Fibronectin in saliva and gingival crevices

Porphyromonas gingivalis Fimbria-Induced Expression of Inflammatory Cytokines and Cyclooxygenase-2 in Mouse Macrophages and Its Inhibition by the Bioactive Compounds Fibronectin and Melatonin

Porphyromonas gingivalis (Pg) fimbriae, in addition to lipopolysaccharide, are involved in the pathogenesis of periodontal disease. At the same time, bioactive compounds such as fibronectin (FN) and melatonin in saliva and gingival crevicular fluid have been reported to exert a preventive effect against periodontitis. Here, we review current knowledge regarding the potent inhibitory effects of FN and melatonin against Pg fimbria-induced induction of proinflammatory cytokines, cyclooxygenase-2 (COX-2) expression, and NF-kappa B activation in mouse macrophages and discuss their possible clinical application for prevention of periodontal diseases induced by oral bacteria.

Effectiveness of oral hygiene interventions against oral and oropharyngeal reservoirs of aerobic and facultatively anaerobic gram-negative bacilli

BACKGROUND

Aerobic and facultatively anaerobic gram-negative bacilli (AGNB) are opportunistic pathogens and continue to cause a large number of hospital-acquired infections. AGNB residing in the oral cavity and oropharynx have been linked to nosocomial pneumonia and septicemia. Although AGNB are not considered members of the normal oral and oropharyngeal flora, medically compromised patients have been demonstrated to be susceptible to AGNB colonization.

METHODS

A literature search was conducted to retrieve articles that evaluated the effectiveness of oral hygiene interventions in reducing the oral and oropharyngeal carriage of AGNB in medically compromised patients.

RESULTS

Few studies have documented the use of mechanical oral hygiene interventions alone against AGNB. Although a number of studies have employed oral hygiene interventions complemented by antiseptic agents such as chlorhexidine and povidone iodine, there appears to be a discrepancy between their in vitro and in vivo effectiveness.

CONCLUSION

With the recognition of the oral cavity and oropharynx as a reservoir of AGNB and the recent emergence of multidrug and pandrug resistance in hospital settings, there is a pressing need for additional high-quality randomized controlled trials to determine which oral hygiene interventions or combination of interventions are most effective in eliminating or reducing AGNB carriage.

Antimicrobial peptides and periodontal disease

AIMS

The goal of this review is to identify the antimicrobial proteins in the oral fluids, saliva and gingival crevicular fluid and identify functional families and candidates for antibacterial treatment.

RESULTS

Periodontal biofilms initiate a cascade of inflammatory and immune processes that lead to the destruction of gingival tissues and ultimately alveolar bone loss and tooth loss. Treatment of periodontal disease with conventional antibiotics does not appear to be effective in the absence of mechanical debridement. An alternative treatment may be found in antimicrobial peptides and proteins, which can be bactericidal and anti-inflammatory and block the inflammatory effects of bacterial toxins. The peptides have co-evolved with oral bacteria, which have not developed significant peptide resistance. Over 45 antibacterial proteins are found in human saliva and gingival crevicular fluid. The proteins and peptides belong to several different functional families and offer broad protection from invading microbes. Several antimicrobial peptides and proteins (AMPs) serve as templates for the development of therapeutic peptides and peptide mimetics, although to date none have demonstrated efficacy in human trials.

CONCLUSIONS

Existing and newly identified AMPs may be developed for therapeutic use in periodontal disease or can serve as templates for peptide and peptide mimetics with improved therapeutic indices.

Specific fibronectin fragments as markers of periodontal disease status

BACKGROUND

The diagnosis of progressing periodontal disease typically relies on retrospective methods that detect changes in the amount of periodontal breakdown. Fibronectin (FN) fragments are found in vivo in association with periodontal disease, and specific FN fragments compromise periodontal ligament cell functions in vitro. The overall goal of this cross-sectional study was to determine whether specific FN fragments are present in gingival crevicular fluid (GCF) and can be used as markers for periodontal disease status. The eventual goal is to test these FN fragments in a longitudinal study as potential markers of disease activity.

