UUL 012004 [9129A OVERDUE F IRES: 25¢ per W per 1:- RETURNIKB LIBRARY MTERIALS: Place in book return to remove charge fro. circulation records 8112079 FARDIAZ, SRIKANDI HEMAGGLUTINATION AND ADHESIVE PROPERTIES OF ESCHERICHIA COLI ISOLATED FROM DISEASED TURKEYS Michigan State University PHD. ' 1980 University Microfilms International 300 N. Zeeb Road, Ann Arbor, MI 48106 HEMAGGLUTINATION AND ADHESIVE PROPERTIES OF Escherichia coli ISOLATED FROM DISEASED TURKEYS BY Srikandi Fardiaz A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science and Human Nutrition 1980 ABSTRACT HEMAGGLUTINATION AND ADHESIVE PROPERTIES OF Escherichia coli ISOLATED FROM DISEASED TURKEYS BY Srikandi Fardiaz This investigation was carried out to isolate and characterize strains of E. gel; from diseased turkeys, and to study the hemagglutination activities (HA) and the adhe- sive properties of the isolates with respect to turkey small intestinal tissue. Hemagglutination and in zitrg ad- hesion tests were done using unheated cultures and cultures that had been heated at 650C for 30 minutes. The ability of various sugars and sugar derivatives, concanavalin A, sodium metaperiodate, and isolated pili to act as inhibi- tors of the in vitrg adhesion was also studied. The strains of E. ggli_isolated from diseased tur- keys represented three serotypes, Ol:Kl, 02:Kl and O78:K80. The cultures tested possessed a mannose-sensitive hemagglu- tinin. Cultures 9 and 119-3 exhibited the HA type III (NNSSS), while culture 94-5 exhibited the HA type IV (SNSSS). Growth at 18°C and heating the cultures at 65°C Srikandi Fardiaz for 30 minutes did not affect the hemagglutination activi- ties, but these treatments affected the adhesive proper- ties. During incubation at room temperature the numbers of bacteria adhering to turkey small intestinal tissue in- creased by ca. 1.25 to 1.75 log cycles for cultures 9 and 119-3, and by ca. 1.0 log cycle for culture 94-5. The presence of D-mannose and a-methyl-D-mannoside at 0.4 and 0.5 percent, respectively, reduced the numbers of bacteria adhering to the intestinal tissue by 1.0 log cycle. Adhesion was completely inhibited by concanavalin A and sodium metaperiodate at concentrations of 0.2 and 0.1 percent, respectively. While isolated pili were less in- hibitory than D-mannose, a-methyl-D-mannoside, concanavalin A, and sodium metaperiodate, adhesion was substantially in- hibited by isolated pili at a concentration of 2.0 mg pro- tein per ml. The approximate molecular weight of protein from the isolated pili was 34,800. To my husband, Dedi, and my daughter, Miri FI— ACKNOWLEDGMENTS The author would like to thank her major professor, Dr. K.E. Stgzggson, for his guidance during the course of the investigation. Thanks are also expressed to the members of my graduate committee: Dr. G.R. Carter, Dr. L.E. Dawson, Dr. J.R. Brunner, and Dr. E.S. Beneke. Special thanks are extended to Dr. K.K. Baker for her help in electron microscopy, and to Marguerite Dynnik for her assistance in the laboratory. The cooperation and help from Bil-Mar Foods Inc., especially Mr. Bill Stout, and the financial support given by MUCIA-AID - Indonesian Higher Agricultural Education Project, are acknowledged. The author is deeply grateful to her parents, her husband, and her daughter for their constant support and encouragement during the course of her studies. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . Pili . . . . . . . . . . . . . . Pili as Adhesive Factors . . Pilus Antigens . . . . . . . Adhesive Properties of E. coli . Adhesion to Human and Animal Factors Affecting Adhesion . Hemagglutination (HA) Activities 0 Mannose-Sensitive (MS) HA . Mannose-Resistant (MR) HA . Cells . . f E. coli Nature, Formation, and Isolation of Pili . Nature and Composition of Pil Methods for Isolation of Pili MATERIALS AND METHODS . . . . . . . . Sources of E. coli Isolates . . . Isolation of E. coli . . . . . . i O O O 0 Biochemical Identification of E. coli . . . iv Page Vii viii 10 10 ll 13 13 15 16 l6 l6 l7 Serological Identification of E. coli . . . O Serotype . . . . . K Serotype . . . . . Electron Microscopy . . . Hemagglutination Tests . In Vitro Adhesion Tests . Inhibition of In Vitro Adhesion Isolation of Pili . . . . Extraction and Precipitation Gel Chromatography . Gel Electrophoresis Quantitation of Protein RESULTS . . . . . . . . . . . Isolation and Identification of Isolated Cultures . O and K Serotypes . Electron Microscopy . . . Hemagglutination Patterns In Vitro Adhesion . . . . Inhibition of In Vitro Adhesion Inhibition by Sugars and Sugar Derivatives . . . Inhibition by Concanavalin A Sodium Metaperiodate Inhibition by Isolated Pili Isolated Pili . . . . . . DISCUSSION . . . . . . . . . . and E. coli Serotypes Found in Diseased Size of E. coli Pili . . Poultry Page 18 18 21 22 23 25 27 28 28 29 32 33 36 36 36 38 41 41 49 61 61 66 66 70 75 75 76 Page Hemagglutination Activities of E. coli . . . . 76 Adhesive Properties of E. coli . . . . . . . . 82 Inhibition of In YIEEQ Adhesion . . . . . . . . 86 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . 90 REFERENCES 0 O O O O O O O O O O O O O O O O O 0 APPENDIX . . . . . . . . . . . . . . . . . . . . . . 102 vi 10. 11. Al. LIST OF TABLES Identification of E. coli by Fisher Entero- setTMzo O O O O O O O O O O O I O O O O O Descriptions of plates inoculated from dis- eased turkeys . . . . . . . . . . . . . . Entero-SetTMZO numbers of isolates obtained from diseased turkeys . . . . . . . . . . O and K titers, and immunodiffusion (ID) of isolates . . . . . . . . . . . . . . . . Appendages of selected isolates, including the approximate diameter of pili . . . HA of different species of red blood cells by strain H10407 (CPA/1+) grown at 37°C HA of different species of red blood cells by strain Moon 263 (K88ab+) grown at 37°C HA of different species of red blood cells by culture 9 . . . . . . . . . . . . . HA of different species of red blood cells by culture 119-3 . . . . . . . . . . . HA of different species of red blood cells by culture 94-5 . . . . . . . . . . . . . HA of different species of red blood cells by culture 102-1 . . . . . . . . . . . . Relative mobilities of standard proteins and pilus protein on SDS-gels at 7.5 and 10.0 percent gel concentrations . . vii Page 19 37 39 40 42 43 44 46 47 48 50 106 LIST OF FIGURES Figure Page 1. Flow chart of the E3 Vitro adhesion tests . . . 26 2. Flow chart of the extraction of pili from bacterial cells . . . . . . . . . . . . . . . 30 3. Flow chart of the precipitation of pili . . . . 31 4. Standard curve for protein assayed by the Lowry method . . . . . . . . . . . . . . 35 5. Adhesion of strain K-12 to turkey small intestinal tissue . . . . . . . . . . . . . . 52 6. Adhesion of culture A20a (02:Kl:H6) iso- lated from human to turkey small intes- tinal tissue . . . . . . . . . . . . . . . . 53 7. Adhesion of culture 9, grown at 37°C, to turkey small intestinal tissue . . . . . . . 55 8. Adhesion of culture 119-3, grown at 37°C, to turkey small intestinal tissue . . . . . . 56 9. Adhesion of culture 9, grown at 18°C, to turkey small intestinal tissue . . . . . . . 57 10. Adhesion of culture 119-3, grown at 18°C, to turkey small intestinal tissue . . . . . . 58 11. Adhesion of culture 94-5, grown at 37°C, to turkey small intestinal tissue . . . . . . 59 12. Adhesion of culture 94-5, grown at 18°C, to turkey small intestinal tissue . . . . . . 60 13. Adhesion of culture 102-1, grown at 37°C, to turkey small intestinal tissue . . . . . . 62 viii Figure 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Adhesion of culture 102-l,grown at 180C, to turkey small intestinal tissue . . . Inhibition of E. coli (culture 119-3) ad- hesion to tuEkey small intestinal tis— sue by sugars after 50 minutes of incu- bation . . . . . . . . . . . . . . . . Inhibition of E. coli (culture 119-3) ad- hesion to turkey small intestinal tis- sue by sugar derivatives after 50 min- utes of incubation . . . . . . . . . . Inhibition of E. coli (culture 119-3) ad— hesion to turkey small intestinal tis- sue by D-mannose and a-methyl—D—man- noside after 50 minutes of incubation Inhibition of E. coli (culture 119-3) ad- hesion to turkey small intestinal tis- sue by concanavalin A and sodium meta- periodate after 50 minutes of incuba— tion . . . . . . . . . . . . . . . . . Inhibition of E. coli (culture 119-3) ad- hesion to turkey small intestinal tis- sue by isolated pili after 50 minutes of incubation . . . . . . . . . . . . Elution pattern obtained from the chroma- tography of ammonium sulfate precipi- tate (culture 119-3 pili) on a Sephadex G-50 column . . . . . . . . . . . . . . SDS-gels of (A) Ammonium sulfate precipi- tate of pili; (B) Pili after chromato- graphy on a Sephadex G-50 column, (C) Standard proteins . . . . . . . . . . . Standard curve for the determination of approximate molecular weight of culture 119-3 pili by SDS-gel electrophoresis . Electron micrographs of (A) Ammonium sul- fate precipitate of culture 119-3 pili; (B) Pili after chromatography on a Se- phadex G-50 column; Concentration, 10 mg/ml; Magnification, x 64,000 . . . . ix Page 63 64 65 67 68 69 71 72 73 74 Figure Page 24. Schematic diagram representing interrela- tionship between ETEC, EPEC, Shigella— like E. coli, and non-EPEC (Evans et gl., I979) . . . . . . . . . . . 7—. . . . . 78 Al. Entero-SetTMZO identification system, using culture 119-3 . . . . . . . . . . . . . 102 A2. Electron micrograph of negatively stained cells of E. coli (culture 9) grown on Peptone agar at 37°C; Magnification, x 103,000 . . . . . . . . . . . . . . . . . . 103 A3. Electron micrograph of negatively stained cells of E. coli (culture 94-5) grown on Peptone agar at 37°C; Magnification, x 103,000 . . . . . . . . . . . . . . . . . . 104 A4. Electron micrographs of negatively stained cells of E. coli (culture 119-3) grown on Peptone agar at 37°C; Magnification, (A) x 60,000, (B) x 103,000 . . . . . . . . . 105 INTRODUCTION Escherichia coli is a lactose-fermenting fecal bac- terium in the family Enterobacteriaceae. It is a small gram-negative rod, aerobic and facultatively anaerobic, either nonmotile or motile with peritrichous flagella. E. 99;; is one of the pathogenic organisms which causes heavy losses in the poultry industry. Certain serotypes of E. 991$ have been reported to be associated with diseases in poultry (Sojka and Carnaghan, 1961; Harry, 1964; Hemsley and Harry, 1965; Heller and Smith, 1973). Pathogenicity of E. ggll in humans and animals is associated with the abili- ty of the bacteria to adhere and colonize in the small in- testine. Adhesion of E. 321$ to intestinal epithelial cells is suggested to be mediated by pili (Jones and Rutter, 1972; Isaacson g£_§E., 1977, 1978; Moon 33 §l°v 1977; Nagy g5 21" 1977). Pili also adhere to red blood cells and cause hemagglutination (Duguid 33 21" 1955). Most enterotoxigenic strains of E. 99;; (ETEC) iso- lated from humans and animals have been found to possess specific pilus antigens, which are also mannose-resistant (MR) hemagglutinins (Jones and Rutter, 1974; Burrows 2E 31" 1976; Evans 3E 21°! 1977; Evans and Evans, 1978). These pilus antigens, which enable the bacteria to attach to the epithelial cells and colonize in the small intes- tine, are considered to be one of the primary determinants of virulence. However, some enterotoxigenic strains from neonatal pigs do not produce any of the known MR hemagglu- tinins (Moon 33 Sl‘! 1980), and some human enteropatho- genic E. 99;; (EPEC) produce only mannose-sensitive (MS) hemagglutinins (Evans g5 EE., 1979). A study by Duguid g3 gi. (1979) showed that many strains of E. 99;; possessed both MS and MR hemagglutinins. The adhesion of some E. Egli strains to human and animal cells has been studied. It has been found that bac- terial binding can occur via pili to the carbohydrate resi- dues of the glycoproteins on mammalian cell surfaces (Jones and Freter, 1976; Salit and Gotschlich, 1977; Ofek 3E 3E., 1977, 1978). This study was carried out to observe the hemagglutination and adhesive properties of E. 99;; strains (that possessed pili) which were isolated from diseased turkeys in Michigan. The hemagglutination pat- terns of the E. 92;; strains to human and animal red blood cells, the characteristics of the 12.21332 adhesion of E. 32;; to turkey small intestinal tissue, and the inhibi- tory effects of certain carbohydrates on the £3 XEEEQ adhe- sion, were determined. The inhibition of adhesion was also tested using isolated pili as an inhibitor. LITERATURE REVIEW Pili Pili are nonflagellar filamentous appendages that are produced by many gram-negative bacteria. The name pili (=hair) was proposed by Brinton (1959), whereas the Latin name fimbriae (=fibers) was given to such structures by Duguid E; El. (1955). The classification of pili has been based on their electron microscopy, function, and the hemag- glutination of human and animal red blood cells by bacte- ria possessing the pili. Brinton (1965) distinguished six types of pili (l to 5, and F pili), and Duguid 32 El' (1966) named F pili and six types of common pili in gram- negative bacteria. Ottow (1975) differentiated six groups of pili and four subtypes within the first group. F pili are sex pili which function in the transport of nucleic acid during bacterial conjugation and male-specific phage infection. Pili as Adhesive Factors The type 1 pili, found in most strains of E. coli are responsible for adhesion of bacteria to fungal, plant, and animal cells, including red blood cells, and epithelial cells of the alimentary, respiratory, and urinary tracts (Duguid, 1968). Type 1 pili are also associated with the pellicle-forming property (Duguid, 1968; Old EE.