Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Signals triggered by enteropathogenic Escherichia coli in epithelial cells Foubister, Vida Joan 1994

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

Item Metadata


831-ubc_1994-0138.pdf [ 2.36MB ]
JSON: 831-1.0087426.json
JSON-LD: 831-1.0087426-ld.json
RDF/XML (Pretty): 831-1.0087426-rdf.xml
RDF/JSON: 831-1.0087426-rdf.json
Turtle: 831-1.0087426-turtle.txt
N-Triples: 831-1.0087426-rdf-ntriples.txt
Original Record: 831-1.0087426-source.json
Full Text

Full Text

SIGNALS TRIGGERED BY ENTEROPATHOGENICESCHERICHIA CULl IN EPITHELIAL CELLSbyVIDA JOAN FOUBISTERB.Sc., The University of AlbertaA THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIESDepartment of Biochemistry and Molecular BiologyWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAFebruary 1994© Vida Joan Foubister, 1994In presenting this thesis in partial fulfilment of the requirements for an advanced degree atthe University of British Columbia, I agree that the Library shall make it freely available forreference and study. I further agree that permission for extensive copying of this thesis forscholarly purposes may be granted by the head of my department or by his or herrepresentatives. It is understood that copying or publication of this thesis for financial gainshall not be allowed without my written permission.Department of BiochemistryThe University of British ColumbiaVancouver, CanadaDate: February 22, 1994DE-6 (2/88)AbstractEnteropathogenic Escherichia coli (EPEC) is a leading cause of infantile diarrhea indeveloping countries. Upon attachment to cultured epithelial cells, EPEC triggers anumber of eucaryotic signals including host cell protein phosphorylation and increases inintracellular calcium. These signals function to initiate cytoskeletal rearrangement andother changes in infected cells. Intracellular and extracellular calcium was shown to beinvolved in bacterial attachment and entry into HeLa cells. In addition the level ofinositol phosphates (IPs) increased in eucaryotic cells following EPEC addition. EPECinduced release of IPs preceded actin rearrangement beneath the attached bacterium andinvasion. Noninvasive EPEC mutants and tyrosine protein kinase inhibitors were used todemonstrate that the formation of IPs depends on tyrosine phosphorylation of eucaryoticcell proteins. Inhibitors of host protein tyrosine phosphatases (vanadate/H20)promoteuptake of EPEC into HeLa cells, supporting a role for host cell tyrosine phosphorylationin EPEC invasion. Collectively, these results highlight the essential role of host cellsecond messengers in EPEC infection.Several EPEC loci have been identified and found to be involved in eventsoccurring prior to invasion, including adherence, cytoskeletal rearrangements, and theactivation of host cell signals. The cfm genes are necessary for the tyrosinephosphorylation of a eucaryotic protein (Hp90). A second chromosomal locus, eaeA,encodes a 94-kDa outer membrane protein (intimin) that is essential for the establishmentof intimate adherence and invasion and contributes to virulence. Recently an additionallocus (eaeB), which maps approximately 5 kb downstream of the eaeA gene, wasidentified. Mutants carrying a deletion in the eaeB gene (UMD864) were deficient forinvasion and induction of actin rearrangement, and did not attach intimately to eucaryotic11cells. The AeaeB mutant was also unable to activate epithelial cell signals, includingtyrosine phosphorylation of Hp90 and the release of IPs. Coinfection of HeLa cells witheaeA and eaeB mutants restored all of the phenotypes associated with attaching andeffacing lesion formation. These data indicate that the EaeB protein is necessary forinduction of host cell signals in infected cells, and that it has a different function than theeaeA gene product, intimin.111Table of ContentsAbstract iiTable of Contents ivList of Tables viiList of Figures viiiList of Abbreviations xAcknowledgment xiiDedication xiiiChapter 1 Introduction 11.1 Three Stage Model of EPEC Pathogenesis 21. 1.1 Localized Adherence 21.1.2 Intimate Attachment 61.1.3 Rearrangement of Host Cell Cytoskeletal Components 81.1.4 Decrease in Transepithelial Electrical Resistance 81.1.5 Invasion 91.2 Host Cell Signal Transduction 101.2.1 Overview of Signal Transduction 101.2.2 Role of Signal Transduction in Bacterial Pathogenesis 201.2.3 Signals Induced by EPEC 291.3 Objectives 35Chapter 2 Experimental Procedures 372.1 Bacterial Strains, Plasmids and Media 372.2 Tissue Culture 372.3 Inhibitors 39iv2.4 Bacterial Internalization and Cell Association 392.5 Invasion Rates with Calcium Chelators 402.6 Measurement of the Release of Inositol Phosphates 402.7H0/vanadate treatment 422.8 Phospholipase C Diagnostic Plates 422.9 Immunofluorescent Microscopy 432.10 Protein Extraction 432.11 Western Immunoblotting 44Chapter 3 EPEC Infection Triggers Host Cell Signals 453.1 Role of Calcium in EPEC Infection 453.2 EPEC Induces a Flux of Inositol Phosphates in Infected HumanEpithelial Cells 503.3 Formation of Inositol Phosphates in not a Consequence of ActinRearrangement or Bacterial Invasion 563.4 EPEC Mutants that Fail to Induce Host Cell Protein TyrosinePhosphorylation do not Trigger a Flux of Inositol Phosphates 633.5 Host Cell Tyrosine Phosphorylation Precedes EPEC InducedFormation of Inositol Phosphates 633.6 Inhibition of Host Cell Tyrosine Specific Phosphatases StimulatesInvasion 653.7 EPEC Does Not Secrete Phospholipase Activity 71Chapter 4 The eaeB gene of EPEC is Necessary for Signal Transductionin Epithelial Cells 744.1 The &aeB Mutant is Deficient in Triggering Host ProteinRearrangement 74v4.2 The AeaeB Mutant Fails to Induce the Tyrosine Phosphorylationof a 90 kDa Host Protein (Hp9O) 794.3 Formation of Inositol Phosphates is Not Triggered by theNoninvasive AeaeB Mutant 794.4 Functional Complementation Between the eaeB and eaeA Mutants 824.4.1 Functional Complementation - Immunofluorescent Microscopy 874.4.2 Functional Complementation - Invasion 92Chapter 5 Conclusions 95Appendix Publications 104Bibliography 105viList of Tables2.1 Bacterial strains and plasmids 383.2 Release of separate inositol phosphates in uninfected and EPEC oreaeA mutant infected HeLa cells 553.3 Release of inositol phosphates in HeLa cells infected with differentnoninvasive EPEC mutants 643.4 Bacterial uptake byV04/H20treated HeLa cells 704.5 Relative bacterial internalization efficiency by HeLa cells 93viiList of Figures1.1 Three stage model of EPEC pathogenesis 41.2 Model of the mechanism of regulation by a G-protein linked receptor 131.3 A model for the interactions between the epidermal growth factorreceptor and cellular signaling molecules 161.4 Mechanism of Ras activation by tyrosine kinase linked receptors 191.5 A common mechanism shared by G-protein and tyrosine kinase linkedreceptors 221.6 Model for the role of protein tyrosine phosphorylation in Yersiniainfection of host cells 271.7 A model for S. typhimurium internalization by epithelial cells 311.8 Induction of eucaryotic cell signals by EPEC 343.9 Effect of calcium chelators on EPEC association with and entryinto HeLa cells 473.10 Reversibility of invasion inhibition by calcium chelators 493.11 Formation of inositol phosphates in HeLa cells infected withEPEC or a cf,n::TnphoA mutant [14-2-1(1)] 523.12 EPEC association with and entry into HeLa cells 543.13 EPEC induced formation of inositol phosphates in Henle-407and Caco-2 human epithelial cells 583.14 EPEC induced formation of inositol phosphates in Rat2fibroblast cells 603.15 Effect of cytochalasin D on EPEC induced flux of inositolphosphates 62viii3.16 Effect of staurosporine and genistein on the formation of inositolphosphates by EPEC in HeLa cells 673.17 Rate of release of inositol phosphates in genistein treated,uninfected, or EPEC infected HeLa cells 693.18 Effect of vanadate/H202 treatment on tyrosine phosphorylation levelsin HeLa cells 734.19 Rearrangement of actin in HeLa cells upon infection with wildtype EPEC, eaeB deletion mutant, and the eaeB deletion mutantcontaining pMSD3 764.20 Rearrangement of eucaryotic tyrosine phosphorylated proteins inHeLa cells infected with wild-type EPEC, eaeB deletion mutant, andeaeB deletion mutant containing pMSD3 784.21 Induction of Hp90 tyrosine phosphorylation in HeLa cells infectedwith different noninvasive EPEC mutants 814.22 Formation of inositol phosphates in HeLa cells infected withEPEC or ieaeB mutant 844.23 Release of inositol phosphates in HeLa cells infected for 3 hwith different noninvasive EPEC mutants 864.25 Rearrangement of actin in HeLa cells infected with wild type EPEC,eaeA::TnphoA [10-5-1(1)], tXeaeB, and a mixture of eaeA::TnphoA andi\eaeB mutants 894.25 Rearrangement of tyrosine phosphorylated proteins in HeLa cellsinfected with wild type EPEC, eaeA::TnphoA [10-5-1(1)], AeaeB,and a mixture of eaeA::TnphoA and LeaeB mutants 915.26 A model of the interaction of EPEC with HeLa cells 103ixList of AbbreviationsA/E attaching and effacingafa operon afimbrial adhesion operon of uropathogenic E. coliATCC American tissue culture collectionBAPTA 1 ,2-bis-(2-aminophenoxy)ethane-N,N,N’ ,N’tetraacetic acidBAPTA/AM 1 ,2-bis-(2-aminophenoxy)ethane-N,N,N’ ,N’tetraacetic acid acetoxy methylesterBFP bundle forming pilusBHI agar brain heart infusion agarCaco-2 human colon adenocarcinoma cellsCFU colony forming unitsDAG diacylglycerolDMEM Dulbecco’s modified Eagle’s mediumEAF EPEC adherence factorECM extracellular mathxEGF epidermal growth factorEGFR epidermal growth factor receptorEIEC enteroinvasive E. coliEPEC enteropathogenic E. coliETEC enterotoxigenic E. coliFCS fetal calf serumF1TC flourescein isothiocyanateHeLa human cervix epithelioid cellsxHenle-407 human embryonic intestine cellsH20 hydrogen peroxideHp90 90 kDa HeLa proteinIP inositol monophosphateJP2 inositol bisphosphate1P3 inositol trisphosphate1P4 inositol tetrakisphosphateIPs inositol phosphatesLA localized adherenceLB agar/broth Luna Bertani agar/brothMDCK Madin-Darby canine kidneyMEM minimal essential mediumOMP outer membrane proteinPBS phosphate buffered salineP1 3-kinase phosphatidylinositol 3-kinasePIP2 phosphatidylinositol 4,5-bisphosphatePKC protein kinase CPLC phospholipase CPTK protein tyrosine kinasePTP protein tyrosine phosphataseRat2 modified Fischer rat fibroblast 3T3-lilce (Rati) cellsSH2/3 Src homology regions 2/3V04 vanadatexiAcknowledgmentsI want to express my thanks to my supervisor, Brett Finlay, for his enthusiasm andsupport throughout my graduate program. The time I have spent working in his lab hastaught me to be an independent worker and thinker. han Rosenshine has been a constantsource of inspiration and ideas. Without his encouragement, many fruitful experimentswould not have been done. I have enjoyed working with the many members of the Finlaylab, and in close contact with the Jeffries and Brock labs down the hail. I am grateful toSharon Ruschowski for her excellent technical assistance, and to Michael S. Donnenbergand James B. Kaper for providing us with the EPEC mutants.Much thanks also goes to my parents, who have continually supported methroughout my education. Tim van Biesen’s encouragement and sense of humor havehelped me through the stressful situations I encountered.The Natural Sciences and Engineering Research Council of Canada is recognizedfor financial support in the form of a studentship.xiiFor my parents,Alice and RonxliiChapter 1IntroductionEnteropathogenic Escherichia coli (EPEC) was the first category of E. coli to beidentified as a causative agent of diarrhea (Bray, 1945; reviewed by Robins-Browne,1987). EPEC causes a persistent, watery diarrhea that can lead to dehydration and death(Rothbaum et al., 1982). Although there are recent reports of diarrheal disease due toEPEC in child-care settings in the United States (Bower et al., 1989; Paulozzi et al.,1986), these outbreaks are rare and the role of EPEC in developed countries is declining.Diarrheal disease causes 4.2 million deaths yearly and is the second leading cause ofhuman death (World Health Organization, 1992). Of these diarrheal diseases, EPEC isthe leading cause of bacteria diarrhea in infants and continues to be a major healthproblem in underdeveloped countries worldwide (Senerwa et a!., 1989). In thesecounthes, nosocomial neonatal diarrhea due to EPEC in community settings still causeshigh mortality. EPEC causes disease in the very young, rarely affecting children over 1year of age and is most closely associated with diarrhea in those under 6 months (Gomeset al., 1991; Levin and Edelman, 1984).Despite considerable progress in understanding some E. coli diarrhea syndromes,the pathogenesis of EPEC infection remains poorly defined. EPEC is distinct from otherE. coli that cause diarrhea. Unlike enterotoxigenic E. coli (ETEC), EPEC strains do notelaborate heat-labile or heat stable enterotoxins. In addition, EPEC does not inducekeratoconjunctivitis on inoculation into the conjunctival sac of Guinea pigs (Sereny test)(Echeverria et a!., 1976; Goldschimdt and DuPont, 1976) seen with enteroinvasive E. coli12(EIEC). Finally, EPEC belongs to a restricted set of Escherichia 0-serotypes which aretraditionally associated with outbreaks of infantile enteritis (Neter, 1959).1.1 Three Stage Model of EPEC Pathogenesis.A genetic analysis of EPEC virulence factors has led to an infection model whichproceeds in three stages (Fig. 1.1, adapted from Donnenberg and Kaper, 1992). 1.Localized adherence, the initial interaction between the bacterium and host cells, occursat a distance from the cell surface. 2. In the second stage of infection, signal transductionevents are initiated by EPEC and result in several characteristic changes in the host cell,including the effacement of microvilli. 3. The third stage of infection occurs followingthe establishment of intimate contacts between EPEC and eucaryotic cells. Intimateattachment initiates further changes in the host cell, including the rearrangement offilamentous actin and other cytoskeletal proteins, and a decrease in transepithelialresistance. A subset of the intimately attached organisms can then invade the epithelialcell.1.1.1 Localized AdherenceSmall bowel biopsies of EPEC infected children reveal discrete clusters of bacteriaattached to the mucous membranes of the intestine. (Ulshen and Rollo, 1980; Rothbaumet al., 1982; 1983). Infection of cultured epithelial cells by EPEC yields a similar patternof adherence in vitro. Rather than uniformly covering the entire surface of epithelialcells, EPEC forms adherent microcolonies in a pattern that is termed localized adherence(LA) (Scaletsky et al., 1984). LA appears to be the initial stage of attachment, in whichthe bacterium adheres at some distance from microvilli. This type of attachment is highly3Figure 1.1EPEC infection of eucaryotic cells proceeds in three stages. (1) In the first stage, EPECattaches to cells at a distance from the microvilli, in distinct microcolonies. This initialadherence, termed localized adherence (LA), is mediated by a plasmid encoded bundle-forming pilus (BFP) and possibly other factors. (2) Following this initial adherence,EPEC triggers several signal transduction events in the host cell that result in severalcharacteristic changes, including the effacement of microvilli. (3) In the third stage ofinfection, intimate contacts are established between the bacterium and the infected cell.The contact lesion underlying adherent bacteria is characterized by the assembly of ahighly organized structure which contains host cytoskeletal and tyrosine phosphorylatedproteins. Subsequent to these events, approximately one percent of the attached bacteriaenter into (invade) the epithelial cells.45correlated with specific EPEC serotypes in strains isolated from patients with diarrhea(Scaletsky et al., 1985).LA is known to be a common property of the so-called classic EPEC serotypesand is dependent on the presence of a 60-megadalton (MDa) EPEC adherence factor(EAF) plasmid (Baldini et a!., 1983; Cravioto et a!., 1979; Nataro et al., 1985; 1987).EPEC strains cured of this plasmid lose the ability to bind to tissue culture cells in alocalized manner (Knutton et a!., 1987; Levine et at., 1985). Moreover, transfer of theEAF plasmid to HB1O1, a non adherent E. coli laboratory strain, confers adherence ofHB 101 to HEp-2 cells (Baldini et at., 1983). However, adherence of HB1O1 transformedwith the EAF plasmid is poor in comparison to that of EPEC (McConnell et a!., 1989).The EAF plasmid appears to have a role in pathogenesis. E. coti strains whichhave been isolated during outbreaks of infantile gastroenteritis and from children withdiarrhea almost invariably possess 55- to 70-MDa EAF plasmids (Nataro et at., 1985)that share significant regions of DNA homology (Nataro et a!., 1987). In addition,Levine et at. (1985) reported that MAR20, an EAF plasmid-cured strain, is significantlyless pathogenic in volunteers than E2348/69, the parental strain from which it wasderived. Thus the EAF plasmid appears to be required for the LA phenotype in vitro andfor infectivity in vivo.LA is an inducible property, as transfer of EPEC from a rich nutrient broth totissue culture media decreases the time required to exhibit this phenotype from 30-60 mmto 15 mm (Vuopio-Varkila and Schoolnik, 1991). Bacteria within these microcoloniesinteract not only with the host cells to which they are bound, but also with each other.Autoaggregation of bacteria is also inducible, occurs at the same rate and in the samegrowth conditions that induce LA, and requires the presence of the EAF plasmid.Induction of LA was found to be associated with de novo protein synthesis and changesin outer membrane proteins, including the production of a 18.5-kDa polypeptide (VuopioVarkila and Schoolnik, 1991). Several lines of evidence suggest that this 18.5-kDa6protein is an EPEC-specific pilus protein (bfp) which constitutes the bundle forming