METHODS

GCF was collected from 94 subjects with untreated periodontitis from clinically healthy, mild/moderate periodontitis, and severe periodontitis sites. Sites were defined on the basis of clinical criteria, including gingival bleeding index, probing depth, and clinical attachment level. Western immunoblotting was used to detect FN fragments in GCF using antibodies to specific FN domains, including the collagen/gelatin-, central cell-, and carboxyl terminal heparin-binding domains, plus the CS-1 site on the alternatively spliced V region and the EIIIA region. FN fragments identified by immunoblotting and analyzed by NIH image software were scored based on pixel intensity and an ordinal grade scale.

RESULTS

We identified several fragments highly associated with severe periodontitis sites, including 40-kDa, 120-kDa, and 68-kDa fragments.

CONCLUSIONS

This study demonstrates that specific FN fragments are markers for periodontal disease status and supports the role of FN fragments as potential components in the pathogenesis of periodontal disease.

Reverse transcription-polymerase chain reaction and immunohistochemical study of the expression of fibronectin mRNA in human submandibular salivary gland

The alternative expression of fibronectin mRNA in submandibular salivary glands was analysed using three groups of polymerase chain reaction probes to amplify the cell-binding domain (CBD), and the extra domain (ED)-A and ED-B exon regions in five normal human adults. The fibronectin CBD region was fairly well expressed in the salivary gland while less than trace amounts of ED-A and ED-B exons were expressed. Immunohistochemical staining with an antihuman plasma fibronectin monoclonal antibody, which recognized specifically the CBD region, demonstrated definite positive cytoplasmic staining in the duct-cell area. On the other hand, an anticellular fibronectin ED-A-specific monoclonal antibody was very low positive in duct cells. The results suggest that salivary fibronectin is synthesized in the submandibular salivary gland only, and does not include the ED-A segment. Furthermore, duct cells also produced fibronectin without ED-A.

Identification of a CD36-related thrombospondin 1-binding domain in HIV-1 envelope glycoprotein gp120: relationship to HIV-1-specific inhibitory factors in human saliva

Human and non-human primate salivas retard the infectivity of HIV-1 in vitro and in vivo. Because thrombospondin 1 (TSP1), a high molecular weight trimeric glycoprotein, is concentrated in saliva and can inhibit the infectivity of diverse pathogens in vitro, we sought to determine the role of TSP1 in suppression of HIV infectivity. Sequence analysis revealed a TSP1 recognition motif, previously defined for the CD36 gene family of cell adhesion receptors, in conserved regions flanking the disulfide-linked cysteine residues of the V3 loop of HIV envelope glycoprotein gp120, important for HIV binding to its high affinity cellular receptor CD4. Using solid-phase in vitro binding assays, we demonstrate direct binding of radiolabeled TSP1 to immobilized recombinant gp120. Based on peptide blocking experiments, the TSP1-gp120 interaction involves CSVTCG sequences in the type 1 properdin-like repeats of TSP1, the known binding site for CD36. TSP1 and fusion proteins derived from CD36-related TSP1-binding domains were able to compete with radiolabeled soluble CD4 binding to immobilized gp120. In parallel, purified TSP1 inhibited HIV-1 infection of peripheral blood mononuclear cells and transformed T and promonocytic cell lines. Levels of TSP1 required for both viral aggregation and direct blockade of HIV-1 infection were physiologic, and affinity depletion of salivary TSP1 abrogated >70% of the inhibitory effect of whole saliva on HIV infectivity. Characterization of TSP1-gp120 binding specificity suggests a mechanism for direct blockade of HIV infectivity that might be exploited to retard HIV transmission that occurs via mucosal routes.

Immunocytochemical localization of fibronectin and a 165-kDa membrane protein in the odontoblast layer under initial carious lesions in man

The possible role of fibronectin in dental tissue repair was investigated by comparing its distribution and that of the 165-kDa fibronectin-binding membrane protein (165 kDa-FnBP) in odontoblasts underlying carious and sound dentine. By immunoperoxidase and light microscopy, fibronectin was localized in the dentine underlying the carious lesion, mainly on the surface of the tubule walls, whereas it could not be detected in neighbouring sound zones. The antibody to the 165 kDa-FnBP strongly reacted with the membrane of odontoblasts underlying the lesion, although those facing sound dentine did not express this antigen. Ultrastructurally the 165 kDa-FnBP was localized in the cell membrane at the apical portion of odontoblasts, including the process membrane, beneath the initial lesion; fibronectin was detected in the dentinal area close to the process, and also in contact with its external surface. By a high-resolution immunogold procedure, the proteins were colocalized at the external surface of odontoblast processes. These data suggest that fibronectin present in human carious dentine could modulate the behaviour of underlying odontoblasts by means of newly expressed 165 kDa-FnBP.

