Principles of Bacterial Pathogenesis

Principles of Bacterial Pathogenesis This Page Intentionally Left Blank Principles of Bacterial Pathogenesis EDITED...

Author: Eduardo A. Groisman

274 downloads 2082 Views 50MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Principles of Bacterial Pathogenesis

This Page Intentionally Left Blank

Principles of Bacterial Pathogenesis EDITED BY

Eduardo A. Groisman Howard Hughes Medical Institute Washington University School of Medicine Department of Molecular Microbiology St. Louis, Missouri

ACADEMIC PRESS San Diego London

Boston

New York

Sydney

Tokyo

Toronto

This book is printed on acid-free paper, (oo)

Copyright © 2001 by ACADEMIC PRESS

All Rights Reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher Requests for permission to make copies of any part of the work should be mailed to: Permissions Department, Harcourt Inc., 6277 Sea Harbor Drive, Orlando, Florida 32887-6777

Academic Press a Harcourt Science and Technology Company 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.academicpress.com Academic Press Harcourt Place, 32 Jamestown Road, London NWl 7BY, UK http://www.academicpress.com

Library of Congress Catalog Card Number:

International Standard Book Number:

00-107418

0-12-304220-8

PRINTED IN UNITED STATES OF AMERICA 00 01 02 03 04 05 06 SB 9 8

7

6

5

Contents

Contributors Preface

xi XV

1. Evolution of Bacterial Pathogens HOWARD OCHMAN

I. II. III. IV. V. VI. VII. VIII. IX.

Introduction 2 The Genetic Basis of Virulence 2 Identification of Sequences Involved in Bacterial Pathogenesis 6 Recovery of Genes Contributing to Virulence 8 The Population Genetics of Pathogens 9 Studying Bacterial Population Genetics 10 The Organization of Genetic Diversity in Pathogenic Microorganisms Population Genetics of Representative Bacterial Pathogens 14 Conclusions 28 References 29

13

2. Germ Warfare: The Mechanisms of Virulence Factor Delivery JILL REISS HARPER AND THOMAS J. SILHAVY

I. II. III. IV. V VI. VII. VIII.

Introduction 43 The General Secretory Pathway 45 Autotransporters: Type V 47 Two-Step Secretion: Type II 49 ABC Transporters: Type I 52 Conjugal Transfer Systems: Type IV 55 Contact-Dependent Secretion: Type III 57 Concluding Remarks 61 References 61

vi

CONTENTS

3. Regulation of Virulence Gene Expression in Bacterial Pathogens CHARLES J. DORMAN AND STEPHEN G . J. SMITH

I. II. III. IV. V. VI. VII. VIII. IX. X. XL XII. XIII. XIV. XV. XVI. XVII. XVIII. XIX. XX. XXI. XXII.

Introduction 76 Transcription Initiation 77 Regulatory Protein Families 79 Covalent Modification of Transcription Factors 82 Regulatory Networks 86 The Oxidative Stress Response 87 The Modular Nature of Bacterial Regulatory Proteins 89 The Overlap between Genome Structure and Gene Regulation 92 Other Classes of Protein Regulators 94 DNA Structure and Gene Regulation 94 Stereotypical and Stochastic Events in the Control of Gene Expression The Switch Controlling Type 1 Fimbrial Expression in E. coli 101 Pap Pilus Gene Transcription 103 Contact-Dependent Gene Regulation 105 The Virulence Gene Regulatory Cascade of 5. y7exA2^n 106 A Thermometer Protein from the Salmonella Virulence Plasmid 109 Cell-Density-Dependent Regulation 110 Adaptive Mutation 114 Rare tRNAs and Translation Modulation 114 Protein Splicing 115 AntisenseRNA 115 Perspective 116 References 117

97

4. Strategies to Identify Bacterial Pathogenicity Factors ANDREW CAMILLI, D . SCOTT MERRELL, AND JOHN J. MEKALANOS

I. II. III. IV. V VI.

Introduction 134 Biochemical Strategies 135 Genetic Screens 149 Genetic Selections 159 Genomic Approaches 165 Concluding Remarks 169 References 170

5. Mechanisms of Bacterial Pathogenesis in Plants: Familiar Foes in a Foreign Kingdom JAMES R . ALFANO AND ALAN COLLMER

I. Introduction 180 II. An Overview of Bacterial Plant Pathogens and Plant Diseases 181 III. Tumorigenic Agrobacterium tumefaciens: Using the Type IV Secretion System to Transform the Host into a Factory for Bacterial Nutrients 186

CONTENTS

Vli

IV. Necrogenic, Stealth Pathogens: Parasites Strongly Dependent on the Hrp (Type III) Protein Secretion System 189 V. Necrogenic, Brute-Force Pathogens: Soft-Rotters Dependent on Type II Secretion of Pectic Enzymes 200 VI. Other Virulence Factors of Gram-Negative Plant Pathogens Compared with Those of Animal Pathogens 201 VII. Host Innate Immune Systems: Common Components in Pathogen Recognition and Defense Signaling 206 VIII. The R Gene Surveillance System: An Innate Immune System with Elaborate Recognition Specificity 207 IX. Pseudomonas aeruginosa: Dual-Kingdom Pathogenesis 209 X. Conclusions 210 References 211

6. Yersinia AoiFE P. BOYD AND GUY R . CORNELIS

I. II. III. IV. V. VI. VII.

Introduction 228 The Adhesive Factors 231 Iron Acquisition 236 Pathogenicity Islands 237 Yst Enterotoxin 238 The Yersinia Virulence Plasmid Conclusion 252 References 253

238

7. Molecular Pathogenesis of Salmonellae CHRISTINA A. SCHERER AND SAMUEL I. MILLER


Introduction 266 History 266 Taxonomy 267 Epidemiology and Clinical Disease 268 Clinical Course and Basic Immunology 272 In Vitro Models of Salmonella Virulence 280 Virulence Factors 290 Antibiotic-Resistant Salmonellae 312 5a/mon^//// 418 E. coli That Cause Sepsis and Meningitis 426 Conclusions 428 References 428

10. Molecular Basis of Vibrio cholerae Pathogenesis VICTOR J. DIRITA

I. II. III. IV. V. VI.

Introduction 457 Vibrio cholerae 458 Cholera 463 Molecular Mechanisms of Disease 465 Natural and Induced Immunity against Vibrio cholerae Infection Future Studies: The Past Is Prologue 493 References 495

\\. H. pylori Pathogenesis TIMOTHY L . COVER, DOUGLAS E . BERG, MARTIN J. BLASER, AND HARRY L . T. MOBLEY

I. II. III. IV. V.

Introduction 510 Epidemiology 510 Gastric Histology and Physiology 512 Clinical Diseases Associated with H. pylori Infection Microbiology 519