§l-' 1968; Ottow, 1975). Type 1 pili from different E. 99;; strains are closely related (Nowotarska and Mulczyk, 1977). Fur- ther classification of pili could be done according to their distinctive physical, chemical, functional, and anti- genic characteristics. Pili of types 2 to 6 are dis- tinguished partly by their adhesiveness or lack of adhe- siveness for a particular surface (Duguid, 1968). Pilus Antigens The groups of pili which facilitate intestinal col- onization by ETEC in humans and animals consist of several antigenic types. These pilus antigens can be used in de- tecting ETEC infections because they occur more commonly in ETEC than in non-ETEC. One of the pilus antigens present in E. gel; is the K88 antigen, described by ¢rskov gE_§E. (1964) and Stirm g3 EL. (1967) in enterotoxigenic strains isolated from swine with diarrhea. This antigen is plasmid determined (Orskov and ¢rskov, 1966). Another pilus anti- gen is the K99 antigen, found in E. gglE_strains isolated from newborn calves and lambs with diarrhea (¢rskov g3 3E., 1975; Burrows E; El., 1976; Guinée 33 3E., 1976; Smith and Huggins, 1978), and also in pigs (Moon 33 21" 1977). A pilus antigen known as 987P, which does not agglutinate animal red blood cells, has been demonstrated also in strains isolated from pigs (Isaacson g2 21" 1977; Nagy g; 21" 1977). Some ETEC have been found that are entero- pathogenic for neonatal pigs but do not produce K88, K99 or 987P (Moon 3E 31-! 1980). Two different pilus antigens known as colonization factor antigens, CFA/I (Evans 3E 21" 1975, 1977, 1978; ¢rskov and ¢rskov, 1977), and CFA/II (Evans and Evans, 1978; Evans 35 3E., 1979), have been de- tected in some human enterotoxigenic strains. Another pi- lus antigen, called F7 antigen (Orskov SE 21" 1980), was found in a strain of E. 99;; isolated from human urinary tract infections. Adhesive Properties of E. coli Adhesion to Human and Animal Cells Adhesion to human and animal cells is a factor in the pathogenicity of E. ggli. Pathogenicity is the ability of an organism to cause disease in animal or plant hosts. To produce disease a pathogenic organism must accomplish the essential steps: enter the host, multiply in the host, resist or do not stimulate host defenses, and damage the host. The process in each step is complex, therefore, several determinants are usually involved in the overall process. Lack of one determinant may result in attenuation of a strain. It means that a component essential for viru- lence may also be present in a strain that is attenuated due to lack of some other determinants (Smith and Pearce, 1972). The determinants of virulence may not be expressed completely by microorganisms grown E3 XEEEQ' It has been recognized that bacterial virulence may be maintained or increased during growth E5 2139' but can be lost on subcul- ture. The conditions of growth E2 yiyg may influence the production of components necessary for pathogenicity (Smith, 1964; Smith and Pearce, 1972). Most infections caused by microorganisms begin on the mucous membranes of the respiratory, alimentary, and urinary tracts. These membranes are protected by moving lumen contents, surface mucus, and often by commensal mi- croorganisms (Smith, 1976; Smith and Pearce, 1972). There are at least three types of early attack on mucous mem- branes: adhesion and multiplication without significant penetration such as in cholera; adhesion and penetration to mucosal cells in which they multiply without or with little spread from the initial site such as in bacillary dysen- tery;<1r,adhesion and penetration into the underlying tis- sues either through or between the mucosal cells such as in salmonellosis, streptococcal infections, and amoebic dysen- tery (Schlesinger, 1975). Certain strains of E. 99;; can cause disease in hu- mans and animals, either by producing a cholera-like en- terotoxin or by invading the intestinal epithelium. ETEC adhere and colonize the small intestine and elaborate an enterotoxin. The toxin(s) can be a heat-stable toxin (ST) that resists heating at 100°C for 15 minutes, a heat-labile toxin (LT) that is antigenic and destroyed at 60°C for 30 minutes, or both types of toxins (Gyles and Barnum, 1969); Gyles, 1971). The strains which produce enterotoxin(s) do not invade, but the toxins they produce cause secretion of electrolytes and water into the lumen. This results in mild to severe diarrhea, and finally dehydration and shock without fever (Smith and Gyles, 1970; DuPont 3E 21" 1971). The invasive or septicemic strains of E. 99;; have the ability to penetrate cells of the intestinal mucosa and cause a febrile illness with chills, fever, headache, myal- gia, abdominal cramps, and watery diarrhea. The invasive strains can also cause hypertension, systemic toxemia, and tenesmus, and the feces sometimes contain blood, mucus, and abnormal numbers of epithelial cells (DuPont 33 21" 1971). The infections of neonatal animals characterized by bacte— remia or septicemia usually are caused by the invasive E. 99;; strains, while the less severe diarrheic infections are caused by ETEC. Some E. coli strains, such as those found in urinary tract infections, have been called oppor- tunistic pathogens because they are generally harmless in their normal habitat, but when they invade other tissues they can cause disease (Carter, 1976). Factors Affecting Adhesion The adhesion of bacteria is promoted by surface components. These surface components, primarily suggested to be the pili, can contribute to mucosal surface infection by penetration and damage of epithelial cells, or by breaching the epithelial surface to facilitate spread of infection to other tissues (Schlesinger, 1975). Factors which affect adhesion and colonization of microorganisms are: resistance to mechanical flushing action of moving lu- men contents, competition with the surface commensals for space on the mucosa and for food materials, resistance to antimicrobial materials, and resistance to humoral and cel- lular antimicrobial mechanisms in the mucous secretions (Smith and Pearce, 1972). Pathogenic organisms adhere to mucous membrane sur- faces with selectivity. Specific interactions between sur- face components of bacteria and host are responsible for the selective adherence. The specificity consists of host specificity which is the ability of microorganisms to attack some animal species in preference to others, and tissue specificity which is the ability of microorganisms to attack some tissues in preference to other tissues (Smith, 1972, 1976). The most important host influences are the nutritional environment, the nature and strength of the humoral and cellular defense mechanisms, the host re- ceptors for initial attachment, and the barriers for spreading (Smith and Pearce, 1972). Some of the specific surface components of bacteria which are responsible for the selective adhesion and have been demonstrated to be im- portant for virulence include K88 antigen (Jones and Rutter, 1972; Sellwood 23 21°! 1975; Nagy 3E 31°! 1978) in E. 921E strains found in piglets, K99 antigen in strains isolated from calves, lambs (Burrows g; 21" 1976; Smith and Huggins, 1978) and pigs (Moon 23 3E., 1977), 987P antigen in strains isolated from pigs (Isaacson 3E 21" 1977; Nagy 23 21" 1977), and CFA/I (Evans 3E 3E., 1975) and CFA/II (Evans g3 21" 1978) in strains isolated from human. The probable sites for bacterial adhesion and sub- sequent colonization are the brush borders of the epithe- lial cells and the overlaying mucous membrane (Jones, 1975). Brush border surfaces are covered with a carbohy- drate-rich layer, called glycocalyx (Ito, 1969). The gly- cocalyx is composed primarily of various glycoproteins (Forstner, 1971). Bacteria probably adhere to the sugar residues of the glycocalyx. Ofek 33 EE. (1977, 1978) described the inhibitory effects of D-mannose and methyl- a-D-mannopyranoside in the adhesion of E. 39;; to epithe- lial cells of the human oral cavity, and Salit and Gotschlich (1977) described the inhibitory effects of the same sugars in the adhesion of E. 221$ to monkey kidney cells. Yeast mannan, a polymer of D-mannose, is also a strong inhibitor of adhesion (Ofek 33 3E., 1978). These reports suggest that saturation of binding sites on the bacterial surface by a sugar prevents the attachment of these organisms to the epithelial membrane receptors which contain D-mannose or a mannose-like structure. A study by 10 Jones and Freter (1976) found that L-fucose, and to some extent D-mannose, inhibited the adhesion of Vibrio cholerae to rabbit intestinal epithelial cells, but the hemaggluti- nation activity was inhibited only by L-fucose, not by D— mannose. The absence of hemagglutination inhibition by D- mannose appeared to distinguish the hemagglutination ac- tivity from adhesion to epithelial cells. Hemagglutination (HA) Activities of E. coli Mannose-Sensitive (MS) HA Three groups of E. 99;; strains were distinguished by Duguid 33 El- (1955) based on their patterns of hemag- glutination activity against red blood cells of different animal species. The HA activity of group I strains is as- sociated with the presence of type 1 pili (Duguid 33 21" 1955; Duguid, 1968; Ottow, 1975). Their HA activity is best developed in stationary-phase cultures in static liquid media (Duguid 25 EE., 1979). Some of these E. gel; Strains do not form the type 1 pili when grown on a solid medium (¢rskov g5 3E., 1980). The HA activity of type 1 pili is inhibited by 0.5 percent (w/v) D-mannose and methyl-a-D-mannoside (Duguid and Gillies, 1957). According to Duguid 33 El: (1979), the MS hemagglutination persisted when the test mixture was warmed to 50°C in the absence of D-mannose, and the HA activity was unaffected when the bac- teria were heated at 65°C for 30 minutes. 11 Mannose-Resistant (MR) HA The HA activities of groups II and III strains are resistant to inhibition by D-mannose (Duguid g; 21" 1955; Duguid and Gillies, 1957). Group II is different from group III in that group II strains possess pili, while group III strains do not (Duguid 35 21" 1955). The HA ac- tivity of the MR groups is destroyed when the bacteria are heated at 65°C for 30 minutes (Duguid _e_t_ 31:, 1979). The reactions between some MR strains and red blood cells of some animal species develop only in the cold (3-5°C), and the HA activities elute when the test mixtures are warmed to 20°C. When an eluted mixture is cooled again to 3-5°C, it will reagglutinate. However, most MR strains aggluti- nate red blood cells at 20°C, and the reaction elutes when the test mixture is warmed to 30, 40 or 50°C (Duguid g; 31-! 1955). The hemagglutinins, which elute when warmed to 30-50°C, were called mannose-resistant and eluting (MRE) hemagglutinins by Duguid (1964). The MRE hemagglutinins are best developed in cultures grown on agar at 37°C, but not at temperatures below 25°C (Duguid 33 21" 1979). Jones (1977) proposed the name "fibrillae" for the MRE-type pili. The HA properties of group II strains have been found to be associated with enteropathogenicity of E. gel; and the presence of pilus antigens such as K88 (Jones and Rutter, 1974; Isaacson §£_§l., 1977), K99 (Orskov §E_§l., 12 1975; Guinée 33 21°! 1976; Burrows 3E 3E., 1976), CFA/I (Evans 2E gl., 1975, 1977, 1979; ¢rskov and ¢rskov, 1977), CFA/II (Evans and Evans, 1978; Evans gE,§E., 1979), and F7 (Orskov §E_EE., 1980). Although the HA activities of these pilus antigens with different species of animal red blood cells are resistant to D-mannose, HA activities of some CFA/I-positive strains with guinea pig red blood cells and HA activities of some CFA/II-positive strains with human and guinea pig red blood cells are sensitive to D-mannose (Evans 32 EE., 1979). Many strains of E. 39;; possess both MS and MRE he- magglutinins (MS+/MRE+), and it is usually impossible to identify separately the two types of pili by electron mi- croscopy. The presence of MRE hemagglutinin in an MS+/MRE+ strain, according to Duguid 2E EE. (1979), could be de- tected by its positive reactions with red blood cells of some animal species in the absence and presence of D-man- nose. The MS hemagglutinin in such strain could be de- tected by warming the test mixture to 50°C, at which the hemagglutination persisted in the absence of D-mannose, but eluted in the presence of D—mannose. Another way to detect the MS activity in such strains is by testing with a culture that has been heated at 65°C for 30 minutes to destroy its MRE hemagglutinin. Based on these methods, E. 22$; strains were classified by Duguid 33 El- (1979) into four types: MS+/MRE+, MS‘VMRE", MS"/MRE+, and MS'/MRE". Evans 912 21- (1979) developed the HA patterns of human enterotoxigenic 13 and enteropathogenic E. coli, and classified them into four types (types I to IV) based on the MS, MR, and negative (N) hemagglutination with human, bovine, chicken, and guinea pig red blood cells. Nature, Formation, and Isolation of Pili Nature and Composition of Pili Like flagella, pili originate from the cytoplasma and penetrate through the peptidoglycan layers of the cell wall (Hoeniger, 1965). Pili are distinguishable from flagella under the electron microscope by their smaller diameter, irregular length, and lack of a wave-like motion (Brinton and Stone, 1961). However, pili are reported to have a rod shaped, rigid, helical structure similar to that of bacterial flagella (Brinton, 1965). Different cells from one culture may carry from one up to a thousand pili (Brinton, 1965; Duguid, 1968). Brinton g2 31. (1964) described the type 1 pili of E. 99;; as having an external diameter of 7 nm and an internal diameter of 2.0-2.5 nm. The K99 pili have a diameter of 7.0-9.8 nm, and a length between 84 and 183 nm (Isaacson, 1977), while the pili pu- rified by Korhonen EE.3$° (1980) have a diameter of 5-7 nm, and a length between 0.5 and 1.0/um. Studies with pure pili have shown that E. ggli pili consist of nearly 100 percent protein. These proteins are 14 called pilin, and contain less than 0.6 percent each of DNA, RNA, phosphate, polysaccharides, and reducing sugars (Brinton and Stone, 1961). Stirm EE.E$° (1967) purified the K88 pilus antigen, and found that K88 was a pure pro- tein which had a sedimentation coefficient ($20,w) of 36.78. This protein contained all the common amino acids except for cysteine and cystine. The K88 pilus antigen is a K antigen of the thermolabile L types because the pro- tein denatures at temperatures above 70°C. A study by Isaacson (1977), which used SDS-gels for the determination of the molecular weight of the K99 pilus protein, showed that there were two different bands from K99 protein, one had a molecular weight of 22,500, and the other had a molecular weight of 29,500. When the gels were scanned at 550 nm, the results indicated that there were five subunits in the first band and the other band was ho- mogeneous. Less than 0.6 percent neutral sugars and 6.6 percent lipid were detected in K99 protein. The K99 pro- tein contained few aromatic amino acids, with no tyrosine or phenylalanine residues. Isoelectric focusing showed that K99 protein had a pI of 4.2 (Morris 23 21°! 