The pathogenesis of ventilator-associated pneumonia: I. Mechanisms of bacterial transcolonization and airway inoculation

Ventilator-associated pneumonia (VAP) is an infection of the lung parenchyma developing in patients on mechanical ventilation for more than 48 h. VAP is associated with a remarkably constant spectrum of pathogenic bacteria, most of which are aerobic Gram-negative bacilli (AGNB) and, to a lesser extent Staphyloccus aureus. Most authorities agree that VAP develops as a result of aspiration of secretions contaminated with pathogenic organisms, which appear to be endogenously acquired. These pathogens gain access to the distal airways by mechanical reflux and aspiration of contaminated gastric contents and also by repetitive inoculation of contaminated upper airway secretions into the distal tracheobronchial tree. Persistence of these organisms in the upper airways involves their successful colonization of available surfaces. Although exogenous acquisition can occur from the environment, the rapidity at which critically ill patients acquire AGNB in the upper airways in conjunction with the low rate of AGNB colonization of health-care workers exposed to the same environment favors the presence of endogenous proximate sources of AGNB and altered upper airway surfaces that are rendered receptive. Proximate sources of AGNB remain unclear, but potential sites harboring AGNB prior to illness include the upper gastrointestinal tract, subgingival dental plaque, and the periodontal spaces. Following illness or antibiotic therapy, competitive pressures within the oropharynx favor AGNB adherence to epithelial cells, which lead to oropharyngeal colonization. Similar dynamic changes in contiguous structures (oropharynx, trachea, sinuses, and the upper gastrointestinal tract) lead to the transcolonization of these structures with pathogenic bacteria. Following local colonization or infection, these structures serve as reservoirs of AGNB capable of inoculating the lower airways. As the oropharynx becomes colonized with AGNB, contaminated oropharyngeal secretions reach the trachea, endotracheal tube, and ventilator circuit. Contaminated secretions pooled above the endotracheal tube cuff gain access to the trachea and inner lumen of the endotracheal tube by traversing endotracheal tube cuff folds. Amorphic particulate deposits containing AGNB form along the endotracheal tube and are capable of being propelled into the distal airways by ventilator-generated airflow or by tubing manipulation. Bacteria embedded within this type of amorphous matrix are particularly difficult for the host to clear. If host defenses fail to clear the inoculum, then bacterial proliferation occurs, and the host inflammatory response progresses to bronchopneumonia.(ABSTRACT TRUNCATED AT 400 WORDS)

Saliva-bacterium interactions in oral microbial ecology

Saliva is thought to have a significant impact on the colonization of microorganisms in the oral cavity. Salivary components may participate in this process by one of four general mechanisms: binding to microorganisms to facilitate their clearance from the oral cavity, serving as receptors in oral pellicles for microbial adhesion to host surfaces, inhibiting microbial growth or mediating microbial killing, and serving as microbial nutritional substrates. This article reviews information pertinent to the molecular interaction of salivary components with bacteria (primarily the oral streptococci and Actinomyces) and explores the implications of these interactions for oral bacterial colonization and dental plaque formation. Knowledge of the molecular mechanisms controlling bacterial colonization of the oral cavity may suggest methods to prevent not only dental plaque formation but also serious medical infections that may follow microbial colonization of the oral cavity.

Characterization of fibronectin and fibrin(ogen) fragments in gingival crevicular fluid

A total of 49 crevicular fluid (CF) samples were collected with paper strips from 12 healthy adults. Each sample was eluted into sterile saline and two aliquots were drawn for SDS-PAGE, one for fibronectin and one for fibrin analysis. Peptides were transferred to nitrocellulose membranes, and fibronectin and fibrin were detected using specific antibodies. The relative amounts of different molecular forms of fibronectin and fibrin were analyzed using a laser densitometer. After the sample collection, Plaque Index, Papilla Bleeding Index and pocket depth were measured. Bone loss was estimated from the orthopanthomograms. Fibronectin fragments were seen in all CF samples. Intact fibronectin was seen in 21 samples, of which 76% were collected from periodontitis-affected sites. There was a positive correlation between the proportion of intact fibronectin and the clinical parameters. Intact fibrin and fibrin fragments were seen in all samples. Fibrin-positive material with larger molecular weight than intact fibrin was also seen in all samples. A negative correlation was found between the proportion of intact fibrin and the clinical parameters. There was no correlation between total amounts and concentrations of fibronectin and fibrin. Molecular forms of fibronectin and fibrin may affect the pathogenesis and healing of periodontal diseases, since the biologic effects of the fragments of these molecules differ from those of the intact molecules.