pilus(BFP). Recently, bfp was cloned from the EAF plasmid of EPEC (Sohel et aL, 1993).Analysis of the amino acid sequence of the 18.5-kDa protein and its migration in SDSPAGE gels showed it was identical to the BFP subunit (Vuopio-Varkila and Schoolnik,1991). This identity was further confirmed by Western blotting, as an antiserum to pilusprotein crossreacts with the 1 8.5-kDa protein. Further, the bundle forming pilus isinduced at the same rate and under the same conditions as the LA phenotype (Sohel et al.,1993). BFP-like structures, shown to contain a BFP-specific antigen byimmunofluorescence microscopy, can be seen by high-resolution scanning electronmicroscopy linking bacteria in adherent microcolonies. And finally, loss of the EAFplasmid abrogates both the LA phenotype and the capacity to express BFP (Giron et al.,1991; Vuopio-Varlika and Schoolnik, 1991). Thus the BFP appears to be directlyresponsible for the LA and autoaggregating phenotypes. Conclusive evidence awaitsdemonstration of a functional correlation between the presence of bfp and the capacity toexhibit the LA phenotype and produce BFP-like filaments.1.1.2 Intimate AttachmentIn the third stage of infection, EPEC adheres intimately to the epithelial membrane(Polotsky et at., 1977; Staley et al., 1969). A chromosomal gene, eaeA, was identified byJerse et a!. (1990) as a locus necessary for intimate attachment. This gene encodes a 94-kDa outer membrane protein (OMP), intimin, which is recognized by sera fromvolunteers convalescing from experimental EPEC infections (Jerse and Kaper, 1991).The predicted amino acid sequence of intimin is similar to those of the invasins ofYersinia pseudotuberculosis and Yersinia enterocolitica (Yu and Kaper, 1992). Invasinsbind with a high affinity to members of the 1 family of integrin receptors to mediate theefficient uptake of bacteria (Van Nhieu and Isberg, 1991). Although intimin is necessary7for EPEC invasion, it alone does not convey the invasive phenotype to E. coli laboratorystrains. Although present data demonstrate that intimin is necessary for intimateadherence, it is not known if intimin is the actual adhesin involved or simply plays anindirect role in the process.Recently an additional mutant was constructed by creating a deletion in a gene,eaeB, which maps approximately 5 kb downstream of the eaeA gene (Donnenberg et al.,1989). This mutant is deficient in its ability to invade and induce actin rearrangementand does not attach intimately to eucaryotic cells (Donnenberg et al., 1993). However, incontrast to the eaeA mutant, the eaeB mutant still produces intimin (Donnenberg et al.,1993). The product of the eaeB locus is a 39 kDa protein with a predicted amino acidsequence that is not closely related to previously identified proteins. However a motifcommon to the pyridoxal phosphate binding sites of several aminotransferase enzymes isfound within the predicted amino acid sequence (Donnenberg et al., 1993). Manyaminotransferase enzymes participate in biosynthetic pathways and catalyze the transferof a small moiety form coenzyme A to an amino acid. Therefore it is possible that theEaeB protein covalently modifies and activates a protein(s) necessary for intimateattachment and/or signal transduction events which mediate attaching and effacing (AlE)lesion formation. The EaeB protein lacks an obvious signal sequence for protein export,suggesting that it may be located in the bacterial cytoplasm and exert its effects onepithelial cells indirectly by acting through another protein(s) (Donnenberg et al., 1993).However, recent observations suggest that the eaeB gene product may be secreted via asignal sequence independent mechanism (Kenny and Finlay, unpublished observations),and thus may act directly.Plasmid cured EPEC strains induce A/E lesions with delayed kinetics and reducedefficiency (Knutton et al., 1987, Tzipori et al., 1989). Knutton et al. (1987) proposed thatthe EAF plasmid may increase the efficiency of intimate attachment by facilitating initialadherence to the cells. However, the expression of the EaeA protein (intimin) has been8shown to increase in the presence of the EAF plasmid, although plasmid cured strains canstill produce low but detectable levels of intimin (Jerse and Kaper, 1991). Recently, twoloci have been cloned from the EAF plasmid that enhance expression of the chromosomaleaeA gene (Gomez and Kaper, 1992). The nucleotide sequence of one gene, designatedper for plasmid encoded regulator, is related to that of envY, a gene involved inthermoregulation of porin expression (Lundrigan et al., 1989). Therefore the EAFplasmid may increase the number of lesions both by increasing expression of intimin andby increasing initial adherence, thus promoting contact so that the A/E activity may occurmore efficiently.1.1.3 Rearrangement of Host Cell Cytoskeletal ComponentsIntimate attachment leads to a localized loss of microvilli and induces profound effects onthe epithelial cytoskeleton. Actin accumulates in the apical cytoplasm beneath theattached bacterium, often forming a pedestal-like structure upon which the bacterium lies(Knutton et al., 1989). These effects have been termed attaching and effacing (A/E)(Moon et a!., 1983) and form the basis of a diagnostic test for EPEC (Knutton et a!.,1989). This test uses flourescein isothiocyanate (FITC)-conjugated phalloidin to identifythe site specific concentrations of actin that are characteristic of the A/E membranelesions. In addition to actin, other cytoskeletal components including myosin (ManjarrezHernandez et at., 1991), cz-actinin, talin, and ezrin (Finlay et at., 1992) accumulatebeneath the attached bacterium in EPEC infected cells.1.1.4 Decrease in Transepithelial Electrical ResistanceFollowing EPEC infection of polarized cell monolayers of Caco-2 and MDCK cells,transepithelial resistance decreases (Canil et a!., 1993). However,[3H]inulin penetration9across MDCK monolayers does not increase, suggesting that damage to intercellular tightjunctions is minimal. This concept is supported by immunofluorescent data, as actinfilaments supporting tight junctions are not noticeably affected in the epithelial cells, noris the distribution of ZO- 1, a tight junction protein. The decrease in transepithelialresistance is therefore thought to be triggered by the disruption of a transcellular(intracellular) pathway rather than by disrupting intercellular tight junctions(paracellular). These disruptions only occur following the formation of wild typeattaching and effacing lesions. Mutants which are deficient in any loci involved inadherence or formation of attaching and effacing lesions are unable to cause a decrease intransepithelial resistance.1.1.5 InvasionEPEC was originally thought to be noninvasive, primarily due to negative resultsobtained with EPEC in the Sereny test (Echeverria et al., 1976; Goldschmidt and DuPont,1976). However more recent studies have demonstrated that once EPEC is associatedwith the eucaryotic cell, approximately one to five percent of the innoculated bacteriainvade cultured cells (Donnenberg et a!., 1989; Francis et a!., 1991). Invasion appears tobe the last step in the process of EPEC infection, as many mutants which are deficient ininitial adherence, intimate attachment, or signal transduction, are also unable to enterepithelial cells. The entry of EPEC can be completely inhibited by cytochalasins whichinhibit actin polymerization (Francis et a!., 1991). In addition, colchicine, a specificinhibitor of microtubule polymerization, also markedly inhibits EPEC invasion (Franciset at., 1991). These data suggest that EPEC entry is mediated by a microfilament andmicrotubule dependent mechanism. Like attaching and effacing activity, the presence ofthe EAF plasmid augments EPEC invasion (Donnenberg et a!., 1989; Francis et a!.,1991). The clinical significance of bacterial entry into host cells in EPEC pathogenesis is10not known. EPEC strains do not cause dysentery or typhoidal syndromes. However, ithad been noted long ago that EPEC can disseminate late in infection (Drucker et al.,1970; Giles et al., 1949) and although it is extremely rare, sepsis due to EPEC has beenreported (Bower et al., 1989).1.2 Host Cell Signal TransductionMany bacteria that cause disease have the capacity to enter into and live within eucaryoticcells such as epithelial cells and macrophages. The mechanisms used by these organismsto achieve and maintain this intracellular lifestyle vary considerably, but mostmechanisms involve subversion and exploitation of host cell functions. This overviewhighlights some the the host cell second messenger pathways that are exploited.1.2.1 Overview of Signal TransductionThe binding of certain growth factors and hormones to their respective receptors inducescellular signals that can ultimately result in cell growth and proliferation (Yarden andUlirich, 1988a; Ullrich and Schlessinger, 1990). The process by which these signals aretransduced from receptors on the cell surface to intracellular second messenger pathwaysis presently the focus of much ongoing research. Two different membrane receptor typeshave been shown to be involved in signal transduction. One is a family of receptorslinked to heterotrimeric G proteins, and the other is a group of receptors linked directly orindirectly to a tyrosine kinase activity. Recent reports highlight the funnelling of theseactivators into the same, highly conserved recipient proteins.G protein linked receptors include those for bombesin, acetyicholine (muscarinic),adrenergic agonists, thrombin, and many other effectors (reviewed by Crouch andHendry, 1993 and Berridge, 1993). These receptors are composed of a single protein11polypeptide which traverses the plasma membrane seven times. Binding of an agonist tothe external domain induces a conformational change in the receptor which leads to theactivation of a G protein, a heterotrimeric protein consisting of a, f3, and y subunits. Inthe inactive state, the a subunit binds the nucleoside diphosphate GDP. After receivingan activating signal, GDP is exchanged for GTP, the aGTP-subunit dissociates from thefry subunit and goes on to interact with its target effector (Fig 1.2, adapted from Stemweisand Smrcka, 1991). There is growing evidence that the fry subunit may also play a role inactivation (Fig. 1.2; Blank et al., 1992; Katz et al., 1992; Sternweis and Smrcka, 1991).The enzyme systems that appear to be affected by G-protein activation includephospholipase A2 (PLA2), adenylate cyclase, phospholipase(s) C f (PLCI31) andchannels of different ions. PLA2 hydrolyzes phospholipids to liberate fatty acids such asarachidonic acid and adenylate cyclase generates cAMP from ATP. PLC1 is one ofseveral PLC isoforms that cleaves the phospholipid phosphatidylinositol 4,5-bisphosphate(PIP2) to the second messengers inositol trisphosphate (1P3) and diacylglycerol (DAG),which in turn mediate the release of Ca2 from intracellular stores and activate proteinkinase(s) C (PKCs) (Rhee et al., 1989).Lymphocyte receptors, including the 1gM receptor in B lymphocytes and T-cellantigen receptor (TCR)-CD3 complex of T cells, are tyrosine kinase receptors which areindirectly linked to a tyrosine kinase activity. These receptors lack intrinsic tyrosinekinase activity, and instead recruit members of the src family of tyrosine kinases such asfyn and ick. These cytoplasmic kinases associate with the cytoplasmic domains of thelymphocyte receptors following activation. The epidermal growth factor receptor(EGFR) and the platelet-derived growth factor receptor (PDGFR) are members of asecond group of tyrosine kinase receptors which possess intrinsic protein tyrosine kinaseactivity (Yarden and Ullrich, 1988b; Gill et al., 1987). The EGFR will be used as anexample to describe a pathway activated by tyrosine kinase linked receptors.12Figure 1.2A model showing two possible pathways for the activation of PLC by heterotrimeric Gproteins. Following receptor stimulation GDP is exchanged for GTP, and the aGTPsubunit dissociates from the fy subunit. In the upper pathway, the aGTP-subunit goes onto activate PLC, which in turn, cleaves PIP2 into 1P3 and DAG. Alternatively, as shownin the lower pathway, the fry subunit may cause the downstream activation of PLC. Thepathway utilized may depend on the subtype of PLC, or the components of the G protein./C,,-DC,)>El14The EGFR is a transmembrane glycoprotein composed of an extracellular domain whichcontains the binding site for epidermal growth factor (EGF), a cytoplasmic domain wherethe active site for the tyrosine kinase is located, and a single membrane spanning segmentwhich links the two domains (Ullrich et a!., 1984). Binding of EGF induces aconformational change in the EGFR monomer that favors the formation of EGFR dimers(Hammacher et al., 1989; Canals et al., 1992). Dimerization actives the kinase andallows the two kinase domains to phosphorylate each other on specific tyrosine residues.Following tyrosine phosphorylation, the EGFR becomes a docking site for signalingproteins, including phospholipase Cyl (PLCy1), the GTPase activating protein of p2lras(GAP), phosphatidylinositol 3-kinase (P1 3-kinase) and an adapter protein involved inRas activation, growth factor receptor-bound protein (Grb2). A model of theseinteractions, adapted from Lowenstein et al. (1992), is shown in Fig. 1.3. All of theseproteins share conserved noncatalytic domains termed Src homology (SH) regions 2 and3 (Koch et a!., 1991). SH2 domains are conserved stretches of —100 amino acids that arethought to function in binding to tyrosine-phosphorylated growth factor receptors or toother non-receptor tyrosine phosphoproteins (Sadowski et al., 1986; reviewed by Pawson,1988). Different SH2 containing molecules recognize and bind to differentphosphotyrosine sites of tyrosine phosphorylated receptors. The identity of thephosphotyrosine sites is determined by short sequence motifs adjacent to eachphosphotyrosine (Fantl et al., 1992). The SH3 domain is a distinct motif of —45 aminoacids that is thought to function as a binding sites for specific proline-rich motifs (Mayeret a!., 1988; Cicchetti et al., 1992). It is thought that once SH2 containing proteinsbecome closely associated with the a tyrosine phosphorylated receptor, they are activatedby phosphorylation. Although such phosphorylation and associated activation has beenreported for PLCy1 (Nishibe et a!., 1990), others believe this correlation has not beenclearly demonstrated (Majerus, 1992). In further contrast to this hypothesis, PT 3-kinaseactivity does not appear to correlate with phosphorylation (Kazlauskas and Cooper,15Figure 1.3Activation of the EGFR with EGF induces tyrosine phosphorylation of the receptor atseveral sites. The phosphorylated receptor then acts as a binding site for a number ofcellular signaling molecules that contains one or two SH2 domains. The signalingmolecules are activated following their association with the receptor, and in turn, directfurther signaling cascades.EGFGAPsSosILOthereffectorsKinaseOthercascadeenzymes,adaptors0171990). Therefore, the phosphotyrosyl binding activity of the SH2 domain may simplyfunction to promote recruitment of cytoplasmic signaling proteins to activated tyrosinekinases, thereby facilitating the transduction of external signals to second messengerpathways.Once activated, PLCy1 hydrolyzes PIP2 into 1P3 and DAG, as described forPLCf31. GAP functions to stimulate the ability of Ras to hydrolyze GTP to GDP, andthereby acts as a negative regulator by turning Ras from the active GTP-bound state tothe inactive GDP-bound conformation (Trahey and McCormick, 1987). P1 3-kinase is alipid kinase that phosphorylates the D3 position of phosphatidylinositol,phosphatidylinositol 4-phosphate, or P1 4,5-P2 (Whitman et a!., 1988).Phosphatidylinositol 3,4,5-trisphosphate has been shown to effectively and selectivelyactivate the isozyme of PKC (Nakanishi et al., 1993). The physiological roles of theother 3-phosphorylated inositol phospholipids are unknown. Although the evidence isnot conclusive, it has been suggested that they may be involved in the control of cellproliferation (Carpenter et al., 1990), and in altering cellular structure throughinteractions with cytoskeletal and associated proteins (Downes and Carter, 1991).There have recently been significant advances in understanding how Ras isactivated and how it transduces signals (Fig. 1.4, adapted from reviews by McCormick,1993 and Egan and Weinberg, 1993). An adapter protein, Grb2, binds to activatedreceptors by its SH2 domain and recruits a Ras activator (Sos) to the receptor throughinteractions with its two SH3 domains (reviewed by McCormick, 1993). Once associatedwith the receptor, Sos functions as a guanine nucleotide releasing agent, converting RasGDP to active Ras-GTP. Activation of Ras triggers a cascade of serine-threonine kinasesthat send trophic signals to the nucleus (reviewed by Egan and Weinberg, 1993). Morespecifically, the principal downstream target is the Raf protein. Once activated by Ras,Raf phosphorylates a second kinase, MAP kinase kinase (MEK). MEK then activatesMAP kinase (MAPK) which also interacts with other effectors, including other kinases.18Figure 1.4The Ras pathway runs from the cell surface to the nucleus, where genes are turned on oroff in response to the incoming stimulus. The binding of a growth factor to its tyrosinekinase receptor triggers activation, dimerization, and autophosphorylation of thereceptor’s tyrosine residues. Grb2 then binds to the activated receptor by its SH2domain, and to Sos, a Ras activator, by its SH3 domain. Receptor-associated Sosprovokes GDP-GTP exchange on Ras, triggering a cascade of serine-threonine kinasesthat send trophic signals to the nucleus.19Grb2‘.4 ‘Inactivereceptors Inactive RasReceptoractivationActive RasPLCOtherenzymes,adaptorsOthereffectors(GAPs)KinaseCascade1’Other20Following mediation