Mediation of adherence of streptococci to human endothelial cells by complement S protein (vitronectin).

The role of S protein in the adherence of group A and G streptococci to human umbilical vein endothelial cells cultivated in 96-well microdilution plates was studied by utilizing fluorescein-labeled streptococci. The assay proved suitable for quantitative determination of bacterial adherence to cultured endothelial cells for all tested strains of streptococci. Only bacterial strains with significant S protein binding but weak fibronectin binding were included in these studies. Fibronectin-mediated adherence to endothelial cells of these streptococci was less than 25% of total and could be blocked by antifibronectin immunoglobulin G. Further treatment of endothelial cell monolayers with anti-S protein immunoglobulin G at concentrations up to 1 mg per well led to an almost complete inhibition of adherence for all tested streptococcal cultures, indicating significant contribution of S protein in the streptococcus-endothelial cell interaction. Blocking of S-protein-binding sites on streptococci by preincubation with exogeneous S protein at a concentration of 10 micrograms per 4 x 10(7) streptococci led to about 75% reduction of S-protein-mediated adherence to endothelial cells. Trypsin pretreatment of group G streptococci and pronase pretreatment of group A and G streptococci, modifications known to destroy the bacterial binding sites of S protein, also inhibited the capacity of the streptococci for S-protein-mediated adherence to endothelial cells by 75 to 80%. These results indicate that S protein plays a mediatory role in adherence of streptococci to endothelial cells and that S-protein-specific binding sites on streptococci are involved in this interaction.     

Ingestion of Staphylococcus aureus, Staphylococcus epidermidis, and Escherichia coli by human peritoneal mesothelial cells.

In the present study we examined whether mesothelial cells can ingest and digest bacteria. The results showed that all strains were ingested. Ingested staphylococci proliferated abundantly, and only a few were digested. Escherichia coli, however, was digested during the first 8 h, whereafter the mesothelial cells disintegrated and proliferation of bacteria could be observed. The clinical implications of these findings are discussed.     

Staphylococcus aureus binding to human nasal mucin.

Colonization of human nasal mucosa with Staphylococcus aureus sets the stage for subsequent systemic infection. This study characterizes S. aureus adhesion to nasal mucosa in vitro and investigates the interaction of S. aureus with human nasal mucin. S. aureus binding to cell-associated and cell-free mucus was greater than to nonmucin-coated epithelial cells. Scanning electron microscopy of S. aureus incubated with human nasal mucosal tissue showed minimal binding to ciliated respiratory epithelium. In a solid-phase assay, S. aureus bound to purified human nasal mucin-coated wells significantly more than to bovine serum albumin-coated microtiter wells. Binding to mucin was saturable in a dose- and time-dependent fashion. Staphylococcal adherence to human nasal mucin was inhibited by bovine submaxillary mucin but not by fibrinogen. Pretreatment of mucin with periodate but not with pronase reduced adherence. Trypsin treatment of the bacteria significantly reduced adherence to mucin. 125I-labelled nasal mucin bound to two surface proteins (138 and 127 kDa) of lysostaphin-solubilized S. aureus. Binding to human nasal mucin occurs in part via specific adhesin-receptor interactions involving bacterial proteins and the carbohydrate moiety in mucin. These experiments suggest that S. aureus binding to mucin may be critical for colonization of the nasopharyngeal mucosa.     

Kinetics of phagocytosis of Staphylococcus aureus and Escherichia coli by human granulocytes.