Interaction of the envelope glycoprotein of human immunodeficiency virus with C1q and fibronectin under conditions present in human saliva

Human saliva has been shown to reduce the infectivity of human immunodeficiency virus (HIV) particles in vitro. The factors in human saliva involved in this inhibition of HIV infectivity are unknown, although the salivary sediment of normal individuals has the major HIV neutralizing activity. Interestingly, the first complement component (C1) has been detected on the surface of the salivary sediment in the whole saliva of normal individuals. At the relatively low ionic strength of saliva, we determined that purified human C1q bound with high affinity to the envelope glycoprotein of HIV. Normally, the interaction of the C1q globular heads with immune complexes causes C1 activation. However, direct interactions between C1 and rgp120 (or rgp160) did not lead to C1 fixation, as determined by hemolytic studies with rate-limiting levels of C1, nor did rgp120 cause C1 activation as determined by activated C1s-mediated C4 conversion in normal human serum. Using ELISA, it was observed that intact C1, with the C1r2C1s2 tetramer associated with the collagen-like stem of C1q, did not bind to immobilized rgp120, whereas free C1q did bind. In addition, digestion of the C1q stem portion with collagenase completely eliminated its binding to rgp120. These findings suggest that the collagen-like stem region of C1q, rather than the globular heads, may participate in the binding to the envelope glycoprotein of HIV. Fibronectin, which is present in submandibular saliva, appeared to bind to rgp120 and to enhance the interaction of C1q with rgp120. It is conceivable that C1q and fibronectin, in binding and sequestering HIV particles (i.e. to the salivary sediment), may play an important role in the reduction of HIV transmission via saliva. Further studies will be needed to test the latter speculation.

The interaction of salivary secretions with the human complement system--a model for the study of host defense systems on inflamed mucosal surfaces

When complement first contacts salivary secretions, as when gingival crevicular fluid first meets saliva at the gingival margin, complement function is enhanced. The immediate potentiation of the complement system at equal volume ratios of serum to saliva is due to several factors, including the lower ionic strength of saliva when compared with serum and the presence of certain salivary glyproteins such as the nonimmunoglobulin agglutinins that appear to simultaneously activate C1 and affect (sequester) certain complement control proteins, such as Factor H. This initial potentiation of the complement cascade by saliva may aid in defending the area immediately above the gingival crevice from oral microbiota that are being coated with a combination of serous exudate components and salivary components. As serum becomes much more diluted with saliva (i.e., crevicular fluid moves away from the supragingival area), the acidic proline-rich salivary proteins (APRP) begin to disrupt the unbound C1q-C1r2-C1s2 macromolecular complexes. Thus, the APRP along with other C1 fixing substances in saliva appear to restrict complement function, but only when the ratios of saliva to serum exceed 250:1. Since certain salivary glycoproteins bind to viruses, the potentiation of the complement system by saliva may also play a role in neutralizing certain viral infections on mucosal surfaces where tissue transudates containing complement begin to contact mucosal secretions such as saliva. Again, the ratio of serous fluid to mucosal secretion appears to be an important factor. This article also discusses some of our preliminary data and speculations concerning the binding of the self-associating high-molecular-weight nonimmunoglobulin salivary agglutinins (NIA) with the envelope of the human immunodeficiency virus (HIV) and the possible cooperative role of C1q and fibronectin in aiding neutralization of HIV infectivity.