by other kinases, signals pass into the cell nucleus thoughphosphorylation of transcription factors that regulate gene expression. Althoughupstream proteins (PT 3-kinase, PLC, and GAP for example) have been left out of thispicture, it does not mean they are not involved.Although the G protein and tyrosine kinase linked receptors have been presentedseparately, they have common features. For example, these two pathways, one initiatedby a family of G protein linked receptors and the other by tyrosine kinase linkedreceptors, both lead to the formation of 1P3 and DAG through the activation of differentisoforms of PLC (Fig. 1.5, adapted from Berridge, 1993). In addition, evidence isaccumulating that the signal transduction pathways activated by tyrosine kinase receptorsand by 7-transmembrane receptors are not exclusive within receptor subtypes, but there isconsiderable crosstalk. There is evidence that G proteins have a direct effect on PT 3-kinase activity (reviewed by Crouch and Hendry, 1993). For example, stimulation of thefMet-Leu-Phe receptor induces the activation of PT 3-kinase in neutrophils, and thisresponse is blocked by pertussis toxin, an effect thought to be mediated by a Gi protein(Traynor-Kaplan et al., 1989). Tn addition, P1 3-kinase activity is enhanced in thrombinstimulated platelets, an effect that is also evoked by direct activation of G-proteins withGTPyS (Kurcera and Rittenhouse, 1990). PLCy stimulation in some cells appears to becontrolled in part by a direct interaction with G in addition to the well characterizedpathway of tyrosine phosphorylation (Yang et aL, 1991). And finally, EGF has beenshown to activate PLA2 to stimulate the production of arachidonic acid and prostaglandinE2 (Teitelbaum, 1990). This stimulation requires the EGFR (Hack et al., 1991a; 1991b;Clark and Dunlop, 1991) and is enhanced by guanosine 5’-(3-thio)-triphosphate (GTP-yS) and eliminated by pertussis toxin which functions to ADP ribosylate G proteins(Teitelbaum, 1990).1.2.2 Role of Signal Transduction in Bacterial Pathogenesis21Figure 1.5Summary of the two major receptor mediated pathways for stimulating the formation ofIF3 and DAG. Following G protein linked receptor stimulation, a G protein activatesPLC, whereas tyrosine kinase linked receptors activated by growth factor bindingactivate PLOy. Both PLCI3 and y function to cleave PIP2, releasing 1P3 and DAG. Othereffectors interact with the tyrosine kinase linked receptors, including P1-3 kinase andGrb2 (see Fig. 1.4).GproteinlinkedreceptorsTyrosinekinaselinkedreceptorsEGFPDGFIPI-3KIPIP3RcGTP VPIP21P4IIP4RIIIP3RI-cD TCRIck1/nDAG‘IPKCIAntigenECa2+I(CellularactivityI&mitogenesisSerineGrb2_____threonine______kinase‘i1:icascade23As described above, cellular processes involved in the normal growth and development ofeucaryotic cells are regulated by a variety of proteins which function to transduce signalsfrom the cell surface to the cytoplasm. The activation of these signaling proteins isprecisely controlled by protein phosphorylation on tyrosine or serine/threonine residues,protein dephosphorylation, and protein-protein interactions. At each step, a smalldisturbance can unhinge the highly coordinated and meticulously timed interactionsbetween signaling proteins. It is becoming increasingly clear that many pathogenicbacteria, including enteropathogenic Escherichia coli (EPEC), Listeria monocytogenes,and species of Salmonella, Shigella, and Yersinia, exploit these host cell signalingpathways to facilitate infection. The pathogenic mechanisms of Yersinia and Salmonellatyphimurium will be described as examples of the involvement of the host cell signals inthe infection process.Yersinia enterocolitica and Yersinia pseudotuberculosis are enteropathogens thatinfect lymphoid follicles and cause a variety of illnesses, ranging from mildgastroenteritis to mesenteric lymphadenitis (Brubaker, 1991; Comelis et al., 1987).These two bacteria encode multiple independent pathways for attachment to and entryinto cultured mammalian cells (Isberg, 1990; Miller et al., 1989).Invasin, the product of the chromosomal mv gene, was identified by its capacity toconfer the ability to enter cultured mammalian cells to a noninvasive strain of E. coli.(Isberg and Falkow, 1985). Invasin has been shown to be solely responsible for theaccumulation of actin and actin-associated proteins (filamin and talin) around theinvading bacterium (Young et al., 1992; Isberg and Leong, 1988). The ability of invasinto trigger cytoskeletal reorganization is consistent with the finding that invasin recognizesmultiple integrins (Isberg and Leong, 1990). integrins belong to a large family ofa/f3 heterodimeric transmembrane receptors and are involved in cell to cell interactionsand cellular adhesion to extracellular matrix (ECM) proteins.24YadA is an adhesin encoded on the Yersinia virulence plasmid (Bolin et a!., 1982)that mediates bacterial attachment (Heesemann et al., 1987) and entry (Yang and Isberg,1993; Bliska et al., 1993a) into mammalian cells. It appears that YadA also mediatesthese effects through interactions with f1 integrins, since monoclonal antibodies againstthis receptor block entry by this pathway (Bliska et al., 1993a). In addition, YadAfacilitates bacterial adherence to the ECM proteins collagen and fibronectin (Emody etal., 1989; Schulze-Koops et a!., 1992; Tertti et al., 1992). It is not known whether theseECM proteins are involved in bridging the interaction between YadA and f3 integrins, orif YadA and invasin recognize the same site on the integrin receptor.Integrins are thought to link extracellular binding events functionally to thecytoskeleton by the transduction of signals into the eucaryotic cell (Desimone et al.,1987). In nucleated cells, clustering of c3I31 or c531 integrins induces protein tyrosinephosphorylation of two proteins localized to focal adhesions, ppl25’ (Juliano andHaskill, 1993) and paxillin (Burridge et al., 1992). Tyrosine phosphorylation alsoappears to play a role in integrin-mediated bacterial entry. For example, protein tyrosinekinase inhibitors have been shown to block invasin mediated bacterial entry into culturedmammalian cells (Rosenshine et al., 1992). In addition, a protein encoded by theYersinia virulence plasmid, YopH, is a protein tyrosine phosphatase (PTP) (Guan andDixon, 1990). PTPs are thought to play specific roles in signal transduction and cellularphysiology and not simply to reverse the action of protein tyrosine kinases (PTKs)(Fischer et a!., 1991). YopH is a secreted protein that appears to function bydephosphorylating host proteins, thereby disrupting normal signal transduction processesin mammalian cells (Bliska et a!., 1991; 1992). Although the host proteinsdephosphorylated by YopH are not known, two macrophage proteins that have beenspecifically coisolated from cell lysates with a catalytically inactive form of YopH(C403A) have been shown to have intrinsic PTK activity (Bliska et al., 1992). Therefore,YopH may function to dephosphorylate host PTKs, reversing the activation induced by25the integrin mediated binding of Yersinia. In support of this hypothesis, YopH has beenshown to contribute to the ability of Yersinia to inhibit uptake by phagocytic cells(Goguen et a!., 1986; Lian et a!., 1987; Rosqvist et a!., 1988). Other Yops are proposedto act synergistically to interfere with other key host functions including actinmicrofilament stability (YopE), serine/threonine phosphorylation (YpkA), and G proteinlinked receptor stimulation (YopM) (Straley et a!., 1993). The expression of thesesecreted proteins is induced in response to extracellular stimuli, calcium and temperature(Bolin et a!., 1985; Straley et a!., 1986). YopE, YopH, YpkA and YopM, are all essentialvirulence determinants which are encoded on the common virulence plasmid of Yersinia(Rosqvist eta!., 1990; 1991; Leung eta!., 1991; Galyov eta!., 1993).Although many Yersinia and host proteins involved in host-parasite “cross-talk”and association have been identified, our understanding of the Yersinia infection strategyis not complete. A model summarizing our present knowledge is shown in Fig. 1.6(adapted from Bliska and Falkow, 1993).Sa!monella typhimurium is an enteric pathogen which invades non-phagocyticepithelial and fibroblast cells via a process resembling phagocytosis. Infection with S.typhimurium induces dramatic changes in the host cell, including disruption of the brushborder microvilli (Takeuchi, 1967; Finlay et a!., 1988), increased permeability ofepithelial tight junctions (Finlay et a!., 1988), membrane ruffling (Francis et a!., 1993),and the rearrangement of host actin filaments and other cytoskeletal associated proteins(a-actinin, tropomyosin, ezrin, and talin) (Finlay et a!., 1991). Presumably, the bindingof S. typhimurium induces an “uptake signal” which generates the above changes in thehost cell.S. typhimurium invasion is known to be associated with the activation of host cellsignals. For example, the concentration of inositol phosphates (IPs) transiently increases20 mm following infection (Ruschkowski et a!., 1992). This increased release of IPs isclosely correlated with the invasion rate and cytoskeletal rearrangement. S. typhimurium26Figure 1.6A model for the role of protein tyrosine phosphorylation in integrin mediated bacterialentry and YopH antiphagocytic activity. Initial interactions between bacterial invasin(mv), the product of the chromosomal mv gene, and epithelial cell integrins result inbacterial attachment to the eucaryotic cells. Bacterial entry is hypothesized to bestimulated by the invasin-integrin complex, which triggers tyrosine phosphorylation(encircled P) and activation (+) of a PTK. Contact between the host cell and bacteriumalso leads to expression and secretion of the product of the yopH gene (YopH). YopH ispresumed to translocate across the host cell membrane and reverse activation (-) of thePTK by tyrosine dephosphorylation, thereby inhibiting integrin mediated bacterialuptake.BacterialentrystimulatedinhibitedYersinia(cR5(+)UBacterialentryHostceIL—28invasion also stimulates a rapid increase in the levels of free intracellular calcium incultured epithelial cells (Pace et a!., 1993) Increased [Ca2j1appears to be important forinvasion, as mutants which can not induce these calcium fluxes fail to enter host cells(Pace et a!., 1993). In addition, chelation of host intracellular, but not extracellular,calcium inhibits S. typhimurium uptake (Ruschkowski et al., 1992). DAG activation ofhost PKC activity is not required for S. typhimurium uptake into HeLa cells as a potentPKC inhibitor, staurosporine, has no effect on invasion (Rosenshine et al., 1992). Theseobservations support the hypothesis that S. typhimurium stimulates host PLC activity tocleave PIP2 to DAG and 1P3, which then mobilizes [Ca2].It was recently reported that infection of Henle-407 cells with S. typhimurium isaccompanied by induction of tyrosine phosphorylation of the host EGF receptor, and thatexogenously added EGF restored the invasiveness of a noninvasive mutant (Galan et at.,1992). However, TPK inhibitors or anti-EGFR antibodies do not reduce the invasionefficiency of S. typhimurium (Rosenshine et a!., 1992; Galan et at., 1992). Moreover, S.typhimurium also invades cells that do not have EGFR (Francis et a!., 1993; Galan et a!.,1992). Bliska et al. (1993b) suggest that these contradictions are the result of the abilityof S. lyphimurium to use alternative uptake signals when invading different cell lines.However, Rosenshine et a!. (submitted) have been unable to demonstrate activation of thehost EGFR by S. typhimurium in HeLa, MDCK, and Henle-407 epithelial cell lines, andA431, an epidermal cell line overexpressing the EGFR. Therefore, the EGFR does notappear to be required for entry of Satmonella into host cells.Additional results by Pace et a!. (1993) suggest that S. typhimurium triggers anelaborate biochemical cascade, including activation of mitogen-activated protein kinase(MAPK). Activation of MAPK by protein tyrosine phosphorylation has been confirmed(Rosenshine et al., submitted), yet the role of this activation in invasion in not known.Listeria species also induce the tyrosine phosphorylation of MAPK (Tang et at., 1994).The role of the other activities induced in this cascade, phospholipase A2 and 5-29lipooxygenase, result in the production of leukotriene D4 (LTD4) (Pace et al., 1993).LTD4 is proposed to be required for bacterial entry, as noninvasive S. typhimuriummutants are internalized by addition of LTD4 to the infected cells.Much of the pathogenic process of S. typhimurium remains to be elucidated, asmost of the genetic determinants of Salmonella virulence discovered thus far areregulatory elements rather than functional determinants of pathogenicity. A summary ofthe infection process of Salmonella, adapted from a model previously proposed byRosenshine and Finlay (1993), is shown in Fig 1.7. Although the exact mechanisms arenot well defined, the role of the host cell in the infection process of both Yersinia speciesand S. typhimurium has been clearly demonstrated.1.2.3 Signals Induced by EPECUpon attachment to cultured epithelial cells, EPEC induces the assembly of a complexcytoskeletal structure localized beneath the adherent bacterium. EPEC infection causesadditional changes to the host cell, including disruption of the brush border microvilli andan increase in transepithelial permeability (Canil et al., 1993). Recent studies suggestthat these changes, which ultimately lead to diarrhea, are induced by the subversion ofhost cell signals by EPEC.As discussed for Yersinia and S. typhimurium, tyrosine phosphorylation of hostcell substrates also appears to play a role in EPEC infection. EPEC induces tyrosinephosphorylation of three eucaryotic proteins, 39 kDa (Hp39), 72 kDa (Hp72) and 90 kDa(Hp9O). The major substrate, Hp90, is a membrane associated protein. In correlationwith Hp90 tyrosine phosphorylation, the EPEC induced cytoskeletal structure alsocontains tyrosine phosphorylated proteins. Tyrosine phosphorylation appears to play a30Figure 1.7A model for S. typhimurium internalization by epithelial cells. Following contact withhost epithelial cells, Salmonella stimulates a PLC to cleave PIP2, releasing 1P3 and DAG.1P3 mobilizes [Ca2]i from intracellular stores, initiating cytoskeletal rearrangement andinternalization of the bacterium. The induced cytoskeletal structure, which surrounds theinvading bacterium, dissociates shortly after bacterial uptake. The role of tyrosinephosphorylation in this infection pathway is not clear.Salmonella DAGHOSTCELLIPs?CytoskeletàlRearrangement32role in EPEC infection, as protein kinase inhibitors (staurosporine and genistein) thatinhibit Hp90 tyrosine phosphorylation also block bacterial uptake. In addition, EPECstrains which have a mutation in the cfin gene(s) [cfm::TnphoA 14-2-1(1) and 27-3-2(1)]fail to induce Hp90 phosphorylation (Rosenshine et at., 1992). These cfm mutantspossess fully functional intimin, a protein involved in EPEC adherence, and are able toadhere intimately to the host cell surface at normal levels (Donnenberg et at., 1990).However, cfm mutants fail to induce the rearrangement of tyrosine phosphorylated hostcell proteins or cytoskeletal components (Rosenshine et at., 1992). Two strains of adifferent noninvasive mutant, eaeA (intimin), cannot establish a tight association withhost cells but are able to induce phosphorylation of Hp90 (Jerse et a!., 1990; Rosenshineet at., 1992; Foubister et at., 1994). Intimin mutants are capable of inducing theaccumulation of filamentous actin, but in comparison to the parental strain, thecytoskeletal rearrangements are less sharply focused beneath the adherent organisms. Anintense but diffuse haze or “shadowt’ of tyrosine phosphorylated proteins alsoaccumulates underneath microcolonies of the eaeA mutant.Interestingly, coinfection of the eaeA mutant with both cJin::TnphoA mutants,which alone demonstrate a minimal rearrangement of host proteins, restores the ability ofmany of the adherent bacteria to induce a wildtype rearrangement of host cytoskeletonand tyrosine phosphorylated proteins (Rosenshine et at., 1992). Additionally, theinternalization of cfm: :TnphoA mutants into HeLa cells is facilitated by the addition ofeaeA mutants in trans. Unlike the cfm mutant, the eaeA function cannot becomplemented. This suggests that for invasion and organization of the cytoskeletalstructure under a specific individual bacterium to occur, intimin has to be present on thebacterial surface. In contrast, the ability of the cfm function to be supplemented in transsuggests that some aspect of the phosphorylation and localization of Hp90 involves adiffusible product, presumably inside the host cell. A model summarizing these results isshown in Fig. 1.8 (adapted from Rosenshine and Finlay, 1993).33Figure 1.8The induction of eucaryotic cell signals occurs during the second stage of the EPECinfection process. Attached bacteria use the cfm gene product to induce specific hostPTK activity. This, possibly in combination with other signals which increase [Ca2]i,causes aggregation of cytoskeletal elements. Simultaneously intimin, which promotesintimate attachment of EPEC to the host membrane, nucleates the cytoskeletal aggregatesinto an organized structure beneath the bacterium.LiiLiiC)35Additional host cell proteins become phosphorylated by calcium dependentserine/threonine protein kinases following EPEC infection (Baldwin et al., 1990;Manjarrez-Hemandez et al., 1992; Baldwin et al., 1993). Of these proteins, the mostprominent phosphorylated species has a molecular weight of 20 kDa. This protein wasidentified by immunoprecipitation as myosin light chain, an important cytoskeletalprotein known to affect actin organization in non-muscle cells (Manjarrez-Hemandez etal., 1992; Keller and Mooseker, 1982). Accumulation of actin and myosin at the