southern, E.M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J . Mol. Biol. 98: 503-517. 227. Starkey, P.M., and A .J . Barrett. 1977. Alpha-2-macroglobulin, a physical regulator of protinase activity. In A.J . Barrett (ed.) Proteinase in mammalian cells and tissues. North-Holland Publishing Co. Amsterdam, p.663-696. 228. Stollerman, G.H. 1975. Rheumatic fever and streptococcal infections. Grune and Stratton, New York, p.47-62. 229. Sundquist, G.K., M.I. Eckerbom, A.P. Larsson, and U.T. Sjogren. 1979. Capacity of anaerobic bacteria from necrotic dental pulps to induce purulent infections. Infect. Immun. 25: 685-693. 230. Sundquist, G.K., J . Carlsson, B. Hermann, and A. Tarnvik. 1985. Degradation of human immunoglogulins G and M and complement factors C3 and C5 by black-pigmented Bacteroides. J . Med. Microbiol. 19: 85-94. 231. Suzuki, Y., E. Yoshimura, K. Takahesti, H. Tarn, and T. Suzuki. 1988. Detection of fimbriae and fimbrial antigens on the oral anaerobe Bacteroides gingivalis by negative staining and serological methods. J . Gen. Micobiol. 134: 2713-2720. 232. Takazoe, I., M. Tanaka, and T. Homma. 1971. A pathogenic strain of Bacteroides melaninogenicus. Arch. Oral Biol. 16: 817-822. 233. Tanner, A.C.R., C. Haffer, G.T. Bratthall, R.A. Visconti, and S.S. Socransky. 1979. A study of the bacteria associated with advancing periodontitis in man. J. Clin. Periodont. 6: 278-307. 234. Thompson, S.S., Y.M. Naidu, and J .J . Pestka. 1985. Ultrastructure localization of an extracellular protease in Pseudomonas tragi by using the peroxidase-antiperoxidase reaction. Appl. Envirron. Microbiol. 50: 1038-11042. 235. Tonzetich, J . , and B.C. McBride. 1981. Characterization of volatile sulphur 164 production by pathogenic and non-pathogenic strains of oral Bacteroides. Arch. Oral Biol. 26: 963-969. 236. Tsai, C M . , and C .E . Frasch. 1982. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119: 115-119. 237. Uitto, V-J . , H.Larjava, T.Heino, and T. Sorsa. 1989. A protease of Bacteroides gingivalis degrades cell surface and matrix of cultured gingival fibroblast and induces secretion of collagenase and plasminogen activators. Infect. Immun. 57: 213-218. 238. van Steenbergen, T.J.M., F. Namevar, and J . de Graff. 1985. Chemiluminescence of human leukocytes by black-pigmented Bacteroides strains from dental plaque and other sites. J . Periodont. Res. 20: 58-71. 239. van Steenbergen, T.J.M., M.D. den Ouden, J.J.A. Touw, and J. de Graff. 1982. Cytotoxic activity of Bacteroides gingivalis and Bacteroides asaccharolyticus. J . Med. Microbiol. 5: 253-258. 240. van Steenbergen, T.J.M., J .J . De Soet, and J . de Graff. 1979. DNA base composition of various strains of Bacteroides melaninogenicus. Ferns. Microbiol. Lett. 5: 127-130. 241. Vieira, J . , and J . Messing. 1980. The pUC plasmids, and M13 mp 7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene. 19: 259-268. 242. Weerkamp, A.H., and B.C. McBride. 1980. Characterization of the adherence properties of Streptococcus salivarius. Infect. Immun. 29: 459-468. 243. Weerkamp, A.H., P.S. Handley, A. Barrs, and J.W. Slot. 1986. Negative staining and immunoelectron microscopy of adhesin-deficient mutants of Streptococcus salivarius reveal that the adhesin protein antigens are separate classes of cell surface fibril. J . Bacteriol. 165: 746-755. 244. Weir, D.M. 1980. Surface carbohydrates and lectins in cellular recognition. Immunol. Today. 1: 45-51. 245. Weiss, E.I., J . London, P.E. Kolenbrander, A.S. Kagermeier, and R.N. Anderson. 