Hemagglutination and adhesiveness of toxigenic Escherichia coli isolated from humans

Mechanisms of association of bacteria with mucosal surfaces

Bacterial association with host mucosal surfaces involves a large number of steps. Successful negotiation of each of these requires -- or is at least facilitated by -- the development of a distinct set of characteristics (virulence factors) by the bacterium. The major steps include: (a) chemotactic attraction of motile bacteria to the surface of the mucus gel, (b) penetration of and trapping within the mucus gel (which may be passive or can be promoted actively by bacterial motility and chemotaxis), (c) adhesion to receptors in the mucus gel or to mucosa-associated layers of the indigenous microflora, (d) adhesion to epithelial cell surfaces, and (e) multiplication of the mucosa-associated bacteria. Each reaction is further modified -- or reversed entirely -- by substances such as taxins, inhibitors of adhesion, and substrates for bacterial growth that are present in the mucosal microenvironment. Association with the mucosa is often important for bacterial colonization but can also lead to more effective elimination of the bacterium by the host. Bacteria lacking one or several of these virulence factors may still be successful colonizers if they show exceptionally high competence in relation to others. Examples are the strong adhesion to epithelial cells by Escherichia coli strains bearing the K88 antigen (such strains need not be motile in order to be pathogenic) or the active chemotactic association with mucus gel by cholera vibrios (some of which do not appear to adhere strongly to epithelial cells). Consequently a single in vitro assay for "adhesion" can be expected to correlate with bacterial pathogenicity only when the assay is based on the same specific mechanism(s) which the bacterium under study actually uses for mucosal association in vivo.

Genetic analysis of Escherichia coli K1 gastrointestinal colonization

Strains of Escherichia coli expressing the K1 polysaccharide capsule colonize the large intestine of newborn infants, and are the leading cause of Gram-negative septicaemia and meningitis in the neonatal period. We used signature-tagged mutagenesis (STM) to identify genes that E. coli K1 requires to colonize the gastrointestinal (GI) tract. A total of 2140 mTn5 mutants was screened for their capacity to colonize the GI tract of infant rats, and 16 colonization defective mutants were identified. The mutants have transposon insertions in genes affecting the synthesis of cell surface structures, membrane transporters, transcriptional regulators, enzymes in metabolic pathways, and in genes of unknown function, designated dgc (defective in GI colonization). Three dgcs are absent from the whole genome sequence of E. coli K-12, although related sequences are found in other pathogenic strains of E. coli and in Shigella flexneri. Additionally, immunohistochemistry was used to define the nature of the colonization defect in five mutants including all dgc mutants. STM was successfully applied to examine the factors involved in E. coli K1 colonization, and the findings are relevant to the pathogenesis of other enteric infections.

Simultaneous expression of CFA/I and CS3 colonization factor antigens of enterotoxigenic Escherichia coli by delta aroC, delta aroD Salmonella typhi vaccine strain CVD 908

Among the known colonization factors of enterotoxigenic Escherichia coli (ETEC), CFA/I and CS3 (the common antigen in the CFA/II family of fimbrial antigens) are two of the most prevalent fimbrial antigens found in clinical isolates but are never expressed by the same wild-type strain. We manipulated the genetic determinants encoding CS3 and CFA/I fimbriae so that these two important colonization factors are expressed simultaneously in attenuated Salmonella typhi live oral vaccine strain CVD 908, including after growth in liquid medium (CFA/I is poorly expressed by wild-type ETEC in broth culture). The recombinant fimbrial structures produced by CVD 908 are morphologically indistinguishable from the CS3 fibrillae and CFA/I rod-like fimbriae produced by ETEC, and are recognized by monospecific CS3 and CFA/I antibodies. This prototype construct may prove useful in investigating the live vector approach to immunoprophylaxis of ETEC diarrheal disease.