sites ofbacterial infection suggest that following adherence, EPEC may directly triggercytoskeletal rearrangement through signal transduction pathways which stimulate proteinkinase activity.Other signaling pathways also appear to be involved in inducing cytoskeletalrearrangement. EPEC infection of HEp-2 cells leads to an elevation of intracellular freecalcium levels (Baldwin et at., 1991). Actin accretion and loss of cell viability haverecently been shown to be calcium dependent processes, as intracellular calcium chelatorsinhibit these responses in EPEC infected cells (Baldwin et at., 1993). The ability ofdantrolene to block EPEC induced calcium elevation suggests that calcium is releasedfrom IF’3 sensitive stores (Baldwin et a!., 1991). The phosphorylation of myosin lightchain is also dependent on the elevation of intracellular calcium, which functions toactivate the calcium-calmodulin-dependent enzyme myosin light chain kinase(Manjarrez-Hemandez et at., 1992). These data suggest that the elevation of intracellularcalcium plays a key role in infection. In addition, it is likely that the intracellular changesinduced by increased calcium levels disrupt the function of the brush border microvilli,causing a reduction in the absorptive capacity of the intestinal mucosa and ultimatelydiarrhea.1.3 Objectives36The goal of my project was to further define the role of host cell signaling pathways inEPEC infection. Using intracellular and extracellular calcium chelators, the role ofcalcium in EPEC adherence and invasion of host cells was investigated. To determine ifinositol phosphates are involved in EPEC induced [Ca2]release, I examined whetherEPEC infection of cultured epithelial cells induces PLC activity. Since an increase in thelevel of inositol phosphates is indicative of PLC activity, the level of IPs (both total andindividual IPs including IP, 1P2, 1P3 and 1P4) was measured in epithelial cells atincreasing times of infection. Tyrosine phosphatase inhibitors were used to verify theimportance of increased host cell tyrosine phosphorylation in EPEC infection. Finally, arecently constructed mutant, AeaeB, was tested for its ability to induce host cell signals.37Chapter 2Experimental Procedures2.1 Bacterial Strains, Plasmids, and MediaEPEC strains E2348/69, JPN15, UMD864, CVD2O6, 10-5-1(1), 14-2-1(1) and 27-3-2(1),described in Donnenberg et at. (1990), Jerse et a!. (1990), Donnenberg and Kaper (1991)and Donnenberg et at. (1993) were grown in LB agar or LB broth at 37°C withoutshaking. Chioramphenicol (20 .tg/ml) or ampicillin (50 .ig/ml) was added as needed.The product of the eaeB gene was provided in trans by plasmid pMSD3 (Donnenberg etal., 1993). The plLl4 plasmid contains the afa operon of uropathogenic E. coli clonedinto pBR322 (Labigne-Roussel et a!., 1985). Table 2.1 summarizes the genotype orrelevant properties of the bacterial strains and plasmids used.2.2 Tissue CultureThe human cervix epithelioid carcinoma HeLa (CCL 2) cells, the human colonadenocarcinoma Caco-2 (HTB 37) cells, and the human embryonic intestine Henle-407(CCL 6) cells were obtained from the American Type Culture Collection (Rockville,MD) and were maintained in MEM (Gibco Laboratories, Gaithersburg, MD), 10% FCS(Gibco Laboratories), nonessential amino acids, penicillin (100 .tg/ml) and streptomycin(100 pg/ml). Rat2 cells (CRL 1764), also obtained from ATCC, were grown in DMEM(Gibco Laboratories), 10% FCS, nonessential amino acids, penicillin (100 .tg/ml) andstreptomycin (100 pg/ml). These cells were derived from a subclone of a 5’-bromo-38Table 2.1 Bacterial strains and plasmids used in this studyStrain or Genotype or relevant properties Source orplasmid referenceStrainsE2348/69 Virulent 0127:H6 EPEC strain; nalidixic acid Levine et al.,resistant 1978; 198514-2-1(1) E2348169 noninvasive mutant containing two TnphoA Donnenberg et at.,insertions in the cfln gene 199027-3-2(1) E2348/69 noninvasive mutant containing two TnphoA Donnenberg et al.,insertions in the cfin gene 199010-5-1(1) E2348/69 noninvasive mutant containing one TnphoA Donnenberg et at.,insertion in the eaeA gene 1990CVD2O6 E2348/69 AeaeA8 Donnenberg andKaper, 1991UMD864 E2348/69 iXeaeB Donnenberg et at.,19931{B 101 E. coli K12JB laboratory strainPlasmidspILl4 afa operon of uropathogenic E. coti cloned into Labigne-RousselpBR322 et at., 1985pMSD3 2.3-kb BglII fragment of the eae gene cluster cloned Donnenberg et at.,into the BamHI site of pACYC184 199339deoxyuridine resistant strain of the Fischer rat fibroblast 3T3-like cell line Rati, and lackappreciable levels of nuclear thymidine kinase. Cells were used below passage 40 andwere grown and assayed at 37°C, 5% Co2.2.3 InhibitorsInhibitors were dissolved in dimethyl sulfoxide. Stock solutions of cytochalasin D(Sigma Chemical Co., St. Louis, MO; 1 mg/mi), staurosporine (Boehringer Mannheim,Mannheim, Germany; 1 mM) and genistein (UBI Inc., Lake Placid, NY; 100 mM) werealiquoted and stored at -20°C. Prior to use, inhibitors were diluted in MEM with 10%FCS and then added to cultured HeLa cells. The kinase inhibitors have no effect onbacterial viability (Rosenshine et a!., 1992).2.4 Bacterial Internalization and Cell AssociationInvasion and cell association levels were determined as described by Finlay and Falkow(1988), with slight modifications. HeLa cells were seeded in 24 well tissue culture platesat a density of 1 x iO cells per well and grown overnight. The cells were then washedwith phosphate buffered saline (PBS) and 500 p.1 of media without antibiotics was added.5 p.1 of a fresh overnight EPEC culture, or 10 p.1 of mixed culture (1:1 mix) for a mixedinfection, was added to each well and the cells were infected for 3 h at 37CC, 5% CO2.To measure invasion, cells were washed 2 times with PBS and fresh medium containing100 p.g/ml gentamicin was added to kill extracellular bacteria. Infected cells wereincubated for an additional 2 h, washed twice and 200 p.1 of 1% Triton X- 100 (BDHChemicals, Darmstadt, Germany) was added to lyse the eucaryotic cells. After 5 mm atroom temperature, 800 p.1 of PBS was added to each well and mixed vigorously.Appropriate dilutions were plated onto LB plates and the number of colony forming units40were used to determine the number of viable intracellular bacteria. To determine thenumber of cell-associated bacteria, the eucaryotic cells were lysed just before thegentamicin step. Thus, the number of cell-associated bacteria represents the number ofbacteria that are either adherent or intracellular.2.5 Invasion Rates with Calcium Chelators:Stocks of 10 mM BAPTA [1 ,2-bis-(2-aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid;Sigma] or BAPTA/AM [1 ,2-bis- (2-aminophenoxy)ethane-N,N,N’ ,N’ -tetraacetic acidacetoxy methylester; CalBiochem] (Weiss and Insel, 1991) were made in calcium-freePBS or dimethysuiphoxide (DMSO), respectively. HeLa cells in MEM with 10% FCSwere plated onto 24-well microtiter plates (1 x 1o cells/well) that had been coated withpoly-L-lysine and allowed to attach for 60 mm. 0.1% w/v poly-L-lysine was diluted 1/50in PBS, and 1 ml was added to each well, incubated overnight, 4°C, and washed 4 timeswith PBS. Pretreatment of wells with poly-L-lysine prevented cell detachment afterchelator addition, as determined by Trypan blue staining and cell counting. Bacteria wereadded to HeLa cells in the presence of an appropriate chelator in serum-free MEM +1-Ca2 for 3 h, followed by gentamicin treatment and quantitation of invasion as described.To determine the reversibility of chelator treatment, cells were washed extensively withcalcium-free PBS and fresh MEM media with calcium was added and incubated forvarious times prior to gentamicin addition.2.6 Measurement of the Release of Inositol Phosphates.Release of IPs was determined as described by Ruschkowski et al. (1992). Cells wereseeded onto 60-mm petri dishes, resulting in a 90-100% confluent culture at the time ofinfection. After 4 h, the medium was removed and MEM with 1% FCS containing 1041pCi of[3H]myoinositol (18.4 Ci/mmol, Amersham Life Science, Arlington Heights, IL)was added and incubated overnight. Before use, the[3H]myoinositol was absorbed withDowex to remove contaminants. The cells were washed with Dulbecco’s PBS + 1 g/lglucose and incubated for 15 mm in HEPES [N-(2-hydroxyethyl) piperazime-N’-(2-ethane sulfonic acid]-saline solution (25 mM NaHepes, pH 7.4, 125 mM NaC1, 5 mMKC1, 1 mM CaC12, 1 mMNa2HPO4,0.5 mMMgSO4, 1 g/L glucose) containing 2 mMglutamine, 1 mM Na pyruvate, 1 mg/mi BSA and 10 mM LiCl. Cells were washed asbefore, 2 ml of HEPES-saline solution was added, and then they were either infected with40 p1 of a fresh overnight bacterial culture (—1.5 x 106 bacteria) or left untreated. Aftervarious times, cells were washed with cold PBS and scraped off into 1 ml PBS. To stopthe reaction, cells were immediately added to 4 ml of 2:1 methanol:chloroform, mixed,and incubated for 30-45 mm at room temperature. 0.125 ml of ethylene-diaminetetraacetic acid (EDTA; pH 8.0), 1.2 ml chloroform and 1 ml H20 were added to eachsample, which was then vortexed and centrifuged at 500xg for 20 mm. The upper(aqueous phase) was loaded onto a 1 ml Dowex anion exchange column (AG 1-X8 resin,100-200 mesh formate form; Bio-Rad Laboratories, Richmond, CA) which had beenwashed with 20 ml of cold 5 mM inositol. The columns were washed again with 20 ml of5 mM inositol, and then with 20 ml of 5 mM Na borate/60 mM Na formate. Inositolphosphates were eluted either collectively with 1 M ammonium formate in 20 ml 0.1 Mformic acid, or separately as follows: inositol monophosphate (IP), 20 ml of 0.1 M formicacid/0.2 M ammonium formate; inositol bisphosphate (1P2), 20 ml of 0.1 M formicacid/0.4 M ammonium formate; inositol trisphosphate (1P3), 30 ml of 0.1 M formicacid/0.8 M ammonium formate; and inositol tetrakisphosphate (1P4), 20 ml of 0.1 Mformic acid/1.0 M ammonium formate. The counts per minute (cpm) represent therelease of inositol phosphates, were determined by counting the eluant of each samplediluted 1 to 4 in Ready Safe liquid scintillation cocktail (Beckman, Fullerton, CA).422.7H0/vanadate treatmentCells were treated withH20 and vanadate as described by Volberg et al. (1992), withslight modifications. Stage 1: infected or uninfected HeLa cells were washed with PBS,then, 500 jii of 2 mMH20and 1 mM NaVO4in MEM + 10% FBS was added and cellswere incubated at 37°C, 5% CO2 for 30 mm. Stage 2: cells were washed once in MEM+ 10% FBS and further incubated in the same medium for 30 minutes at 37°C, 5% CO2.Different bacterial strains exhibit different kinetics of cell binding and invasion and weretherefore used in different ways to evaluate the influence of theVO4/H20treatment onbacterial uptake. EPEC bind to and invade host cells very slowly. Thus wild type EPECand the noninvasive mutant, cfin 14-2-1(1) were preincubated with the host cells in MEM+ 10 % FBS at 37°C for 2.5 h prior toVO4/H20treatment. The VO4/H20treatmentwas followed by gentamicin treatment and an invasion assay as described.E. coli HB 101 was added to the cells after stage 1 of the VOdH2treatment andincubated with the cells during stage 2 of the treatment. E. coli I{B 101 containing thepILl4 plasmid were incubated with cells on ice for one hour prior to VO41H20treatment. This allowed bacterial binding to the epithelial cells, but not uptake. Theseinfections were followed by the two stages ofVO4JH2Otreatment, gentamicin treatmentand an invasion assay.2.8 Phospholipase C Diagnostic PlatesEgg yolk agar was made by preparing a 1:2 dilution of fresh egg yolk in 150 mM NaC1solution and adding 12.5 ml of this mixture to 250 ml of BHI agar at 56°C (Kocks et al.,1992). Bacteria were grown on egg yolk agar plates and analyzed for the presence of azone of egg yolk opacification. Egg yolk opacification results from the activity of aphosphatidylcholine phospholipase C (lecithinase) (Fuzi and Pillis, 1962). Listeria43monocytogenes, which is known to contain a secreted PLC activity, was used as apositive control.2.9 Immunofluorescent MicroscopyHeLa cells were seeded at 1 x iO cells per well onto glass coverslips in a 24 well tissueculture plate and grown overnight. The cells were washed with PBS and infected with 10tl (—3.75 x 10 bacteria) of a late logarithmic EPEC culture. Following a 3 h infection,cells were washed and fixed for 30 mm in 2% paraformaldehyde in PBS. The cells werewashed three times with PBS, permeabilized with 0.1% Triton X-100 in PBS and washedas before. Cells were incubated with 50 pi of primary antibodies (monoclonal antiphosphotyrosine, 4Gb, UBI Inc., rhodamine-phalloidin or anti-a-actinin, BM-75.2,Sigma) for 60 mm, washed and incubated with 100 .tl of secondary antibody (anti-mouseIgG or 1gM, FITC conjugated, both from ICN Inc.), for 60 mm or washed and mounteddirectly (samples labeled with rhodamine-phalloidin). Samples were washed, mounted,examined using a microscope and photographed as described (Finlay eta!., 1991).2.10 Protein ExtractionHeLa cells were seeded in 100 mm tissue culture plates at a density of 2 x 106/plate. Thenext day, cells were infected with 100 p1 of a late logarithmic bacterial culture (0D600nm —1.5) for 3 h. Infected cells were washed three times with cold PBS, scraped into 1.0ml of PBS, pelleted (60 s @ 7000 rpm) and lysed in 0.1 ml of lysis solution (1% TritonX-100, 50 mM Tris-HCL, pH 7.6, 0.4 mM NaVO4, 1mM NaF, 0.1 mg/mlphenylmethylsulfonyl fluoride and 10 p.g/ml leupeptin). The lysate was spun (60 S14000 rpm) and the 1% Triton X-100 soluble supernatant was mixed with 20 p1 of a 5 xloading buffer. The insoluble pellet was dissolved in 100 of a 2.5 x loading buffer.44The lysates were boiled for 7 mm and cleared (5 mm at 14 000 r.p.m.) prior to resolutionby gel electrophoresis.2.11 Western ImmunoblottingProteins were separated by SDS - PAGE as described by Laemmli (1970), and thentransferred to nitrocellulose membrane (AB-S 83 Schleicher and Schuell Inc.) using aNovaBlot electrophoretic transfer unit (LKB). Monoclonal anti-phosphotyrosine (4Gb,UBI Inc.) was diluted in TBS (150 mM NaC1, 20 mM Tris-HC1, pH 7.5) containing 1%BSA (Sigma) and used at a concentration of 0.5 .tg/ml. Anti-mouse IgG alkalinephosphatase conjugate (Calbiochem) was used similarly at a final concentration of 3EPEC Infection Triggers Host Cell Signals3.1 Role of Calcium in EPEC infectionThe role of intracellular and extracellular calcium in EPEC adherence and invasion tocultured epithelial cells was analyzed. HeLa cells were grown on poly-L-lysine treatedmicrotiter plates to prevent cell detachment in the presence of calcium chelators. Bacteriawere added and allowed to infect either in MEM, Ca2 free MEM, or in Ca2 free MEMcontaining Ca2 chelators. BAPTA was used to chelate extracellular calcium andBAPTA/AM was used to chelate intracellular calcium. EPEC adhered to human cells at—50% efficiency in Ca2 free MEM and failed to invade host cells at wild type levels.Treatment with BAPTA greatly decreased levels of bacterial adherence to host cells, andalso decreased invasion efficiency (Fig. 3.9a and b). In contrast, BAPTA/AM treatmentonly slightly decreased the attachment of EPEC, while invasion was profoundly inhibited(Fig. 3.9a and b). These results suggest that both intracellular and extracellular calciumchelators inhibit EPEC adherence and invasion in a dose dependent manner.The reversibility of Ca2 chelator treatment on bacterial invasion was examined.HeLa cells were infected for 3 h in the presence of Ca2 free MEM, 133 .tM BAPTA or100 .tM BAPTA/AM. After washing to remove unbound bacteria and Ca2 chelators,calcium containing MEM medium was added and the number of intracellular bacteriawere quantitated at increasing times. Reversibility was extremely rapid with Ca2 freeMEM infected cells, but was slower for BAPTA treated cells (Fig 3.10). In addition,BAPTA/AM treated cells exhibited low levels of reversibility. The lack of reversibility46Figure 3.9The number of cell associated and intracellular bacteria per well (24 well plate) weremeasured following a 3 h infection in the absence or presence of the calcium chelators.Both intracellular (BAPTA/AM) and extracellular (BAFI’A) calcium chelators decreasedEPEC association with (a) and entry into (b) HeLa cells in a dose-dependent manner.Values are the average of four samples ± SD and represent the results of two independentexperiments.477e+6°— BAPTA4)6e+6Cu \NT.BAPTNAMNAM. 5e+64)5 4e+60Cl)3e+62e+6le÷6CuOe+0- • • •0 50 100 150 200Drug Concentration (pM)12000Cu—0—— BAPTA4)O 10000 • BAPTNAMCu80006000Cu.1-•400020000•- I • I •0 50 100 150 200Drug Concentration (pM)48Figure 3.10Reversibility of calcium chelator treatment on EPEC invasion of HeLa cells. Afterinfecting cells for 3 h in calcium free MEM, 133 p.M BAPTA, or 100 p.M BAPTA/AM,the chelators were removed and replaced with calcium containing medium. The numberof internalized bacteria were quantitated at increasing times following chelator removal.Values are the average of quadruplicate samples ± SD.49250000Ca2+ free MEM—0-— BAPTA200000 A BAPTA’AMCu0150000= 100000•a)C.)Cut— 500004t:00 30 60 90 120Time (mm) After Chelator Removal50with BAPTA/AM treated cells is likely due to the cytolethal effect of prolongedintracellular calcium buffering, since uninfected cells loaded with BAPTA/AM round upand lose viability after about 5 h (Baldwin et al., 1993).3.2 EPEC induces a flux of inositol phosphates in infected human epithelial cells.To determine if EPEC triggers a flux of IPs, HeLa cells were labeled with[3H]myoinositol and infected with bacteria for increasing times. The level of total[3HIIPs began increasing approximately 2 h after EPEC addition, reaching a maximum of2.6 fold over uninfected at 3 h (Fig. 3.11). Addition of 200 nglml of EGF (as a positivecontrol) for 10 mm prior to IP extraction caused a 1.9 fold increase in[3H]JPs release.Similar to the flux of IPs, the rate of association (adherence and invasion) of EPEC toHeLa cells peaked at 3-3.5 h (Fig. 3.12a), and bacterial entry into HeLa cells reached amaximum rate at 3.5 h (Fig. 3. 