1978). The type 1 pili of a substrain of E. 99;; K—12 were reported to consist of subunits with apparent molecular weights of 15,500, 17,000, or 19,000, depending on the de- gree of separation and age of preparation (McMichael and On, 1979). The pili purified by Korhonen 3E El: (1980) from E. coli strains isolated from patients with urinary 15 tract infections had a single subunit with a molecular weight of 17,000, a pI of 4.9, and 43 percent of the amino acids in the subunit protein were hydrophobic. Methods for Isolation of Pili Separation of pili from the bacterial cells is easily achieved by blending of cultures in the logarithmic growth phase (Brinton and Stone, 1961; Brinton, 1965; Novotny 2E 3E., 1969). Novotny 33 3E. (1969) reported that blending at a speed of 4,220 rpm for 2 minutes was effec- tive in removing flagella and F pili of E. ggli, but nOt type 1 pili, while blending at 11,650 rpm for 2 minutes completely removed type 1 pili. The extracted pili can be precipitated with ammonium sulfate (Brinton, 1959; Isaacson, 1977; McMichael and Cu, 1979; Korhonen 33 gl., 1980) or by using ultracentrifugation (Stirm gE 21" 1967; Evans g3 21" 1978; Freer 33 EE., 1978). If the sediment from the ultracentrifugation is suspended in a neutral so- lution and left for several days at 4°C, the aggregated pili can be separated from the remaining impurities by cen- trifugation at a higher speed (Brinton and Stone, 1961). Further purification of pili can be accomplished by using gel filtration chromatography (Isaacson, 1977; Korhonen 33 31" 1980), ion exchange gel chromatography (Isaacson, 1977), or ultracentrifugation in sucrose gradients (Isaacson, 1977; Evans 33 EE., 1978; Korhonen 33 El., 1980). MATERIALS AND METHODS Sources of E. coli Isolates E. 999; isolates were obtained from Bil-Mar Foods Inc., Zeeland, Michigan, from diseased turkeys during several E. 9919 outbreaks that occurred in 1978 and 1979. The E. 9999 outbreaks occurred in young turkeys which were three to five weeks old. During these outbreaks between 10 and 100 birds per day were lost in houses containing ap- proximately 10,000 birds. E. 9999 isolates were also ob- tained from normal turkeys, feed, water, and litter in Au- gust 1979. Isolation of E. coli Plates were inoculated by a veterinarian using swabs from different parts of the body of diseased turkeys such as lung, liver, heart, spleen, and yolksac. Each plate contained three separated media, Eosin methylene blue (EMB) agar for the detection and isolation of E. 9999, Brain heart infusion (BHI) agar for the cultivation of fas- tidious pathogenic bacteria, and Brilliant green (BG) agar which is a highly selective medium for the isolation of Salmonella. After incubation at 37°C for one day, the 16 17 plates were sent to Michigan State University. Several colonies from each plate were streaked on EMB plates, and the plates were incubated at 37°C for 20-24 hours. Typical E. 9999 colonies were picked from EMB plates, and purified by repeated streaking on EMB plates. After incubation on EMB plates at 37°C for 20-24 hours, the pure cultures were then streaked onto slants of Blood agar base (BAB) which were incubated at 37°C for 20-24 hours. The slants were stored at 4°C for a short period, or inocu- lated into Casamino yeast extract (CYE) broth and frozen or lyophilized for long-term storage. Biochemical Identification of E. coli The cultures were screened based on their gram stain reaction (Frazier 99 99., 1968), and the results of oxidase tests using Kovacs' method (Kovacs, 1956). Oxi- dase-negative, gram-negative bacilli were identified using the Fisher Entero-SetTMZO identification system. The growth from a BAB slant, which had been incu- bated at 37°C for 18 hours, was suspended in 5 ml of 0.15M sterile saline. Water (9 ml) was added to the bottom of the incubation chamber, and a Fisher Entero-SetTMZO card which consisted of 20 test capillaries was placed in the bottom of the chamber. Each capillary was filled with 5 drops of bacterial suspension using a sterile Pasteur pi- pette, and the plastic cover was sealed. After incubation for 24 hours at 37°C, FeCl3 solution was dropped into the 18 PHE (phenylalanine deaminase) capillary, and Kovac's rea- gent was dropped into the IND (indole production) capillary. Reactions were read in the center of the capillaries, un- less the capillaries were bracketed at the top. The or- ganisms were identified using the Entero-TrakTMZO manual, TM or using the Entero—Set 20 identification table (Table l). Serological Identification of E. coli O Serotyp9 Preparation of O Antigen. Serological identifica- tion of 0 antigen was done based on the methods described by Mehlman 99 99. (1974) and ¢rskov and ¢rskov (1975). A pure culture was inoculated into a BAB slant and incubated at 37°C for 24 hours. Using the growth from the BAB slant, a suspension was prepared in 15 ml of 0.15M saline with turbidity corresponding to McFarland standard number 3 (9.0 x 108 cells/ml). The suspension was boiled at 100°C for one hour (to inactivate the capsular antigen), cooled, and 0.07 ml of 37 percent formaldehyde (0.5 percent formalin) were added as a preservative. Presumptive Identification. The O antisera used were obtained from Difco Laboratory (Detroit, Michigan), and represent the O antigens of E. 9999_strains found in dis- eased poultry, i.e. Ol, 02, O8 and 078 (Sojka and Carnaghan, 1961; Harry, 1964; Hemsley and Harry, 1965; Heller and Smith, 1973). The presumptive identification of O antigens l9 Table 1. Identification of E. coli by Fisher Entero- TM —’ “—— Set 20 . Entero-SetTMZO Typical Test Description géaggii Total no? tion No. (typical E. coli) MAL Malonate utilization - l GLU Glucose-Resazurin + 2 2 PHE Phenylalanine deaminase - 4 ONPG ONPG-beta-galactosidase : IND Indole production + 2 or 3 H28 H28 production - 4 LYS Lysine decarboxylase : 1 ORN Ornithine decarboxylase + 0' 1' - 2 or 3 UR Urease - SUC Sucrose fermentation : l . . . o 1 ARG Ar inine dlh drolase + ' ' g y - 2 or 3 CIT Citrate utilization - SAL Salicin fermentation : 1 AD Adonitol fermentation - 2 0 or 1 IN Inositol fermentation - 4 SOR Sorbitol fermentation + l ARB Arabinose fermentation + 2 3 or 7 MLT Maltose fermentation : 4 TR Trehalose fermentation + 1 1 or 3 XY Xylose fermentation : aTotal number represents the total number of positive re- actions. 20 by tube agglutination was done by adding into a small tube 0.125 ml 0.15M saline, 0.125 ml monovalent 0 serum (1:10), and 0.250 ml culture suspension. After mixing, the tube was incubated at 37°C for 24 to 48 hours. A tube which contained 0.250 ml 0.15M saline and 0.250 ml culture sus— pension served as a control for autoagglutination. A nega- tive agglutination reaction in the control tube and a posi- tive agglutination reaction in the other tube containing a specific 0 antiserum (the minimum titer was 40) indicated a positive presumptive reaction. Cultures which showed nega- tive presumptive reactions were autoclaved at 120°C for 2 hours, and tested again for presumptive identification. The autoclaved cultures were discarded if they still had not shown any positive agglutination reaction with the above antisera. Determination of O Titer. A row of 10 unscratched 13x10 mm tubes were aligned in a rack, and 0.5 ml 0.15M sa- line was added to all tubes. Then 0.5 m1 of the monovalent O antiserum (1:10) showing a positive presumptive reaction was added to the first tube, mixed, and 0.5 m1 from the first tube was transferred to the second and mixed. Dilu- tions were continued through 9 tubes, and 0.5 ml from tube 9 was discarded. Tube 10 served as a control for autoag- glutination. Then 0.5 m1 of culture suspension was added to each tube and mixed by gentle agitation, so that the first tube contained a 1:40 final dilution and tube 9 21 contained a 1:10.240 final dilution of the antiserum. The O-tube agglutination reaction was read after 24 to 48 hours at 37°C. Many cross-reactions might exist between 0 anti- gens, therefore a culture suspected to belong to a given 0 serotype should give a titration value similar to, or higher than, the known control culture. The cultures used as controls for O antigens were U5—41 (Ol:Kl:H7) for 01 an- tigen, A20a (02:Kl:H6) for 02 antigen, 63404-41 (08:K8:H4) for 08 antigen, and E38 (O78:K80) for 078 antigen. These cultures were obtained from Drs. I. ¢rskov and F. ¢rskov, Center for Reference and Research on Escherichia, Copen- hagen, Denmark. K Serotype Immunodiffusion. Immunodiffusion was used as a screening method to detect the K serotypes of E. 9999. Double diffusion was done on agarose-coated glass slides (Cahill and Glantz, 1978). The bacterial suspension was prepared using a procedure similar to that used for the identification of O antigens, with Kl, K80, K88ab, K88ac, and K99 antisera purchased from the Veterinary Science De- partment, Pennsylvania State University. The differences in the preparation of bacterial suspension for immunodiffu- sion were: a Trypticase soy agar (TSA) slant was used to grow the bacteria, the bacterial suspension was heated in a water bath at 600C for 20 minutes to release the heat-labile 22 antigen into the supernatant, and the heated suspension was centrifuged at 20,000 x g for 20 minutes at 40C. The su- pernatant was removed and stored at 40C. Three glass slides were coated with 11 m1 of one percent agarose in Veronal buffer, pH 8.6, ionic strength 0.025’p. Five wells were prepared, and an antiserum was placed in the center well while the antigen extracts were placed in the outer wells. The slides were stored for 4 days to two weeks in a walk-in refrigerator at 4°C, or at room temperature for 4- 10 days. The results were recorded as negative or positive precipitation. The cultures used for controls of K anti- gens were U5-4l for K1 antigen, E38 for K80 antigen, Moon 263 (08:K87:K88ab:H19) for K88ab antigen, 61253 (Ol47:K88ac: H19) for K88ac antigen,-and 1474 (0-:K99) for K99 antigen. Cultures Moon 263 and 1474 were obtained from Dr. H.W. Moon, National Animal Disease Center, Ames, Iowa, while culture 61253 was obtained from Drs. I. ¢rskov and F. ¢rskov. Determination of K Titer. The determination of K titer was done as in the determination of O titer, except that the bacteria were grown on a TSA slant, the bacterial suspension was not heated, and a series of 14 tubes were used for agglutination reaction (final dilution l:l63,840). Electron Microscopy Cells from a culture grown on a Peptone agar slant were harvested and suspended in distilled water. Negative staining was done by floating a carbon-coated collodion grid 23 with a drop of cell suspension for 30 seconds, and removing the excess suspension by adsorption onto filter paper. The grid was then stained with 2 percent phosphotungstic acid for 30 seconds. After the excess of stain was removed, the grid was examined in a Philips EM 200 electron microscope to detect the presence of pili on the bacterial cells. The approximate diameter of the pili was determined by measuring the diameter of the pili from electron photomicrographs, using a micrometer (l division = 0.01 mm). Data reported are the average of ten observations. Hemagglutination Tests The hemagglutination (HA) tests were done using hu- man type A red blood cells obtained from The American Red Cross (Lansing, Michigan), and bovine, chicken, guinea pig and Rhesus monkey red blood cells obtained from Flow Labo- ratories Inc. (Rockville, Maryland). Red blood cells, which were preserved in Alsever‘s solution, were centrifuged and washed three times with 0.15M phoSphate buffer, pH 7.2, containing 0.15M saline (PBS), immediately before use, and made up to a 3 percent (v/v) suspension in PBS. HA tests were carried out by a slide agglutination procedure, which was a combination of methods described by Duguid 99 99. (1979) and Evans 99 99. (1979). Bacterial cells were obtained from cultures grown at 370C for 18 hours or at 180C for 72 hours, on a Colonization 24 factor antigen (CFA) agar slant.and in CYE broth. CFA agar consists of 1 percent Casamino acids (Difco), 0.15 percent yeast extract (Difco), 0.005 percent MgSO 0.0005 percent 4' MnClZ, and 2 percent agar, pH 7.4 (Evans 99 99., 1979). CYE broth consists of 2 percent Casamino acids (Difco), 0.6 percent yeast extract (Difco), 0.25 percent NaCl, 0.871 per- cent KZHPO4 (0.05M), and 0.1 percent trace salts solution, pH 8.5 (adjusted with 0.1N NaOH). The trace salts solution consists of 5.0 percent MgSO 0.5 percent MnCl and 0.5 2! SO4 (Evans 99 99., 1973). 