Detection of attachment of enterotoxigenic Escherichia coli (ETEC) to human small intestinal cells by enzyme immunoassay

Simple immunoassays were developed to study the binding between enterocytes of the small intestine and other cell types, and enterotoxigenic Escherichia coli (ETEC). CFA/I or CFA/II pilus protein or CFA-positive E. coli bacteria were immobilised in wells of microtitre plates and incubated with vesicles or crude mucus prepared from human brush border enterocytes. Binding of the cell preparations was detected by adding specific rabbit anti-brush border IgG followed by urease-labelled goat anti-rabbit IgG and urea substrate. The binding of purified CFA/I to human or rabbit small intestine, human oral epithelial cells or Caco-2 cells was detected with specific anti-CFA/I IgG. Both human brush border and mucus-derived preparations were able to attach to ETEC. The binding was CFA-specific and strong enough to withstand several washings. In contrast, CFA/I did not bind to small intestinal cells of non-human small intestinal origin, indicating that there may be important differences in affinity between receptors present on human small intestinal cells and cells of non-human small intestinal origin. Antibodies directed against human small intestinal and non-small intestinal cells did not cross-react with either preparation, indicating that receptors between these different cell sources are different. The EIA proved useful during the identification of a newly-recognised 15 kDa bacterial surface component of ETEC strain H10407P, which may function as a putative attachment factor. The EIAs developed in this study were easy to perform and multiple tests could be performed on small samples, including biopsy samples obtained during endoscopy.

In vitro adhesion of K88ab-, K88ac- and K88ad-positive Escherichia coli to intestinal villi, to buccal cells and to erythrocytes of weaned piglets

The phenotype of 21 weaned piglets, concerning adhesion of Escherichia coli possessing K88ab, K88ac or K88ad fimbriae to pig cells, was determined in an in vitro assay. Comparison was made with adhesion of these three K88 variant strains to buccal mucosal epithelial cells and to erythrocytes (haemagglutination) in the same piglets. Whereas adhesion of the three K88 variant strains to intestinal villi was piglet specific, buccal cell adhesion (BCA) and haemagglutination (HA) were not. The K88ab strain was weakly adhesive or non-adhesive in the BCA and negative in the HA test. K88ac strains consistently gave negative and K88ad consistently gave positive results in both assays. After washing the bacteria with phosphate-buffered saline, the K88ab strain revealed a positive HA test. Neither the BCA, nor HA test can be used to determine the pig intestinal adhesive phenotype.

Virulence characteristics of Aeromonas spp. in relation to source and biotype

The significance of Aeromonas spp. as potential water-borne enteric pathogens in Tasmania, Australia, an area with a mild climate and comparatively low year-round water temperatures, was investigated in view of the reported marked peak of Aeromonas-associated gastroenteritis in the summer and the apparent influence of temperature on levels of potentially pathogenic species in water supplies. Biochemical characteristics and virulence-associated properties--exotoxin production (hemolysin, enterotoxin), ability to grow at 43 degrees C, and possession of pili--were determined for 105 Tasmanian isolates of Aeromonas spp.; 43 isolates were from clinical specimens (greater than 75% diarrhea associated) and 62 were from water. Current classification schemes were evaluated for these isolates. A. sobria comprised 35% of the clinical isolates and 16% of the water isolates, A. hydrophila comprised 56 and 79%, and A. caviae comprised 9 and 5%. A total of 42% of the clinical isolates and 15% of the environmental isolates were enterotoxigenic (by the suckling mouse assay); these levels were significantly lower than those found in warmer environments. The majority (74%) of enterotoxigenic isolates were A. sobria. Enterotoxin-producing isolates possessed three or more of the following properties. They were Voges-Proskauer positive, did not hydrolyze arabinose, were positive for lysine decarboxylase, were able to grow at 43 degrees C, and produced large amounts of hemolysin (titer, greater than 128). Thus, the biochemical scheme proposed by Burke et al. (V. Burke, J. Robinson, H.M. Atkinson, and M. Gracey, J. Clin. Microbiol. 15:48-52, 1982) for identifying enterotoxigenic isolates appears to have widespread applicability. Environmental enterotoxigenic isolates possessed numerous pili, but these appeared to be lost once infection was established, as a similar isolates from patients with diarrhea were poorly piliated.

Enzyme-linked immunosorbent assay for adherence of bacteria to animal cells

Epithelial cells scraped from human oral mucosa and from pig intestines were immobilized onto the flat bottom surfaces of microtiter plates to study the adherence of various bacterial species to host cells. Bacterial adherence was quantitated either by an enzyme-linked immunosorbent assay technique with specific antibacterial serum as the first antibody followed by peroxidase-conjugated second antibody or by using biotinylated bacteria and avidin-peroxidase as the detecting agent. Unlabeled Escherichia coli and purified E. coli 987P fimbriae inhibited the adherence of biotinylated E. coli to immobilized enterocytes. The adherence of a mannose-sensitive strain of E. coli to immobilized oral epithelial cells was inhibited by mannose derivatives. The adherence of fimbriated E. coli 987P to immobilized enterocytes was approximately four times higher than the adherence of a nonfimbriated variant of the same strain. The adherence of Streptococcus pyogenes to oral cells was detected in the range of 10 to 150 bacteria per cell and was inhibited by lipoteichoic acid and albumin. The data suggest that the putative receptors which bind bacteria on the immobilized cells retain a functional form similar to that of native cells in suspension. The proposed adherence assay is easy to perform, allows the detection of specific adherence of test bacteria, and provides objective quantitation of adherence with a sensitivity of 10 bacteria per cell. Most importantly, the assay allows the testing of many variables in the same day.