516

489

CONTENTS

VI. VII. VIII. IX. X. XI. XII.

ix

Genetic Diversity and Population Structure of H. pylori 521 Initial Gastric Colonization 524 Gastric Inflammation 529 Interactions of H. pylori with the Gastric Epithelium 531 Vacuolating Cytotoxin 532 Persistence of H. pylori Infection 536 Factors Influencing Development of Clinically Evident Disease References 542

12. Neisseria SCOTT D . GRAY-OWEN, CHRISTOPH DEHIO, THOMAS RUDEL, MICHAEL NAUMANN, AND THOMAS F. MEYER


Introduction 559 Natural Competence for Transformation Surface Structures 567 Tissue Colonization 570 PorB31 586 IgAl Protease 590 Iron Acquisition in Vivo 592 Immune Response 594 Summary 599 References 600

566

13. Bordetella PEGGY A. COTTER AND JEFF F. MILLER


Introduction 620 Respiratory Infections by B6>rj£'r6'//rt Species 621 Evolutionary Relationships among Bordetella Subspecies 624 Bordetella Virulence Factors 628 The Bordetella-Host Interaction 639 The BvgAS Sensory Transduction System 642 Phenotypic Modulation 646 Transcriptional Control of Bvg-Regulated Genes 640 The Role of Bvg-Mediated Signal Transduction in the Bordetella Life Cycle 654 References 658

14. Pathogenesis of Haemophilus influenzae Infections CHRISTOPH M . TANG, DEREK W . HOOD, AND E . RICHARD MOXON

I. II. III. IV. V.

Introduction 676 Population Biology 680 Molecular Determinants of Pathogenicity Pathogenesis 699 Conclusions 705 References 705

682

539

CONTENTS

15. Pathogenic Mechanisms in Streptococcal Diseases MICHAEL CAPARON

I. II. III. IV. V. VI. VII.

Introduction 717 Three Basic Mechanisms of Pathogenesis: Example of 5. pyogenes Steps Common to All Three Pathogenic Mechanisms 721 First Mechanism: Invasion and Multiplication in Tissue 728 Second Mechanism: Toxin-Mediated Disease 739 Third Mechanism: Immunopathological-Based Diseases 742 Concluding Remarks 743 References 743

719

16. Listeria monocytogenes HAFIDA FSIHI, PIERRE STEEPEN, AND PASCALE COSSART

I. General Overview of Listeria monocytogenes and Listeriosis II. Genetic Tools and Cell Biology Techniques to Study L. mono cytogenes Infection 755 III. Molecular Mechanisms for Entry and Spread of L. monocytogenes in Nonphagocytic Cells 758 IV. Regulation of L. monocytogenes Virulence Gene Expression V. Conclusion 787 References 787

Index

805

752

782

Contributors

Numbers in parentheses indicate the pages on which the authors' contributions begin.

(179), Department of Biological Sciences, University of Nevada, Las Vegas, Nevada 89154-4004

JAMES R. ALFANO

(509), Departments of Molecular Microbiology and Genetics, Washington University Medical School, St. Louis, Missouri 63110-1093

DOUGLAS E. BERG

J. BLASER (509), Department of Medicine, New York University School of Medicine, New York, New York 10016

MARTIN

AoiFE P. BOYD (227), Microbial Pathogenesis Unit, Universite Catholique de Louvain, ICP et Faculte de Medecine, Brussels 1200, Belgium (133), Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts 02111

ANDREW CAMILLI

(717), Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, Missouri 63110-1093

MICHAEL CAPARON

(179), Department of Plant Pathology, Cornell University, Ithaca, New York 14853-4203

ALAN COLLMER

(227), Microbial Pathogenesis Unit, Universite Catholique de Louvain, ICP et Faculte de Medecine, Brussels 1200, Belgium

GUY R. CORNELIS

(751), Unite de Interactions Bacteries-Cellules, Institut Pasteur, Paris, Cedex 15, 75724, France

PASCALE COSSART

A. COTTER (619), Department of Microbiology, Immunology, and Molecular Genetics, UCLA School of Medicine, Los Angeles, CA 90095-1747

PEGGY

(509), Department of Medicine and Microbiology and Immunology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232

TIMOTHY L. COVER

(559), Abteilung Infektionsbiologie, Max-Planck-Institut-fur Biologic, Tubingen 72076, Germany

CHRISTOPH DEHIO

XI

xii

CONTRIBUTORS

J. DIRITA (457), Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Michigan 48109

VICTOR

J. DORMAN (75), Department of Microbiology, Moyne Institute of Preventive Medicine, Trinity College, Dublin 2, Ireland

CHARLES

COUMARAN EGILE (335), Unite de Pathogenic Microbienne Moleculaire & Unite INSERM 389, Institut Pasteur, 75724 Paris, Cedex 15, France B.

(387), Biotechnology Laboratory, and the Departments of Biochemistry & Molecular Biology and Microbiology & Immunology, University of British Columbia, Vancouver, BC V6TIZ3, Canada

BRETT FINLAY

(751), Unite des Interactions Bacteries-Cellules, Institut Pasteur, Paris, Cedex 15, 75724, France

HAFIDA FSIHI

(559), Department of Medical Genetics and Microbiology, University of Toronto, Toronto M5S 1A8, Canada

SCOTT D . GRAY-OWEN

(43), Department of Molecular Biology, Princeton University, Princeton, New Jersey 08544

JILL REISS HARPER

DEREK W. HOOD (675), Molecular Infectious Diseases Group, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, 0X3 9DU, United Kingdom J. MEKALANOS (133), Department of Microbiology and Molecular Genetics, Shipley Institute of Medicine, Harvard Medical School, Boston, Massachusetts 02115

JOHN

D.

(133), Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts 02111

SCOTT MERRELL

(559), Max-Planck-Institut fur Infektionsbiologie, Department of Molecular Biology, 10117 Berlin, Germany

THOMAS F. MEYER

(619), Department of Microbiology, Immunology, and Molecular Genetics, UCLA School of Medicine, Los Angeles, CA 90095-1747

JEFF F. MILLER

I. MILLER (265), Departments of Medicine and Microbiology, University of Washington, Seattle, Washington 98195

SAMUEL

(509), Department of Microbiology and Immunology, University of Maryland School of Medicine, Baltimore, Maryland 212011559

HARRY L.T. MOBLEY

E.