Although phagocytosis of micro-organisms by granulocytes is one of the most important defence mechanisms against infection, little is known about the kinetics of this process. The present study showed that the rate of ingestion of Staphylococcus aureus and Escherichia coli depends on the concentrations of the granulocytes and bacteria. Phagocytosis of bacteria at a bacteria-to-cell ratio in the range between 100:1 and 1:10 showed an exponential course during the first 30 min. At a bacteria-to-cell ratio of 1:1, application of a correction for the outgrowth of extracellular bacteria gave an exponential course of ingestion over the first 90-min period. Since it was found that the phagocytosis of bacteria by granulocytes at various bacteria-to-cell ratios can be described with Michaelis-Menten kinetics, we studied the kinetics of phagocytosis on the basis of the initial rate for the first 30-min period. The rate of phagocytosis and the maximal degree of ingestion of bacteria by granulocytes proved to be related to the concentration of serum used in the assay. The minimal serum concentration required for maximal ingestion was 2.5% for Staphylococcus aureus and 5% for Escherichia coli. When bacteria were pre-opsonized, the duration of pre-opsonization proved to be limiting for the rate of phagocytosis in dependence on the serum concentration. The effect of temperature on the phagocytosis of micro-organisms proved to be two-fold. First, at temperatures between 4 and 33 degrees a decrease in the functioning of the cells leads to a decrease in the rate of phagocytosis. Above 42 degrees, the temperature affects mainly the opsonization of the micro-organisms and has only a slight influence on the ingestion process. From the data obtained in this study, maximal rates of 6.3 X 10(6) Staphylococcus aureus/5 X 10(6) granulocytes/min and of 7.1 X 10(6) Escherichia coli/5 X 10(6) granulocytes/min were calculated for phagocytosis at a bacteria-to-cell ratio of 100:1 at 37 degrees, i.e. on average about one bacterium per granulocyte per min. The maximum calculated number of bacteria ingested by one granulocyte lies between 40 and 50.     

Specific binding of lactoferrin to Escherichia coli isolated from human intestinal infections.

The degrees of human lactoferrin (HLf) and bovine lactoferrin (BLf) binding in 169 Escherichia coli strains isolated from human intestinal infections, and in an additional 68 strains isolated from healthy individuals, were examined in a 125I-labelled protein binding assay. The binding was expressed as a percentage calculated from the total labelled ligand added to bacteria. The HLf and BLf binding to E. coli was in the range 3.7 to 73.4% and 4.8 to 61.6%, respectively. Enterotoxigenic strains demonstrated a significantly higher HLf binding (median = 19%) than enteropathogenic, enteroinvasive, enterohaemorrhagic strains or normal intestinal E. coli isolates (medians 6 to 9). Enteropathogenic strains belonging to serotypes O44 and O127 demonstrated significantly higher HLf binding compared to O26, O55, O111, O119 and O126. No significant differences in the degree of HLf or BLf binding were found between aerobactin-producing and non-producing strains. The interaction was further characterized in a high Lf-binding EPEC strain, E34663 (serotype O127). The binding was stable in the pH range 4.0 to 7.5, did not dissociate in the presence of 2M NaCl or 2M urea, and reached saturation within two h. Unlabelled HLf and BLf displaced the 125I-HLf binding to E34663 in a dose-dependent manner. Apo- and iron-saturated forms of Lf demonstrated similar binding to E34663. Among various unlabelled subepithelial matrix proteins and carbohydrates tested (in 10(4)-fold excess) only fibronectin and fibrinogen caused a moderate inhibition of 125I-HLf binding. According to Scatchard plot analysis, 5,400 HLf-binding sites/cell, with an affinity constant (Ka) of 1.4 x 10(-7) M, were estimated in strain E34663. These data establish the presence of a specific Lf-binding mechanism in E. coli.     

Collagen binding to Staphylococcus aureus.

Staphylococcus aureus can bind soluble collagen in a specific, saturable manner. We have previously shown that some variability exists in the degree of collagen binding between different strains of heat-killed, formaldehyde-fixed S. aureus which are commercially available as immunologic reagents. The present study demonstrates that live S. aureus of the Cowan 1 strain binds amounts of collagen per organism equivalent to those demonstrated previously in heat-killed, formaldehyde-fixed bacteria but has an affinity over 100 times greater, with Kd values of 9.7 X 10(-11) M and 4.3 X 10(-8) M for live and heat-killed organisms, respectively. Studies were also carried out with S. aureus killed by ionizing radiation, since this method of killing the organism seemed less likely to alter the binding moieties on the surface than did heat killing. Bacteria killed by exposure to gamma radiation bound collagen in a manner essentially indistinguishable from that of live organisms. Binding of collagen to irradiated cells of the Cowan 1 strain was rapid, with equilibrium reached by 30 min at 22 degrees C, and was fully reversible. The binding was not inhibited by fibronectin, fibrinogen, C1q, or immunoglobulin G, suggesting a binding site for collagen distinct from those for these proteins. Collagen binding was virtually eliminated in trypsin-treated organisms, indicating that the binding site has a protein component. Of four strains examined, Cowan 1 and S. aureus ATCC 25923 showed saturable, specific binding, while strains Woods and S4 showed a complete lack of binding. These results suggest that some strains of S. aureus contain high-affinity binding sites for collagen. While the number of binding sites per bacterium varied sixfold in the two collagen-binding strains, the apparent affinity was similar. The ability of S. aureus to bind collagen with high affinity may provide a mechanism for bacterial adhesion to host tissue and thereby play a role in the invasive characteristics of this organism.     