Enterotoxigenic Escherichia coli in a population of infants with diarrhea in Chile

The incidence of enterotoxigenic Escherichia coli (ETEC) was investigated in 95 E. coli strains isolated from 48 infants with diarrhea in Santiago, Chile. By using standard biological assays and DNA-DNA hybridization procedures, ETEC was found in 31.2% of the cases: 14 strains produced heat-stable enterotoxin (ST) only, three strains produced heat-labile enterotoxin (LT) and ST, and two strains produced LT only. DNA probes detected all enterotoxin producers except one ST-producing strain. The ST strains hybridized with one or both of the human ST probes (ST Ib and ST A2). Two of the LT-ST strains hybridized with the ST Ia and ST Ib probes, and the third strain did not hybridize with any of the ST probes. Only the ST group expressed multiple resistance (85.7%) and colonization factor antigen I (CFA I) (92.8%); CFA II was found in two of three LT-ST strains. The O153:H45 serotype was found in 10 of 14 ST strains, and O6:K15:H16 was found in one LT strain and in two LT-ST strains. These findings suggest that ETEC, especially strains that produce ST, may be an important cause of diarrhea among Chilean infants.

Frequency and properties of naturally occurring adherent piliated strains of Haemophilus influenzae type b

We found that 41 of 75 (55%) children with Haemophilus influenzae type b disease (70 cases of meningitis, 2 of cellulitis, 2 of septic arthritis, and 1 of epiglottitis) and 2 of 120 (1.7%) children with upper respiratory infection were colonized with H. influenzae type b in the nasopharynx (NP). Of these 43 NP strains from children with systemic H. influenzae type b disease, 7 (16%) adhered to human buccal epithelial cells. The strains isolated from the systemic site of all children, including children from whose NP adherent bacteria were isolated, did not adhere to buccal epithelial cells in vitro. Each adherent NP strain had biotype (I), serotype (b), and antibiotic susceptibility (sensitive) similar to that of the corresponding nonadherent systemic isolate. With one exception, all NP-systemic pairs had similar major outer membrane proteins. Six of the seven NP strains had a protein band in the whole cell lysate preparation with a molecular weight between 22,000 and 23,000, which could not be seen in the nonadherent cerebrospinal fluid strains. Electron micrographs of all adherent strains showed that more than 95% of the organisms examined were highly piliated, whereas the nonadherent strains were not piliated. All piliated strains agglutinated human erythrocytes. Adherence to buccal epithelial cells and agglutination of erythrocytes could not be blocked by mannose or alpha-methyl-D-mannoside. We speculate that piliation is not important for NP colonization by H. influenzae type b and that the loss of pili may be required for host invasion.

Infectious diarrhea. Pathogenesis and risk factors

Our understanding of the pathogenesis of infectious, especially bacterial, diarrhea has increased dramatically. New etiologic agents, mechanisms, and diseases have become known. For example, Escherichia coli serogroup O157 is now known to cause acute hemorrhagic colitis. Also, E. coli serogroups that produce Shiga toxin are recognized as etiologic agents in the hemolytic-uremic syndrome. The production of bacterial diarrhea has two major facets, bacterial-mucosal interaction and the induction of intestinal fluid loss by enterotoxins. Bacterial-mucosal interaction can be described in stages: (1) adherence to epithelial cell microvilli, which is often promoted by or associated with pili; (2) close adherence (enteroadherence), usually by classic enteropathogenic E. coli, to mucosal epithelial cells lacking microvilli; and (3) mucosal invasion, as with Shigella and Salmonella infections. Further large strides in understanding infectious diarrhea are likely with the cloning of virulence genes if additional host-specific animal pathogens become available for study.