(675), Molecular Infectious Diseases Group, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, 0X3 9DU, United Kingdom

RICHARD MOXON

(559), Department of Molecular Biology, Max-Planck-Institut-fur Infektionsbiologie, Berlin 10117, Germany

MICHAEL NAUMANN

CONTRIBUTORS

Xili

(1), Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona 85721

HOWARD OCHMAN

(387), Department of Molecular Microbiology, Instituto de Biotecnologia, Universidad Nacional Autonoma de Mexico, Cuemavaca 62210, Mexico

JOSE L. PUENTE

(559), Department of Molecular Biology, Max-Planck-Institutfiir Infektionsbiologie, Berlin 10117, Germany

THOMAS RUDEL

J. SANSONETTI (335), Unite de Pathogenic Microbienne Moleculaire & Unite INSERM 389, Institut Pasteur, 75724 Paris, Cedex 15, France

PHILIPPE

A. SCHERER (265), Departments of Medicine and Microbiology, University of Washington, Seattle, Washington 98195

CHRISTINA

J. SILHAVY (43), Department of Molecular Biology, Princeton University, Princeton, New Jersey 08544

THOMAS

(75), Department of Microbiology, Moyne Institute of Preventive Medicine, Trinity College, Dublin 2, Ireland

STEPHEN G J . SMITH

(751), Unite des Interactions Bacteries-Cellules, Institut Pasteur, Paris, Cedex 15, 75724, France

PIERRE STEFFEN

(675), Molecular Infectious Diseases Group, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, 0X3 9DU, United Kingdom

CHRISTOPH M . TANG

(335), Department of Medical Microbiology and Immunology, University of Goteborg, S41345 Goteborg, Sweden

CHRISTINE WENNERAS


Preface

Infectious diseases are the leading cause of morbidity and mortality worldwide. While some of these infections are caused by eukaryotic parasites, bacterial pathogens continue to present a threat to the well-being of humans and animals both in the developing and developed worlds. The use of vaccines and antibiotics, together with changes in sanitary practices, has contributed to an important increase in the life span of humans in the last century. However, these significant improvements are now challenged by the appearance of microbes resistant to multiple antibiotics, the emergence of new bacterial pathogens, and the use of health care treatments that, while prolonging life, render individuals susceptible to opportunistic pathogens. Then, what can be done to sustain the improvements in health care developed during the last 100 years? Novel strategies are currendy being tested to prevent and/or treat bacterial infections. An increasing number of these strategies are based on our understanding of the mechanisms by which pathogenic microorganisms cause disease. This is possible due to exciting developments in the field of bacterial pathogenesis, which started some 20 years ago with the use of molecular genetics to investigate the microorganisms responsible for causing disease, and are now complemented with cell biological and biochemical approaches aimed at unraveling the consequences that infection by such microorganisms has on their hosts. We now have a basic understanding not only of the varied nature of virulence determinants but also of their origin and acquisition by pathogenic microbes. We appreciate that expression of virulence determinants is most often regulated in response to host signals and that microbes use different devices to deliver toxic products to host cells. These studies have revealed a set of principles that govern bacterial pathogenesis and, as the tide indicates, constitutes the subject matter of this book. The motivation for Principles in Bacterial Pathogenesis was twofold: first, to provide in-depth coverage of the best-characterized bacterial pathogens, with the goal of uncovering the salient features that these microbes have in common which allow them to conquer new niches and to circumvent host defense mechanisms, and, second, to group contributions by the world experts in bacterial pathogenesis

XV

XVI

PREFACE

in which they present a general discussion of the subject beyond the work performed in their own laboratories. This book is divided in two parts that comprise a total of 16 chapters, each of which can be read independent of the rest. The first part consists of five chapters, three of which discuss aspects of bacterial pathogenesis that are common to all pathogens: evolution, secretion, and regulation of virulence determinants. The fourth chapter presents a thorough description of the strategies currently used to identify virulence determinants. The fifth chapter discusses bacterial pathogens of plants, highlighting the similar mechanisms that bacterial pathogens of animal and plants employ when interacting with their respective hosts. These first 5 chapters serve as a general introduction to the 11 pathogen-based chapters that comprise the second part of the book. Each of the latter chapters provides a broad discussion of the best-understood human pathogens. In sum, while novel aspects of pathogenic organisms will continue to be discovered, a basic understanding of the principles governing bacterial pathogenesis will not only allow us to appreciate the sophisticated mechanisms used by microbes in their pathogenic lifestyle, but will also be essential in beginning to understand the plethora of information emerging from genomics, and to develop new rational approaches to the treatment and prevention of infectious diseases. Eduardo A. Groisman Department of Molecular Microbiology Howard Hughes Medical Institute Washington University School of Medicine St. Louis, Missouri

CHAPTER 1

Evolution of Bacterial Pathogens HOWARD OCHMAN

I. II. HI. IV. V. VI.

Introduction The Genetic Basis of Virulence Identification of Sequences Involved in Bacterial Pathogenesis Recovery of Genes Contributing to Virulence The Population Genetics of Pathogens Studying Bacterial Population Genetics A. Multilocus Enzyme Electrophoresis B. DNA Sequencing C. Multilocus Sequence Typing vn. The Organization of Genetic Diversity in Pathogenic Microorganisms VIII. Population Genetics of Representative Bacterial Pathogens A. Bordetella B. Borrelia C. Escherichia coli and Shigella D. Haemophilus E. Helicobacter F. Listeria G. Mycobacterium H. Neisseria I. Salmonella J. Staphylococcus K. Streptococcus L. Vibrio IX. Conclusions References

Principles of Bacterial Pathogenesis Copyright © 2001 by Academic Press All rights of reproduction in any form reserved. ISBN 0-12-304220-8

2 2 6 8 9 10 10 12 12 13 14 14 16 17 19 19 20 21 22 23 25 26 27 28 29

2

HOWARD OCHMAN

/. Introduction In what ways do pathogenic microorganisms differ from nonpathogenic forms? For an organism to be considered a pathogen, it must, during some phase of its life-cycle, advance disease and alter the health or behavior of another organism, that is, its host. Every organism can serve as a host for pathogens, and pathogens come into contact with a very large number of species; however, most pathogenic microorganisms are virulent to relatively few, and often only one, host species, and infection may cause disease in only a limited segment of the host population. So despite the wide range of mechanisms deployed by pathogens to disable their hosts (and to promote their own replication and transmission), there is one common theme: virulence depends upon the susceptibility of a host. Therefore, the identification of pathogens, the differences between pathogenic and nonpathogenic microorganisms, and the specific factors required for virulence must each be defined with regard to its relevance to the host. This chapter addresses two general issues about the evolution of microbial pathogenesis. First, we consider the differences between the genomes of pathogenic and nonpathogenic bacteria, the specific types of genes that contribute to the virulence phenotype, and the evolutionary history of these sequences in the genome of pathogens. Next, we discuss the molecular population genetics of microbial pathogens and the factors that govern the organization of genetic diversity within these populations. Because the origins and genetic bases of virulence influence the population structure of pathogens, these topics are interconnected and broadly relevant to the emergence, outbreak, control, and prevention of the diseases caused by microbial pathogens.