4: percent FeCl3 dissolved in 0.001N H2 Growth from the CFA slant was suspended in PBS and growth from CYE broth was centrifuged at 17,000 x g for 10 minutes and suspended in PBS. The suspensions in PBS were adjusted to a turbidity corresponding to McFarland standard number 5 (1.5 x 109 cells/m1). One drop of bacterial sus— pension, unheated or heated at 65°C for 30 minutes, one drop of PBS or PBS containing 1.5 percent (w/v) D-mannose, and one drop of blood suspension were mixed on a glass slide which was placed on the surface of crushed ice. Af- ter observation for at least 2 minutes in a walk-in re- frigerator (temperature 4°C) with intermittent mixing by rotation of the slide, the slide was placed at room tem— perature with continued rotation. Complete HA was recorded as 4+, the lesser degrees of HA were recorded as 3+, 2+, or 1+. The HA reaction was considered as a positive re— action if the reaction wasZ>2+. HA was reported as mannose- 25 resistant (R) if the same degree of HA occurred with and without D-mannose, and as mannose-sensitive (S) if the HA was eluted or prevented in the presence of D-mannose. The persistence or elution of HA reactions at elevated tempera- tures was demonstrated by warming the slides to 400C and 500C (Duguid 99 99., 1979) in different incubators with in- termittent rotation for no longer than 2 minutes at each temperature to prevent drying of the test mixture. Each test was done in duplicate. Culture H10407 (CPA/1+), ob- tained from Drs. D.G. Evans and D.J. Evans, The University of Texas Medical School, Houston, Texas, and culture Moon 263 (K88ab+) were used as controls for hemagglutination re- actions. In Vitro Adhesion Tests The 99 99999 adhesion tests were done by a modifi- cation of procedures described by Jones and Rutter (1972) and McNeish 99 99. (1975) which used piglet small intestin- al tissue. The modified procedure, outlined in Figure 1, used turkey small intestinal tissue obtained from Bil-Mar Foods Inc. The tissue preparation was freshly used, or stored for a maximum of 3 weeks at 4°C in 5 mM EDTA (pH 7.4) containing 10 percent (v/v) formalin. According to Jones 99 99. (1976) and Isaacson 99 99. (1978), a tissue extract could be stored by this method for at least 3 months after preparation, with no difference in the extent 26 Turkey small E. coli culture intestine \L ,¢ Cut Grow in CYE broth, (the middle third) 37°C 24 hrs Open longitudinally, Grow on CF agar: 1) 37°C 18 hrs 2) 18°C 72 hrs $ Cut into discs Harvest (McFarland (S-23211 cutting tube turbidity no.4) into: no.3, ¢ 6 mm) 1) PBS 2) PBS + 0.5% D-mannose \v ,w””//7<\\\‘\\e_ clean & wash with 0.05M PBS (pH 7.2) I Wash with PBS 1) Unheated 2) Heated 65°C 30 min 2 discs 1 m1 Incubate with rotation at room temp. for 2 sec.(0 time), 10, 20, 30, 40, 50 min Wash 3x with 3 x 5 m1 PB Homogenize in 10 m1 PB with sand (0.11-0.12 mm) Dilute with PB Pour plate in TSYA, 37°C 48 hrs Figure 1. Flow chart of the 99 Vitro adhesion tests. 27 of adhesion of bacteria between the formalin-preserved tis- sue and the freshly prepared tissue. However, in these ex- periments most of the tissues used were freshly prepared. When tested, the adhesive properties of E. 9999 to forma- lin-preserved turkey small intestinal tissue was similar to that which occurred with the freshly prepared tissue. The viable counts of the 99 99999 adhesion tests were deter- mined using pour plates of Trypticase soy agar containing 0.3 percent yeast extract (TSYA) (Ray and Speck, 1973ab)- The results are reported as the average of duplicate plates of four tissue samples, using two incubation tubes per treatment and two tissue samples per tube. Plate counts of the uninoculated turkey small intestinal tissues were also made to determine the number of organisms present in the original tissues. Inhibition of In Vitro Adhesion D-glucose, D-galactose, D-mannose, L-fucose, N-ace- tyl—D-glucosamine, N-acetyl-D-galactosamine, N-acetyl~neur- aminic acid, and a-methyl-D-mannoside (Sigma Chemicals, St. Louis, Missouri), were each used at 0.5 and 1.0 percent (w/v) concentrations to test their effects on 99 99999 adhesion. Six tubes were used for each sugar or sugar de- rivative; 0.1 ml buffer (PBS) was added to the first two tubes, 0.05 ml PBS and 0.05 ml of 10 percent (w/v) sugar solution were added to the next two tubes, and 0.1 ml of 28 10 percent sugar solution was added to the other two tubes. Then 0.9 m1 of bacterial suspension was added to all tubes, so that each tube contained the same dilution of bacterial suspension, and the first two tubes, which served as con- trols, contained no sugar, the next two tubes contained 0.5 percent sugar, and the other two tubes contained 1.0 per- cent sugar. The next steps were the same as in the 99 vitro adhesion tests (Figure 1), using only an incubation time of 50 minutes. A similar procedure was also applied using D-mannose and a-methyl-D-mannoside at 0.1, 0.2, 0.3, 0.4, 0.5, 1.0, 1.5, 2.0, and 2.5 percent (w/v) concentra- tions. In some experiments, concanavalin A (Sigma), sodium metaperiodate (MCB), and isolated pili were added to the intestinal tissue discs 30 minutes prior to the addition of the bacterial suspension. After this preincubation, the bacterial suspension was added to yield a final concentra- tion of 0.05, 0.10, 0.15, 0.20, and 0.25 percent (w/v) of concanavalin A and sodium metaperiodate, and 0.5, 1.0, 1.5, 2.0, 2.5, 5.0, and 7.5 mg pilus protein per m1. Isolation of Pili Extraction and Precipitation Extraction of pili from the cells was done by a method described by Novotny 99 99. (1969). Bacterial cells were first blended at lower speed (4,500 rpm for 2 minutes) 29 to break the flagella that might be present, without af- fecting the pili. After removal of the flagella by cen- trifugation at 14,350 x g for 30 minutes, the cells were blended at higher speed (12,000 rpm for 4 minutes) to break the pili, followed by centrifugation at 14,350 x g for 30 minutes to separate the pili from the cells. Precipitation of the pili was done by adding ammonium sulfate to 65 per- cent saturation (Isaacson, 1977). The detailed procedures for the extraction and precipitation of pili are outlined in Figures 2 and 3, respectively. The ammonium sulfate precipitate was lyophilized and stored in a freezer at -18°C. Gel Chromatography The lyophilized ammonium sulfate precipitate was resuspended in 5 ml 0.05M sodium phosphate buffer (PB), pH 7.2, and applied to a Sephadex G-50 (Pharmacia Fine Chemicals) column, 2.5 by 40 cm, that had been equilibrated with PB. Blue dextran 2000 (Pharmacia Fine Chemicals) was added to the solution before chromatography as a marker to indicate the void volume (V0), and the protein was eluted with PB at 1.5 ml per minute. The protein was detected using an Isco Model UA-2 Ultraviolet Analyzer connected to a recorder (Isco Instrumentation Specialties Co., Inc.; Lincoln, Nebraska). Fractions of 3 ml were collected using a fraction collector (Rinco Instruments Co.; Greenville, 30 E. coli culture Grow in CYE broth, 37°C 24 hrs $ Grow on CF agar in 12 Roux culture bottles, 37°C 24 hrs Harvest with 0.05M PB, pH 7.2 containing 1.0M NaCl Homogenize in Sorval Omnimixer, 4,500 rpm for 2 min. Centrifuge, 14,350 x g for 30 min. \I/ v Supernatant Cells=< (flagella) L V Discard Homogenize with 0.05M PB + 1.0M NaCl, 12,000 rpm for 4 min. Repeat Centrifuge, 2x 14,350 x g for 30 min. Supernatant l' i (pili) Cells Discard Figure 2. Flow chart of the extraction of pili from bac— terial cells. 31 Supernatant (pili) 1 Filter through 0. 45 /um membrane ¢ 7. Supernatant Cells Precipitate with Discard 65% saturation of (NH4)ZSO4, overnight at 4°C 1 Centrifuge, 20,280 x g for 45 min. | 9 ¢ Supernatant Precipitate (pili) . .\/ Discard Suspend 1n 0.05M PB \/ Dialyze overnight in 0.05M PB at 4°C through membrane (mw cut off 6,000-8,000) Lyophilize Store in freezer Figure 3. Flow chart of the precipitation of pili. 32 ) Illinois), and were assayed for absorbance at 280 nm (A280 using a Beckman DU Model 2400 Spectrophotometer (Beckman Instruments, Inc.; Fullerton, California). Fractions which belonged to the primary peak were pooled and the mixture was lyophilized, and stored in a freezer at -18°C. Gel Electrophoresis The gel for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was prepared by the method of Weber and Osborn (1969) as described by Cooper (1977), using 7.5 percent acrylamide (T). The gel was prepared by mixing 15 m1 gel buffer (0.2M phosphate, pH 7.2, 0.2 per- cent SDS), 10.1 ml acrylamide solution (22.2 g acrylamide and 0.6 g N,N'-methylene-bis(acrylamide) in a final solu- tion of 100 ml), 3.4 ml water, 45’pl TEMED (N,N,N',N'- tetramethylethylenediamine), and 1.5 ml ammonium persulfate solution (35 mg ammonium persulfate in 10 ml water). For electrophoresis the lyophilized sample was dis- solved in buffer (0.01M phosphate, pH 7.0, 1.0 percent SDS, 1.0 percent mercaptoethanol), and the solution was heated in a boiling water bath for 5 minutes. After cooling, crys— talline sucrose was added to increase the density, and a drop of 1.0 percent bromophenol blue solution was added as a marker dye. The upper and lower chambers were filled with res- ervoir buffer which consisted of one part gel buffer and 33 one part water. Then 25 to 50/u1 of 0.5 percent of the sample solution were applied to the gels. A current of 8 mA per tube was applied, and the electrophoresis was run for 4 to 5 hours in a water-cooled cell. The gels were stained in a 3.5 percent perchloric acid solution contain- ing 0.04 percent (w/v) Coomassie brilliant blue G-250 (Sig- ma) for 5 hours, and destained in a 7.5 percent acetic acid solution containing 5 percent methanol. The gels were stored in 7.5 percent acetic acid. The approximate molecular weight of the isolated pili was determined from a standard curve prepared using a mixture of proteins (Pharmacia Fine Chemicals) containing a-lactalbumin (mw. 14,400), soybean trypsin inhibitor (mw. 20,100), carbonic anhydrase (mw. 30,000), ovalbumin (mw. 43,000), bovine serum albumin (mw. 67,000), phosphorylase b (mw. 94,000), and sucrose. Relative mobilities were calcu- lated as follows: (Protein ) ( Gel length ) Relative _ distance before staining mobility _ Marker dye Gel length (- )X( --) distance after destaining The standard curve was a plot of the relative mobilities vs log molecular weights of the standard proteins. Quantitation of Protein Protein concentration was determined by the method of Lowry 99 99. (1951) as described by Cooper (1977). 34 Absorbance at 660 nm was measured on a Spectronic 20 (Bausch & Lomb Inc.; Rochester, New York). A standard curve was developed, using bovine serum albumin (Sigma) as the standard protein. The curve was linear (r = 0.9981) up to 180/ug protein (Figure 4) according to the formula y = 0.0037 x + 0.0097, where y was absorbance at 660 nm (A660) and x was the amount of protein in Pg. 35 I l l 0'71 0.6- 0.5? E C) $3 0.4~ '2 g 0.3- Si :3 0.2~ 0.1- 0 0 FIGURE 4. 30 60 90 120 150 PROTEIN./UG STANDARD CURVE FOR PROTEIN ASSAYED BY THE LOWRY METHOD. 180 RESULTS Isolation and Identification of E. coli Isolated Cultures Plates from various fatal infections of poultry that occurred in 1978 and 1979 were obtained from Bil-Mar Foods Inc. (Table 2). The plates were inoculated with swabs from parts of the bodies of diseased turkeys such as lung, liver, heart, spleen, and yolksac. Due to the diffi- culties in differentiating between the commensal and the infectious E. 9999 in the intestinal tract, strains were not isolated from the intestines or feces. Preliminary da- ta indicated that the intestines of normal young turkeys contained 105 to 106 fecal coliforms per gram, while the litter obtained from different flocks contained 106 to 107 fecal coliforms per gram. After streaking on EMB plates, all colonies showed the typical green metallic sheen of E. 9999 colonies. Two different separated colonies were isolated from each plate, however, the two cultures from the same plate invariably had the same Entero-SetTMZO numbers and O titers, there- fore, only one culture from each plate was reported. As 36 37 Table 2. Descriptions of plates inoculated from diseased turkeysa Condition No. of Date Source of turkey plates Code Aug.28, 1978 Lung Consolidated 2 l, 4 Liver Perihepatitis 3 2, 5, 8 Heart Pericarditis 3 3, 6, 9 Oct. 4, 1978 Heart Pericarditis 2 94-1 94-5 Spleen ? 1 94-3 Feb. 7, 1979 ? ? 4 102-1, 102-2, 102-3, 102-4 Sep. 4, 1979 Heart Pericarditis 2 119-1, 119-3 Liver Normal 2 119-2, 119-4 Sep.10, 1979 Yolk- Omphalitis 3 120-1, sac 120-2, 120-3 Total 22 a . . . . Descriptions were obtained from Bil-Mar Foods Inc. 38 many as 21 cultures were identified as E. 9999 by the Ente- ro-SetTMZO identification system (Appendix 1). All cultures from the first, second, and third outbreaks gave the same Entero-SetTMZO results (2331073), while the cul- tures from subsequent outbreaks yielded different results (Table 3). One culture (120-3) was a gram-positive coccus and was identified as Streptococcus. O and K Serotypes The majority of E. 9999 strains which are patho- genic to poultry belong to only a few serotypes. There- fore, several investigators have screened the E. 9999 strains that caused fatal infections in poultry using anti- sera against certain O antigens of E, 9999 that were most frequently encountered in diseased poultry, e.g. Ol, 02, 08, and 078 (Sojka and Carnaghan, 1961; Harry, 1964; Hemsley and Harry, 1965; Heller and Smith, 1973). In this research, a similar O-antigen screening procedure was applied to the isolates obtained from diseased turkeys. Cultures 9, 94-5, 102-1, and 119-3 were selected for further study, including electron microscopy to deter- mine the presence of pili. Subsequently, screening for K serotypes was done on these cultures. The results are listed in Table 4. Cultures 9 and 102-1 were identified as O78:K80, and cultures 94-5 and 119-3 were identified as Ol:K1 and 02:K1, respectively. 