Salivary fibronectin in man: an immunoblotting, radioimmunoassay and immunohistochemical study

An antiserum and monoclonal antibody against plasma fibronectin recognized a 230-kDa, intact fibronectin molecule when 1000-fold diluted human plasma was subjected to immunoblot analysis. Immunoblot analysis of the parotid saliva demonstrated that the rabbit antiserum to human plasma fibronectin bound to five molecules (200, 110, 85, 75 and 65 kDa), other than the 230-kDa, intact fibronectin molecule, while the mouse monoclonal antibody recognized only the 230-kDa molecule. The 230-kDa type was not found in whole saliva with either the antiserum or monoclonal antibody. The antiserum reacted with 85-, 75- and 33-kDa molecules, and the monoclonal antibody recognized 75-, 33- and 20-kDa molecules in the whole saliva. Radioimmunoassay revealed that the mean +/- SD of fibronectin concentration in parotid saliva was 2.5 +/- 1.4 ng/ml (n = 20) and 149.8 +/- 46.2 ng/ml (n = 30) for whole saliva. Immunoperoxidase staining with rabbit antiserum and the mouse monoclonal antibody showed positive cytoplasmic staining of cells in the intralobular and interlobular ducts in parotid, submandibular, and sublingual glands. No acinar cells were stained.

Gingival crevicular fluid fibronectin degradation in periodontal health and disease

The molecular forms of fibronectin (FN) in gingival crevicular fluid of five subjects with at least two sites exhibiting clinical signs of inflammation and pockets of at least 4 mm (test group) and five subjects with clinically healthy periodontium (control group) were investigated. Samples were collected with standard filter paper strips. In the test group samples from both diseased and healthy sites were collected. After collection the test group received one episode of periodontal treatment (scaling and root planning). The sampling and clinical recording were repeated for the diseased sites after about 2 wk. The crevicular fluid FN was analyzed using sodium dodecyl sulphate gel electrophoresis followed by western blotting with polyclonal antibodies against FN. Both intact FN and FN fragments were found in all samples. A larger proportion of FN was in degraded form in the diseased sites than in the healthy or the treated sites. FN was also degraded into smaller peptide fragments in the diseased than in the treated sites. These results suggest that crevicular fluid FN is partially degraded both in periodontal health and disease and that the degree of degradation of FN increases with periodontal inflammation and decreases with periodontal treatment.

Fibronectin levels of unstimulated saliva from naval recruits with and without chronic inflammatory periodontal disease

Past studies have suggested that gingival crevicular fluid is produced more readily in persons with severe periodontal diseases than in persons with healthy gingivae. In this study, salivary fibronectin was selected as an index of total gingival crevicular fluid flow. Our purpose was to determine whether a relationship could be found between salivary fibronectin level and periodontal disease status. Unstimulated saliva was collected from 20 healthy and 20 periodontally-diseased naval recruits. The periodontally-diseased subjects included 10 with localized juvenile periodontitis and 10 with moderate to severe periodontitis. Mean subject ages and salivary flow rates were similar for the 2 groups. Although 2 of the periodontally-diseased subjects showed unusually high fibronectin levels, the mean level for the remaining 18 subjects did not differ significantly from the mean of the healthy group, and no association of periodontal disease status with salivary fibronectin content was seen. Consequently, it was not evident from salivary fibronectin levels that the content of gingival crevicular fluid in unstimulated whole saliva differed significantly for persons with or without severe periodontal disease, except possibly for extreme cases of disease.

Concentrations of fibronectin in the sera and crevicular fluid in various stages of periodontal disease

While fibronectin (FN) has previously been demonstrated to be present in gingival crevicular fluid (GCF), its quality and quantity has not been reported. Since this information is relevant for ongoing studies on the use of FN for gingival reattachment, we performed these measurements and compared plasma levels in healthy subjects, patients with gingivitis and periodontitis, and in patients undergoing maintenance therapy. Plasma and GCF samples were obtained from 4 sites in each subject using a Periotron to permit quantification of samples. FN concentrations were determined in a microELISA using hyperimmune anti-FN antibody. Purified FN served as a reference for quantification. The functional activity of each sample was assessed by examining the natural affinity of FN for gelatin. Subjects with gingivitis and those in maintenance had significantly depressed levels of plasma fibronectin. While little fibronectin could be detected in the GCF of healthy sites regardless of patient category, examination of the most diseased sites in each group revealed that the concentration of FN in the GCF was highest in health and reduced when there was gingival inflammation. In no case was GCF FN found to be biologically active.