//. Thie Genetic Basis of Virulence One way to gain insight into the specific factors contributing to virulence has been to identify the genetic differences between pathogens and closely related nonpathogenic bacteria. In this regard, there are potentially four types of genetic events that might be responsible for the differences in pathogenic potential among related bacteria: (1) virulence as a result of genes specific to the pathogen; (2) virulence as a result of the absence of a suppressor locus in the pathogen; (3) virulence as a result of allelic differences between genes shared by the pathogen and nonpathogen; and (4) virulence as a result of the differential regulation of the same complement of genes in the pathogen and nonpathogen.

1.

VIRULENCE AS A RESULT OF SPECIES- OR STRAIN-SPECIFIC GENES

The most common approach to investigate virulence genes—on both analytical and technical grounds—has been to search for sequences that are restricted to

1.

THE EVOLUTION OF BACTERIAL PATHOGENS

3

pathogenic organisms. Virulence determinants are often acquired through horizontal transfer, which may explain why virulence traits are distributed sporadically among bacterial taxa (Fig. 1) or within some bacterial species (Fig. 2). In many bacterial species, there is an association between species-specific genes and virulence. Pathogenicity islands (or Pai)—i.e., segments of the chromosome that encode virulence genes and are absent from related nonpathogenic bacteria [1-3]—have been identified in pathogenic strains of E. coli [4-6], Shigella flexneri [7, 8, 24], Salmonella enterica [9], Yersinia pestis [10-12], Vibrio cholerae [13], Haemophilus influenzae [14, 15], Helicobacter pylori [16], and Staphylococcus aureus [17]. In each of these cases, a specific chromosomally encoded gene, or cluster of genes, has been implicated in the virulence of the microorganism, and the corresponding region was not present in avirulent strains or related species. Although it has long been known that species-specific regions confer traits that are unique to a bacterial species, the term "pathogenicity island" was first used to describe a DNA segment harbored by uropathogenic strains of E. coli [18]. Further characterization of pathogenicity islands revealed that many were situated at transfer RNA loci, which are commonly used as integration sites for foreign sequences [19]. For example, certain phages detected in E. coli, such as the retronphage (t)R73 [20] and the bacteriophage P4 [21] insert at, or near, tRNA genes, suggesting that pathogenicity islands are often transferred and acquired through phage-mediated events [1, 22]. The frequent insertion of foreign DNA sequences at tRNA genes is presumably due to the high degree of sequence conservation of tRNA genes across species. In fact, there is recurrent use of the

H:

Escherichia Shigella Salmonella Citrobacter Klebsiella Serratia Yersinia Proteus

Fig. 1 Phylogenetic relationships among enteric bacteria showing taxa that are normally capable of invading eukaryotic cells (denoted with darkened branches).

HOWARD OCHMAN

E. CO//K-12 ETEC0159:H4 EPEC0in:H12 ETEC078:Hn EIEC0n2:NM EPEC0m:H12 ECOR 5 EC0R6 ECOR 10 ECOR 14 (UTI)

-

ECOR27 ETEC0148:H28

EIEC0124:NM • Shigella flexneri Shigella boydii Shigella flexneri ETEC0159:H4 I ECOR 69 ' ECOR 30

LT ^

J ECOR 70 I ECOR 58 EIEC028:NM

ECOR 38

ECOR 61 ECOR 62 (UTI) ECOR 52 ECOR 64 (UTI) ECOR 59 ECOR 66 EPEC055:H6 Shigella sonnei I ECOR 50 (UTI) ^ 1 FPOR 4q ECOR 49

EHEC0157:H7 ECOR 37 Fig. 2 Relationships among commensal and pathogenic strains of Escliehchia coli and Shigella spp. based on nucleotide sequences of the gene encoding malate dehydrogenase (adapted with permission from Pupo et al. (1997) [56]). Abbreviations are as follows: UTI = urinary tract infection; EHEC = enterohemorrhagic E. coli; EIEC = enteroinvasive E. coli; EPEC = enteropathogenic E. coli; ETEC = enterotoxigenic E. coli. ECOR strains are from the E. coli reference collection [100], and the 0:H serotypes of pathogenic E. coli are noted.

tRNA^^^^ locus, which is targeted by (\>R\13 and as the integration site for several pathogenicity islands: Pai-1 of uropathogenic E. coli [4], the LEE island of enteropathogenic E. coli [23], the SHI-2 island of Shigella flexneri [8, 24], and the SPI-3 island of Salmonella enterica [25] each represent independent insertions of different virulence gene clusters into similar chromosomal locations. Flanking many pathogenicity islands there are signature sequences, such as short direct repeats, reminiscent of the integration of mobile elements (or even, in the case of Yersinia pestis, copies of the IS elements themselves) further attesting that these species- or strain-specific regions can be acquired laterally through a variety of transfer mechanisms. Although research on pathogenicity islands focuses on chromosomally encoded regions, genes involved in bacterial virulence are also carried on extrachromosomal elements that are maintained within the genome of pathogens. For example.

1.


5

many of the genes required for Shigella virulence reside on a 220-kb plasmid [26, 27], and, similarly, all virulent strains of Yersinia harbor a 70- to 75-kb plasmid that encodes proteins necessary for their antihost properties [28, 29]. In this regard, the acquisition of plasmid-borne antibiotic resistance genes will also allow previously sequestered pathogens to exploit new hosts. Other virulence determinants in these enteric species have been acquired by the organism in phage-mediated events. For example, the cytotoxins first characterized in Shigella are encoded on a bacteriophage that has subsequently been transferred to enterohemorrhagic strains of E. coli [30, 98]. In the case of Vibrio cholerae, two coordinately regulated factors contribute to virulence: cholera toxin, which is encoded by a filamentous bacteriophage (termed CTX(t)) related to the coliphage Ml3 [31], and the toxin-coregulated pili, which is encoded within a large pathogenicity island. This pathogenicity island of Vibrio cholera is in itself another filamentous bacteriophage [32], and this demonstrates a novel case where one horizontally acquired phage encodes the receptor for the products specified by a second phage, both of which are required for full virulence.

2.