12b). The rate of increase in intracellular calcium alsopeaks at 3 h (Baldwin et al., 1991).Individual inositol phosphates were also measured and the levels of IP, 1P3, and1P4 all peaked approximately 3 h following EPEC infection and 10 mm followingtreatment with lOOng/mi EGF (Table 3.2). The inability to detect an increase in 1P2, andthe small increases in 1P3 and IP.4 are probably a result of the rapid conversion of mostforms of these different species to IP (reviewed by Majerus, 1992), although theconversion of IP to inositol is prevented by lithium in the assay buffer. Alternatively,since the levels of IP, IP, 1P3 and 1P4 are known to increase significantly followingstimulation with EGF (Wahl et al., 1987), it suggests that the assay used to measure theindividual IPs is not sensitive enough to demonstrate the differences between values arequantitatively significant.It has previously been reported that preinducing EPEC can shorten the timerequired to establish LA (Scaletsky et al., 1985; Vuopio and Schoolnik, 1991). We tested51Figure 3.11EPEC triggers the formation of IPs in HeLa cells. Release of[3H]IPs was measured inHeLa cells that were prelabeled with[3H]myoinositol and infected with either EPEC or acJin::TnphoA mutant [14-2-1(1)] for increasing times. The concentration of[3H]IPs wasmeasured in uninfected cells and used to determine the level of release of IPs beforebacterial addition (at time zero). Infection timepoints were an average of two samples ±error and are representative of one of six different experiments.520i0. 6000C.)E2348/6914-2-1(1)5000Co004000Cl)0C3000C0EI..0 2000 •LI 0 60 120 180 240Infection Time (mm)53Figure 3.12EPEC association with (a) and entry into (b) HeLa cells. The number of cell associatedand intracellular bacteria per well (24 well plate) were measured at half hour intervals for4 h. Values are the average of four samples ± SD and represent the results of twoindependent experiments.54Cua)C.)Cua).1-’CUC.)0Cl)C’)a)C)CUCUC.)CUL.CUa)C.)CU4.’C.3000000250000020000001500000100000050000000700060005000400030002000100000S I • I •60 120 180 240Infection Time (mm)NI•60 120Infection1 80Time (mm)24055Table 3.2 Release of separate inositol phosphates* in HeLa cells followingdifferent treatmentsTreatment Time IP 1P2 ‘P3 ‘P4Untreated 100% 100% 100% 100%±53% ±21% ±20% ±16%EGF,lOOngfml 10mm 328% 99% 115% 108%±41% ±15% ±34% ±36%EPEC 3h 374% 99% 115% 107%± 104% ±17% ±36% ±16%eaeA::TnphoA 3h 269% 100% 106% 92%± 30% ± 30% ± 36% ± 35%*To facilitate comparison between individual experiments, the the % release of IPs(release of ‘Ps/total incorporation of[3Hjmyoinositol) in untreated cells was set at100% in each experiment. Untreated and EPEC infected values are the average of5 samples (4 separate experiments), and EGF treated and eaeA::TnphoA infectedvalues an average of 4 samples (3 separate experiments).56whether preinducing the bacteria in tissue culture medium, or spinning them onto the cellmonolayers and allowing them to adhere on ice prior to warming and measurement of IPswould decrease the time required by EPEC to induce a flux of IPs. These conditionsdecreased the time needed for EPEC induced release of IPs by approximately 2 h.However, even under these conditions, EPEC induced formation of IPs still required aminimum of 60 minutes before increasing, a much longer time than seen with an EGFinduced flux (10 minutes).The ability of EPEC to induce a flux of IPs in two other human epithelial celllines was examined. Formation of IPs was found to increase 1.3 fold in Caco-2 and 1.9fold in Henle-407 cells (Fig. 3.13). The kinetics of these fluxes of IPs were similar tothat observed for EPEC infected HeLa cells, with maximal levels of IPs induced 3 h post-infection. In addition, it was found that EPEC could induce formation of IPs in Rat2fibroblast cells (Fig. 3.14), suggesting that this response is not limited to human cells.HeLa cells were used for the remainder of the experiments.3.3 Formation of Inositol Phosphates is Not a Consequence of Actin Rearrangement orBacterial Invasion.To determine if EPEC induced release of IPs triggers, or is a result of bacterialinternalization, we analyzed the effect of blocking EPEC invasion on levels of IPsfollowing infection. Cytochalasin D inhibits actin polymerization and blocks entry ofEPEC into human cells, without affecting adherence (Donnenberg et al., 1990; Francis etal., 1991). We found that 2.5 .tg/ml cytochalasin D decreased invasion by 99.5 %. 2.5pg/ml cytochalasin D was added to[3H]myoinositol labeled HeLa cells eitherindependently or simultaneously with EPEC and incubated for 3 h. Formation of[3HIIPswas induced by EPEC in the presence of cytochalasin D (Fig. 3.15), demonstrating thatcytoskeletal rearrangement and bacterial invasion are not prerequisites for formation of57Figure 3.13Formation of IPs is induced by EPEC in the human epithelial cell lines, Henle-407 andCaco-2. The levels of IPs were measured at 60 mm intervals for 4 h following EPECinfection. Values are the average of 4 samples (Henle-407) or two samples (Caco-2) ±error and represent the results of two different experiments.580. o.C.)50000______23000Henle-407CO COa) Caco-222000 C450000.21000 0.C•C 40000 0.0.!?20000 — CO35000COOCO19000 0 C•Cd)C.)— 2 3000018000 -oO0-CC 25000 17000 00=cuE 20000 • • •0 60 120 180 240 0Infection Time (mm)59Figure 3.14EPEC induces formation of IPs in Rat2 fibroblast cells. Release of[3H]IPs was measuredin Rat2 cells at increasing times after EPEC infection. Values are the average ofduplicate samples ± SD.603000C.)(I)a)I 25000a)C)2000C‘4-o 1500a)U)a)a)1000• I • I • I •0 60 120 180 240 300Time (mm)61Figure 3.15The effect of cytochalasin D on the EPEC induced flux of IPs.[3H]myoinositol labeledHeLa cells were either left uninfected, infected with EPEC, treated with 2.5 p.g/mlcytochalasin 0, or infected with EPEC in the presence of 2.5 .tg/m1 cytochalasin D.Following a 3 h treatment, the level of[3H]JPs was measured. Values are the average oftwo samples ± error and represent the results of two independent experiments.0C)Cl)Cu0.Cl)0.c0.0Cl)0CI.0C0CuE0U-628000700060005000400030002000Uninfected EPECInfected2.5 tgImI 2.5 jig/mICytochalasiri D Cytochalasin D+ EPEC63Ps.3.4 EPEC Mutants that Fail to Induce Host Cell Protein Tyrosine Phosphorylation do notTrigger a Flux of Inositol PhosphatesNoninvasive mutants of EPEC, which differ in their ability to stimulate tyrosinephosphorylation of a 90 kDa host protein, Hp90, were examined for their capacity toinduce a flux of IPs. Several classes of these mutants were examined: (i) two strains ofeaeA mutants including a TnphoA insertion [10-5-1(1)] and an internal deletion(CVD2O6); (ii) two different TnphoA class 4 insertion mutations (cfm) [14-2-1(1) and 27-3-2(1)]; and (iii) JPN15, a non-adherent EPEC strain that is cured of its large plasmid.The eaeA::TnphoA and AeaeA mutants, which still induce tyrosinephosphorylation of Hp90 (Rosenshine et al., 1992), also induced the flux of IPs at 3 h(Table 3.3). Although the amount of released IPs was lower than wild type EPEC, theincrease in P and 1P3 was proportionately similar (Table 3.3). JPN15, the plasmidlessstrain, which displays reduced adherence and induction of Hp90 phosphorylation(Rosenshine et al., 1992; Baldini et al., 1983), did not induce release of IPs (Table 3.3).Both cfin::TnphoA mutants, which adhere normally but do not induce tyrosinephosphorylation, did not trigger a flux of IPs (Fig. 3.11, Table 3.3).3.5 Host Cell Tyrosine Phosphorylation Precedes EPEC Induced Formation of InositolPhosphates.EPEC induced tyrosine phosphorylation of Hp90 and bacterial invasion are bothattenuated following treatment with the protein kinase inhibitors genistein andstaurosporine. Staurosporine, however, inhibits Hp90 tyrosine phosphorylation lessefficiently than genistein. We examined whether these drugs also inhibited the EPEC64Table 3.3 Release of IPs in HeLa cells infected for 3 h with different noninvasiveEPEC mutants*Strain Genotype Localized A/E Hp90 Release of InositolAdherence Lesions tyrosine phos- Phosphates at 3 hphorylation post-infection (cpm)uninfected 3156 ±533E2348/69 wildtype ++ ++ ++ 8440 ± 650CVD2O6 AeaeA ++ -- ++ 6074 ±76410-5-1(1) eaeA::TnphoA ++ -- ++ 6387 ±39614-2-1(1) cfm::TnphoA ++ -- 4144 ± 68927-3-2(1) cf,n::TnphoA ++ +- -- 4467 ± 272JPN15 plasmid cured -- +- +- 3117 ± 146*average of a minimum of 3 assays65induced flux of IPs. 1 j.LM staurosporine or 250 pM genistein was added eitherindependently or simultaneously with bacteria to[3H]myoinositol labeled HeLa cells.Following a 3 h infection, formation of[3HJIPs was measured. Genistein completelyinhibited the formation of EPEC induced IPs, whereas treatment with staurosporine hadonly a small effect (Fig. 3.16).Formation of IPs was also measured at increasing times in the presence of 250.tM genistein with or without EPEC. Unexpectedly, genistein alone induced a flux of IPsthat peaked 2 h after drug addition and then gradually declined (Fig 3.17). The level ofIPs in cells infected with EPEC in the presence of genistein increased in a manner similarto that induced by the drug alone, but after 2 h the level of IPs declined sharply (Fig.3.17). In contrast, EPEC infection without genistein caused the level of IPs to increaseafter 2 h, reaching a even higher level at 3 h (Fig. 3.17). These results suggest thatgenistein does not inhibit the EPEC induced peak of IPs at 3 h by a direct mechanismsuch as phospholipase C (PLC) inhibition, as this drug alone induces the level of IPs toincrease at 2 h. Instead, it suggests that genistein inhibits a protein tyrosinephosphorylation event(s) that is required for downstream activation by EPEC.3.6 Inhibition of Host Cell Tyrosine Specific Phosphatases Stimulates InvasionThe levels of tyrosine phosphorylation in eucaryotic cells is controlled by a dynamicbalance between protein tyrosine kinases (PTK) and protein tyrosine phosphatases (PTP).Using potent inhibitors of tyrosine-specific phosphatases (vanadate and H20), Iexamined whether increased host cell tyrosine phosphorylation could induce the uptakeof EPEC and a mutant deficient in signal transduction [cJin 14-2-1(1)1. HeLa cells wereinfected with bacteria for 2.5 hours, and treated with 1 mM vanadate and 2 mMH20 asdescribed (Volberg et al., 1992). Levels of EPEC and cfm invasion were measured andfound to increase by 5 or 12 fold, respectively (Table 3.4). Since PTP inhibitors cause66Figure 3.16Effect of staurosporine and genistein on formation of IPs by EPEC in HeLa cells. 250i.LMgenistein or 1 p.M staurosporine were added alone or with bacteria to[3H]myoinositollabeled HeLa cells and incubated for 3 h. The release of IPs was measured and comparedto the release of iPs in uninfected cells and in cells infected for 3 h with EPEC alone.Samples are the average of three values ± SD and represent the results of fourindependent experiments.670.0Cl)a)Cu.c0.Cl)0.ca0Cl)0CI.0C0CuE0Li.7000600050004000300020001000Uninfected EPEC 250 1iMGenistein+ EPECInfected Staurosporine+ EPEC68Figure 3.17[3H]myoinositol labeled HeLa cells were either infected with EPEC, treated with 250 tMgenistein, or infected with EPEC in the presence of 250 .tM genistein. The release of IPswas measured at increasing times following each treatment. Release of IPs is shown as apercentage of the total[3Hjmyoinositol incorporation per 60 mm plate. All samples werethe average of 2 values.693.5.EPEC0 250 uM GenisteinI—0-— EPEC + 250 uM GenisteiniI • I • I • I0 60 120 180 240Time (mm)70Table 3.4 Bacterial uptake byV04/H20treated HeLa cellsBacterial strain Uptake byV041j12Qtreated cells Invasion EfficiencyUptake by untreated cells CPU/wellEPEC E2348/69 5.3 ± 1.6 2.7 ± 0.2 x 10cfrn 14-2-1(1) 12.8±4.6 3.6±l.0x102E. coli HB 101 Bacteria were not recovered <10E. coli HB1O1/pIL14 2.6 ± 1.1 3.1 ± 0.5 x 102* CPU/well represents the number of intracellular bacteria in a 24-well tissue cultureplate well containing —2 x iO HeLa cells.71strong tyrosine phosphorylation of many eucaryotic proteins, a particular protein(s) couldnot be implicated as having a role in the invasion process (Fig. 3.18).To determine ifV04/H20treatment enhances EPEC and cfm invasion in aspecific manner, the ability of this treatment to stimulate uptake of noninvasive E.coliHB 101 was examined. Since EPEC adheres to the epithelial cells more efficiently thanE. coli HR 101, attachment efficiency was increased by introducing a plasmid that carriesthe afimbrial adhesin (afa) operon into these strains. These transformed bacteriaexhibited a high level of binding to the HeLa cell surface. V04/H20treatment wasunable to enhance uptake of E. coli HB1O1, both with or without AFA expression.Therefore it appears as that the increase in the invasion efficiency of cfm and EPECfollowing this treatment is specific, as uptake is only enhanced if the bacteria areassociated with the host cell in a unique manner.3.7 EPEC Does Not Secrete Phospholipase ActivitySeveral pathogenic bacteria are known to have endogenous PLC activity, includingListeria monocytogenes, Bacillus cereus, and Pseudomonas aeruginosa (reviewed byTitball, 1993). Phosphatidylinositol-hydrolyzing PLCs are involved in the pathogenesisof these organisms. To determine if EPEC secretes a PLC which may be responsible forproduction of IPs, EPEC was grown overnight on egg yolk agar plates. L.monocytogenes, which contains phospholipase C activity, produced a zone of egg yolkopacification surrounding the growing colony (Fuzi and Pillis, 1962). EPEC was unableto produce a zone of opacity, suggesting that EPEC does not contain a secreted PLCactivity. EPEC may possess a cytoplasmic PLC enzyme, however an unsecreted enzymeis unlikely to be directly involved in triggering host cell signals. Alternatively, thegrowth conditions used in this analysis may not have been suitable for inducing theactivity of a secreted PLC.72Figure 3.18Tyrosine phosphorylated protein profile of HeLa cells treated withV041H20. HeLacells were either treated withV04/H20or infected for 3 h with EPEC or the noninvasivecfm [14-2-1(1)]. Proteins were extracted with TxlOO, separated by SDS - PAGE,transferred to nitrocellulose and probed with anti-phosphotyrosine antibodies. V041H20treatment induces strong tyrosine phosphorylation of many eucaryotic proteins. EPECtriggers tyrosine phosphorylation of one host 90 kDa protein (Hp90) (arrow), while thecfm mutant is unable to induce tyrosine phosphorylation of Hp90 or any other hostprotein.—,—.-EL74Chapter 4The eaeB gene of EPEC is Necessary for Signal Transduction in Epithelial Cells4.1 The AeaeB Mutant is Deficient in Triggering Host Protein RearrangementIt has been previously demonstrated that the AeaeB mutant, which contains an internaldeletion in the eaeB gene (UMD864), is unable to attach intimately to epithelial cells andinduce changes in host cell actin (Donnenberg et al., 1993). We compared the ability ofthe noninvasive AeaeB mutant and the wild type strain to induce HeLa epithelial cellcytoskeletal rearrangements by staining infected cells with rhodamine-phalloidin andanti-c-actinin antibodies as cytoskeletal markers. In addition, we examined whetherthese bacteria recruited and nucleated tyrosine phosphorylated proteins.In contrast to parental EPEC (Fig. 4.19a and b; 4.20a and b), the AeaeB mutantcaused only a minimal rearrangement of actin and a—actinin and did not cause anydetectable change in the distribution or concentration of host cell tyrosine phosphorylatedproteins (Fig. 4.19c and d; 4.20c and d). The ability to recruit and organize tyrosinephosphorylated proteins, actin, and x-actinin was restored to the AeaeB mutant when thecloned eaeB gene was reintroduced on the pMSD3 plasmid (Fig. 4. 19e and f; 4.20e andf). Although the cfm::TnphoA mutants have a phenotype indistinguishable from theAeaeB mutant with respect to rearrangement of tyrosine phosphorylated and cytoskeletalproteins, the introduction of the eaeB gene on the pMSD3 plasmid did not restore theability of the cfm mutants to rearrange host proteins.75Figure 4.19Rearrangement of actin in HeLa cells infected for 3 h with wild type EPEC stainE2348/69 (a, b), eaeB deletion mutant UMD864 (c, d, e), and UMD864 containing thecloned eaeB gene on the pMSD3 plasmid (f, g). Rhodamine-phalloidin was used tolocalize host cell actin by immunofluorescent microscopy and was compared tocorresponding phase contrast images. The unfocused immunofluorescent image shown ind was taken at the same focal distance as the corresponding phase image in c. The sizebaringis 10p.M.Np77Figure 4.20Rearrangement of eucaryotic tyrosine phosphorylated proteins in HeLa cells infected for3 h with wild type EPEC stain E2348/69 (a, b), eaeB deletion mutant UMD864 (c, d) andUMD864 containing the cloned eaeB gene on the pMSD3 plasmid (e, f). Antiphosphotyrosine antibodies were used to localize tyrosine phosphorylated proteins byimmunofluorescent microscopy and was compared to corresponding phase contrastimages. The size bar in f is 10 .tM.78, I* \.!