Colonization antigens and haemagglutination patterns of human Escherichia coli

The haemagglutinating properties of 223 (35 enterotoxigenic and 188 non-enterotoxigenic) Escherichia coli strains with nine erythrocyte types were investigated; 153 strains were also tested for beta-haemolysis and colicin production and for the presence of CFA/I, CFA/II, K88 and K99 antigens. A selected group of strains was also examined by electron microscopy to determine the presence of fimbriae or fibrils and to establish the relationship between these, the haemagglutinating properties and the presence of colonization antigens. Generally, the haemagglutinating patterns yielded by the same strains grown in Mueller Hinton broth and on CFA agar differed considerably. Mannose-sensitive haemagglutinating (MSHA) patterns were more homogeneous than mannose-resistant haemagglutinating (MRHA) patterns. Forty-seven percent of the non-enterotoxigenic MRHA+ strains were haemolytic while only 6% of the remaining strains were (chi2 correction = 34.01; p less than 0.001). CFA/I was only detected in the four enterotoxigenic MRHA+ strains which were positive only with human and calf erythrocytes when grown on CFA agar. CFA/II was detected in three of a total of six enterotoxigenic strains which were MRHA+ only with calf erythrocytes when grown on CFA agar. K88 and K99 antigens were not detected. All strains in which bacteria with fimbriae or fibrils were observed showed haemagglutinating activity. Thus, 18 (66.7%) of the 27 haemagglutinating strains grown on CFA agar showed fimbriae or fibrils while none of the 19 non-haemagglutinating strains did (chi2 correction = 18.10; p less than 0.001).

In vitro adhesion of enterotoxigenic Escherichia coli to human intestinal epithelial cells from mucosal biopsies

An adhesion assay with isolated human enterocytes prepared from duodenal biopsies has been developed and tested by using human enterotoxigenic Escherichia coli expressing colonization factor antigens I and II (CFA/I and CFA/II) and type 1 fimbriae. Enterotoxigenic E. coli strains H10407 (CFA/I) and B2C (CFA/II) bound to duodenal enterocytes to a much greater extent (mean of 4.6 and 4.0 bacteria per brush border) than did strain H10407P, a CFA/I- mutant of H10407 (mean of 0.1 bacteria per brush border). Type 1 fimbriae also promoted adhesion of strain H10407P to duodenal enterocytes but attachment was to basolateral rather than brush border surfaces. CFA/I and CFA/II, on the other hand, promoted adhesion only to human enterocyte brush borders.

Attachment of enterotoxigenic Escherichia coli to human intestinal cells

An adherence test was developed with human ileal cells isolated from patients with a long-standing ileostomy by saline lavage through the stoma. Enterotoxigenic Escherichia coli isolated from humans bound to the human ileal cells to a greater extent (1.2 x 10(6) bacteria bound) than did control E. coli strains, including K99 pili antigen-producing strains (whether originally isolated from pigs or calves), the rabbit pathogen RDEC-1, the laboratory-derived nontoxigenic strain 334LL, or human normal fecal strains. However, K88 strains, either K88ab or K88ac, bound to the human ileal cells as well as did enterotoxigenic E. coli. Ileal cells isolated from two donors with different blood types behaved similarly. These cells remained viable and retained their binding ability for at least 3 days when stored in tissue culture medium at 4 degrees C.

Hemagglutination and adhesiveness of epidemiologically distinct strains of Vibrio parahaemolyticus

Twelve strains of Vibrio parahaemolyticus from four epidemiologically distinct groups were examined for their ability to hemagglutinate human, bovine, chicken, guinea pig, and rabbit erythrocytes and to adhere to human buccal mucosal epithelial cells in the presence and absence of mannose. Four of six Kanagawa-positive but none of six Kanagawa-negative strains showed mannose-sensitive hemagglutination with erythrocytes of rabbits and of one or more additional species. Mannose-resistant hemagglutination was shown by one strain in each group with no apparent relationship to strain source or hemolytic capability. All strains adhered to human buccal mucosal cells, with but a single strain showing significant difference in adherence at the alpha = 0.05 level. The adherence pattern had no relationship to the four epidemiological groups. Although adhesive processes may well be involved in disease caused by V. parahaemolyticus, our results do not support a role for adherence as a predictor of pathogenicity.