VIRULENCE RESULTS FROM THE ABSENCE OF A SUPPRESSOR LOCUS

Similar to the mechanisms described above—whereby a microbe has acquired certain genes that render it virulent—it is also possible that the pathogen has either lost a gene encoding a product capable of diminishing its virulence potential, or that such a determinant was acquired by the related avirulent forms. An early example of a virulence suppressor in enteric bacteria is the surface protease OmpT, which is absent from the genomes of Shigella and enteroinvasive E. coli (EIEC). The presence of ompT results in attenuation of virulence because the encoded protease interferes with expression of the VirG protein, which is required for intercellular spread [33]. The ompT gtUQ is probably not ancestral to enteric bacteria: it is located within the 21-kb cryptic lambdoid phage, suggesting that avirulent strains of E. coli acquired ompT through horizontal gene transfer. In addition to lacking ompT, Shigellae are also devoid of genes whose products suppress virulence. Representatives of the four species of Shigella, as well as enteroinvasive strains of E. coli (Fig. 2), harbor deletions for the region containing cadA, which encodes lysine decarboxylase. When the cadA gene from an avirulent strain of E, coli was introduced into Shigella flexneri, the resulting strain was still able to invade cells in tissue culture, but did not exhibit the toxic effect that induces the fluid secretion normally associated with infection. In contrast to the situation where a microorganism gains genes that enhance its pathogenic potential (i.e., pathogenicity islands), these regions have been termed "black holes" to denote deletion of genes that reduce the pathogenic potential of an organism [34].

6

HOWARD OCHMAN

3.

VIRULENCE RESULTS FROM ALLELIC DIFFERENCES

BETWEEN HOMOLOGOUS GENES

Because pathogenic and nonpathogenic have often diverged in sequence, it is possible that the differences in pathogenic properties result from allelic variation in homologous genes due to nonsense or missense mutations. For example, point mutations in the fimH gene of E. coli can change the binding of fimbrial adhesins and confer increased virulence in the mouse urinary tract [35, 36].

4.

VIRULENCE RESULTS FROM THE DIFFERENTIAL REGULATION OF THE SAME COMPLEMENT OF GENES

In addition to genetic polymorphisms among bacterial strains and species, it is possible that the differences in pathogenic properties are caused by differential regulation of essentially the same set of genes. For example, the invasion gene complexes of S. enterica and S. flexneri are largely homologous, but controlled by very different environmental signals: invasion in S. enterica is regulated by oxygen tension [37], whereas the expression of virulence genes by Shigella is controlled by temperature [38]. The origin of virulence properties in many pathogenic species is likely to result from a combination of the factors presented above. For Shigella and enteroinvasive E. coli, it is clear that virulence is the result of the incorporation of a large virulence plasmid into a strain lacking the ompT gene and the deletion of cadA from their genomes. And while the virulence genes on the Shigella and EIEC plasmids are 99% identical, these species exhibit large differences in their median infective doses, which could be due to allelic variation or to differential regulation of homologous genes or to species-specific chromosomal genes. Also note that these types of genetic changes do not pertain to the analysis of most opportunistic or newly emerging pathogens. Because these microbes are displaced to nonstandard hosts, tissues or environments, the genes contributing to virulence are not preadapted to the host and are not likely to differ from the repertoire required for growth in their customary environments.

///. Identification of Sequences Invoived in Bacterial Pattiogenesis From the previous discussion, it is obvious that most of the differences in pathogenic potential among related bacteria are due to changes in gene content (mechanisms 1 and 2) rather than to changes in the ancestral genes themselves (mechanisms 3 and 4). Although the specific approach, as well as technical

1.


7

considerations, bias the identification of the particular genetic events and the recovery of genes involved in virulence (see below), the vast majority of traits that are unique to a species are encoded on segments of the genome that arose through horizontal transfer [39]. Stepwise mutational changes in existing genes only rarely confer novel functions, whereas traits encoded by acquired DNA will occasionally confer the ability to explore new hosts or environments and can have a large impact on bacterial evolution [40]. As a result, none of the phenotypic characteristics that distinguish E. coli from S. enterica are attributable to the divergence of homologous genes by mutation; instead, all of the species-specific traits derive from functions encoded by horizontally transferred genes (e.g., lactose utilization, citrate utilization, indole production) or from the loss of ancestral DNA (e.g., alkaline phosphatase) [39]. The broad association of species-specific traits with unique portions of the chromosome does not imply that all (or even the majority of) genes contributing to the virulence phenotype are restricted to pathogens and absent from the related nonpathogenic bacteria. The classification of genes as being involved in pathogenesis depends largely on the particular approach used to define and identify these factors [41]. The traditional approach to recognizing virulence determinants was purification of microbial products, which, upon introduction into a susceptible host, produced some of the symptoms advanced by the whole organism. This biochemical approach resulted in identification of a variety of factors, usually toxins, produced by several pathogens such as Vibrio cholerae and Clostridium botulinum. The classical bacterial genetics approach defines virulence genes as those that, on mutation, give rise to strains with median lethal doses (LD50) higher than that corresponding to the wild-type parent [42]. This interpretation of virulence includes all relevant loci—apart from those essential for growth under laboratory conditions—without assumptions about the precise role that particular virulence determinants play in the pathogenicity of the microorganism. The molecular genetic approach is used to isolate virulence genes based on their capacity to confer certain pathogenic properties on normally benign strains, such as E. coli. A prime example of this approach was the recovery of DNA segments contributing to the invasive character of Yersinia by selecting for clones that could render an E. coli laboratory strain capable of eliciting its uptake by epithelial cells [43]. Similarly, the introduction of a plasmid containing the LEE island into a laboratory isolate of E. coli creates strains that produce attachment and effacing lesions in host cells [23]. It is not surprising that these three strategies have led to the recovery of somewhat different subsets of genes involved in pathogenesis. However, if virulence is to be characterized in terms of the consequences of bacterial infection on the health of the host, the classical bacterial genetics approach—whose aim is to identify all genes that affect host fitness—provides the most comprehensive definition of virulence genes.

8

HOWARD OCHMAN

Given this perspective on bacterial pathogenicity, many of the "virulence" genes required for propagation of pathogens within a host would be identical to those required in commensal or benign interactions with hosts. In fact, the molecular genetic experiments, which attempt to convert nonpathogenic E. coli into pathogens, suggest that E. coli, as normal constituents of the human intestinal flora, already contain genes necessary for interaction with animal cells and are, thus, predisposed to become pathogens on acquisition of a particular virulence gene cluster. In addition, many of the genes implicated in Salmonella virulence are also present in nonpathogenic strains of E. coli [9]. These genes encode enzymes responsible for the biosynthesis of nutrients that are scarce within host tissues, transcriptional and posttranscriptional regulatory factors, proteins necessary for the repair of damaged DNA, and products necessary for defense against host microbicidal mechanisms. The presence of these genes in nonpathogenic species suggests that they promote survival within the nutritionally deprived and/or potentially lethal environments that microbes encounter inside and outside animal hosts.