39 Table 3. Entero-SetTMZO numbers of isolates obtained from diseased turkeys Isolate Entero-SetTMZO . strain no. numbers Organism 1, 2, 3. 4, 5: . 6, 8, 9 2331073 E. C011 94‘;;_§4‘3' 2331073 9. coli 102-1, 102-2, . 102_3, 102_4 2331073 E. coli 119-1 2335073 E. coli 119-2, 119-3 2331033 E. coli 119-4 2233033 E. coli 120-1 2336073 E. coli 120-2 2334073 E. coli 120-3 Gram-positive Streptococcus coccus aStrain no. corresponds to the code of the plate (Table 2) from which the organism was isolated. 40 Table 4. O and K titers, and immunodiffusion (ID) of isolates 0 K Cultures Serum Titer Serum Titer ID Controls: 05-41 (Ol:Kl:H7) 01 1,280 K1 20,480 + A20a a E38 (O78:K80) 078 1,280 K80 2,560 + Isolates: 1. 2. 3, 4 - 5, 6' 8 078 1,280 ND ND 9b 078 5,120 K80 5,120 + 94-1 02 1,280 ND - ND 94-3 01 1,280 ND - ND 94-5b 01 1,280 K1 81,920 + 102-1b 078 1,280 K80 40,960 + 102-2, 102-3 078 1,280 ND - ND 102-4 078 2,560 ND - ND 119-1 0? - ND - ND 119-2 02 640 ND - ND 119-3b 02 640 K1 81,920 + 119-4 02 320 ND - ND 120-1 0? - ND - ND 120-2 0? - ND - ND aND, not done. Cultures selected for further study. 41 Electron Microscopy Electron microscopy of the cells grown on Peptone agar was used to observe pili and flagella on the bacterial cells (Appendices 2, 3 and 4). The average diameter of pili of each culture was calculated and the results are listed in Table 5. The pili of culture 9 had an average diameter of 6.2 nm, while the pili of cultures 94-5 and 119-3 had average diameters of 9.5 and 8.9 nm, respectively. Hemagglutination Patterns Two strains possessing mannose-resistant and eluting (MRE) hemagglutinins, H10407 (CFA/I+) and Moon 263 (K88ab+), were used as controls for hemagglutination (HA). The HA activities of MRE hemagglutinins were better devel- oped when bacteria were grown on CF agar than in CYE broth (Tables 6 and 7). According to Duguid 99 99. (1955), the optimal condition for the development of MRE hemagglutinins was an overnight incubation on agar at 37°C. The reactions between CFA/I+ strain and bovine red blood cells (Table 6), and between K88ab+ strain and guinea pig red blood cells (Table 7), developed only at 4°C. All the positive reac- tions were resistant to D-mannose, but disappeared or were eluted when the bacteria were heated at 65°C for 30 min- utes, or when the test mixtures were warmed to 25, 40 and 50°C. These characteristics indicated the presence of a MRE hemagglutinin(s) (Duguid 99 99., 1979; Evans 99 99., 42 Table 5. Appendages of selected isolates, in- cluding the approximate diameter of pili a Appendages Approximate Isolates diameterbof . . pili Pili Flagella (nm) H10407 + + 9.0 + 0.2 (CPA/1+) — 9 + - 6.2 i 0.1 94-5 + - 9.5 i 0.4 102-l - - - 119-3 + - 8.9 i 0.5 a +, present; -, absent (as determined by electron microscopy). b . . . - Means of 10 microsc0pic observations; x 9 SEM (=standard error of means). 43 Table 6. HA of different species of red blood cells by strain H10407 (CPA/1+) grown at 37°C Media Treat- Temp. HA with r.b.c. of:a HA to ment of pattern grow test Hu Bv Ck Gp Rm mixt. CFA Unheated 4°C R(4+) R(2+) R(3+) N N RRRNN 25°C R(4+) N R(3+) N N 40°C N N N 50°C N N Heatedb 4°C R(2+) N N N RNNNN 25°C N N N N 40°C N CYE Unheated 4°C R(l+) N R(2+) N N RNRNN 25°C R(l+) N R(1+) N N 40°C N N 50°C N N Heated 4°C N N N N N NNNNN 25°C N N N N N aHu, human type A; Bv, bovine; Ck, chicken; Gp, guinea Pig; Rm, negative HA. Rhesus monkey; resistant to D-mannose; N, b . . . Bacterial suspenSion was heated at 65°C for 30 minutes. 44 Table 7. HA of different species of red blood cells by strain Moon 263 (K88ab+) grown at 37°C a Media Treat- Temp. HA with r.b.c. of: HA to ment of pattern grow test Hu Bv Ck Gp Rm ‘ mixt. CFA Unheated 4°C N N R(3+) R(2+) N NNRRN 250C N N R(2+) N N 40°C N N 50°C N b O Heated 4 c N N N N N NNNNN 25°C N N N N N CYE Unheated 4°C N N R(1+) 5(1+) s<1+) NNRSS O 25 C N N N N N 40°C N N N Heated 4°C N N N N N NNNNN 25°C N N N N N a b . . ' See Table 6; S, senSitive to D-mannose. 45 1979). The HA patterns using red blood cells of human type A, bovine, chicken, guinea pig, and Rhesus monkey were RRRNN for the CFA/I+ strain (Table 6) and NNRRN for the K88ab+ strain (Table 7), grown on CFA. The HA pattern of the CFA/I+ strain was similar to the observation of Evans 99 99. (1979), which was RRRN without using Rhesus monkey red blood cells. According to Jones and Rutter (1974), the HA activities of K88+ strains with red blood cells of some animal species were not inhibited by D-mannose, but eluted at 37°C. Heated and unheated cultures 9 and 119-3 exhibited the same HA patterns (NNSSS) when grown on CFA and in CYE broth. Weaker reactions developed when the bacteria were grown in CYE broth (Tables 8 and 9). The HA reactions were eluted or disappeared when the test mixtures were warmed to 25, 40 and 50°C. The HA pattern of culture 94-5 after growth on CPA was the same as that after growth in CYE broth (Table 10). In both media, the HA pattern developed by unheated culture was SNSSS. After the bacteria were heated at 65°C for 30 minutes, the MS HA with human red blood cells was prevented and the HA pattern was NNSSS. The reactions were eluted when the temperature of test mixtures was raised to 40 and 50°C. 46 Table 8. HA of different species of red blood cells by culture 9 Media Treat- Temp. HA with r.b.c. of:a HA to ment of pattern grow test Hu Bv Ck Gp Rm mixt. CFA Unheated 4°C N N s<4+) 5(4+) 5(4+) NNSSS 25°C N N s<3+) 3(3+) 5(3+) 40°C s<2+) s<2+) 3(2+) 50°C 5(2+) 5(2+) 3(2+) Heatedb 4°C N N s<4+) s<4+) 3(4+) NNSSS 25°C N N 3(4+) 3(4+) 5(3+) 40°C 5(2+) s<2+) N 50°C s<2+) s<2+) N CYE Unheated 4°C N N s<2+) 5(2+) 5(2+) NNSSS 25°C N N s<2+) s<2+) s<2+) 40°C 5(1+) 5(1+) s<1+) 50°C s<1+) 5(1+) s<1+) 4°C N N 5(2+) 5(2+) 5(2+) NNSSS 25°C N N 3(2+) 5(2+) 5(2+) 40°C s<1+) 5(1+) 5(1+) 50°C 3(1+) 3(1+) S(1+) a, bSee Table 6; 3. sensitive to D-mannose. 47 Table 9. HA of different species of red blood cells by culture 119-3 Media Treat- Temp. HA with r.b.c. of:a HA to ment of pattern grow test Hu Bv Ck Gp Rm mixt. CFA Unheated 4°C N N 5(4+) 5(4+) 5(3+) NNSSS 25°C N N s<4+) s<4+) 5(2+) 40°C s<2+) s<1+) N 50°C 5(2+) 5(1+) N Heatedb 4°C N N 3(4+) s<3+) 5(3+) NNSSS 25°C N N s<3+) 5(2+) s<2+) 40°C s<2+) N N 50°C 5(2+) N N CYE Unheated 4°C N N 3(3+) 5(3+) 5(3+) NNSSS 25°C N N 5(3+) s<3+) 3(3+) 40°C 3(2+) 5(2+) 5(1+) 50°C s<2+) 5(1+) 5(1+) Heated 4°C N N 3(3+) 5(3+) 5(2+) NNSSS 25°C N N s<3+) S(3+) s<2+) 40°C s<2+) s<2+) S(1+) 50°C s<2+) s<1+) S(1+) a, b See Table 6; S, sensitive to D-mannose. 48 Table 10. HA of different species of red blood cells by culture 94-5 Media Treat- Temp. HA with r.b.c. of:a HA to ment of pattern grow test Hu Bv Ck Gp Rm mixt. CPA Unheated 4°C 5(3+) N s<4+) 5(4+) s<4+) SNSSS 25°C 5(2+) N 5(4+) 3(4+) 5(4+) 40°C N» s<2+) 5(2+) 3(2+) 50°C N N N N b o Heated 4 c N N s<4+> 5(4+) s<4+) NNSSS 25°C N S(4+) 5(3+) 5(3+) 40°C s<4+) s<3+) s<2+) 50°C s<4+) S(2+) s<2+) CYE Unheated 4°C 5(2+) N 5(3+) 5(4+) s<3+) SNSSS 25°C s<1+) N 3(3+) s<4+) s<3+) 40°C N 5(2+) s<3+) s<1+) 50°C N S(1+) S(l+) N Heated 4°C N N s<4+) 5(4+) 5(4+) NNSSS 25°C N N s<3+) 5(4+) 5(4+) 40°C 5(2+) 5(2+) 3(2+) 50°C s<1+) s<1+) 5(1+) a. bSee Table 6; 5: sensitive to D-mannose. 49 The HA patterns exhibited by culture 102-1 were NNNSS when the bacteria were grown on CFA, NNNSN when the bacteria were grown in CYE broth, and NNNNN when the bacteria were heated after growing in both media (Table 11). All cultures grown at 18°C, except the controls, exhibited the same HA patterns as those grown at 37°C. No HA reaction was developed by the controls (CFA/I+ and K88ab+ strains) grown at 18°C, which indicated the inhibi- tion of MRE hemagglutinin formation at 18°C. Several in- vestigators have demonstrated that the formation of MRE hemagglutinins such as K88, K99 and CPA was depressed at a growth temperature of 180C (¢rskov 99 99., 1975; Burrows 99 99., 1976; Evans 99 99., 1978). In Vitro Adhesion The control levels for the adhesion tests were determined based on the viable counts of a tissue disc after mixing with the bacterial suspension for two seconds. This method ensured the measurement of a nonspecific carry- over of bacteria on the tissue after washing (Jones and Rutter, 1972). The increase in the viable counts of adhering bacteria between two seconds and certain times of incubation were indicative of mucosal adhesiveness. 50 Table 11. HA of different species of red blood cells by culture 102-1 a Media Treat- Temp. HA with r.b.c. of: HA to ment of pattern grow test Hu Bv Ck Gp Rm mixt. CPA Unheated 4°C N N N s<2+) 5(2+) NNNSS N 25°C N N N N N (NNNN ) 40°C N N b o Heated 4 C N N N N N NNNNN 25°C N N N N N CYE Unheated 40C N N N S (2+) N NNNSN 25°C N N N N N (NNNNN) 40°C N o Heated 4 C N N N N N NNNNN 25°C N N N N N a, bSee Table 6; 5. sensitive to D-mannose. 51 Strain K-12, which possesses type 1 pili, and cul- ture A20a (02:Kl:H6), which was isolated from a human ap- pendix, were used as controls. As shown in Figure 5, D- mannose did not cause a substantial decrease in adhesion. However, the increase in adhesion of an unheated culture of strain K-12 after 50 minutes of incubation was less than one log cycle, therefore, the effect of D-mannose on the adhesion was not clear. Strain K-12 was very sensitive to heat, as indicated by the inability to detect viable cells in the heated culture suspension. However, some of cells appeared to be heat-injured and recovered rapidly during incubation with intestinal tissue. There was a concomi- tant, rapid increase in the adhesion of viable bacteria to the tissue. The presence of D-mannose appeared to have some effect on the viability of the injured cells. Most of the bacterial cells were destroyed during heating, thus, the adhesion of the heated bacteria was less than that of the unheated bacteria, even after 50 minutes of incubation. The increase in the adhesion of heated and unheated cells of culture A20a was also less than one log cycle (Figure 6). However, culture A20a was more heat resistant than strain K-12 as indicated by the relatively high counts of the heated culture suspension (1.0 x 107 cells/ml). In the absence of D-mannose, heated and unheated cultures 9 and 119-3 showed an increase of about 1.25 to 52 UNHEATED BACTERIAL SUSPENSION/ML ---——-----—_----_---_—- 0—0 PBS. UNHEATED H PBS, HEATED 5 H PBS+D"MANNOSE: UNHEATED H PBS+D'MANNOSE: HEATED LOG VIABLE COUNTS PER TISSUE DISC 4 3 / 2 / SMALL INTESTINAL TISSUE 0 HEATED BACT.SUSP.:3.0x10l (0) CELLS/ML 2 SEC. 10 20 30 40 50 INCUBATION TIME I MINUTES FIGURE 5. ADHESIoN 0F STRAIN K12 T0 TURKEY SMALL INTESTINAL TISSUE. 53 10 __u_~+iE_AT_ED_EI.~9_TsI_AL__s_US.PEE§m(14L_._. 9 o—-o PBS. UNHEATED o—o PBS. HEATED H PBS+D-MANNOSE. UNHEATED 8 s—s PBS+D-MANNOSE. HEATED 7 _ _ HEATELBACIEBJALSB EL” El! 2 9594;. _ __ LOG VIABLE COUNTS PER TISSUE DISC 0 1‘ 10 20 30 A0 50 2 SEC. INCUBATION TIME. MINUTES FIGURE 6. ADHESION OF CULTURE A20a (02:Kl:H6) ISOLATED FROM HUMAN T0 TURKEY SMALL INTESTINAL TISSUE. 54 1.75 log cycles in the viable bacteria adhered on tissue after 50 minutes of incubation (Figures 7 and 8). However, in the presence of 0.5 percent D-mannose, the increase in adhesion after 50 minutes of incubation was less than ten- fold; the greatest increase occurred during the first 20 minutes of incubation. There were large decreases in the viable counts of the bacterial suspension and tissue con- taining bacteria after heating at 650C for 30 minutes. Af- ter 50 minutes of incubation in the absence of D-mannose the log increases in the adhesion of heated bacteria were close to those of the unheated bacteria. When cultures 9 and 119-3 were grown at 180C (Figures 9 and 10), fewer bac- teria adhered to the tissue than when the bacteria had been grown at 37°C. During incubation there was virtually no increase in the adhesion of heated cells of cultures 9 and 119-3 grown at 180C, except for culture 9 in the presence of D-mannose. In the presence or absence of 0.5 percent D-man- nose, there was a tenfold increase in unheated cells of culture 94-5 adhered to the tissue after 50 minutes of in- cubation (Figure 11). However, in the absence of D-man- nose, the adhesion of cells increased during the entire in- cubation period, while maximum adhesion occurred after 20 minutes of incubation in the presence of D-mannose. In the absence of D-mannose, cells of culture 94-5 grown at 18°C (Figure 12) had a lower rate of adhesion than those grown LOG VIABLE COUNTS PER TISSUE DISC 55 10 _ 989411.69. LACEMESEEEEE-ME _ 9 8 Jflwflmuwufl_ ___I 7 6 o—o PBS. UNHEATED 5 PBS. HEATED 0—0 H PBS+D'MANNOSE I UNHEATED H PBS+D’MANNOSE I HEATED ........................ , ................................. .. 2."_fl%uflgfl&fl§£___ 1 0 . 2 SEC. 10 20 30 40 50 INCUBATION TIME, MINUTES FIGURE 7. ADHESION OF CULTURE 9. GRONN AT 37°C. T0 TURKEY SMALL INTESTINAL TISSUE. 56 10 9 _nyEuaJ_§A.ETERIAL_SLLSEENs__Io_~/JL__ 8 7 fiQELflflfififimL ______ 6 .___ ::::::::::::::.::::::::::::.X 0-—O PBS, UNHEATED PBS, HEATED 0—0 5 H PBS+D'MANNOSE, UNHEATED H PBS‘I'D-MANNOSE, HEATED LOG VIABLE COUNTS PER TISSUE DISC q 3 ............................... g 2 _ _ _~°flAL|-_1N_T.E§1_N_AL my; _ _ 1 o 2 SEC. 10 20 30 40 50 INCUBATION TIME I MINUTES FIGURE 8. 99HESION OF CULTURE 119-3. GROWN AT 0 . To TURKEY SMALL INTESTINAL TISSUE. LOG VIABLE COUNTS PER TISSUE DISC 57 10 UNHEATED BACTERIAL SUSPENSION/ML o-o PBS, UNHEATED 5 o-—0 PBS. HEATED 6-¢ PBS+D-MANNOSE, UNHEATED t-t PBS+D~MANNOSE, HEATED 4 2 ------------------------------- 1 _ _ _3_U£\L.L JfilEfiIUiAL. HSiUL ._ _. _ l 2 SEC_ 10 20 30 40 50 INCUBATION TIME: MINUTES FIGURE 9. ADHESION OF CULTURE 9, GRowN AT 18°C, TO TURKEY SMALL INTESTINAL TISSUE. 58 10 _.UlflliDiwflW-Jlfl’fl ELOMLL _.. 9 8 _ .5312) .993 ELSE "_ENiI QE/AL 8 7 Q . Lu " D A v U) 32 6 """ -‘------ .::::::;::::::a *— CZ DJ 0. E 5 0—0 PBS. UNHEATED g H PBS. HEATED m H PBS‘I'D-MANNOSE. UNHEATED _l 2° 1; H PBS+D-MANNOSE. HEATED S: (D 3 3 2 .______§fléEE_IEIE§IINAE_II§§2E_H_.