‘— •_1b,—I 7r —-4.794.2 The AeaeB Mutant Fails to Induce the Tyrosine Phosphorylation of a 90 kDa HostProtein (Hp90)The failure of the AeaeB mutant to cause a detectable increase in the concentration ofeucaryotic tyrosine phosphorylated proteins suggests that the eaeB gene may be neededfor activation of a host cell tyrosine kinase(s). We examined this question by testing theability of the zleaeB mutant to induce tyrosine phosphorylation of a 90 kDa host protein,Hp90, which is normally phosphorylated in response to EPEC infection (Rosenshine etal., 1992). HeLa cells were infected with the AeaeB mutant for 3 h and total protein wasextracted and analyzed by Western blot for tyrosine phosphorylated proteins.The LeaeB mutant was unable to induce Hp90 phosphorylation (Fig. 4.21). Asexpected, parental EPEC triggered tyrosine phosphorylation of Hp90 (Rosenshine et al.,1992; Fig. 4.21), whereas a tyrosine phosphorylated Hp90 was not visible in uninfectedcells. As previously reported, two cfm mutants [14-2-1(1) and 27-3-2(1)] were unable totrigger phosphorylation of Hp90, although the AeaeA mutant still stimulated Hp90phosphorylation (Rosenshine et al., 1992; data not shown).A plasmid containing a cloned eaeB gene (pMSD3) was capable ofcomplementing the 1eaeB mutation, since the AeaeB mutant harboring pMSD3 wascapable of triggering Hp90 tyrosine phosphorylation (Fig. 4.21). However, when thisplasmid was transformed into the two cfm mutants, these strains did not induce Hp90tyrosine phosphorylation.4.3 Formation of Inositol Phosphates is Not Triggered by the Noninvasive /xeaeBMutantWild type EPEC induces the release of IPs 3 h after infection of cultured epithelial cells(Foubister et al., 1994; Chapter 3). A correlation between EPECs ability to induce80Figure 4.21Induction of Hp90 tyrosine phosphorylation in HeLa cells infected with differentnoninvasive EPEC mutants. HeLa cells were infected with the different strains for 3 hbefore extraction. Samples were resolved by electrophoresis, using a 8% SDSpolyacrylamide gel and analyzed by immunoblotting with anti-phosphotyrosineantibodies. Wild type EPEC, E2348/69, induces tyrosine phosphorylation of Hp90.However, three noninvasive EPEC mutants, cfm [14-2-1(1) and 27-3-2(1)] and AeaeB(UMD864), fail to induce tyrosine phosphorylation. The pMSD3 plasmid, whichcontains the cloned eaeB gene, complements the tXeaeB mutation, but not the two cfmmutations.00r82tyrosine phosphorylation of host proteins and to trigger the release of IPs has beenpreviously demonstrated (Foubister et al., 1994; Chapter 3). To determine if the eaeBgene function is also necessary for activation of host cell signals downstream of Hp90tyrosine phosphorylation, the AeaeB mutant was tested for its ability to trigger theformation of IPs.To measure release of IPs, HeLa cells were labeled with[3H]myoinositol andinfected with bacteria for increasing times. There are significant differences in levels ofIPs released among the parental EPEC, cfm and eaeB strains (P < .000001). The AeaeBmutant was unable to trigger the release of IPs following addition to HeLa cells for up to4 h (Fig. 4.22). As expected, there was no difference between eaeB and the two cfmmutants (P = .055), which are also unable to stimulate release of IPs (Fig. 4.23; Foubisteret al., 1994). However parental EPEC, which triggers a 2 fold increase in IPs, and iXeaeBmutants were significantly different (P = .006), as were parental EPEC and the two dinmutants.The AeaeB mutant’s inability to stimulate release of IPs could be complementedby a cloned eaeB gene, since a AeaeBIpMSD3 strain triggered release of LPs (Fig. 4.23).The level of release induced by the complemented AeaeB mutant corresponds to thelevels observed for wild type EPEC (P = .493). When the pMSD3 plasmid wasintroduced into the two cfm mutants, it was unable to complement these mutations anddid not trigger release of IPs (Fig. 4.23).4.4 Functional Complementation Between the eaeB and eaeA MutantsWe have previously shown that coinfection of the eaeA mutant with either cfm::TnphoAmutant restores the ability of many of the adherent bacteria to induce a wildtyperearrangement of host cytoskeleton and tyrosine phosphorylated proteins (Rosenshine etal., 1992). Additionally, the internalization of cJIn::TnphoA mutants into HeLa cells is83Figure 4.22The /xeaeB mutant fails to trigger the release of IPs. The level of[3H]IPs was measuredat increasing times in HeLa cells that were prelabeled with[3H]myoinositol and infectedwith wild type EPEC or a &aeB mutant. The concentration of[3H]IPs was measured inuninfected cells and used to determine the level of IPs before bacterial addition (at timezero). Infection timepoints were an average of two samples ± SD.846000AeaeB—0-- EPECCl)500040003000Cl)Cua)• •0 60 120 180 240Time (mm)85Figure 4.23Release of IPs in HeLa cells infected for 3 h with different noninvasive EPEC mutants.As expected, EPEC triggers a flux of IPs, while the eaeB (UMD864) mutant and twodifferent strains of cfin mutants [14-2-1(1) and 27-3-2(1)1 fail to induce the release of IPs.By providing the cloned eaeB gene (pMSD3 plasmid) in trans. the ability to inducerelease of IPs was restored to the eaeB mutant, but not to the two cfm mutants. Thefraction of total IPs released from cells following infection was adjusted by subtractingthe fraction released from uninfected cells (background). Values are the mean (± SD) ofat least two experiments, each conducted with duplicate samples.86E2348/69 1 1UMD864UMD864(pMSD3)14-2-1(1)14-2-1(1)(pMSD3)27-3-2(1)27-3-2(1 )(pMSD3) HI I •-1 0 1 2 3 4% Release of Inositol Phosphates87facilitated by the addition of eaeA mutants in trans. Unlike the cfm mutant, the eaeAfunction cannot be complemented. This suggests that, for invasion and organization ofthe cytoskeletal structure under a specific individual bacterium to occur, intimin must bepresent on the bacterial surface. In contrast, the ability of the cfm function to besupplemented in trans suggests that some aspect of the phosphorylation and localizationof Hp90 involves a diffusible product, presumably inside the host cell.In this study we examined whether an eaeA mutant could also complement the/eaeB mutant. HeLa cells were infected with a mixed culture of AeaeB and eaeAmutants, and the ability of the bacteria to complement in trans was analyzed by: (i)immunofluorescent microscopy (IFM) and (ii) a differential invasion assay.4.4.1 Functional Complementation - Immunofluorescent MicroscopyThe distribution of actin, cc-actinin, and tyrosine phosphorylated proteins was examinedin HeLa cells coinfected with the eaeA::TnphoA and AeaeB mutants. Infection with theeaeA mutant alone caused an increase in the intensity of tyrosine phosphorylated proteinsand a recruitment of these proteins to the area of bacterium-host cell contact, but theproteins were not tightly focused around the bacteria (Rosenshine et at., 1992; Fig. 4.24cand d). The effect of the eaeA mutant on the distribution of actin and x-actinin closelyresembled the distribution of the tyrosine phosphorylated proteins (Rosenshine et at.,1992; Fig 4.25c and d). In contrast, the eaeB mutant only induced a marginalrearrangement of actin and a-actinin, and did not cause any change in the concentrationor distribution of host tyrosine phosphorylated proteins (Fig. 4.25e and f; 4.24e and 0.However, a mixed infection of the eaeA::TnphoA and AeaeB mutants restored the abilityof many of the adherent bacteria to efficiently recruit and organize tyrosinephosphorylated proteins, actin, and x-actinin into lesions indistinguishable from thosecaused by parental type EPEC (Fig. 4.24a, b, g, and h; 4.25 a, b, g, and h). Therefore the88Figure 4.24Rearrangement of actin in HeLa cells infected for 3 h with: a pure culture of wild typeEPEC (a, b); a pure culture of eaeA::TnphoA [10-5-1(1)] (c, d); a pure culture of Z\eaeB(e, f); a mixed culture of eaeA::TnphoA and /xeaeB mutants (g, h). Rhodamine-phalloidinwas used to localize host cell actin by immunofluorescent microscopy. This wascompared with the phase contrast images. The size bar in h is 2.5 I.LM.6890Figure 4.25Rearrangement of host tyrosine phosphorylated proteins in HeLa cells infected for 3 hwith: a pure culture of wild type EPEC (a, b); a pure culture of eaeA::TnphoA [10-5-1(1)] (c, d); a pure culture of AeaeB (e, f); a mixed culture of eaeA::TnphoA and &aeBmutants (g, h). Anti-phosphotyrosine antibodies were used to localize tyrosinephosphorylated proteins by immunofluorescent microscopy. This was compared with thephase contrast images. The size bar in h is 2.5 tM.-a0\4‘4.4.I92functions of the eaeA and eaeB proteins do not need to be present on the same bacterialcell, but one of these functions can be supplied in trans by helper bacteria.The cfm mutant was also tested for its ability to restore the function of the eaeBgene. Infection with a mixture of cfm 114-2-1(1) or 27-3-2(1)1 and AeaeB mutants didnot induce wild type rearrangement of host proteins. In addition, coinfection of HeLacells with the cfln and eaeB mutants did not trigger tyrosine phosphorylation of Hp90 asanalyzed by immunoblotting (data not shown).4.4.2 Functional Complementation - InvasionSince invasion is the last step in the process of EPEC infection, mutants that are deficientin initial adherence, intimate attachment, or signal transduction are also unable to enterhost cells. In this study we measured invasion levels to determine whether coinfectingbacteria can complement the function of events prior to bacterial uptake. These studiesalso allow us to determine whether complementation is reciprocal or unidirectional.Following a 3 h infection with a single strain (wild type EPEC, AeaeA,cfm::TnphoA, eaeA::TnphoA or AeaeB) or with a mixture of two strains (AeaeA andcfln::TnphoA, cfm::TnphoA and AeaeB, or eaeA::TnphoA and AeaeB), the number ofbacteria within HeLa cells was measured. The coinfecting bacteria were distinguished bytheir sensitivity or resistance to neomycin, as only the TnphoA mutants are neomycinresistant (Manoil and Beckwith, 1985).As demonstrated by Rosenshine et al. (1992), the AeaeA mutant restored theinvasive capacity of the cfln::TnphoA mutant (Table 4.5). Further, the eaeA::TnphoAmutant was also able to promote uptake of the AeaeB mutant into HeLa cells (Table 4.5).In contrast, neither the cfm::TnphoA nor the AeaeB mutant was able to increase theinvasion efficiency of each other or of the eaeA mutants. Thus in each case a mutantcapable of producing intimin, but deficient in signal transduction, could be93Table 4.5 Relative bacterial internalization efficiency by HeLa cells% Internalization efficiency when infected asStrain Genotype Pure Mix with Mix with Mix with Mix withculture CVD2O6 10-5-1(1) 14-2-1(1) UMD864E2348/69 WT EPEC 100 - - - -CVD2O6 ieaeA 6.7 - - 2.4 -10-5-1(1) eaeA::TnphoA 3.0 - - - 2.514-2-1(1) cJin::TnphoA 0.5 31.7 - - 0.4UMD864 AeaeB 0.8 - 14.7 0.2 -Internalization efficiency of teaeA (CVD2O6), eaeA: :TnphoA [10-5-1(1)], cfm: : TnphoA[14-2-1(1)] and &aeB (UMD864) mutants in pure and mixed culture infection iscompared to wild type E2348/69 values.- not an eaeA mutant, which is competent for signal transduction but unableto produce intimin.95Chapter 5ConclusionsEPEC is one of several pathogenic bacteria which subvert host cell signaling pathways topromote infection. While there are some similarities in the mechanisms used to exploitthe host eucaryotic cell, there are also differences unique to each type of bacteria. Thecommon theme emerging is that host cells play an active role in bacterial infection. Bystudying the role of eucaryotic cells in the infection process, we will be able to gainfurther insight into the pathogenicity of these bacteria.Intracellular calcium, a key player in many cellular signaling pathways, isinvolved in the EPEC infection process. Following EPEC adherence to cells, there is arelease of [Ca2] in the eucaryotic cell. Dantrolene treatment abolishes the ability ofEPEC to induce an increase in [Ca2], suggesting that calcium may be released frominositol 1 ,4,5-trisphosphate (1P3) sensitive stores (Baldwin et al., 1991). In support ofthis hypothesis, I found that EPEC infection triggers an increase in the level of inositolphosphates within cultured epithelial cells. The inositol phosphates released followingEPEC infection probably contains many species of phosphorylated inositol including IP,1P3, and 1P4. The different inositol phosphates are proposed to have roles as secondmessengers, although the role of 1P3 is the best characterized since binding of 1P3 to itsreceptor releases calcium from intracellular stores (Berridge, 1993). Therefore, ourresults suggest that the EPEC induced increase in IPs may mediate the reportedintracellular calcium flux (Baldwin et al., 1993). In contrast to the rapid formation of IPsthat occurs within seconds or minutes in response to many growth factors and hormones,96EPEC induced the release of IPs only after several hours of infection. This long delay ininduction time is not solely dependent on the time needed for EPEC to establishadherence, as release of IPs does not commence until at least 60 mm following bacterialattachment. Interestingly, other host signals triggered by EPEC, including Hp90 tyrosinephosphorylation and increased [Ca2]1,are also induced at similar times (Rosenshine etal., 1992; Baldwin et al., 1991). These signals do not appear to be transient, but insteadare relatively stable and continuous.Using an intracellular calcium chelator (BAPTA/AM), intracellular calcium wasfound to play a role in EPEC attachment and entry to host cells. Intracellular calcium isknown to activate protein kinases, affect protein-protein interactions and cause alterationsin actin filament structure through proteins such as villin and gelsolin (Korn, 1982).Released intracellular calcium may ultimately cause depolymerization of microfilamentsthat form the structural support for the microvilli, providing free actin monomers for theformation of other cytoskeletal structures. Effacement of host cell microvilli by attachedbacteria reduces the absorptive capacity of the intestinal mucosa. It is possible thatintracellular calcium is also involved in decreasing transepithelial electrical resistance inEPEC infected cells (Canil et a!., 1993). Taken together, these effects likely contribute todiarrhea.Although increased [Ca2]1has been shown to play an important role in EPECinfection, a role for extracellular calcium has not been previously demonstrated. Usingan extracellular calcium chelator (BAPTA), I found that extracellular calcium is essentialfor EPEC association to HeLa cells. It is possible that extracellular calcium may bedirectly involved in the establishment of adherence, or alternatively may functionindirectly by opening ion channels.EPEC’s ability to trigger the release of IPs correlates with its ability to inducetyrosine phosphorylation of Hp90. For example, cfm::TnphoA mutants fail to induce bothtyrosine phosphorylation of Hp90 and formation of IPs, suggesting that Hp90 tyrosine97phosphorylation may be involved in the release of IPs. The role of tyrosinephosphorylation is further supported by the differential effectiveness of two kinaseinhibitors in blocking an EPEC induced flux of Ps (Rosenshine et al., 1992). Genisteinefficiently inhibits both Hp90 tyrosine phosphorylation and release of IPs, whilestaurosporine, which does not completely inhibit Hp90 tyrosine phosphorylation, haslittle effect on formation of IPs. Since staurosporine only partially inhibits Hp90phosphorylation but does not effectively block release of Ps, the phosphorylation eventmay not be a rate limiting step. Alternatively, genistein may specifically inhibit thetyrosine phosphorylation of a substrate that participates in the release of IPs by EPEC.Staurosporine treatment, in contrast, does not affect the tyrosine phosphorylation state ofthis substrate. Both of these hypotheses suggest that host tyrosine phosphorylation isinvolved in the release of IPs by EPEC.Genistein was originally defined to be a highly specific inhibitor of tyrosinespecific protein kinases that scarcely inhibits the activity of serine and threonine kinasesand other ATP analogue-related enzymes in vitro (Alciyama et at., 1987). Recent studies,however, demonstrate that genistein also has inhibitory effects on fatty acid synthesis,lactate transport, mitochondrial oxidative phosphorylation, and aldehyde dehydrogenase(Young et al., 1993). Although it is possible that genistein is inhibiting a non-proteintyrosine kinase dependent event, the role of tyrosine phosphorylation in EPEC inducedformation of IPs is supported by additional data. For example, genistein treatmentspecifically decreases EPEC induced tyrosine phosphorylation of Hp90 by —90 %(Rosenshine et at., 1993). In addition, the ability of EPEC mutants to trigger tyrosinephosphorylation of Hp90 correlates with their ability to induce release of IPs.Vanadate/H20treatment of HeLa cells, which inhibits PTPs, was found tospecifically stimulate the uptake of EPEC and a noninvasive signal transduction mutant(cfm). This treatment did not trigger the uptake of a noninvasive lab strain of E. coli,HB1O1, suggesting that an invasion specific pathway is being activated. Increased levels98of tyrosine phosphorylation are known to be a prerequisite to the rearrangement ofcytoskeletal proteins and bacterial uptake. It is possible that tyrosine phosphorylation ofHp90 or an unidentified protein(s) plays some role in EPEC invasion. Alternatively,EPEC may inhibit tyrosine protein dephosphorylation of some host protein(s) to induceuptake.Salmonella lyphimurium and Helicobacter pylon also induce the formation of IPsin infected cells (Ruschkowski et al., 1992; Dytoc et al., 1993). S. typhimurium rapidlystimulates release of IPs in epithelial cells, with maximal formation occurring 30 minutesafter infection, simultaneous with its invasion rate (Ruschkowski et al., 1992). However,in contrast to EPEC, tyrosine kinase inhibitors do not decrease entry of S. typhimuriuminto host cells (Rosenshine et al., 1992). Although both EPEC and S. typhimuriumthgger a flux of IPs, different host cell mechanisms appear to be activated bothdownstream and upstream of this common signal messenger.Several loci in EPEC have been shown to be important in invasion. The eaeAmutant is able to trigger host cell signals, but fails to associate intimately with the hostcells and to tightly aggregate host cell cytoskeletal and tyrosine phosphorylated proteins.The product of the eaeA gene, intimin, is an outer membrane protein (OMP) that isnecessary for invasion. However, intimin, unlike invasin of Yersinia species (Isberg etal., 1987), cannot alone