IV. Recovery of Genes Contributing to Virulence Although the identification and isolation of virulence genes largely depends on how these genes are defined, many pathogens require genes that are absent from related nonpathogenic bacteria. Therefore, several researchers have applied molecular and genetic techniques to recover segments of the genome that are specific to particular bacterial lineages. These procedures yield anonymous DNA fragments and have been typically employed to obtain diagnostic probes for the identification of particular bacterial strains or species [44-49]. However, in a few cases, these techniques have been exploited to find new pathogen-specific genes having a potential role in virulence. A subtractive hybridization procedure [50] was used to recover DNA sequences present in an avian pathogenic strain of Escherichia coli but absent from a nonpathogenic laboratory strain of Escherichia coli K-12 [51]. The pathogen-specific sequences recovered by this method mapped to at least 12 positions in the chromosome. Subsequently, the phenotype of mutant strains harboring deletions for each of these unique fragments was tested, and two were found to contain genes required for virulence in avian hosts. Other studies have examined the unique DNA in the genome of pathogens, but have yet to directly assess the function of these sequences. For example, a subtractive hybridization procedure used to examine the differences in gene content among strains of Helicobacter pylori yielded 18 clones, several of which were presumed to have a role in the

1.


9

specific virulence characteristics of H. pylori strains [52]. Although genome subtraction and physical mapping techniques allow one to identify and clone the differences between two genomes, there is usually no rapid way to determine if these pathogen-specific sequences are indeed relevant to pathogenesis. A thorough review of the strategies used to identify virulence determinants is presented in the chapter by Camilli et al. in this book (see Chapter 4).

V. The Population Genetics of Pattiogens What is the genetic structure of pathogen populations, and how much genetic variation is present in these populations compared with that in related nonpathogenic microorganisms? Moreover, what is the apportionment of genetic diversity among pathogenic strains in relation to that in species at large? Although there are several genetic mechanisms by which microorganisms change their pathogenic potential, it has become evident that the bacteria responsible for disease outbreaks are distinct clones that are frequendy characterized by unique combinations of virulence genes or of alleles at virulence genes. The situation is not as clear for pathogenic species, principally because the classification of a pathogenic "species" is somewhat arbitrary, and historically reflects the ability of epidemiologists to classify strains. On one hand, pathogens have been overclassified—that is, they are typed into a multitude of genetically narrow groups or species—compared to nonpathogens. This is probably judicious, given the importance of assigning an identity of each isolate implicated in human disease. For example, based on serological characteristics, the Salmonellae were assorted into thousands of distinct species (now termed serovars), but a recent taxonomic revision based on DNA-DNA hybridization reclassified these strains into a single species. Salmonella enterica [53, 54]. Similarly, the Shigellae have traditionally been subdivided into four species—Shigella boydii, S.flexneri, S. dysenteriae, and S. sonnei—although the total amount of genetic variation within this genus is contained within E. coli [55-58]. The classification of Shigellae based on serologic and metabolic characteristics illustrates two additional points. First, the amount of diversity can vary widely among species: while Shigella sonnei consists a single genetically uniform clone, each of the other three species comprises a heterogeneous array of clones. Second, the classification of strains does not always reflect their true genetic similarities or relationships. As shown in Figure 2, certain strains of S. flexneri can be more closely related to S. boydii than to any other strains typed S. flexneri, and evidence from other studies have shown that Shigella boydii and S. dysenteriae have multiple origins from within E. coli [55-58].

10

HOWARD OCHMAN

In contrast to these enteric pathogens, the classification of strains into other pathogenic species is more liberal and diffuse, due, in part, to the inability to rapidly differentiate among isolates that display a diagnostic phenotype or produce a specific pathology in hosts. However, the subsequent analysis of such species using additional genetic and/or phenotypic markers often reveals the true nature of the diversity and the relationships among strains.

VL Studying Bacterial Popuiation Genetics Studies addressing the genetic structure of bacterial populations began relatively recently [59, 60] by evolutionary geneticists who had originally examined the amount and distribution of genetic variation among natural populations of eukaryotes. Bacteria, particularly E. coli and Salmonella, were an attractive group of organisms on account of their phenotypic diversity, haploid chromosomes, large populations sizes, short generation times, and ease of propagation and experimental manipulation. Hence, bacterial population genetic research was originated by population geneticists interested in bacteria, rather than by bacterial geneticists or medical microbiologists interested in population genetics. The information gathered by population geneticists has broad implications for medicine and epidemiology, and these studies typically go beyond the simple identification of strains and address questions pertaining to their genetic relationships and their levels of allelic diversity [61, 62]. Numerous molecular techniques—such as RAPDs, IS (and other repetitive element) fingerprinting, ribotyping, phage typing, macrorestriction mapping by pulsed-field gel electrophoresis, and plasmid profile analysis—have been applied to establish the identity of strains for the epidemiological purposes. But these methods typically do not supply the information necessary to establish the relationships among strains, infer the genetic structure of natural populations, or assess the relative roles of natural selection, random drift (i.e., the change in gene frequencies caused by the chance event of random sampling in small populations), new mutations, and horizontal gene transfer (including intragenic and intergenic recombination) on the organization of allelic diversity [63]. Once the evolutionary relationships of clones of a species is available, it is possible to examine the manner in which the total genetic diversity is apportioned with respect to host species, geographic populations, and the specific disease pathology. The following methods have been applied to establish the evolutionary relationships and genetic structure of bacterial populations.

A.

Muitilocus Enzyme Electrophoresis

Multilocus locus enzyme electrophoresis (or MLEE) has been the primary method used to assess genetic variation in bacterial populations [64]. The main advantage

1.

11


of this technique is that many genes can be readily examined in hundreds, if not thousands, of isolates. However, this method rehes on the discrimination of alleles of distinct electrophoretic mobilities—also called allozymes or electromorphs in this context—and, therefore, can detect only a portion of the sequence variation at a locus (Fig. 3). The key concept underlying the use of MLEE in population genetics is that the electromorphs can be direcdy equated with alleles of the corresponding structural gene and that electromorph profiles over the sample of different enzymes (frequently termed electrophoretic types or ETs) correspond to multilocus chromosomal genotypes [65]. The proteins assayed by this method are usually metabolic enzymes, such as those involved in glycolysis, which are expressed in all isolates of strains, and the allelic variation is unaffected by environmental conditions, including host, culture medium, or laboratory storage. Moreover, the allelic variation detected at these enzyme loci is selectively neutral, or nearly so, such that there is minimal convergence to the same allele through adaptive evolution [66-68]. Hence, this technique provides a rapid way to index the overall levels of genetic diversity at numerous loci throughout the chromo-

B

ISOLATE

Electrophoretic types (ETs) ET

ABCDEFGHI PGI

1

A

3

1

2

5

3

2

H

3

1

2

6

3

c D E F 6

2 2 2 2 2

1 1 1 1 1

3 3 3 3 3

2 2 2 2 2

1 1 1 1 1

4

B

1

1

4

4

2

5

I

2

1

1

1

1

3

G3P

AGO

IDE

PEP

PGI G3P ACO IDH PEP

1 2 3 4 5

Fig. 3 Multilocus enzyme electrophoresis (MLEE). (A) Supematants extracted from individual bacterial cultures (A-I) are extracted, electrophoresed through a support matrix, and selectively stained to detect the product of a specific enzyme locus (e.g., PGI). (B) Allelic variants (a.k.a. electromorphs) at each locus are numbered and cumulatively form the basis of the electrophoretic type (ET) of the isolate. (C) Relationships among isolates can be visualized as a dendrogram, constructed from a matrix of pairwise differences between the allelic profiles of the isolates. Reprinted with permission from and through the courtesy of Dr. Thomas Whittam.