Antibodies to a range of Staphylococcus aureus and Escherichia coli heat shock proteins in sera from patients with S. aureus endocarditis.

Antibodies to a range of Staphylococcus aureus and Escherichia coli heat shock proteins were present in sera from patients with S. aureus endocarditis. This suggests the highly immunoreactive nature of a range of heat shock proteins in addition to the GroEL equivalent (common antigen) protein. In one case, antibodies to three proteins unique to the infecting S. aureus strain, which were more prominent in heat-shocked cells, were also detected.     

Streptococcal protein G, expressed by streptococci or by Escherichia coli, has separate binding sites for human albumin and IgG

Protein G is expressed at the cell surface of certain group C and group G streptococcal strains. The protein shows a unique and specific affinity for the Fc region of mammalian polyclonal and monoclonal immunoglobulin G (IgG). We have cloned the streptococcal gene coding for protein G into E. coli, using phage lambda as the vector. The protein G produced by E. coli infected with this phage was detected and analysed in Western blot experiments using radiolabelled IgG Fc fragments as a probe. Three major IgG Fc-binding bands were obtained corresponding to apparent mol. wts of 47,000, 57,000 and 65,000, respectively. Analysis of the expression in E. coli indicates that this heterogeneity is caused by a post-translational degradation of the molecule before lysis of the lambda infected E. coli cells occurred. The protein G produced in E. coli was purified by affinity chromatography on IgG-Sepharose followed by gel-filtration on Sephadex G-200. This highly purified E. coli-produced protein G was compared to protein G solubilized by papain from streptococci, in direct binding experiments and in a competitive binding assay. The two protein G variants were found to interact with polyclonal IgG from different species in a similar way. Streptococcal strains expressing protein G also show affinity for human albumin, and at the molecular level protein G was found to be responsible also for the binding of albumin. Thus, both E. coli-produced protein G and the proteolytic fragment of protein G obtained from streptococci, bound albumin. On the protein G molecule, two different and separate sites were found to bind IgG and albumin. Finally, when whole streptococci were incubated with human plasma, the interactions with protein G caused a coating of the bacteria with albumin and IgG, whereas other plasma proteins showed no affinity for protein G.     

Kinetics of intracellular killing of Staphylococcus aureus and Escherichia coli by human granulocytes

The intracellular killing of Staphylococcus aureus and Escherichia coli by human granulocytes was investigated independently of the ingestion of these bacteria. Granulocytes were allowed to phagocytose preopsonized bacteria for only 3 min, after which the noningested bacteria were removed by differential centrifugation and 2 washes. With this technique, the number of viable cell-associated bacteria at the start of the assay, determined after lysis of the granulocytes, includes about 80% intracellular bacteria. Intracellular killing depends on (a) the temperature (no killing occurring at 4°C, maximal killing at 33–39 °C, and a decrease in the capacity of the granulocytes to kill ingested bacteria at temperatures above 42 °C), and (b) the number of bacteria ingested (after phagocytosis at bacteria-to-cell ratios of 100 : 1 and 1000 : 1, not all of the ingested bacteria are killed, whereas after phagocytosis at lower bacteria-to-cell ratios, almost all ingested bacteria are killed). To determine the maximum number of bacteria that can be killed by granulocytes, intracellular killing was measured after phagocytosis of bacteria at various bacteria-to-granulocyte ratios in the presence of phenylbutazone, a drug which inhibits killing during the ingestion period. Phenyl-butazone proved to be a useful tool in the study of intracellular killing, since this drug provides a reversible inhibition of the killing when granulocytes are incubated in its presence for up to 3 min, whereas after longer incubation, the inhibitory effect is irreversible. Calculation based on the data obtained in this study gave maximum rates of intracellular killing amounting to 3.7 × 10⁵ bacteria/5 × 10⁶ granulocytes/min for Staph. aureus and 8.5 × 10⁵ bacteria/5 × 10⁶ granulocytes/min for E. coli. Using the rate of intracellular killing after phagocytosis at a bacteria-to-granulocyte ratio of 1 : 1 and the rate of ingestion obtained in an earlier study, we were able to compute the theoretical numbers of viable extracellular not (yet) ingested, viable intracellular, killed intracellular and total intracellular bacteria. The theoretical curves fit well with the experimental data.     