confer the invasive phenotype to E. coli laboratory strains. Twostrains of a different noninvasive mutant, cfm::TnphoA [14-2-1(1) and 27-3-2(1)], whichexpress intimin, retain the ability to adhere intimately with the host cell. The cfmmutants, however, fail to induce host cell signals, including tyrosine phosphorylation ofHp90 and the release of IPs. A recently constructed EPEC mutant (UMD864), whichcarries an in-frame deletion of the eaeB gene, was found to be deficient in the ability totrigger host cell signal transduction pathways. The eaeB deletion mutant is also unable toattach intimately or induce the accumulation of filamentous actin in eucaryotic cells,suggesting that the eaeB gene product is involved in establishing intimate attachment99(Donnenberg et al., 1993). However, unlike eaeA mutants, the 1eaeB mutant is capableof producing intimin and exporting it to the cell surface. In addition, the protein encodedby the eaeB gene contains no signal sequence for protein export, suggesting that, unlikeintimin, the EaeB protein may be located in the cytoplasm (Donnenberg et al., 1993). Inthis study I demonstrate that a functional eaeB locus is required, not only for intimateattachment to epithelial cells, but also for several processes that can be conceptualized assignal transduction events: the eaeB mutant is deficient in inducing rearrangement in hostepithelial cells of tyrosine phosphorylated and cytoskeletal proteins, tyrosinephosphorylation of Hp90, and the release of inositol phosphates. These signaltransduction events distinguish the eaeB mutant from the eaeA mutant, which is deficientonly in intimate attachment.The phenotype of the eaeB mutant is similar to cfm mutants that are deficient insignal transduction. Like the cfm::TnphoA mutants [14-2-1(1) and 27-3-2(1)], we foundthat the AeaeB mutant is unable to trigger the tyrosine phosphorylation of Hp90.Tyrosine kinase activity appears to be essential for initiating downstream host signalsincluding release of inositol phosphates (IPs) (Foubister et al., 1994), elevation ofintracellular calcium (Baldwin et al., 1991; Berridge et a!., 1993), rearrangement ofcytoskeletal proteins, and bacterial uptake (Rosenshine et a!., 1992). Both the AeaeB andcfin::TnphoA mutants are deficient in these subsequent signals. Despite these phenotypicsimilarities, transformation with the cloned eaeB gene does not restore the wild typephenotype to the cfm mutants. Thus the cfm and eaeB loci are genetically distinct andmust encode separate factors in the pathway leading to signal transduction.The internalization efficiency of the cfm mutants increased 30 to 50-fold whencoinfected with the eaeA strain CVD2O6 (Table 4.5, Rosenshine et a!., 1992). Incontrast, the EeaeB mutant demonstrated a 15-fold increase in invasion efficiency whencoinfected with the eaeA : :TnphoA mutant. This difference in complementation efficiencymay be due to the low binding capacity of the AeaeB mutant. Whereas the cfm mutants100perform localized adherence similar to wild type EPEC, the /xeaeB mutant is not onlyattenuated in intimate adherence (Donnenberg et at., 1992) but also forms smallermicrocolonies on the host cell surface (Donnenberg et al., 1993).The function of the eaeB gene may be to covalently modify a signaling proteinupstream of two pathways which independently are responsible for promoting intimateattachment and activating host cell kinases. Undoubtedly other unidentified genes couldalso be involved in these pathways. Most of the signals which have been studied so farare activated 2-3 h after infection. The eaeB gene could function prior to these signals,thereby affecting two different infection phenotypes. Support for the speculation that theeaeB gene may perform a role in protein modification includes the identification in theeaeB sequence of a motif common to certain aminotransferase enzymes (Donnenberg etal., 1993). In addition, the absence of an obvious signal sequence for protein exportsuggests that the EaeB protein may be located in the bacterial cytoplasm, where it can actonly indirectly through another protein(s) to cause host cell changes. One possible targetof the EaeB protein may be intimin as the expression of both eaeA and eaeB is requiredfor intimate attachment. Alternatively, since recent observations suggest the EaeBprotein may be secreted (Kenny and Finlay, unpublished observations), it may actdirectly to exert its effect on host epithelial cells.Recent advances in the study of EPEC pathogenesis suggest that several differentbacterial factors are required to establish intimate attachment to epithelial cells, activatehost signals, disrupt the cytoskeleton, efface microvilli and promote uptake into hostcells. Based on my findings, I propose a model in which the eaeA and eaeB genesfunction to establish intimate contact between EPEC and the host cell (Fig. 5.26).Following this association, the cfm and eaeB genes trigger tyrosine phosphorylation ofHp90 and possibly other host cell substrates. A host PLC is activated, leading to therelease of inositol phosphates. Increased 1P3 levels trigger the release of calcium fromintracellular stores. The increased calcium levels could then induce cytoskeletal101rearrangements and activate calcium-dependent kinase(s) resulting in morphologicalchanges in the host cell. The eaeA gene product, intimin, is involved in assembling hostcytoskeletal elements and tyrosine phosphorylated proteins into an organized structure.Though several genes involved in this pathway have been identified, many other loci mayyet be discovered. The eaeB gene is the first locus identified whose product appears toaffect two different pathways, the establishment of intimate adherence and the activationof host cell kinases, suggesting that the expression of invasion-related loci may be undercommon control.102Figure 5.26A model of the interaction of EPEC with HeLa cells. EPEC attaches to the cell surfaceand uses the cfm and eaeB gene products to trigger Hp90 tyrosine phosphorylation byinducing a host tyrosine protein kinase (TPK) activity. This, possibly with other signals,activates a host PLC to cleave PIP2 to DAG and 1P3 1P3 mobilizes [Ca2ijwhich initiatescytoskeletal rearrangement beneath the attached bacterium. The eaeA gene product,intimin, and the EaeB protein are involved in nucleating the sub-bacterial structure thatcontains both signaling and cytoskeletal proteins.actinc-actininCCfmEaeBAGVTPK2\‘P3\IP3R104AppendixPublicationsFoubister, V., Rosenshine, I., Donnenberg, M.S., and Finlay, B. B. (1994). The eaeB geneof enteropathogenic Escherichia coil (EPEC) is necessary for signal transduction inepithelial cells. Submitted, Infect. Immun.Rosenshine, I., Ruschowski, S., Foubister, V. and Finlay, B. B.. (1994). Host proteintyrosine phosphorylation during Salmonella typhimurium invasion into epithelialcells. Submitted, Infect. Im,nun.Foubister, V., Rosenshine, I., and Finlay, B. B. (1994). A diarrheal pathogen,enteropathogenic Escherichia coil (EPEC), triggers a flux of inositol phosphates ininfected epithelial cells. J. Exp. Med. In press.Tang, P., Foubister, V., Pucciarelli, M.G., and Finlay, B. B. (1993). Methods to studybacterial invasion J. Micro. Methods 18,227-240.105BibliographyAkiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S., Itoh, N., Shibuya, M.,and Fukami, Y. (1987). Genistein, a specific inhibitor of tyrosine-specific proteinkinases. J. Biol. Chem. 262, 5592-5.Baldini, M. M., Kaper, J. B., Levine, M. M., Candy, D. C., and Moon, H. W. (1983).Plasmid-mediated adhesion in enteropathogenic Escherichia coli. J. Pediatr.Gastroenterol. Nutr. 2, 534-8.Baldwin, T. J., Brooks, S. F., Knutton, S., Manjarrez, H. H. A., Aitken, A., and Williams,P. H. (1990). Protein phosphorylation by protein kinase C in HEp-2 cells infectedwith enteropathogenic Escherichia coli [published erratum appears in Infect. Immun.58, 2024]. Infect. Immun. 58, 761-5.Baldwin, T. J., Lee, D. M. B., Knutton, S., and Williams, P. H. (1993). Calciumcalmodulin dependence of actin accretion and lethality in cultured HEp-2 cellsinfected with enteropathogenic Escherichia coli. Infect. Immun. 61,760-3.Baldwin, T. J., Ward, W., Aitken, A., Knutton, S., and Williams, P. H. (1991). Elevationof intracellular free calcium levels in HEp-2 cells infected with enteropathogenicEscherichia coli. Infect. Immun. 59, 1599-604.Berridge, M. J. (1993). Inositol trisphosphate and calcium signalling. Nature 361, 3 15-25.Bliska, 3. B., Clemens, J. C., Dixon, 3. E., and Falkow, 5. (1992). The Yersinia tyrosinephosphatase: specificity of a bacterial virulence determinant for phosphoproteins inthe J774A.1 macrophage. J. Exp. Med. 176, 1625-30.Bliska, J. B., Copass, M. C., and Falkow, S. (1993a). The Yersinia pseudotuberculosisadhesin YadA mediates intimate bacterial attachment to and entry into HEp-2 cells.Infect. Immun. 61, 39 14-21.Bliska, 3. B., and Falkow, S. (1993). The role of host tyrosine phosphorylation inbacterial pathogenesis. Trends Genet. 9, 85-9.Bliska, J. B., Galan, J. E., and Falkow, S. (1993b). Signal transduction in the mammaliancell during bacterial attachment and entry. Cell 73, 903-20.Bliska, J. B., Guan, K. L., Dixon, 3. E., and Failcow, 5. (1991). Tyrosine phosphatehydrolysis of host proteins by an essential Yersinia virulence determinant. Proc. Nati.Acad. Sci. USA 88, 1187-91.Bolin, I., Norlander, L., and Wolf, W. H. (1982). Temperature-inducible outer membraneprotein of Yersinia pseudotuberculosis and Yersinia enterocolitica is associated withthe virulence plasmid. Infect. Immun. 37, 506-12.106Bolin, I., Portnoy, D. A., and Wolf, W. H. (1985). Expression of the temperature-inducible outer membrane proteins of yersiniae. Infect. Immun. 48, 234-40.Bower, J. R., Congeni, B. L., Cleary, T. G., Stone, R. T., Wanger, A., Murray, B. E.,Mathewson, 3. 1., and Pickering, L. K. (1989). Escherichia coli 0114: nonmotile as apathogen in an outbreak of severe diarrhea associated with a day care center. J. Infect.Dis. 160, 243-7.Bray, J. (1945). Isolation of antigenically homogeneous strains of Bact. colineopolitanum from summer diarrhoea of infants. J. Path. Bacteriol. 57, 239-247.Brubaker, R. R. (1991). Factors promoting acute and chronic diseases caused byyersiniae. Clin. Microbiol. Rev. 4, 309-24.Burridge, K., Turner, C. E., and Romer, L. H. (1992). Tyrosine phosphorylation ofpaxillin and pp125FAK accompanies cell adhesion to extracellular matrix: a role incytoskeletal assembly. J. Cell. Biol. 119, 893-903.Canals, F. (1992). Signal transmission by epidermal growth factor receptor: coincidenceof activation and dimerization. Biochemistry 31, 4493-501.Canil, C., Rosenshine, I., Ruschkowski, S., Donnenberg, M. S., Kaper, 3. B., and Finlay,B. B. (1993). Enteropathogenic Escherichia coli decreases the transepithelialelectrical resistance of polarized epithelial monolayers. Infect. Immun. 61, 2755-2762.Carpenter, G. L., and Cantley, L. C. (1990). Phosphoinositide kinases. Biochemistry 29,11147-56.Cicchetti, P., Mayer, B. 3., Thiel, G., and Baltimore, D. (1992). Identification of a proteinthat binds to the SH3 region of Abi and is similar to Bcr and GAP-rho. Science 257,803-6.Clark, S., and Dunlop, M. (1991). Modulation of phospholipase A2 activity by epidermalgrowth factor (EGF) in CHO cells transfected with human EGF receptor. Role ofreceptor cytoplasmic subdomain. Biochem. J. 274, 7 15-21.Cornelis, G., Laroche, Y., Balligand, G., Sory, M. P., and Wauters, G. (1987). Yersiniaenterocolitica, a primary model for bacterial invasiveness. Rev. Infect. Dis. 9, 64-87.Cravioto, A., Gross, R. 3., Scotland, S. M., and Rowe, B. (1979). An adhesive factorfound in strains of Escherichia coil belonging to the traditional infantileenteropathogenic serotypes. Curr. Microbiol. 3, 95-99.Crouch, M. F., and Hendry, I. A. (1993). Growth factor second messenger systems:oncogenes and the heterotrimeric GTP-binding protein connection. Med. Res. Rev. 13,105-23.DeSimone, D. W., Stepp, M. A., Patel, R. S., and Hynes, R. 0. (1987). The integrinfamily of cell surface receptors. Biochem. Soc. Trans. 15, 789-9 1.Donnenberg, M. S., Calderwooci, S. B., Donohue, R. A., Keusch, G. T., and Kaper, J. B.(1990). Construction and analysis of TnphoA mutants of enteropathogenicEscherichia coil unable to invade HEp-2 cells. Infect. Immun. 58, 1565-7 1.107Donnenberg, M. S., Donohue, R. A., and Keusch, G. T. (1990). A comparison of HEp-2cell invasion by enteropathogenic and enteroinvasive Escherichia coli. FEMSMicrobiol. Lett. 57, 83-6.Donnenberg, M. S., Donohue, R. A., and Keusch, G. T. (1989). Epithelial cell invasion:an overlooked property of enteropathogenic Escherichia coli (EPEC) associated withthe EPEC adherence factor. J. Infect. Dis. 160, 452-9.Donnenberg, M. S., Giron, J. A., Nataro, J. P., and Kaper, J. B. (1992). A plasmidencoded type IV fimbrial gene of enteropathogenic Escherichia coli associated withlocalized adherence. Mo!. Microbiol. 6, 3427-37.Donnenberg, M. S., and Kaper, J. B. (1991). Construction of an eae deletion mutant ofenteropathogenic Escherichia coli by using a positive-selection suicide vector. Infect.Immun. 59, 43 10-7.Donnenberg, M. S., and Kaper, J. B. (1992). Enteropathogenic Escherichia coli. Infect.Immun. 60, 3953-61.Donnenberg, M. S., Yu, J., and Kaper, J. B. (1993). A second chromosomal genenecessary for intimate attachment of enteropathogenic Escherichia coli to epithelialcells. J. Bacteriol. 175, 4670- 80.Downes, C. P., and Carter, A. N. (1991). Phosphoinositide 3-kinase: a new effector insignal transduction? Cell. Signal. 6, 501-13.Dytoc, M., Gold, B., Louie, M., Huesca, M., Fedorko, L., Crowe, S., Lingwood, C.,Brunton, J., and Sherman, P. (1993). Comparison of Helicobacter pylon andattaching-effacing Escherichia coli adhesion to eukaryotic cells. Infect. Immun. 61,448-56.Echeverria, P. D., Chang, C. P., and Smith, D. (1976).Enterotoxigenicity and invasivecapacity of “enteropathogenic” serotypes of Escherichia coli. J. Pediatn. 89, 8-10.Egan, S. E., and Weinberg, R. A. (1993). The pathway to signal achievement. Nature365, 78 1-2.Emody, L., Heesemann, J., Wolf, W. H., Skurnik, M., Kapperud, G., OToole, P., andWadstrom, T. (1989). Binding to collagen by Yersinia enterocolitica and Yersiniapseudotuberculosis: evidence for yopA-mediated and chromosomally encodedmechanisms. J. Bacteriol. 171, 6674-9.Fainl, W. J., Escobedo, J. A., Martin, G. A., Turck, C. W., del Rosario, M., McCormick,F., and Williams, L. T. (1992). Distinct phosphotyrosines on a growth factor receptorbind to specific molecules that mediate different signaling pathways. Cell 69, 413-23.Finlay, B. B., and Falkow, S. (1988). Comparison of the invasion strategies used bySalmonella cholenae-suis, Shigella flexneri and Yersinia enterocoiltica to entercultured animal cells: endosome acidification is not required for bacterial invasion orintracellular replication. Biochimie 70, 1089-99.Finlay, B. B., Rosenshine, I., Donnenberg, M. S., and Kaper, J. B. (1992). Cytoskeletalcomposition of attaching and effacing lesions associated with enteropathogenicEscherichia coli adherence to HeLa cells. Infect. Immun. 60, 2541-3.108Finlay, B. B., Ruschkowski, S., and Dedhar, S. (1991). Cytoskeletal rearrangementsaccompanying Salmonella entry into epithelial cells. .1. Cell. Sd. 99, 283-96.Fischer, E. H., Charbonneau, H., and Tonks, N. K. (1991). Protein tyrosine phosphatases:a diverse family of intracellular and transmembrane enzymes. Science 253, 401-6.Foubister, V., Rosenshine, I., and Finlay, B. B. (1994). A diarrheal pathogen,enteropathogenic Escherichia coli (EPEC), triggers a flux of inositol phosphates ininfected epithelial cells. J. Exp. Med. In press.Francis, C. L., Jerse, A. E., Kaper, J. B., and Falkow, S. (1991). Characterization ofinteractions of enteropathogenic Escherichia coli 0127:H6 with mammalian cells invitro. J. Infect. Dis. 164, 693-703.Francis, C. L., Ryan, T. A., Jones, B. D., Smith, S. J., and Falkow, S. (1993). Rufflesinduced by Salmonella and other stimuli direct macropinocytosis of bacteria. Nature364, 639-42.Fuzi, M., and Pillis, I. (1962). Production of opacity in egg-yolk medium by Listeriamonocytogenes. Nature 13, 195.Galan, J. E., Pace, J., and Hayman, M. J. (1992). Involvement of the epidermal growthfactor receptor in the invasion of cultured mammalian cells by Salmonellatyphimurium . Nature 357, 588-9.Galyov, E. E., Hakansson, S., Forsberg, A., and Wolf, W. H. (1993). A secreted proteinkinase of Yersinia pseudotuberculosis is an indispensable virulence determinant.Nature 361, 730-2.Gill, G. N., Bertics, P. J., and Santon, J. B. (1987).Epidermal growth factor and itsreceptor. Mol. Cell. Endocrinol. 51, 169-86.Giron, 3. A., Ho, A. S., and Schoolnik, G. K. (1991). An inducible bundle-forming pilusof enteropathogenic Escherichia coli. Science 254, 7 10-3.Goguen, J. D., Walker, W. S., Hatch, T. P., and Yother, J. (1986). Plasmid-determinedcytotoxicity in Yersinia pestis and Yersinia pseudotuberculosis. Infect. Immun. 51,788-94.Goldschmidt, M. C., and DuPont, H. L. (1976). Enteropathogenic Escherichia coli: lackof correlation of serotype with pathogenicity. J. Infect. Dis. 133, 153-6.Gomes, T. A., Rassi, V., MacDonald, K. L., Ramos, S. R., Trabulsi, L. R., Vieira, M. A.,Guth, B. E., Candeias, J. A., Ivey, C., and Toledo, M. R. et al. (1991).Enteropathogens associated with acute diarrheal disease in urban infants in Sao Paulo,Brazil. J. Infect. Dis. 164, 331-7.Gomez, 0., and Kaper, 3. B. (1992). Plasmid-encoded factors regulate the expression ofeae gene of enteropathogenic E. coli. Abstra. 92nd Gen Meet. Am. Soc. Microbiol.,American Society for Microbiology, Washington, D.C., abstr. B-179, p. 55.109Guan, K. L., and Dixon, 3. E. (1990). Protein tyrosine phosphatase activity of an essentialvirulence determinant in Yersinia. Science 249, 553-6.Hack, N., Margolis, B. L., Ulirich, A., Schiessinger, J., and Skorecki, K. L. (1991a).Distinct structural specificities for functional coupling of the epidermal growth factorreceptor to. calcium-signalling versus phospholipase A2 responses. Biochem. J. 275,563-7.Hack, N., Margolis, B., Schiessinger, J., and Skorecki, K. (1991b). Interaction ofepidermal growth factor with vasoactive hormones in the regulation of phospholipaseA2. J. Basic Clin. Physiol. Phar,nacol. 