12

HOWARD OCHMAN

some and to infer genetic relationships among strains. Because the number of alleles at a locus is fairly large in bacterial populations, we expect that recombination would, by chance, only rarely generate strains of identical multilocus chromosomal genotypes. Therefore, strains of the same ET are considered to be similar by descent and shared ancestry. Over the past two decades, MLEE has become established as the standard procedure for assessing genetic variation in bacterial populations and the one against which the discriminatory power of all other techniques is measured.

B.

DNA Sequencing

Although MLEE offers a rapid and inexpensive way to appraise the genetic variation in bacterial populations, many laboratories have turned to directly sequencing the genes specifying several of the enzymes originally indexed by MLEE or the genes encoding proteins involved in virulence. In contrast to MLEE, nucleotide sequencing offers a means of uncovering all of the allelic variation at a locus and detecting events of intragenic recombination. Moreover, nucleotide sequences are composed of discrete characters—i.e., the four bases—as opposed to MLEE, which can only provide the relative mobilities of electromorphs. This allows for the unambiguous identification of alleles, and for comparison and portability of data, from different studies and laboratories. Furthermore, the use of PCR to generate sequencing templates permits the analysis of noncultivable organisms. Although nucleotide sequencing provides the most complete information about the genetic variation and relationships among strains, it is still relatively cosdy and time consuming, especially for the analysis of variation at several loci in a large number of strains. And in many applications, the level of variation detected by MLEE has been sufficient to answer all but the subdest questions about the genetic diversity and structure of natural populadons.

C.

Multilocus Sequence l o p i n g

Maiden et al [69] have devised a method for the identification and typing of bacterial clones based on the determination of sequences of several gene fragments (Fig. 4). Multilocus sequence typing (MLST) exploits the advantages of nucleodde sequence data, but also constructs chromosomal genotypes, which can be used to detect intergenic recombination in the manner of MLEE, through an analysis of multiple chromosomally encoded loci [70]. In the initial application of MLST, the nucleodde sequences of PCR-amplified fragments from 11 housekeeping genes were obtained for more than 100 isolates oi Neisseria meningitis. For MLST, the gene fragments (i.e., alleles) are 400 to 500 nucleoddes in length—a convenient size for the direct automated sequencing of a DNA fragment

I.

13


Chromosomal DNA

Amplify -450-bp fragments of several (7-10) house-keeping genes

Sequence gene fragments on both strands

Compare sequences of each gene fragment with the known alleles at the locus

Assign alleles at the loci to give the allelic profile

Compare the allelic profile with those of isolates within a central database via the internet

Fig. 4 Multilocus sequence typing (MLST). The method for allocation of the allelic profile, or sequence type (ST), of a bacterial isolate is shown. As in MLEE, the relationships among isolates can be visualized as a dendrogram, constructed from a matrix of pairwise differences between the allelic profiles of the isolates. Reprinted with permission from Spratt (1999) [70].

with a single primer—and each unique combination of alleles over loci is referred to as a sequence type (ST). As the number of sequencing facilities increase, and the costs of DNA sequencing fall, MSLT is certain to become the method of choice for assessing variation in bacterial populations.

VIL The Orgonization of Genetic Diversity in Pattiogenic Microorganisms Early studies on microbial pathogens, particularly enteropathogenic E. coli [71, 72], suggested that very few clones, as identified by serotyping, were associated with disease outbreaks. Subsequent analysis using MLEE provided the first indication that the species E. coli as a whole was clonal, as evident from the repeated recovery of strains of the same chromosomal genotypes from different

14

HOWARD OCHMAN

times and geographic locations [60,73]. Three generalizations have emerged from the broad-scale application of MLEE to the study of common human pathogens [61, 63]. First, most species of bacteria are highly polymorphic for electrophoretically detectable alleles at each enzyme locus, such that a typical locus may have 10 to 20 electromorphs. Second, despite harboring large amounts of genetic variability, most bacterial species are clonal and consist of a relatively small number of genotypes. This implies that rates of recombination between genetically distinct clones, which would serve to generate new combinations of alleles over loci, must be very low. Finally, in most pathogenic species, only a very small proportion of clones promotes most of the disease worldwide (which indicates large differences in virulence among strains). Furthermore, the same clones that cause disease can be recovered over long periods of time.

VIIL Population Genetics of Representative Bacterial Pattiogens Having considered the connections among the molecular, genetic, and evolutionary perspectives on bacterial virulence, we can now turn our attention to the population genetic analysis of specific pathogens. Most of the information summarized below is based on data obtained through the application of MLEE; however, for several organisms, we can also integrate information on the relationships and diversity among strains, as achieved through the analysis of DNA sequences. The species are discussed in alphabetical order (with the exception of Shigella, which, based on its proper taxonomic position, is included within E. coli), and the key characteristics of their genetic variation and population structure are summarized in Table I.

A.

Bordetella

Bacteria belonging to the genus Bordetella are of primary importance in pediatric and veterinary medicine because of their ability to colonize the epithelium of the respiratory tract of a variety of vertebrate hosts, there causing bronchial and pulmonary pathology (see Chapter 13 herein, by Cotter and Miller). Based on phenotypic characteristics, the genus Bordetella nominally consists of four species: B. pertussis, an obligate human pathogen causing whooping cough, B. parapertussis, which has been isolated from humans and sheep, B. bronchiseptica, which is the etiologic agent of canine kennel cough, and B. avium, which causes respiratory disease in fowl. These four species have been further subdivided on the basis of serology and biotyping; however, evidence from MLEE indicates that

in

O

C

6^ •is

o

S -B

^

•6 £

d d

O

OO lO

'^ O m m o r^

( ^ Os 'd^ fS lO Tt Tt ON

o d d

m v^ \D

_

.