__ 1 0 2 SEC. 10 20 30 40 50 INCUBATION TIME , MINUTES FIGURE 10. A§BESION OF CULTURE 119-3. GROWN AT 1 . TO TURKEY SMALL INTESTINAL TISSUE. 59 1° _ lNEEEEJL E91539; EESENEIMEL. _ 9 8 _ £6151 IECLEELAL Eusrgwgv/DL _ g; 7 / Q ......................... .0 LL! :3 22 :z 6 ---------------------------- '4 m o—o PBS. UNHEATED .“f 0—0 PBS. HEATED g 5 H RBS+D~MANNOSE. UNHEATED g H PBS+D-MANNOSE. HEATED 8 DJ 3% 4 :5 > L9 0 _J 3 i """""""""" —————————————————————————————— i 2 _..___§MAEL_IEIE§IINAL.IL§§Q§._.__ 2 SEC. 10 20 30 40 50 INCUBATION TIME . MINUTES FIGURE 11. ngESION OF CULTURE 94-5. GROWN AT 0 . TO TURKEY SMALL INTESTINAL TISSUE. 6O UNHEATED BACTERIAL SUSPENSION/ML— °-° PBS. UNHEATED 5 0—0 PBS. HEATED a-HA PBS+D'MANNOSE. UNHEATED .b-nl PBS+D'MANNOSE. HEATED LOG VIABLE COUNTS PER TISSUE DISC 4 3 __'_____ 2 ._ _... _.. EILALLJ NlEfl I_N_AL.T_I_s 8.9.5. _ _.. _ 2 SEC. 10 20 30 no 50 INCUBATION TIME . MINUTES FIGURE 12. AgHESION OF CULTURE 94-5. GROWN AT . T0 TURKEY SMALL INTESTINAL TISSUE. 61 at 370C. This result is similar to those reported for cul- tures 9 and 119-3. Culture 102-l had weak or negative hemagglutination reactions with animal red blood cells (Table 11). To some extent the unheated cells adhered to the epithelial tissue of turkey small intestine (Figures 13 and 14), although the increase in adhesion after 50 minutes of incubation was much lower than that of the other cultures. Inhibition of In Vitro Adhesion Inhibition by Sugars and Sugar Derivatives Culture 119-3 was used to study the effects of va- rious sugars and sugar derivatives on the adhesion of E. coli to turkey small intestinal tissue. This culture had the highest degree of adhesion among all the cultures tested, and was identified as 02:Kl, a serotype commonly found in fatal infections in poultry. All sugars and sugar derivatives, except D-glucose at 0.5 percent and N-acetyl— neuraminic acid at 0.5 and 1.0 percent, caused a statisti- cally significant, but partial inhibition (P<:0.0l) of E. coli adhesion to turkey small intestinal tissue (Figures 15 and 16). However, only inhibition by D-mannose and a- methyl-D-mannoside at 0.5 and 1.0 percent concentrations resulted in more than one log cycle decrease in the adhe- sion. Adhesion was not inhibited completely even by 1.0 percent D-mannose or a-methyl-D-mannoside. 62 10 UNHEATED BACTERIAL SUSPENSION/ML 0--0 PBS. UNHEATED H PBS. HEATED H PBS‘I’D'MANNOSE. UNHEATED H PBS+D'MANNOSE. HEATED LOG VIABLE COUNTS PER TISSUE DISC U'l L} 3 I --------------- 1 ..................... _. 2 _ _ _SML-l-JNLESJ I_NAL LISEUE _ _ l 0 2 ggc. 10 20 30 no 50 INCUBATION TIME. MINUTES FIGURE 13. éyHESION OF CULTURE 102-l. GROWN AT 0 . TO TURKEY SMALL INTESTINAL TISSUE. 10 LOG VIABLE COUNTS PER TISSUE DISC 63 UNHEATED BACTERIAL SUSPENSION/ML °—° PBS. UNHEATED H PBS. HEATED b—A PBS'I'D'MANNOSE. UNHEATED H PBS‘I’D'MANNOSE . HEATED SMALL INTEST I NAL TISSUE I 2 SEC. FIGURE 14. 10 20 30 H0 50 INCUBATION TIME. MINUTES AgHESION OF CULTURE 102-l. GROWN AT 1 0 . TO TURKEY SMALL INTESTINAL TISSUE. 64 -z_ Lo mmhzziz mm mmhu< wmm mammih D<2~Hmmhz~ 44mxm2h Oh onm .zo~h_p<>_amn mm mammap szspmmpzi 44<2m >mx 2H 0H zo_mm:a< Am-mfia waspsan saga .m no zo_H~m~:z_ .ma mazes; .zIm H o m . . nu . Nd w . m . 2 w. . . R .. . 3.0 0 N .. . W . .. m5 Figure 21. 72 SDS-gels of (A) Ammonium sulfate precipitate of pili; (B) Pili after chromatography on a Sepha- dex G-SO column; (C) Standard proteins. 73 10 PHOSPHORYLASE B 8 :T BOVINE SERUM ALBUMIN c: 6 H x . " OVALBUMIN 55 4 L9 PI ——5 E (34'560) CARBONIC ANHYDRASE a: 15 :3 2 SOYBEAN TRYPSIN INHIBITOR t“). _I O E a'LACTALBUMIN 0 0.2 0.4 0.6 0.8 1.0 RELATIVE MOBILITY FIGURE 22. STANDARD CURVE FOR THE DETERMINATiiN 9E APPROXI- MATE MOLECULAR WEIGHT OF CULTURE 9- PILI BY SOs-GEL ELECTROPHORESIS. 74 Figure 23. Electron micrographs of (A) Ammonium sulfate precipitate of culture 119-3 pili; (B) Pili after chromatography on a Sephadex G-50 column; Concentration, 10 mg/ml; Magnification, x 64,000. DISCUSSION E. coli Serotypes Found in Diseased Poultry Sojka and Carnaghan (1961), and Heller and Smith (1973) have reported that over 60 percent of the E. coli strains causing fatal infections in chickens belonged to OlzKl, 02:Kl, and O78:K80 serotypes. Sojka and Carnaghan (1961) also isolated strains in the 08 group from cases of coli-septicemia in chickens. However, Harry (1964) found that more than half of the strains from coli-septicemia outbreaks belonged to the 02 group, about 25 percent be— longed to the 078 group, and 8 percent belonged to the 01 group. According to Hemsley and Harry (1965), 01, 02 and 078 groups of E. coli were involved in outbreaks of ompha- 1itis and coliform pericarditis, with 65 percent belonging to 02 group. The same serotypes were detected in E. coli cultures isolated from diseased turkeys in Michigan; they were identified in the 01, 02 and 078 groups. No culture was identified in the 08 group. Cultures 9 and 102-1 were identified as O78:K80, and cultures 94-5 and ll9-3 were identified as Ol:Kl and 02:Kl, respectively. Cultures 9, 94-5, and 119-3 were isolated from the hearts of turkeys with pericarditis. The screening procedure for serotyping 75 76 of E. Egll was a rapid method for the presumptive identifi- cation of O and K serotypes of E. gel; isolated from diseased poultry, however, it might not give a complete identification of O and K serotypes due to the possibility of cross-reactions between 0 antigens of E. ggli, and the selected 0 and K antisera used. Size of E. coli Pili Several investigators have reported pili from E. 99;; of varying diameter, including type 1 pili which had a diameter of 7 nm (Brintongg gl., 1964), K99 with a diameter between 7.0 and 9.8 nm (Isaacson, 1977), and pili from other E. 99;; strains which had a diameter between 5 and 7 nm (Korhonen 3E gl., 1980). Electron photomicrographs from this inVestigation show that culture 9 had the smallest pili, with an average diameter of 6.2 nm, while culture 94-5 had the largest pili with an average diameter of 9.5 nm. The length of the pili was not measured due to the irregular length of the pili. Hemagglutination Activities of E. coli The HA pattern (NNSSS) exhibited by cultures 9 and 119-3 indicated the presence of a mannose-sensitive (MS) hemagglutinin, and was one of the HA type III patterns (NNSS without Rhesus monkey red blood cells) described by Evans £3.3l- (1979). However, the HA reactions were eluted or disappeared when the test mixtures were warmed to 40 and 77 500C, which disagreed with one of the characteristics of MS hemagglutinins described by Duguid 33 3;. (1979). Of the 170 HA type III cultures tested by Evans 33 EE. (1979), 82 percent belonged to EPEC serogroups, 5 percent were CFA- negative ETEC, 2 percent were Shigella-like serogroups, and 8 percent were isolated from children with diarrhea but did not belong to the known EPEC serogroups. The majority of the HA type III strains which did not belong to the above serogroups have been reported to be associated with sporadic diarrhea and extraintestinal infections in humans (Czirok 33 gl., 1976, 1977; ¢rskov gE gl., 1977; Evans 3E gl., 1979), including some strains which were cross-reactive with Salmonella and EPEC serogroups (Evans 33 El., 1979) The HA type III phenotype might be responsible for the virulence of E. Egll- Evans 35 3;. (1979) suggested that EPEC might possess a specific virulence factor known as MS hemaggluti- nin. No individual somatic antigen or serogroup has been reported to relate consistently to the HA type III pheno- type, and the surface factors responsible for this HA pat- tern have not been determined. The HA type III pattern could be related to the possession of type 1 pili (SNSS; HA type IV), and another factor which masked MS HA of human red blood cells so that a NNSS (type III) pattern was ob- served rather than a SNSS (type IV) pattern. Figure 24 shows a schematic diagram representing the interrelationship among ETEC, EPEC, Shigella-like strains and non-EPEC. Region 78 Enterotoxigenic Shigella genera (non-coli) Salmonella \ \‘ EPEC \Shi Non-EPEC or normal flora (HA type IV) Figure 24. Schematic diagram representing inter- relationship between ETEC, EPEC, Shi ella-like E. coli, and non-EPEC (Evans §E_gl., 1979 . 79 A represents the ETEC isolates which belong to EPEC sero- groups and those ETEC which exhibit the HA type III pheno- type. Region B represents the Shigella-like isolates which exhibit the HA type III phenotype. Region C represents the facultatively enteropathogenic E. 22$; (FEEC) groups which are recognized as normal flora but exhibit type III pheno- type and may cause sporadic diarrhea, enteritis, and extra- intestinal infections. Culture 9 was identified as O78:K80, whereas cul- ture 119-3 was identified as 02:Kl. 02 serotypes were found as normal flora in humans and were not recognized as EPEC (¢rskov SE §l°' 1977). In addition, Evans 2E.2l- (1979) found that 02 strains were associated with sporadic diarrhea and extraintestinal infections. 02 and 078 sero- types were two of the 18 O-serotypes proposed by Evans 3E EE. (1979) as FEEC. These two O-serotypes have also been reported as strains which commonly cause fatal infections in poultry, including coli-septicemia, omphalitis, and pericarditis (Sojka and Carnaghan, 1961; Harry, 1964; Hemsley and Harry, 1965; Heller and Smith, 1973). Evans 3E il- (1979) suggested that non-EPEC serogroups could be con- sidered as enteropathogens on the basis of at least two of the following criteria: (1) possession of HA type III phe- notype, (2) cross-reaction with Salmonella serogroups, (3) cross-reaction with EPEC serogroups, (4) association with sporadic diarrhea or enteritis, with or without septicemia, 80 and (5) association with extraintestinal infections. The presence of a certain serogroup in stools or intestine is not a primary indication that the culture is an EPEC since these E. ggli serogroups are known to be members of the normal flora. Initiation of disease by EPEC, according to Sakazaki 23 ii‘ (1974), was dependent on the special abili- ty of the bacteria to multiply in the intestine of the host and thereby develop a large population. Although cultures 9 and 119-3 possessed HA type III phenotype and associated with extraintestinal infections (pericarditis), these cul- tures should not be considered as human enteropathogens based on the criteria given by Evans 3E El- (1979) since these cultures were isolated from turkeys. The HA pattern (SNSSS) exhibited by culture 94—5 was similar to the HA type IV phenotype, and it was exhib- ited by many strains including strain K412 which was known to possess type 1 pili (Evans g3 El°r 1979). Of the 74 cultures tested by Evans 33 EE. (1979) and known to possess this HA type IV phenotype, 31 percent belonged to EPEC serogroups, 12 percent were CPA-negative ETEC, and 16 per- cent were isolated from children with diarrhea but did not belong to the known EPEC serogroups. The HA pattern of heated culture 94-5 was NNSSS, similar to the pattern of the HA type III. Thus, this culture may possess only type 1 pili. However, the nega- tive HA of the heated culture with human red blood cells 81 and the eluted reactions when the temperature of test mix— tures was raised to 40 and 500C were unexpected (Duguid 35 3E., 1979). Another possible explanation of the result is that culture 94-5 possessed two MS hemagglutinins, a type III hemagglutinin which was heat resistant and responsible for MS HA with chicken, guinea pig, and Rhesus monkey red blood cells, and another responsible for MS HA of human red blood cells and possibly chicken, guinea pig and Rhesus monkey red blood cells, but its HA activity was destroyed when the culture was heated at 65°C for 30 minutes. Evans 33 3;. (1979) suggested that at least two factors, the type 1 pili and an unidentified factor, were associated with MS HA of EPEC. MS, as well as MRE, hemagglutinins of E. 99;; might function in adhesion of bacteria to host cells (Salit and Gotschlich, 1977; Evans 33 §l°r 1979). The HA patterns NNNSS and NNNSN (or NNNS without Rhesus monkey red blood cells) exhibited by unheated cul- ture 102-l after growth on CFA and in CYE broth, respec- tively, were not included in the HA patterns described by Evans §E_§E. (1979). The reactions were very weak (2+) in all MS HA. These results could be expected since electron microscopy of culture 102-l after growth on Peptone agar revealed no pili were present. 