2, 16 1-82.Hammacher, A., Melistrom, K., Heldin, C. H., and Westermark, B. (1989). Isoformspecific induction of actin reorganization by platelet-derived growth factor suggeststhat the functionally active receptor is a dimer. EMBO .1. 8, 2489-95.Heesemann, J., Gross, U., and Gruter, L. (1987). Genetic manipulation of virulence ofYersinia enterocolitica. Contrib. Microbiol. Im,nunol. 9, 3 12-6.Isberg, R. R. (1990). Pathways for the penetration of enteroinvasive Yersinia intomammalian cells. Mo!. Biol. Med. 7, 73-82.Isberg, R. R., and Falkow, S. (1985). A single genetic locus encoded by Yersiniapseudotuberculosis permits invasion of cultured animal cells by Escherichia coli K-12. Nature 317, 262-4.Isberg, R. R., and Leong, J. M. (1988). Cultured mammalian cells attach to the invasinprotein of Yersinia pseudotuberculosis. Proc. Nati. Acad. Sci. U S A 85, 6682-6.Isberg, R. R., and Leong, J. M. (1990). Multiple beta 1 chain integrins are receptors forinvasin, a protein that promotes bacterial penetration into mammalian cells. Cell 60,861-7 1.Isberg, R. R., Voorhis, D. L., and Falkow, S. (1987). Identification of invasin: a proteinthat allows enteric bacteria to penetrate cultured mammalian cells. Cell 50, 769-78.Jerse, A. E., and Kaper, 3. B. (1991). The eae gene of enteropathogenic Escherichia coliencodes a 94-kilodalton membrane protein, the expression of which is influenced bythe EAF plasmid. Infect. Immun. 59, 4302-9.Jerse, A. E., Yu, J., Tall, B. D., and Kaper, J. B. (1990). A genetic locus ofenteropathogenic Escherichia coli necessary for the production of attaching andeffacing lesions on tissue culture cells. Proc. Natl. Acad. Sci. U S A 87, 7839-43.Juliano, R. L., and Haskill, S. (1993). Signal transduction from the extracellular matrix. J.Cell. Biol. 120, 577-85.Kazlauskas, A., and Cooper, I. A. (1990). Phosphorylation of the PDGF receptor betasubunit creates a tight binding site for phosphatidylinositol 3 kinase. EMBO J. 9,3279-86.Keller, T. C., and Mooseker, M. S. (1982). Ca++-calmodulin-dependent phosphorylationof myosin, and its role in brush border contraction in vitro. J. Cell. Biol. 95, 943-59.110Knutton, S., Baldwin, T., Williams, P. H., and McNeish, A. S. (1989). Actinaccumulation at sites of bacterial adhesion to tissue culture cells: basis of a newdiagnostic test for enteropathogenic and enterohemorrhagic Escherichia coli. Infect.Immun. 57, 1290-8.Knutton, S., Baldini, M. M., Kaper, J. B., and McNeish, A. S. (1987). Role of plasmidencoded adherence factors in adhesion of enteropathogenic Escherichia coil to HEp-2cells. Infect. Immun. 55, 78-85.Koch, C. A., Anderson, D., Moran, M. F., Ellis, C., and Pawson, T. (1991). SH2 and SH3domains: elements that control interactions of cytoplasmic signaling proteins. Science252, 668-74.Kocks, C., Gouin, E., Tabouret, M., Berche, P., Ohayon, H., and Cossart, P. (1992). L.monocytogenes-induced actin assembly requires the actA gene product, a surfaceprotein. Cell 68, 521-531.Kucera, G. L., and Rittenhouse, S. E. (1990). Human platelets form 3-phosphorylatedphosphoinositides in response to aipha-thrombin, U46619, or GTP gamma S. J. Biol.Chem. 265, 5345-8.Labigne-Roussel, A., Schmidt, M. A., Wa;z, W., and Falkow, 5. (1985) Geneticorganization of the afimbrial adhesion operon and nucleotide sequence from auropathogenic Escherichia coil gene encoding an afimbrial adhesin. J. Bacterioi. 162,1285-92.Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the headof bacteriophage T4. Nature 227, 680-5.Leung, K. Y., Reisner, B. S., and Straley, S. C. (1991). YopM inhibits plateletaggregation and is necessary for virulence of Yersinia pestis in mice. Infect. Immun,58, 3262-327 1.Levine, M. M., Bergquist, E. J., Nalin, D. R., Waterman, D. H., Homick, R. B., Young,C. R., Sotman, S., and Rowe, B. (1978). Escherichia coli strains that cause diarrhoeabut do not produce heat-labile or heat-stable enterotoxins and are non-invasive.Lanceti, 1119-22.Levine, M. M., and Edelman, R. (1984). Enteropathogenic Escherichia coli of classicserotypes associated with infant diarrhea: epidemiology and pathogenesis. Epidemiol.Rev. 6, 31-51.Levine, M. M., Nataro, J. P., Karch, H., Baldini, M. M., Kaper, 3. B., Black, R. E.,Clements, M. L., and OBrien, A. D. (1985). The diarrheal response of humans tosome classic serotypes of enteropathogenic Escherichia coil is dependent on aplasmid encoding an enteroadhesiveness factor. J. Infect. Dis. 152, 550-9.Lian, C. 3., Hwang, W. S., and Pai, C. H. (1987). Anti-phagocytic role of plasmidassociated outer membrane proteins of Yersinia enterocolitica. Contrib. Microbioi.Immunol. 9, 186-9.Lowenstein, E. 3., Daly, R. 3., Batzer, A. G., Li, W., Margolis, B., Lammers, R., Ulirich,A., Skolnik, E. Y., Bar-Sagi, D., and Schiessinger, 3. (1992). The SH2 and SH3111domain-containing protein Grb2 links receptor tyrosine kinases to ras signaling. Cell70, 43 1-442.Lundrigan, M. D., Friedrich, M. J., and Kadner, R. J. (1989). Nucleotide sequence of theEscherichia coli porin thermoregulatory gene envY. Nucleic Acids Res. 17, 800.Majerus, P. W. (1992). Inositol phosphate biochemistry. Annu. Rev. Biochem. 61, 225-50.Manjarrez, H. H. A., Baldwin, T. J., Aitken, A., Knutton, S., and Williams, P. H. (1992).Intestinal epithelial cell protein phosphorylation in enteropathogenic Escherichia colidiarrhoea. Lancet 339, 52 1-3.Manoil, C., and Beckwith, J. (1985). TnphoA: a transposon probe for protein exportsignals. Proc. Nat!. Acad. Sci. US A 82, 8 129-33.Mayer, B. J., Hamaguchi, M., and Hanafusa, H. (1988). Characterization of p47gag-crk, anovel oncogene product with sequence similarity to a putative modulatory domain ofprotein-tyrosine kinases and phospholipase C. Cold Spring Harb. Symp. Quant. Biol.53, 907-14.McConnell, M. M., Chart, H., Scotland, S. M., Smith, H. R., Wilishaw, G. A., and Rowe,B. (1989). Properties of adherence factor plasmids of enteropathogenic Escherichiacoli and the effect of host strain on expression of adherence to HEp-2 cells. J. Gen.Microbiol. 135, 1123-34.McCormick, F. (1993). Signal transduction. How receptors turn Ras on [news;comment]. Nature 363, 15-6.Miller, V. L., Farmer, 3. J. 3., Hill, W. E., and Falkow, S. (1989). The ail locus is founduniquely in Yersinia enterocolitica serotypes commonly associated with disease.Infect. Immun. 57, 121-3 1.Moon, H. W., Whipp, S. C., Argenzio, R. A., Levine, M. M., and Giannella, R. A.(1983). Attaching and effacing activities of rabbit and human enteropathogenicEscherichia coli in pig and rabbit intestines. Infect. Immun. 41, 1340-5 1.Nakanishi, H., Brewer, K. A., and Exton, J. H. (1993). Activation of the isozyme ofprotein kinase C by phosphatidylinositol 3,4,5-trisphosphate. J. Biol. Chem. 268, 13-6.Nataro, J. P., Maher, K. 0., Mackie, P., and Kaper, 3. B. (1987). Characterization ofplasmids encoding the adherence factor of enteropathogenic Escherichia coli. InfectImmun. 55, 2370-7.Nataro, 3. P., Scaletsky, I. C., Kaper, J. B., Levine, M. M., and Trabulsi, L. R. (1985).Plasmid-mediated factors conferring diffuse and localized adherence ofenteropathogenic Escherichia coli. Infect. Immun. 48, 378-83.Neter, E. (1959). Enteritis due to enteropathogenic Escherichia coli. Present-day statusand unsolved problems. J. Pediatr. 55, 223-239.112Nishibe, S., Wahi, M. I., Hemandez, S. S. M., Tonks, N. K., Rhee, S. G., and Carpenter,G. (1990). Increase of the catalytic activity of phospholipase C-gamma 1 by tyrosinephosphorylation. Science 250, 1253-6.Nishizuka, Y. (1988). The molecular heterogeneity of protein kinase C and itsimplications for cellular regulation. Nature 334, 661-5.Pace, J., Hayman, M. J., and Galan, J. E. (1993). Signal transduction and invasion ofepithelial cells by S. zyphimurium. Cell 72, 505-14.Paulozzi, L. J., Johnson, K. E., Kamahele, L. M., Clausen, C. R., Riley, L. W., andHelgerson, S. D. (1986). Diarrhea associated with adherent enteropathogenicEscherichia coli in an infant and toddler center, Seattle, Washington. Pediatrics 77,296-300.Pawson, T. (1988). Non-catalytic domains of cytoplasmic protein-tyrosine kinases:regulatory elements in signal transduction. Oncogene 3, 49 1-5.Polotsky, Y. E., Dragunskaya, E. M., Seliverstova, V. G., Avdeeva, T. A.,Chakhutinskaya, M. G., Ketyi, I., Vertenyl, A., Ralovich, B., Emody, L., Malovics, I.,Safonova, N. V., Snigirevskaya, E. S., and Karyagina, E. I. (1977). Pathogenic effectof enterotoxigenic Escherichia coli and Escherichia coli causing infantile diarrhoea.Acta Microbiol. Acad. Sci. Hung. 24, 22 1-36.Rhee, S. G., Suh, P. G., Ryu, S. H., and Lee, S. Y. (1989). Studies of inositolphospholipid-specific phospholipase C. Science 244, 546-50.Robins, B. R. M. (1987). Traditional enteropathogenic Escherichia coli of infantilediarrhea. Rev. Infect. Dis. 9, 28-53.Rosenshine, I., and Finlay, B. B. (1993). Exploitation of host signal transductionpathways and cytoskeletal functions by invasive bacteria. BioEssays 15, 17-24.Rosenshine. I., Ruschkowski, S., Foubister, V., and Finlay, B. B. (1994). Host proteintyrosine phosphorylation during Salmonella typhimurium invasion into epithelialcells. Submitted.Rosenshine, I., Donnenberg, M. S., Kaper, 3. B., and Finlay, B. B. (1992). Signaltransduction between enteropathogenic Escherichia coli (EPEC) and epithelial cells:EPEC induces tyrosine phosphorylation of host cell proteins to initiate cytoskeletalrearrangement and bacterial uptake. EMBO J. 11, 355 1-60.Rosenshine, I., Duronio, V., and Finlay, B. B. (1992). Tyrosine protein kinase inhibitorsblock invasin-promoted bacterial uptake by epithelial cells. Infect. Immun. 60, 2211-7.Rosqvist, R., Bolin, I., and Wolf, W. H. (1988). Inhibition of phagocytosis in Yersiniapseudotuberculosis: a virulence plasmid-encoded ability involving the Yop2b protein.Infect. Immun. 56, 2139-43.Rosqvist, R., Forsberg, A., Rimpilainen, M., Bergman, T., and Wolf, W. H. (1990). Thecytotoxic protein YopE of Yersinia obstructs the primary host defence. Mol.Microbiol. 4, 657-67.113Rosqvist, R., Forsberg, A., and Wolf, W. H. (1991). Intracellular targeting of the YersiniaYopE cytotoxin in mammalian cells induces actin microfilament disruption. Infect.Im,nun. 59, 4562-9.Rothbaum, R. J., Giannella, R. A., and Partin, J. C. (1982). Diarrhea caused by adherententeropathogenic E. coil [letter]. J. Pediatr. 101, 486.Rothbaum, R. J., Partin, J. C., Saalfield, K., and McAdams, A. 3. (1983). Anultrastructural study of enteropathogenic Escherichia coil infection in human infants.Ultrastruct. Pathol. 4, 29 1-304.Ruschkowski, S., Rosenshine, I., and Finlay, B. B. (1992). Salmonella typhimuriuminduces an inositol phosphate flux in infected epithelial cells. FEMSMicrobiol. Lett.74, 121-6.Sadowski, I., Stone, J. C., and Pawson, T. (1986). A noncatalytic domain conservedamong cytoplasmic protein-tyrosine kinases modifies the kinase function andtransforming activity of Fujinami sarcoma virus Pl3Ogag-fps. Mo!. Cell. Biol. 6,4396-408.Scaletsky, I. C., Silva, M. L., Toledo, M. R., Davis, B. R., Blake, P. A., and Trabulsi, L.R. (1985). Correlation between adherence to HeLa cells and serogroups, serotypes,and bioserotypes of Escherichia coli. Infect. Immun. 49, 528-32.Scaletsky, I. C., Silva, M. L., and Trabulsi, L. R. (1984). Distinctive patterns ofadherence of enteropathogenic Escherichia coil to HeLa cells. Infect. Immun. 45, 534-6.Schuize-Koops H., Burkhardt, H., Heesemann, J., von der Mark, K., and Emmrich, F.(1992). Plasmid-encoded outer membrane protein YadA mediates specific binding ofenteropathogenic yersiniae to various types of collagen. Infect. Immun. 60, 2153-9.Senerwa, D., Olsvilc, 0., Mutanda, L. N., Lindqvist, K. J., Gathuma, J. M., Fossum, K.,and Wachsmuth, K. (1989). Enteropathogenic Escherichia coli serotype 0111 :HNTisolated from preterm neonates in Nairobi, Kenya. J. Clin. Microbiol. 27, 1307-11.Sohel, I., Puente, J. L., Murray, W. J., Vuopio, V. J., and Schoolnik, G. K. (1993).Cloning and characterization of the bundle-forming pilin gene of enteropathogenicEscherichia coil and its distribution in Salmonella serotypes. Mo!. Microbiol. 7, 563-75.Staley, T. E., Jones, E. W., and Corley, L. D. (1969). Attachment and penetration ofEscherichia coil into intestinal epithelium of the ileum in newborn pigs. A m. J.Pathol. 56, 37 1-92.Sternweis, P. C., and Snircka, A. V. (1992). Regulation of phospholipase C by Gproteins. TIBS 17, 502-6.Straley, S. C., and Bowmer, W. S. (1986). Virulence genes regulated at the transcriptionallevel by Ca2+ in Yersinia pestis include structural genes for outer membrane proteins.Infect. Immun. 51,445-54.Straley, S. C., Skrzypek, E., Plano, G. V., and Bliska, J. B. (1993). Yops of Yersinia spp.pathogenic for humans. Infect. Immun. 61, 3 105-10.114Takeuchi, A. (1967). Electron microscope studies of experimental Salmonella infection.I. Penetration into the intestinal epithelium by Salmonella typhimurium. Am. J.Pathol. 50, 109-136.Tang, P., Rosenshine, I., and Finlay, B. B. (1994). Listeria monocytogenes, an invasivebacterium, stimulates MAP kinase during entry into epithelial cells. Mo!. Biol. Cell.In press.Teitelbaum, I. (1990). The epidermal growth factor receptor is coupled to aphospholipase A2-specific pertussis toxin-inhibitable guanine nucleotide-bindingregulatory protein in cultured rat inner medullary collecting tubule cells. J. Biol.Chem. 265, 4218-22.Tertti, R., Skurnik, M., Vartio, T., and Kuusela, P. (1992). Adhesion protein YadA ofYersinia species mediates binding of bacteria to fibronectin. Infect. Immun. 60, 302 1-4.Titball, R. W. (1993). Bacterial phospholipases C. Microbiol. Rev. 57, 347-66.Trahey, M., and McCormick, F. (1987). A cytoplasmic protein stimulates normal N-rasp21 GTPase, but does not affect oncogenic mutants. Science 238, 542-5.Traynor-Kaplan A. E., Thompson, B. L., Harris, A. L., Taylor, P., Omann, G. M., andSklar, L. A. (1989). Transient increase in phosphatidylinositol 3, 4-bisphosphate andphosphatidylinositol trisphosphate during activation of human neutrophils. J. Biol.Chem. 264, 15668-73.Tzipori, S., Gibson, R., and Montanaro, J. (1989). Nature and disthbution of mucosallesions associated with enteropathogenic and enterohemorrhagic Escherichia coli inpiglets and the role of plasmid-mediated factors. Infect. Immun. 57, 1142-50.Ulirich, A., Coussens, L., Hayflick, J. S., Dull, T. J., Gray, A., Tam, A. W., Lee, J.,Yarden, Y., Libermann, T. A., and Schiessinger, J. et al. (1984). Human epidermalgrowth factor receptor cDNA sequence and aberrant expression of the amplified genein A431 epidermoid carcinoma cells. Nature 309, 4 18-25.Ullrich, A., and Schlessinger, J. (1990). Signal transduction by receptors with tyrosinekinase activity. Cell 61, 203-12.Ulshen, M. H., and Rob, J. L. (1980). Pathogenesis of Escherichia coli gastroenteritis inman--another mechanism. N. Engl. J. Med. 302, 99-101.Van Nhieu, G. T., and Isberg, R. R. (1991). The Yersinia pseudotuberculosis invasinprotein and human fibronectin bind to mutually exclusive sites on the alpha 5 beta 1integrin receptor. J. Biol. Chem. 266, 24367-75.Volberg, T., Zick, Y., Dror, R., Sabanay, I., Gilon, C., Levitzki, A., and Geiger, B.(1992). The effect of tyrosine-specific protein phosphorylation on the assembly ofadherens-type junctions. EMBO J. 11, 1733-42.Vuopio, V. J., and Schoolnik, G. K. (1991). Localized adherence by enteropathogenicEscherichia coli is an inducible phenotype associated with the expression of newouter membrane proteins. J. Exp. Med. 174, 1167-77.115Wahi, M. I., Sweatt, J. D., and Carpenter, G. (1987). Epidermal growth factor (EGF)stimulates inositol trisphosphate formation in cells which overexpress the EGFreceptor. Biochem. Biophys. Res. Commun. 142, 688-95.Weiss, B. A., and Insel, P. A. (1991). Intracellular Ca2+ and protein kinase C interact toregulate alpha 1-adrenergic- and bradykinin receptor-stimulated phospholipase A2activation in Madin-Darby canine kidney cells [published erratum appears in J. Biol.Chem. 266, 17720]. J. Biol. Chem. 266, 2126-33.Whitman, M., Downes, C. P., Keeler, M., Keller, T., and Cantley, L. (1988). Type Iphosphatidylinositol kinase makes a novel inositol phospholipid,phosphatidylinositol-3-phosphate. Nature 332, 644-6.World Health Organization (1992). Global estimates for health situation assessment andprojections 1992. WHO, Geneva, Switzerland.Yang, L. J., Baffy, G., Rhee, S. G., Manning, D., Hansen, C. A., and Williamson, J. R.(1991). Pertussis toxin-sensitive Gi protein involvement in epidermal growth factor-induced activation of phospholipase C-gamma in rat hepatocytes. J. Biol. Chem. 266,2245 1-8.Yang, Y., and Isberg, R. R. (1993). Cellular internalization in the absence of invasinexpression is promoted by the Yersinia pseudotuberculosis yadA product. Infect.Immun. 61, 3907-13.Yarden, Y., and Ulirich, A. (1988b). Growth factor receptor tyrosine kinases. Annu. Rev.Biochem. 57, 443-78.Yarden, Y., and Ulirich, A. (1988a). Molecular analysis of signal transduction by growthfactors. Biochemistry 27, 3113-9.Young, S. W., Poole, R. C., Hudson, A. T., Halestrap, A. P., Denton, R. M., and Tavare,J. M. (1993). Effects of tyrosine kinase inhibitors on protein kinase-independentsystems. FEBS 316, 278-82.Young, V. B., Failcow, S., and Schoolnik, G. K. (1992). The invasin protein of Yersiniaenterocolitica: internalization of invasin-bearing bacteria by eukaryotic cells isassociated with reorganization of the cytoskeleton. J. Cell. Biol. 116, 197-207.Yu, J., and Kaper, J. B. (1992). Cloning and characterization of the eae gene ofenterohaemorrhagic Escherichia coli O157:H7. Mol. Microbiol. 6, 411-7.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items