Aeromonas sobria in chlorinated drinking water supplies

Aeromonas species were recovered from over 27% of 183 chlorinated drinking water samples collected during an 18-month period. Sixteen of 20 isolates tested elicited a cytotoxic response by Y-1 mouse adrenal cells. None of the strains was either enterotoxigenic by the rabbit ligated ileal loop assay, exhibited piliation, or showed significant mannose resistant adherence to human buccal cells. TheAeromonas isolates were further identified to beA. sobria and were resistant to ampicillin and susceptible to chloramphenicol, kanamycin, streptomycin, and tetracycline. Total coliform levels did not correlate withAeromonas densities in distribution water. With 85% of the samplings,Aeromonas occurred in distribution water when no coliforms were detectable by either the membrane filter or most-probable-number techniques. A significant correlation (P<.001) existed between standard plate count levels andAeromonas.

Host-specific fimbrial adhesins of noninvasive enterotoxigenic Escherichia coli strains

Images

Anatomical and immunological responses of rabbit gallbladders to bacterial infections

To study the sequential morphological and immunological response of the rabbit gallbladder to bacterial infection and to compare the inflammatory responses with different pathogens, gallbladders were infected with Streptococcus faecalis and two strains of Escherichia coli, one of which produced enterotoxin. Gallbladder infection was produced either by intravenously injecting bacteria into rabbits with a small liver infarct or by injecting bacteria directly into the gallbladder of normal rabbits. The percentage of gallbladders infected intravenously with a nonenterotoxigenic E. coli strain was 86% at 1 week, 70% at 3 weeks, and 15% at 6 weeks. Epithelial necrosis and leukocyte infiltration were prominent 1 week after infection. At 3 and 6 weeks after infection, there was crypt distortion and increased mucus secretion in the epithelium as shown by periodic acid-Schiff staining. The lamina propria was infiltrated with mononuclear cells, many of which were plasma cells. Myofibroblasts (contractile fibroblasts) were also identified on transmission microscopy, In addition to these changes, toxigenic E. coli produced subepithelial capillary dilation in the villus core. Morphological changes (excluding toxin-associated changes) were related to the duration of infection rather than to the specific species of infecting bacteria. Infected gallbladders studied by immunofluorescence had a greater than 50-fold increase in plasma cells compared with control cells. In addition, the number increased with the duration of infection. Immunoglobulin A cells were the major cell type in gallbladders infected by intravesical injection, whereas immunoglobulin G cells predominated in gallbladders infected intravenously. The gallbladder appears to mount a local immune response to bacterial infection.

An assay for measuring specific adhesion of an Escherichia coli strain to tissue culture cells

Escherichia coli SS142 has been found to adhere specifically to the human epithelioid tissue culture cell line Intestine 407, but not to other tissue culture cells. This paper describes an accurate, reproducible and objective method of assessing the rate of adhesion of radiolabelled bacteria to these cellular monolayers. Adhesion was found to be linear with time for 60 min and with bacterial concentrations up to 10(9) bacteria/ml. The binding appeared to be irreversible. Adhesion was not affected by changes in the composition of the medium, its pH or ionic strength, or by the assay temperature within physiological limits, but was diminished at very high ionic strength or low temperature. It increased with increasing cell density of the monolayers. Under appropriate conditions the assay could be used for comparative determinations of the rate of adhesion of different, or differently treated bacteria.

Adherence of Haemophilus influenzae to buccal epithelial cells

The role of adherence of Haemophilus influenzae to epithelial surfaces in the pathogenesis of infection is unknown. Fluorescent-antibody and radiolabeled adherence methods were adapted to study H. influenzae adherence to human buccal epithelial cells. By the fluorescent-antibody method, 19 of 21 (90%) nontypable H. influenzae strains were found to be adherent compared with 2 of 42 (5%) type b strains (P less than 0.0001). Using a radiolabeled adherence method, we found that 9 of 12 (75%) nontypable H. influenzae strains were adherent to buccal epithelial cells whereas only 3 of 32 (9%) type b strains were adherent (P = 0.001). Results of H. influenzae adherence examined by both methods correlated significantly (P = 0.01). H. influenzae adherence to adult pharyngeal, nasal, and buccal epithelial cells was comparable. Type b H. influenzae did not adhere to the buccal epithelial cells of well children, children with H. influenzae type b disease, or children with upper respiratory infections. In contrast, nontypable H. influenzae did adhere to the buccal epithelial cells of well children and children with upper respiratory infections. These observed in vitro differences in adherence between nontypable and type b H. influenzae strains may explain differences in colonization, pathogenesis, and types of infection due to nontypable and type b H. influenzae.