Specific attachment of Staphylococcus aureus to immobilized fibronectin.

Staphylococcus aureus cells have been shown to possess surface-associated proteins with affinity for soluble fibronectin. We have investigated the ability of these surface proteins to mediate attachment to immobilized fibronectin and collagen. Attachment was quantified by determination of bacterial ATP in a bioluminescence assay. The ability to attach to fibronectin- or collagen-coated plastic surfaces was investigated for four S. aureus strains: Cowan 1, Newman, SA113(83A), and Wood 46. Cells from the different strains varied in their attachment properties, but all cells except those of strain Wood 46 attached readily to substrates coated with fibronectin. Only cells from strain Cowan 1 attached reproducibly to collagen-coated substrates in the absence of fibronectin. The attachment of cells from strain SA113(83A) to fibronectin-coated surfaces was shown to be dependent on time, fibronectin concentration, and bacterial growth phase. Soluble fibronectin or NH2-terminal fibronectin fragment (Mr, 29,000) disturbed the attachment to surfaces coated with fibronectin bound to denatured collagen type I. The attachment process to such substrates was also effectively inhibited by preincubating the substrate with fibronectin-binding proteins isolated from S. aureus Newman and SA113 (83A) and purified with affinity chromatography.     

Cloning, expression, and mapping of the Staphylococcus aureus alpha-hemolysin determinant in Escherichia coli K-12.

A fragment of Staphylococcus aureus DNA encoding the alpha-hemolysin determinant was cloned from strain Wood 46 by inserting Sau3A-generated genomic DNA fragments between the BamHI sites of the lambda replacement vector L47.1. Phages expressing alpha-hemolysin were detected by overlaying plaques formed from several thousand independent recombinant phage with erythrocytes and looking for zones of hemolysis. One phage expressing alpha-hemolysin was purified and named lambda w alpha 3. This was subsequently shown to contain a 10.2-kilobase pair insert of S. aureus DNA. A 7.6-kilobase pair HindIII fragment encoding the alpha-hemolysin was subcloned from lambda w alpha 3 into the plasmid vector pACYC184 to form the hybrid plasmid pDU1148. Escherichia coli K-12 cells harboring pDU1148 synthesized a low level of alpha-hemolysin which remained associated with the cells and was not secreted into culture supernatants. When the same strain was stabbed onto blood agar plates, no zones of hemolysis were detected after overnight growth at 37 degrees C but hemolysis developed if the plates were left at room temperature for 48 h. By introducing specific deletions or Tn5 insertions into plasmid pDU1148, the alpha-hemolysin gene was mapped to a region within a 3.3-kilobase pair EcoRI-HindIII fragment which was subcloned onto the vector plasmid pBR322. A specific enzyme-linked immunosorbent assay with peroxidase-labeled rabbit anti-alpha-hemolysin antibodies was used to measure the levels of alpha-hemolysin antigen expressed in E. coli K-12 cells harboring pDU1148 or a variety of pDU1148::Tn5 and pDU1148 deletion mutants.     

Specific and cross-reacting antigens of Staphylococcus aureus of human and canine origins.

Biotype -specificity of Staphylococcus aureus of human and canine origins has been found to be associated with thermolabile agglutinogens represented in S. aureus strains 17 and 61218, respectively. Both strains also have exhibited a common thermostable antigen. On that basis, absorbed antisera have been developed for the differentiation of S. aureus of the two biotypes. In the present study, still another thermostable agglutinogen was established, shared by strain 17 and some S. aureus strains of canine origin, as represented by S. aureus strain 887. These findings led to modification and enhanced specificity of the serological method of distinguishing S. aureus of the human biotype from S. aureus of the canine biotype.     

Regulation of human neutrophil type 3 complement receptor (iC3b receptor) expression during phagocytosis of Staphylococcus aureus and Escherichia coli.