82 Adhesive Properties of E. coli Jones and Rutter (1972) reported an increase of 1.6 to 2.0 log cycle in the viable counts of piglet intestinal tissue incubated for 30 minutes with a K88—positive E. ggii, and less than a tenfold increase in the viable counts when the tissue was incubated with a K88-negative strain. In this research, cultures 9 and 119-3 showed an increase of about 1.25 to 1.75 log cycles in the viable bacteria adhered on turkey small intestinal tissue after 50 minutes of incubation. D-mannose caused partial inhibition in the adhesion of cultures 9 and 119-3 to turkey small in— testinal tissue. This result could be indicative of the fact that more than one carbohydrate-containing receptor for E. gel; was present in the intestinal epithelial cells, or that the concentration of D-mannose was not high enough to cause complete inhibition. Another possibility is that there was nonspecific adhesion of bacteria, which was not mediated by pili, in the presence of D-mannose. The hemag- glutination reactions of these cultures with animal red blood cells were completely inhibited by D—mannose at a concentration of 0.5 percent, indicating that mannose-con- taining receptors were the only receptors or were adjacent to the receptors in the red blood cells for the bacterial pili. Culture A20a, which was isolated from human appendix and had the same O:K serotype (02:K1) as culture 119-3, adhered to turkey small intestinal tissue at much lower 83 rate than culture ll9-3, indicating that the host and/or tissue specificities influenced the adhesion of E. 99;; to animal cells, as was suggested by Smith (1972, 1976). Schaeffer 33 El“ (1979) considered two different attachment mechanisms occurred in the adhesion of bacteria to human or animal epithelial cells, one was the temporary binding which happened as soon as bacteria were in contact with the epithelial cells, and the second was a more stable form of adhesion. The stable adhesion was related to the limited number of receptors on the epithelial cells for the bacterial pili (Jones 33 El°r 1976; Isaacson g5 §l°1 1978; Schaeffer 3E El., 1979). Accordingly, in the absence of D-mannose, the adhesion of heated and unheated cultures 9 and 119-3 during the first 20-30 minutes of incubation would be due to the temporary binding. A more stable adhe- sion occurred upon prolonged incubation. However, in the absence of D-mannose, some of the bacteria, such as the heated culture 119-3, could not develop a more stable adhe- sion, and bacteria were readily detached from the epithelial cells. In the presence of D-mannose, the competition be- tween the receptors and D-mannose for the bacterial pili also resulted in a decrease in the amount of bacteria adhered to the tissue. Jones £2 3;. (1976) considered such a reduction in the number of bacteria attached to epithe- lial cells to be associated with the spontaneous loss or denaturation of the adhesive component (adhesin) on the 84 bacterial surface. The thermal death times of E. 92;; cells generally are between 20 and 30 minutes at 57.30C (Frazier and Westhoff, 1978). Thus, a large portion of cultures 9 and 119-3 were killed or possibly injured during heating at 650C for 30 minutes, and would not be detected even if they had adhered to the epithelial cells. The HA of these cul- tures were resistant to heating at 65°C for 30 minutes. After 50 minutes of incubation in the absence of D-mannose, the increases in the adhesion of heated bacteria were close to those of the unheated bacteria, indicating the repair of some of the injured cells. Ray and Speck (l973ab) re- ported that repair of E. 99;; cells injured during freezing occurred in foods during one hour at 250C; most of the in- jured cells could grow on a non-selective medium such as Trypticase soy agar containing 0.3 percent yeast extract (TSYA), but not in selective media such as Violet red bile agar (VRBA) or Deoxycholate-lactose agar (DLA). A similar condition might occur with the heat-injured bacteria incu- bated with the intestinal tissue, however, there might be a difference in the injuries and repair mechanisms between heat-injured cells and freezing-injured cells. Partial inhibition in the pili formation might have occurred when cultures 9 and 119-3 were grown at 18°C since fewer bacteria adhered to the tissue than when the bacteria had been grown at 37°C. However, growth at 18°C did not 85 affect the hemagglutination reactions. Interestingly, upon prolonged incubation, the viable counts of bacteria adhering to the tissue increased. The increase in the amount of adhesion might be due to reformation of pili when the culture was brought to a higher temperature, i.e. incu- bation at room temperature during the adhesion tests. The presence of D-mannose appeared to affect the attachment, since larger increases in the adhesion of bacteria occurred in the presence of D-mannose. Novotny EE.§l° (1969) studied the reformation of E. 92;; pili after the pili were removed by blending; they reported that the length of type 1 pili increased from 35 to 90 percent of the control length in 30 minutes at 37°C. Cultures 9 and 119—3 were more sensitive to heat when grown at 18°C, and no apparent repair of the cells occurred even during incubation with the tissue. Frazier and Westhoff (1978) stated E. ggli was more heat- resistant when grown at 38.5°C than at 280C. Culture 94-5 was more heat sensitive than cultures 9 and 119-3 as indicated by the relatively low numbers of heated bacteria adhering to the tissue. However, the pres- ence of D-mannose seemed to have some effect(s) on the re- pair and/or adhesion of this culture since the apparent adhesion of heated bacteria increased in the presence of D- mannose. When culture 94-5 had been grown at 18°C, the presence of D-mannose also might have some effect(s) on the reformation of pili during incubation at room temperature. 86 The absence of adhesion of cells of this culture heated at 650C for 30 minutes was also indicative of their increased sensitivity to heat when the culture was grown at 180C for 72 hours. Electron microscopy showed the absence of pili in culture 102-1, therefore, it was concluded that the in- crease in the adhesion of this culture to turkey small in- testinal tissue, which was much lower than that of the other cultures, was due to nonspecific adhesion, i.e. was not mediated by pili. This suggestion was supported by the fact that the presence of D-mannose did not affect the ad- hesion of cells grown at 37 and 180C, indicating a mannose- containing receptor was not involved in the adhesion of these bacteria. Inhibition of In Vitro Adhesion Recent reports (Jones and Freter, 1976; Ofek 3E El-r 1977, 1978; Salit and Gotschlich, 1977) indicate that some bacteria may bind to carbohydrate molecules as recep- tors on epithelial cell surfaces. Thus, if specific carbo- hydrates are added to the incubation mixture, adhesion is inhibited due to competition between carbohydrates and car- bohydrate-containing receptors for the adhesion, in this case, pili. These studies have shown that hemagglutina- tion, caused by adhesion of cultures 9, 94-5 and 119-3 to animal red blood cells, was completely inhibited by 87 D-mannose. D-mannose and a—methyl-D-mannoside resulted in more than one log cycle decrease in the adhesion of E. 39;; (culture 119-3) to turkey small intestinal tissue. The results suggested that saturation of binding sites on the bacterial surface, in this case pili, by D-mannose and a- methyl-D-mannoside prevented the adhesion of the bacteria to the epithelial membrane receptors, which presumably con- tained mannose or a mannose-like structure. The inhibition of adhesion by D-mannose and a-methyl-D-mannoside is simi- lar to the results reported by Ofek 32 2;. (1977, 1978) who suggested that the binding of E. gel; to human epithelial cells was mediated by a mannose-specific, lectin-like sub- stance present in the pili which bound to mannose or man- nose-like receptor sites on the mucosal cells. However, adhesion of culture ll9-3 to turkey small intestinal tissue was not completely inhibited up to a concentration of 2.5 percent sugars. The incomplete inhibition probably was due to the reversible binding of D-mannose and a-methyl-D-man- noside to the pili, so that some bacteria were able to re- place the sugars and bind to the epithelial cells. The inhibition of adhesion by other sugars, which resulted in less than one log cycle decrease in the adhe- sion, might be due to a less powerful binding of the sugars or sugar derivatives to the bacterial surface. Jones (1975) suggested that the fi-D-galactosyl residue of the mucosal 88 glycoprotein in the pig intestine might be involved in the adhesion of E. ggli, but the adhesion could not be inhibit— ed by a simple sugar or by a small oligosaccharide. {In this experiment, N-acetyl-D—galactosamine at 1.0 percent concentration caused a partial (0.6 log cycle) decrease in the adhesion of E. gel; to turkey small intestinal tissue. Concanavalin A and sodium metaperiodate greatly in- hibited the adhesion of E. 92;; (culture ll9-3) to turkey small intestinal tissue. Concanavalin A is a lectin that binds to D-mannose and D-glucose residues (Sharon and Lis, 1972), while sodium metaperiodate is an oxidative compound known to cleave the C-C bond between the vicinal hydroxyl groups of sugars. The inhibition of adhesion caused by these compounds lends credence to the hypothesis that man- nose or mannose-like receptors are important in the adhe- sion of E. 99;; to turkey small intestinal tissue. Since the ig XiEEQ adhesion of E. Egli to intesti- nal epithelial cells has been suggested to be mediated by pili (Jones and Rutter, 1972; Isaacson g; El-r 1977, 1978; Moon 33 3E., 1977; Nagy 3E EE., 1977), isolated pili should act as competitive inhibitors of bacterial adhesion. The presence of isolated pili decreased the adhesion of E, 99;; (culture 119-3) to turkey small intestinal tissue, however, the adhesion was not completely inhibited. The incomplete inhibition of adhesion by isolated pili might be due to a decrease in the adhesive activity of the isolated pili as 89 compared to the native, attached pili. Isaacson 33 E£° (1978) reported that purified homologous pili at a concen- tration of 2000 ug per ml successfully competed for the ad- hesion of E. 991E strains possessing 987P or type 1 pili on pig intestinal epithelial cells; however, addition of puri- fied K99 had a variable effect on the adhesion of K99+ strain, including either competition or enhancement of ad- hesion. These investigators suggested that some specific effects, which were variable from experiment to experiment, resulted from the addition of purified, homologous pili. CONCLUSIONS The E. 22$; cultures, isolated from diseased tur- keys during several disease outbreaks in Michigan, consist- ed primarily of three O-serogroups, Ol, 02 and 078 groups. Culture 9 and 102-1 were identified as O78:K80, while cul- tures 94-5 and ll9-3 were identified as Ol:Kl and 02:K1, respectively. These bacteria possessed pili with a diame- ter between 6.2 and 9.5 nm. The cultures isolated from diseased turkeys exhib- ited mannose-sensitive hemagglutination activities (MS HA), indicating that they possessed MS hemagglutinins. Cultures 9 and 119-3 exhibited one of the type III HA patterns (NNSSS), which has been found in some human EPEC strains. Culture 94-5 exhibited one of the type IV HA patterns (SNSSS), which also has been found in some human EPEC strains and in E. 921$ strains that possessed type 1 pili. Although HA activities of the culture isolated from diseased turkeys were substantially affected by heating at 650C for 30 minutes, the i3 XEEEQ adhesive properties were greatly affected by heating. Heating at 65°C for 30 min- utes caused reductions of more than two log cycles in the numbers of bacteria adhering to turkey small intestinal 90 91 tissue. Growth at 18°C rather than 370C also resulted in reductions in adhesion, although growth at that temperature did not affect the HA activities. During incubation at room temperature the numbers of bacteria adhering to turkey small intestinal tissue increased by ca. 1.25 to 1.75 log cycles for cultures 9 and 119-3, and by ca. 1.0 log cycle for culture 94-5. Among the sugars tested, only D—mannose and a- methyl-D-mannoside had a marked inhibitory effect on the adhesion of culture 119-3 to turkey small intestinal tis- sue. D-mannose and a-methyl-D-mannoside at concentrations of 0.4 and 0.5 percent, respectively, reduced the numbers of bacteria adhering to the turkey small intestinal tissue by 1.0 log cycle; however, adhesion was not completely in- hibited at concentrations of these compounds as high as 2.5 percent. Concanavalin A and sodium metaperiodate at con- centrations of 0.2 and 0.1 percent, respectively, complete- ly inhibited the adhesion of culture 119-3 to turkey small intestinal tissue. Adhesion was substantially inhibited by isolated pili, however, at concentrations higher than 2.0 mg protein per m1, no further inhibition was detected. The isolated pili were less inhibitory than D-mannose, a-methyl -D-mannoside, concanavalin A, and sodium metaperiodate. SDS gel electrophoresis indicated protein from the isolated pili had a molecular weight of ca. 34,800. 92 Further research should be directed towards the study of i5 yiyg adhesion of the cultures isolated from diseased turkeys to turkey small intestinal tissue. Studies of the effects of different sugars, sugar deriva- tives, and isolated pili on the £2 vivo adhesion and other factors which affect adhesion could yield information which may be used to reduce the incidence and/or severity of E. coli infections in poultry. REFERENCES REFERENCES Brinton, C.C., Jr. 1959. Non-flagellar appendages of bac- teria. Nature (London) 183: 782-786. Brinton, C.C., Jr. 1965. The structure, function, synthe- sis and genetic control of bacterial pili and a mo- lecular model for DNA and RNA transport in gram- negative bacteria. Trans. N.Y. Acad. Sci. 27: 1003-1054. Brinton, C.C., Jr., P. Gemski, Jr., and J. Carnahan. 1964. A new type of bacterial pilus genetically controlled by the fertility factor of E. coli K12 and its role in chromosome transfer. Proc. Nat. Acad. Sci. U.S.A. 52: 776-783. Brinton, C.C., Jr., and M.J. Stone. 1961. Chemical compo- sition of bacterial pili - a fibrous protein. Bac- teriol. Proc., 1961: 96. Burrows, M.R., R. Sellwood, and R.A. Gibbons. 1976. Hae- magglutinating and adhesive properties associated with the K99 antigen of bovine strains of Escheri- chia coli. J. Gen. Microbiol. 96: 269-275. Cahill, E.E., and P.J. Glantz. 1978. Demonstration of the K88ac and K88ab antigens of Escherichia coli by means of immunoelectrophoresis and immunodiffusion. Infect. Immun. 20: 811-815. Carter, G.R. 1976. Essentials of Veterinary Bacteriology and Mycology. Michigan State University Press. East Lansing. Cooper, T.G. 1977. The Tools of Biochemistry. John Wiley and Sons. New York. Czirok, E., J. Madar, J. Budai, Z. Stverteczky, A. Kertesz, I. Dombi, and L. Gyengesi. 1976. The role in spo- radic enteritis of facultatively enteropathogenic Escherichia coli serogroups. Acta Microbiol. Acad. Sci. Hung. 23: 359-369. 93 94 Czirok, E., H. Milch, J. Madar, and G. Semjen. 