Participation of pili and cell wall adhesion in the yeast agglutination activity of Escherichia coli

Escherichia coli strain 2699 (O6:K13) which had been isolated from a case of urinary tract infection exhibited pili during the stationary phase (24 to 40 h), but not during the exponential phase (4 h), when grown in static broth culture. The bacteria were also piliated when grown for 24 h on agar. They agglutinated Saccharomyces cerevisiae (baker's yeast) in the piliated as well as in the nonpiliated state. The agglutinations were mannose sensitive, i.e., they could be inhibited with 50 mM methyl-alpha-mannoside. The bacteria were first depiliated by shearing and then used for the isolation of outer membrane vesicles with an Omnimixer. Purified pili and outer membranes were characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electron microscopy. The pili could be demonstrated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis only after treatment at low pH or in saturated guanidine hydrochloride which is typical of the common type 1 pili. Depiliated bacteria, purified pili, and purified outer membranes gave mannose-sensitive agglutination of S. cerevisiae. The findings are discussed with respect to possible mechanisms of cell agglutination.

Serotypes of attachment pili of enterotoxigenic Escherichia coli isolated from humans

Pili from enterotoxigenic Escherichia coli isolated from humans have been partially purified, and antisera have been prepared. These pili were initially attached to erythrocytes and then removed by thermal elution for purification. Three distinct antigenic types of pili have been identified. Antisera against these three pili types reacted with 60 of 106 (56%) enterotoxigenic E. coli isolated from humans but not with nontoxigenic, normal human fecal isolates of E. coli nor with enterotoxigenic E. coli strains isolated from animals. There was no correlation between pili serogroup and any of the following toxin production (heat labile, heat stable, or both), O antigenic type, geographical source of isolation, or mannose-resistant hemagglutination patterns of various erythrocyte types.

Association of Aeromonas sobria with human infection

Fifteen Aeromonas isolates from various human infections and nine isolates from polluted water were identified as either Aeromonas hydrophila or Aeromonas sobria and examined for cytotoxigenicity, enterotoxigenicity, adherence to epithelial cells, and other virulence-associated factors, including proteases, lipases, elastases, and hemolysins. Two groups of organisms (I and II) were distinguishable based on differences in median lethal doses in mice and cytotoxicity for Y-1 adrenal cells. Group I clinical and environmental strains had median lethal doses of less than 10(7) colony-forming units, were cytotoxic, frequently possessed several virulence-associated factors, and had lysine decarboxylase-positive or Voges-Proskauer-positive phenotypes or both. Piliation of Aeromonas was associated strongly with ability to adhere to human buccal cells, and these characteristics were associated with group I strains. Group II clinical and environmental strains had median lethal doses of greater than or equal to 10(7) colony-forming units, were not cytotoxic, and usually were lysine decarboxylase negative or Voges-Proskauer negative or both. Clinical strains in group II exhibited enterotoxigenicity, which was not detected in group II environmental strains. A sobria was more frequently associated with human infections; 13 of the 15 clinical strains were A. sobria, and 2 were A. hydrophila. On the other hand, the majority of the environmental strains (seven of nine) were A. hydrophila.

Enterotoxins from gram-negative bacteria relevant for veterinary medicine

The chemistry, mechanism of action, assay methods, pharmacology, and prevention and treatment of diarrhoea due to toxins of gram-negative microbes are discussed. Other virulence factors are mentioned briefly. Special emphasis is placed on non-specific treatment by oral rehydration.

Attachment of bacteria to mammalian surfaces

Much attention has been devoted to the study of bacterial adherence to mammalian surfaces in vitro during the past several years. Some in vivo evidence also suggests that this process may indeed be an integral part of the pathogenesis of colonization and certain infections. The biochemical basis of attachment and definition of the actual receptor sites involved are just starting to become known and seem to be different amongst individual bacteria genera. However, pili may mediate attachment of a variety of gram-negative organisms to receptor cells, and streptococcal lipoteichoic acids probably serve a similar function. Some recent study methods and results in this field are reviewed.

Attachment pili from enterotoxigenic Escherichia coli pathogenic for humans

Pili from enterotoxigenic Escherichia coli pathogenic for humans have been isolated by adsorption to the surface of erythrocytes followed by thermal elution. The pili are composed of two protein subunits with molecular weights of 13,100 and 12,500 as determined by sodium dodecyl sulfate-gel electrophoresis. These pili also bind to human buccal cells under temperature conditions (37 degrees C) which prevent the binding of these pili to the erythrocytes. Analogous temperature effects on binding have previously been observed with whole bacterial cells. This binding can be inhibited by antiserum prepared against the isolated pili.