Human neutrophils (PMN) express a receptor for iC3b, a cleavage product of C3b. CR3 is an important receptor for phagocytosis of opsonized bacteria and its expression is enhanced by cell activation. We examined PMN CR3 expression during phagocytosis using flow cytometry and a CR3-specific monoclonal antibody. After 30 min phagocytosis of opsonized S. aureus and E. coli, CR3 expression increased to 151% and 221% of controls, respectively. Unopsonized S. aureus had no effect on CR3; however, unopsonized E. coli enhanced CR3 expression despite not being phagocytosed. Time-kinetic studies indicated a rapid initial fall in CR3 after addition of bacteria to PMN, followed by enhanced expression within 5-10 min. The initial fall in CR3 probably represented CR3 internalization rather than receptor destruction, as superoxide dismutase, catalase and protease inhibitors had no effect on this. Correlation of CR3 expression with the PMN oxidative response, measured with the intracellular fluorescent probe DCF-DA, demonstrated a dichotomy. Opsonized S. aureus and E. coli caused an oxidative response from PMN but unopsonized E. coli, which caused significant CR3 up-regulation, did not. CR3 up-regulation with unopsonized and opsonized E. coli was markedly inhibited by Polymyxin B, suggesting a role for endotoxin. These experiments indicate that CR3 expression can be regulated during phagocytosis, and the mechanisms responsible are distinct from those involved in the oxidative burst. CR3 up-regulation following exposure to bacteria in vivo may enhance neutrophil function at sites of infection.     

Participation of immunoglobulins and complement components in the intracellular killing of Staphylococcus aureus and Escherichia coli by human granulocytes.

Immunoglobulins and complement components are required for optimal ingestion and optimal killing of microorganisms by granulocytes. The degree of opsonization of microorganisms necessary for their ingestion was lower than that required for the killing of these bacteria during the ingestion phase. Killing during this phase was found to depend mainly on the presence of heat-labile opsonins, probably C3b, present on the microorganisms. Extracellular immunoglobulin G (IgG) and C3b were indispensable for optimal intracellular killing after ingestion was complete. This was established with an assay permitting assessment of the course of the number of viable intracellular bacteria independent of the ingestion of new live bacteria. Maximal intracellular killing by human granulocytes of ingested catalase-positive (Staphylococcus aureus and Escherichia coli) or catalase-negative (Streptococcus pyogenes and S. pneumoniae) microorganisms was found only when fresh serum was present extracellularly. Killing was suboptimal in the absence of serum. With heat-inactivated serum, the killing index lay between the indices obtained in the presence and absence of fresh serum. The stimulatory activity of heat-inactivated serum was most probably due to the interaction of IgG with the Fc receptor on the granulocyte membrane, since IgG subclasses IgG1 and IgG3 as well as pFc fragments of IgG stimulated the intracellular killing to the same degree as heat-inactivated serum did. In addition, (Fab1)2 fragments of IgG did not stimulate killing, and reduced killing was observed in the presence of heat-inactivated serum after reduction of the number of Fc receptors. The extra stimulation of the killing process in the presence of fresh serum compared with heat-inactivated serum was due to the interaction between membrane receptors and complement--most probably C3b generated by both the classical and the alternative pathways of complement activation. This conclusion is based on results obtained with sera in which one or both complement pathways were blocked, on the restoration of the killing-stimulatory activity of C3-deficient serum after addition of fresh C3, and on the reduced killing observed in the presence of fresh serum after reduction of the number of C3 receptors by the use of pronase or antigranulocyte serum.     

Enhancement of Escherichia coli and Staphylococcus aureus antibiotic susceptibility using sesquiterpenoids

The present work examines the potential of sesquiterpenoids to sensitize Escherichia coli and Staphylococcus aureus, and modulate their susceptibility to the standard antibiotics ciprofloxacin, erythromycin, gentamicin and vancomycin. It was tested samples of three sesquiterpenoids: guaiazulene, nerolidol (racemic mixture of the cis and trans isomers) and germacrene D enriched natural extract. Experiments were conducted aiming to assess the antimicrobial effects of the antibiotic-sesquiterpenoid combination on bacterial growth inhibition, by the disc diffusion assay and the minimum inhibitory concentration (MIC) assessment, the bactericidal effects, the post-antibiotic effect (PAE) and the effect on membrane permeability. The data related with the antimicrobial activity evidenced, through the disc diffusion assay, an antibiotic S. aureus antimicrobial activity enhancement by sesquiterpenoids presence. The MIC value for E. coli decreased significantly by sesquiterpenoids combination with ciprofloxacin, erythromycin and gentamicin, and for S. aureus, with all four selected antibiotics. This combination also increased the PAE, with the exception of guaiazulene, which seemed to quench antibiotic antimicrobial action. A moderate correlation between antimicrobial action and impairment of cell membrane function was detected for germacrene D enriched extract, and nerolidol, as single treatments and in combination with antibiotic, while a poor correlation was obtained for guaiazulene. This study provides basis for the evaluation of sesquiterpenoids as alternative or possible synergistic compounds for current antimicrobial chemotherapeutics, showing the practical utility of natural derived products to increase the susceptibility of E. coli and S. aureus.     