Inhibition of adhesion of viridans streptococci to fibronectin-coated hydroxyapatite beads by lipoteichoic acid

Fibronectin-coated hydroxyapatite (FnHA) beads were used in a model adhesion assay to isolate the lipoteichoic acid (LTA) mediated adhesion of oral streptococci. Representative strains of the commonly isolated viridans streptococci were incubated with FnHA beads in the presence and absence of exogenous LTA. The LTA inhibited the adhesion of all strains to a greater or lesser extent, but only a very few strains were inhibited by more than 90%. Strains of Streptococcus sanguis Type II and Streptococcus mitis which synthesize an amphiphile other than LTA were also inhibited. The findings provided circumstantial evidence for the involvement of LTA in the adhesion of this group of oral bacteria.

Crevicular and salivary fibronectin before and after gingivitis treatment

The aims of this study were to assess crevicular fibronectin concentrations before and after conventional gingivitis treatment and to determine whether fibronectin concentration in whole saliva would be affected concomitantly. 10 subjects with generalized gingivitis were selected. Examinations were made before and after treatment and included measurements of clinical parameters as well as collection of samples of unstimulated saliva, stimulated saliva and crevicular material. The concentration of fibronectin was studied by an ELISA assay. Fibronectin was found in gingival crevices and in unstimulated as well as paraffin-stimulated whole saliva in pre- and post-treatment samples. There were no statistically significant differences between pre- and post-treatment concentrations of fibronectin, whether expressed as micrograms fibronectin/micrograms protein or as micrograms fibronectin/ml saliva.

The degradation of type I collagen and human plasma fibronectin by the trypsin-like enzyme and extracellular membrane vesicles of Bacteroides gingivalis W50

A soluble trypsin-like enzyme (STE) was purified from a cell- and particle-free culture supernatant of this bacterium by a combination of ultra-centrifugation, ammonium-sulphate precipitation and gel-filtration chromatography on Sephacryl S-200. Trypsin-like activity in the culture supernatant was associated with a 58 kDa peptide and also with a higher molecular-weight complex. The STE and extracellular vesicle (ECV) fraction of B. gingivalis W50 rapidly degraded human plasma fibronectin in the presence and the absence of 10 mM dithiothreitol (DTT). The STE yielded a range of lower molecular-weight fibronectin digestion products. Under conditions where little activity was expressed by mammalian trypsin, both STE and ECV depolymerized a denatured and a native type I collagen substrate. Quantitative and qualitative differences were observed in the patterns of digestion products generated by both STE and ECV fraction following incubation with and without 10mM DTT. Inclusion of DTT appeared to reduce the degradative effect of both ECV and STE towards the type I collagen and plasma fibronectin substrates.

Effects of fibronectin and other salivary macromolecules on the adherence of Escherichia coli to buccal epithelial cells

The effect of saliva and fibronectin (Fn) on the adherence of a type 1 fimbriated strain of Escherichia coli to human buccal epithelial cells was studied. Saliva pretreatment of epithelial cells led to a dose-dependent increase in adherence that was inhibited by alpha-methyl mannoside, which is typical of a type 1 fimbria-mediated event. The molecules responsible for affecting this increased adherence were nondialyzable and were recovered after lyophilization. E. coli adherence was stimulated by individual saliva samples from each of 11 volunteers. Fn inhibited E. coli adherence to saliva-treated buccal cells by more than 60%. Biotinylated E. coli and Fn were reacted with Western blots of whole saliva to identify the receptors that might explain the phenomenon described above. Both E. coli and Fn bound to 57- and 62-kilodalton (kDa) protein bands in Western blots of sodium dodecyl sulfate gels of whole saliva. The binding of E. coli to these bands was inhibited by pretreatment with unlabeled Fn. To study these salivary components, samples of saliva were electrophoresed on sodium dodecyl sulfate gels, strips corresponding to the appropriate molecular weights were cut out, and the proteins were eluted electrophoretically. Material that eluted from strips at 57 and 62 kDa, but not that from a control strip, stimulated E. coli adherence to buccal cells. Alternatively, saliva was fractionated over 100- and 50-kDa cutoff filters. Of the three fractions obtained, only the fraction passing through the 100-kDa filter and retained by the 50-kDa filter stimulated E. coli adherence to buccal cells. This fraction also increased the binding of Fn to buccal cells. These observations suggest the possibility that one or more salivary components bind to the surface of buccal cells and serve as receptors for type 1 fimbriated E. coli. Fn also binds to this isolated material; and it is apparently by these interactions, at least in part, that saliva stimulates and Fn inhibits E. coli adherence. The way in which these interactions may affect bacterial adherence in vivo remains to be elucidated.