1987. Characterization of lectin like surface components on Capnocyophaga ochracea A T C C 3359 that mediate coaggregation with gram-positive oral bacteria. Infect. Immun. 55: 1198-1202. 246. Weiss, E.I., P.E. Kolenbrander, J . London, A.R. Hand and R.N. Anderson. 1987. Fimbrin-associated proteins of Bacteroides loeschii PK125 mediate intergeneric coaggregation. J . Bacteriol. 169: 4215-4222. 247. Weiss, C. 1937. Observation on Bacterium melaninogenicus: demonstration of fibrinolysis pathogenicity and serological types. Proc. Soc. Exp. Biol. Med. 37: 473-476. 165 248. Westenfelder, M., C. Galanos, P.O. Madsen, and W. Marget. 1977. Pathologic activities of lipid A: experimental studies in relation to chronic pyelonephritis. In. Schlessinger, D., (ed.) Microbiology. Washington, D . C , American Society for Microbiology, p. 277-279. 249. Whatley, M.H., J.S. Bodwin, B.B. Lippincott and J .J . Lippincott. 1976. Role of Agrobacterium cell envelope lipopolysaccharide in infection site attachment. Infect. Immun. 13: 1080-1083. 250. White, D., and D. Mayrand. 1981. Association of oral Bacteroides with gingivitis and adult periodontitis. J . Periodont. Res. 16: 259-265. 251. Williams, G.D., and S . C Holt. 1985. Characteristics of the outer-membranes of selected oral Bacteroides species. Can. J . Microbiol. 31: 238-250. 252. Williams, P.H., S. Knutton, M.G.M. Brown, D.C.A. Candy, and A. McNeish. 1984. Characterization of non-fimbrial mannose-resistant protein haemagglutinin of two Escherichia coli strains isolated from infants with enteritis. Infect. Immun. 44: 5 9 2 - 5 9 8 . 253. 'Winkler, J.R., S.R. John, R.H. Kramer, C.I. Hoover, and P.A. Murray. 1987. Attachment of oral bacteria to basement membrane-like matrix and to purified matrix proteins. Infect. Immun. 55: 2721-2726. 254. Wollenweber, H.W., K. Brady, O. Luderitz, and E.T. Rietschel. 1982. The chemical structures of lipid A, demonstration of amide-linked-3-acyloxyacyl-residues in Salmonella minnesota Re. lipopolysacchgaride. E. J . Biochem. 124: 1 9 1 - 1 9 8 . 255. Woo, D.L., S . C . Holt, and E.R. Leadbetter. 1979. Ultrastructure of Bacteroides species: Bacteroides asaccharolyticus, Bacteroides fragilis, Bacteroides melaninogenicus subspecies melaninogenicus, and Bacteroides. melaninogenicus subspecies intermedius. J . Infect. Dis. 319: 534-546. 256. Wray, K.S., S.I. Hull, R.G. Cook, J . Barrish, and R.A. Hull. 1986. Identification and characterization of a uroepithelial cell adhesin from a uropathogenic isolate of Proteus mirabilis. Infect. Immun. 54: 43-49. 166 257. Wright, B.G., and P.A. Rebers. 1972. Procedure for determining heptose and hexoses in lipopolysaccharides-modification of the cysteine-sulphuric acid method. Anal. Biochem. 49: 307-319. 258. Yamazaki, Y., S. Ebisu, and H. Okada. 1981. Eikenella corrodens adherence to human buccal epithelial cell. Infect. Immun. 41: 21-27.244. 259. Yanish-Perron, C , J . Viera, and J . Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13 mp 18 and pUC 19 vectors. Gene. 33: 103-119. 260. Yoshimura, F., T. Takasawa, M. Yoneyama, T. Tamaguchi, H. Shiokawa, and T. Suzuki. 1985. Fimbriae from the oral anaerobe Bacteroides gingivalis: Physical, chemical and immunological properties. J . Bacteriol. 163: 730-734. 261. Yoshimura, F., K. Takahasti, Y. Nodasaka, and T. Suzuki. 1984. Purification and characterization of a novel type of fimbriae from the oral anaerobe Bacteroides gingivalis. J . Bacteriol. 160. 949-952. 262. Zambon, J .J . , H.S. Reynolds, and J . Slots. 1981. Black- Pigmented Bacteroides in the human oral cavity. Infect. Immun. 32: 198-203. 167 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items