1977. Characterization of Escherichia coli serogroups causing meningitis, sepsis and enteritis. I. Sero- logical properties and animal pathogenicity of 018, 078 and 083 isolates. Acta Microbiol. Acad. Sci. Hung. 24: 115-126. Duguid, J.P. 1964. Functional anatomy of Escherichia coli with special reference to enteropathogenic E. coli. Rev. Lat. Amer. Microbiol. 7: 1-16. Duguid, J.P. 1968. The function of bacterial fimbriae. Arch. Immunol. Ther. Exp. 16: 173-188. Duguid, J.P., E.S. Anderson, and I. Campbell. 1966. Fim- briae and adhesion in salmonellae. J. Pathol. Bac- teriol. 92: 107-138. Duguid, J.P., S. Clegg, and M.I. Wilson. 1979. The fim- brial and non—fimbrial haemagglutinatinins of Escherichia coli. J. Med. Microbiol. 12: 213-227. Duguid, J.P., and R.R. Gillies. 1957. Fimbriae and adhe- sive properties in dysentery bacilli. J. Pathol. Bacteriol. 74: 397-411. Duguid, J.P., I.W. Smith, G. Dempster, and P.N. Edmunds. 1955. Non-flagellar filamentous appendages ("fim- briae") and hemagglutinating activity in Bacterium coli. J. Pathol. Bacteriol. 70: 335-348. DuPont, H.L., S.B. Formal, R.B. Hornick, M.J. Snyder, J.P. Libonati, D.G. Sheahan, E.H. LaBrec, and J.P. Kalas. 1971. Pathogenesis of Escherichia coli diarrhea. N. Engl. J. Med. 285: 1-9. Evans, D.G., and D.J. Evans, Jr. 1978. New surface-asso- ciated heat-labile colonization factor antigen (CFA/II) produced by enterotoxigenic Escherichia coli of serogroups O6 and 08. Infect. Immun. 21: 638-647. Evans, D.G., D.J. Evans, Jr., and W.S. Tjoa. 1977. Hemag- glutination of human group A erythrocytes by en- terotoxigenic Escherichia coli isolated from adults with diarrhea: correlation with colonization fac- tor. Infect. Immun. 18: 330-337. 95 Evans, D.G., D.J. Evans, Jr., W.S. Tjoa, and H.L. DuPont. 1978. Detection and characterization of coloniza- tion factor of enterotoxigenic Escherichia coli isolated from adults with diarrhea. Infect. Immun. 19: 727-736. Evans, D.G., R.P. Silver, D.J. Evans, Jr., D.G. Chase, and S.L. Gorbach. 1975. Plasmid-controlled coloniza- tion factor associated with virulence in Escheri- chia coli enterotoxigenic for humans. Infect. Immun. 12: 656-667. Evans, D.J., Jr., D.G. Evans, and H.L. DuPont. 1979. He- magglutination patterns of enterotoxigenic and en- teropathogenic Escherichia coli determined with hu- man, bovine, chicken, and guinea pig erythrocytes in the presence and absence of mannose. Infect. Immun. 23: 336-346. Evans, D.J., Jr., D.G. Evans, and S.L. Gorbach. 1973. Production of vascular permeability factor by en- terotoxigenic Escherichia coli isolated from man. Infect. Immun. 8: 725-730. Forstner, G.G. 1971. Release of intestinal surface-mem- brane glycoproteins associated with enzyme activity by brief digestion with papain. Biochem. J. 121: 781-789. Frazier, W.C., E.H. Marth, and R.H. Deibel. 1968. Labora- tory Manual for Food Microbiology. Burgess Publ. Co. Minneapolis, Minnesota. Frazier, W.C., and D.C. Westhoff. 1978. Food Micro- biology. McGraw-Hill Book Co. New York. Freer, J.H., A. Eliss, T. Wadstrom, and C.J. Smyth. 1978. Occurrence of fimbriae among enterotoxigenic intes- tinal bacteria isolated from cases of human infan- tile diarrhoea. FEMS Microbiol. Lett. 3: 277-281. Guinée, P.A.M., W.H. Jansen, and C.M. Agterberg. 1976. Detection of the K99 antigen by means of agglutina- tion and immunoelectrophoresis in Escherichia coli isolates from calves and its correlation with en- terotoxigenicity. Infect. Immun. 13: 1369-1377. Gyles, C.L. 1971. Discussion: heat-labile and heat-stable forms of the enterotoxin from Escherichia coli strains enteropathogenic for pigs. Ann. N.Y. Acad. Sci. 176: 314-322. 96 Gyles, C.L., and D.A. Barnum. 1969. A heat-labile entero- toxin from strains of Escherichia coli enteropatho- genic for pigs. J. Infect. Dis. 120: 419-426. Harry, E.G. 1964. A study of 119 outbreaks of coli-septi- caemia in broiler flocks. Vet. Rec. 76: 443-448. Heller, E.D., and H.W. Smith. 1973. The incidence of an- tibiotic resistance and other characteristics amongst Escherichia coli strains causing fatal in- fection in chickens: the utilization of these char- acteristics to study the epidemiology of the infec- tion. J. Hyg. Camb. 71: 771-781. Hemsley, L.A., and E.G. Harry. 1965. Coliform pericardi- tis (coli septicaemia) in broiler chickens: a three -year study on one farm. Vet. Rec. 77: 103-107. Hoeniger, J.F.M. 1965. Development of flagella by Proteus mirabilis. J. Gen. Microbiol. 40: 29-42. Isaacson, R.E. 1977. K99 surface antigen of Escherichia coli: purification and partial characterization. Infect. Immun. 15: 272-279. Isaacson, R.E., P.C. Fusco, C.C. Brinton, and H.W. Moon. 1978. E2 vitro adhesion of Escherichia coli to porcine small intestinal epithelial cells: pili as adhesive factors. Infect. Immun. 21: 392-397. Isaacson, R.E., B. Nagy, and H.W. Moon. 1977. Coloniza- tion of porcine small intestine by Escherichia coli: colonization and adhesion factors of pig en- teropathogens that lack K88. J. Infect. Dis. 135: 531-539. Ito, S. 1969. Structure and function of the glycocalyx. Fed. Proc. Fed. Am. Soc. Exp. Biol. 28: 12-25. Jones, G.W. 1975. Adhesive Properties of Enteropathogenic Bacteria. E3: Microbiology - 1975 (D. Schlesinger, ed.), pp. 137-142. Am. Soc. Microbiol., Washing- ton, D.C. Jones, G.W. 1977. The Attachment of Bacteria to The Sur- faces of Animal Cells. In: Microbial Interactions (J.L. Reissig, ed.), seriEs B, vol.3, pp. 139 London. 97 Jones, G.W., G.D. Abrams, and R. Freter. 1976. Adhesive properties of Vibrio cholerae: adhesion to isolated rabbit brush border membranes and hemagglutinating activity. Infect. Immun. 14: 232-239. Jones, G.W., and R. Freter. 1976. Adhesive properties of Vibrio cholerae: nature of the interaction with isolated rabbit brush border membranes and human erythrocytes. Infect. Immun. 14: 240-245. Jones, G.W., and J.M. Rutter. 1972. Role of the K88 anti- gen in the pathogenesis of neonatal diarrhea caused by Escherichia coli in piglets. Infect. Immun. 6: 918-927. Jones, G.W., and J.M. Rutter. 1974. The association of K88 with hemagglutinating activity in porcine strains of Escherichia coli. J. Gen. Microbiol. 84: 135-144. Korhonen, T.K., E.L. Nurmiaho, H. Ranta, and C.S. Eden. 1980. New method for isolation of immunologically pure pili from Escherichia coli. Infect. Immun. 27: 569-575. Kovacs, N. 1956. Identification of Pseudomonas pyocyanea by the oxidative reaction. Nature 178: 703. Lowry, O.H., N.J. Rosebrough, A.L. Farr, and R.J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265-275. McMichael, J.C., and J.T. Ou. 1979. Structure of common pili from Escherichia coli. J. Bacteriol. 138: 969-975. McNeish, A.S., J. Fleming, P. Turner, and N. Evans. 1975. Mucosal adherence of human enteropathogenic Esche- richia coli. Lancet 2: 946-948. Mehlman, I.J., N.T. Simon, A.C. Sanders, and J.C. Olson,Jr. 1974. Problems in the recovery and identification of enteropathogenic Escherichia coli from foods. J. Milk Food Technol. 37: 350-356. Moon, H.W., E.M. Kohler, R.A. Schneider, and S.C. Whipp. 1980. Prevalence of pilus antigens, enterotoxin types, and enteropathogenicity among K88-negative enterotoxigenic Escherichia coli from neonatal pigs. Infect. Immun. 27: 222-230. 98 Moon, H.W., B. Nagy. R.E. Isaacson, and I. ¢rskov. 1977. Occurrence of K99 antigen on Escherichia coli iso- lated from pigs and colonization of pig ileum by K99+ enterotoxigenic E. coli from calves and pigs. Infect. Immun. 15: 614-620. Morris, J.A., A.E. Stevens, and W.J. Sojka. 1978. Iso- electric point of cell-free K99 antigen exhibiting hemagglutinating properties. Infect. Immun. 19: 1097-1098. Nagy, B., H.W. Moon, and R.E. Isaacson. 1977. Coloniza- tion of porcine small intestine by enterotoxigenic Escherichia coli: selection of piliated forms lg vivo, adhesion of piliated forms to epithelial cells 23 vitro, and incidence of a pilus antigen among porcine enteropathogenic E. coli. Infect. Immun. 16: 344-352. Nagy, B., H.W. Moon, R.E. Isaacson, C.C. To, and C.C. Brin- ton. 1978. Immunization of suckling pigs against enteric enterotoxigenic Escherichia coli infection by vaccinating dams with purified pili. Infect. Immun. 21: 269-274. Novotny, C., J. Carnahan, and C.C. Brinton, Jr. 1969. Me- chanical removal of F pili, type I pili, and fla- gella from Hfr and RTF donor cells and the kinetics of their reappearance. J. Bacteriol. 98: 1294- 1306. Nowotarska, M., and M. Mulczyk. 1977. Serologic relation- ship of fimbriae among Enterobacteriaceae. Arch. Immunol. Ther. Exp. 25: 7-46. Ofek, I., E.H. Beachey, and N. Sharon. 1978. Surface su- gars of animal cells as determinants of recognition in bacterial adherence. Trends Biochem. Sci. 3: 159-160. Ofek, I., D. Mirelman, and N. Sharon. 1977. Adherence of Escherichia coli to human mucosal cells mediated by mannose receptors. Nature 265: 623-625. Old, D.C., I. Corneil, L.F. Gibson, A.D. Thomson, and J.P. Duguid. 1968. Fimbriation, pellicle formation and the amount of growth of salmonellas in broth. J. Gen. Microbiol. 51: 1-16. 99 Orskov, F., and I. ¢rskov. 1975. Escherichia coli O:H se- rotypes isolated from human blood: prevalence of the K1 antigen with technical details of 0 and H antigen determination. Acta Pathol. Microbiol. Scand. 83: 595-600. Orskov, I., and F. ¢rskov. 1966. Episome-carried surface antigen K88 of Escherichia coli. I. Transmission of the determinant of the K88 antigen and the in- fluence on the transfer of chromosomal markers. J. Bacteriol. 91: 69-75. Orskov, I., and F. ¢rskov. 1977. Special O:K:H serotypes among enterotoxigenic E. coli strains from diarrhea in adults and children. Med. Microbiol. Immunol. 163: 99-110. ¢rskov, I., F. ¢rskov, and A.E. Andersen. 1980. Compari- son of Escherichia coli fimbrial antigen F7 with type 1 fimbriae. Infect. Immun. 27: 657-666. Orskov, I., F. Orskov, B. Jann, and K. Jann. 1977. Sero- logy, chemistry, and genetics of 0 and K antigens of Escherichia coli. Bacteriol. Rev. 41: 667-710. ¢rskov, I., F. ¢rskov, H.W. Smith, and W.J. Sojka. 1975. The establishment of K99, a thermolabile, transmis- sible Escherichia coli K antigen, previously called "Kco", possessed by calf and lamb enteropathogenic strains. Acta Pathol. Microbiol. Scand. 83: 31-36. Orskov, I., F. Orskov, W.J. Sojka, and W. Wittig. 1964. K antigens K88ab (L) and K88ac (L) in E. coli. A new 0 antigen: 0141 and a new K antigen: K89 (B). Acta Pathol. Microbiol. Scand. 62: 439-447. Ottow, J.C.G. 1975. Ecology, physiology, and genetics of fimbriae and pili. Ann. Rev. Microbiol. 29: 79- 108. Ray, B., and M.L. Speck. 1973a. Discrepancies in the enu- meration of Escherichia coli. Appl. Microbiol. 25: 494-498. Ray, B., and M.L. Speck. 1973b. Enumeration of Escheri- chia coli in frozen samples after recovery from in- jury. Appl. Microbiol. 25: 499-503. Sakazaki, R., K. Tamura, A. Nakamura, T. Kurata, A. Gohda, and S. Takeuchi. 1974. Enteropathogenicity and enterotoxigenicity of human enteropathogenic Esche- richia coli. Jpn. J. Med. Sci. Biol. 27: 19-33. 100 Salit, I.E., and E.C. Gotschlich. 1977. Type 1 Escheri- chia coli pili: characterization of binding to mon- key kidney cells. J. Exp. Med. 146: 1182-1194. Schaeffer, A.J., S.K. Amundsen, and L.N. Schmidt. 1979. Adherence of Escherichia coli to human urinary tract epithelial cells. Infect. Immun. 24: 753- 759. Schlesinger, D. 1975. Microbiology - 1975. Am. Soc. Mi- crobiol. Washington, D.C. 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. Sharon, N., and H. Lis. 1972. Lectins: cell-agglutinating and sugar-specific proteins. Science 177: 949-959. Smith, H. 1964. Microbial behaviour in natural and arti- ficial environments. Symp. Soc. Gen. Microbiol. 14: 1-29. Smith, H. 1972. Mechanisms of virus pathogenicity. Bac- teriol. Rev. 36: 281-310. Smith, H. 1976. Survival of vegetative bacteria in ani- mals. Symp. Soc. Gen. Microbiol. 26: 299-326. Smith, H., and J.H. Pearce. 1972. Microbial pathogenicity in man and animals. 22nd Symp. Soc. Gen. Micro- biol. Cambridge Univ. Press. Cambridge. Smith, H.W., and C.L. Gyles. 1970. The relationship be- tween two apparently different enterotoxins pro- duced by enteropathogenic strains of Escherichia coli of porcine origin. J. Med. Microbiol. 3: 387- 401. Smith, H.W., and M.B. Huggins. 1978. The influence of plasmid determined and other characteristics of en- teropathogenic Escherichia coli on their ability to proliferate in the alimentary tract of piglets, calves and lambs. J. Med. Microbiol. 11: 471-492. Sojka, W.J., and R.E.A. Carnaghan. 1961. Escherichia coli infection in poultry. Res. Vet. Sci. 2: 340-352. 101 Stirm, S., F. ¢rskov, I. Orskov, and A.E. Andersen. 1967. Episome-carried surface antigen K88 of Escherichia coli. II. Isolation and chemical analysis. J. Bacteriol. 93: 731-739. Weber, K., and M. Osborn. 1969. The reliability of mole- cular weight determinations by dodecyl sulfate- polyacrylamide gel electrophoresis. J. Biol. Chem. 244: 4406-4412. APPENDIX 102 . a .mumHH mnsuaso mswms Ewum m newumoamflucmpfl omzspwmloumucm .H< musmHm H xflpcmmmfi 103 Appendix 2 Figure A2. Electron micrograph of negatively stained cells of E. coli (culture 9) grown on Peptone agar at 37°C; Magnification, x 103,000. 104 Appendix 3 Figure A3. Electron micrograph of negatively stained cells of E. coli (culture 94-5) grown on Peptone agar at 37°C; Magnification, x 103,000. 105 Figure A4. Electron micrographs of negatively stained cells of E. coli (culture 119-3) grown on Peptone agar at 37°C; Magnification, (A) x 60,000, (B) x 103,000. 106 Appendix 5 Table A1. Relative mobilities of standard proteins and pilus protein on SDS-gels at 7.5 and 10.0 percent gel concentrations Relative mobility Proteins 7.5% gel 10.0% gel Phosphorylase b 0.11 0.05 Bovine serum albumin 0.28 0.10 Ovalbumin 0.41 0.20 Carbonic anhydrase 0.56 0.40 Soybean trypsin inhibitor 0.74 0.63 a-lactalbumin 0.85 0.74 Pilus protein 0.52 0.36