Clumping factor A mediates binding of Staphylococcus aureus to human platelets

The direct binding of bacteria to platelets may be an important virulence mechanism in the pathogenesis of infective endocarditis. We have previously described Staphylococcus aureus strain PS12, a Tn551-derived mutant of strain ISP479, with reduced ability to bind human platelets in vitro. When tested in an animal model of endocarditis, the PS12 strain was less virulent than its parental strain, as measured by bacterial densities in endocardial vegetations and incidence of systemic embolization. We have now characterized the gene disrupted in PS12 and its function in platelet binding. DNA sequencing, Southern blotting, and PCR analysis indicate that PS12 contained two Tn551 insertions within the clumping factor A (ClfA) locus (clfA). The first copy was upstream from the clfA start codon and appeared to have no effect on ClfA production. The second insertion was within the region encoding the serine aspartate repeat of ClfA and resulted in the production of a truncated ClfA protein that was secreted from the cell. A purified, recombinant form of the ClfA A region, encompassing amino acids 40 through 559, significantly reduced the binding of ISP479C to human platelets by 44% (P = 0.0001). Immunoprecipitation of recombinant ClfA that had been incubated with solubilized platelet membranes coprecipitated a 118-kDa platelet membrane protein. This protein does not appear to be glycoprotein IIb. These results indicate that platelet binding by S. aureus is mediated in part by the direct binding of ClfA to a novel 118-kDa platelet membrane receptor.     

Identification of a human lactoferrin-binding protein in Staphylococcus aureus.

Human lactoferrin (HLf) is an iron-binding protein with antimicrobial activity that is present in high concentrations in milk and various exocrine secretions. HLf is also an acute-phase protein secreted by polymorphonuclear leucocytes, and its binding to a large number of clinical isolates of Staphylococcus aureus has been described recently from our laboratory. We have now characterised the HLf-staphylococcal interaction in S. aureus strain MAS-89. The binding of 125I-HLf to strain MAS-89 reached saturation in less than 90 min and was maximal between pH 4 and 9. Unlabelled HLf displaced 125I-HLf binding. Various plasma and subepithelial matrix proteins, such as IgG, fibrinogen, fibronectin, collagen and laminin, which are known to interact specifically with S. aureus, did not interfere with HLf binding. A Scatchard plot was non-linear; this implied a low affinity (1.55 x 10(7) L/mol) and a high affinity (2.70 x 10(8) L/mol) binding mechanism. We estimated that there were c. 5700 HLf binding sites/cell. The staphylococcal HLf-binding protein (HLf-BP) was partially susceptible to proteolytic enzymes or periodate treatment and was resistant to glycosidases. An active HLf-BP with an apparent Mr of c. 450 Kda was isolated from strain MAS-89 cell lysate by ion-exchange chromatography on Q-sepharose. In SDS-PAGE, the reduced HLf-BP was resolved into two components of 67 and 62 Kda. The two components demonstrated a positive reaction with HLf-HRPO in a Western blot. These data establish that there is a specific receptor for HLf in S. aureus.     

Binding of Escherichia coli S fimbriae to human kidney epithelium.

Purified S fimbriae and an Escherichia coli strain carrying the recombinant plasmid pANN801-4 that encodes S fimbriae were tested for adhesion to frozen sections of human kidney. The fimbriae and the bacteria bound to the same tissue domains, and in both cases the binding was specifically inhibited by the receptor analog of S fimbria, sialyl(alpha 2-3)lactose. S fimbriae bound specifically to the epithelial elements in the kidneys; to the epithelial cells of proximal and distal tubules as well as of the collecting ducts and to the visceral and parietal glomerular epithelium. In addition, they bound to the vascular endothelium of glomeruli and of the renal interstitium. No binding to connective tissue elements was observed. The results suggest that the biological function of S fimbriae is to mediate the adhesion of E. coli to human epithelial and vascular endothelial cells.