. _

r-oor-o

?3 a

-C2 ^

;^

CJ-"^

CO ^

(N Q

—' —« Tj-

0 o -^ c -^ 01 (N 95%) of sequence variation within this species is attributable to nonsynonymous substitutions or other mutations in chromosomal loci that confer resistance to antibiotics. Despite very low levels of neutral sequence variation—in fact, the M. tuberculosis species has the lowest level of nucleotide diversity of any bacterial pathogen—strains could be assigned to three genotypic groups based on combinations of polymorphisms at two sites: codon

22

HOWARD OCHMAN

463 in katG, and codon 95 in gyrA. (These two sites are not involved in antibiotic resistance and, hence, serve as useful genetic markers to examine the evolutionary history and relationships of strains.) All isolates of the predominandy nonhuman pathogens, M microti and M. bovis, and the human pathogen, M. africanum, have the same combination of polymorphisms characteristic of M. tuberculosis genotypic group 1, suggesdng that this genotypic group is ancestral to groups 2 and 3. Because the most ancient group is expected to contain the most variadon, this view of the relationships among the genotypic groups is further supported by the fact the genotypic group 1 contains the most variation in IS6100 copy numbers and in the nucleodde sequences at other loci [158]. Despite the low rate amount of nucleotide sequence diversity, the three genotypic groups of M. tuberculosis have diverged with respect to IS6110 copy numbers. \S6110 profiles can change reladvely rapidly, and in many cases isolates resampled after 90 days from the same patients displayed changes in IS6100 genotype, particularly among strains with greater numbers of these IS elements. And due to reladvely minor changes in the \S6110 fingerprint patterns of strains from single individuals, this variation could not have been produced by reinfecdon by a different strain [159].

H.

Neisseria

Due to the widespread and recurrent epidemics of meningococcal disease, there has been extensive work on the population biology of Neisseria meningitidis (see [160-163] for comprehensive reviews and Chapter 12 herein, by Meyer et al). In general, the variation among strains of Neisseria is generated through horizontal exchange, but epidemics are often caused by the spread of specific genedc variants, which results in clonal replacement. After their descent for a common ancestor, strains rapidly diversify through mutation and recombination [162]. A^. meningitidis are conventionally typed on the basis of capsular polysaccharides, and serogroups A, B, and C account for more than 90% of the cases of meningococcal disease worldwide ([164]). In an early study, it was demonstrated that the European epidemic of serogroup B disease that began in the 1970s was caused by a group of 22 very closely related clones, designated as the ET-5 complex, that have no close genedc relationship to other groups of clones [165]. Clones of the ET-5 complex were also the causative agents of later outbreaks in Africa, South America, Cuba, and the United States (where it was likely to have been introduced by Cuban immigrants). N. meningitidis is carried asymptomadcally in the upper respiratory tract by about 15% of the human population, and the clones isolated from carriers are only rarely represented among those causing disease, showing that certain complexes of clones have a low virulence potendal [166].

1.


23

Unlike most other serogroups of N. meningitidis, which are generally associated with endemic disease, isolates of serogroup A are unusual in that they may cause very large epidemics. Among the serogroup A organisms responsible for 23 epidemics between 1915 and 1983, there were 34 distinctive ETs constituting four clone complexes, each representing a group of related clones [167]. Most epidemics were caused by a single clone, and the same clone was responsible for concurrent epidemics in different countries. Clonal analysis has also demonstrated that serogroup A isolates are a restricted phylogenetic subpopulation of the species [168], which probably arose no more than a few hundred years ago [169]. Although recombination produces genetic variation within clone complexes, these results would indicate that only limited amounts of genetic exchange occur between phylogenetically unrelated strains of A^. meningitidis. Sequence analysis of meningococcal genes provides clear evidence that the evolution of N. meningitidis has been characterized by high levels of intra- and intergenic recombination, which is perhaps not surprising for a naturally transformable species. For example, allelic variants of the gene encoding adenylate kinase (adk), whose variation is regularly assayed by MLEE, have a mosaic structure produced by recombination between genes from different strains [170]. Moreover, within and among species of Neisseria, patterns of sequence divergence for adk, recA, aroE (shikimate dehydrogenase), and glnA (glutamine synthetase) are very different [171, 172], and the phylogenetic trees based on each of these genes are not congruent, as expected for species that undergo frequent recombination. Note that even in recombining species, such as A^. meningitidis, the epidemic increase of certain ETs can give the appearance of a clonal population structure despite high levels of gene exchange [130, 173].

I.

Salmonella

Under the original serotyping schemes of White [174] and Kauffman [175], the Salmonellae were assigned to nearly 3000 serotypes or serovars, with each considered a distinct species. However, based on the biotyping and molecular genetic evidence, all strains were subsequently typed as single species, S. enterica, which comprises eight subspecific groups (designated I, II, Ilia, Illb, IV, V, VI, and VII) [176, 177]. Hence, the nomenclature has changed such that Salmonella typhimurium would now be referred to as Salmonella enterica serovar Typhimurium or, simply, Typhimurium. Over 60% of the serotypes belong to subspecies I, including those strains causing >99% of the cases of human salmonellosis (see Chapter 7 herein, by Scherer and Miller). Due to its genetic relationships to the other subspecific groups, subspecies V has recently been reclassified as a separate species. Salmonella bongori [54], and MLEE as well as nucleotide sequence analysis of several genes have confirmed its divergent phylogenetic position [178, 179, 63, 180-182].

24

HOWARD OCHMAN

MLEE was originally used to investigate the allelic variation in genes in large samples of Salmonella [183-189]. The total genetic diversity in Salmonella, as assessed by MLEE, is among the highest reported for any bacterial species (Table I)—nearly twice that observed in E. coli. Clonal aspects of the genetic structure of S. enterica are well illustrated by the serovars Typhi and Paratyphi A, B, and C, all of which are agents of enteric fever in humans. MLEE analysis has demonstrated that there are no close relationships among these serovars, implying independent evolutionary derivation [187, 188]. Typhi is an unusually distinctive and homogenous serovar, and over 80% of the worldwide isolates are of a single ET (with a second ET comprising 16% of strains, all from West Africa). Paratyphi B consists of a large and heterogenous group of lineages that are closely related to Typhimurium. However, the ability of Paratyphi B to cause human enteric fever arose in a single globally distributed clone, and only recendy, since it is only weakly differentiated. The genetic variation and relationships among strains have also been assessed by the nucleotide sequencing of several housekeeping genes, including proline permease (putP) [178], glyceraldehyde-3-phosphate dehydrogenase (gapA) [190], malate dehydrogenase (mdh) [180], 6-phosphogluconate dehydrogenase (gnd) [179], isocitrate dehydrogenase (icd) [191], and isocitrate dehydrogenase kinase/phosphatase (aceK) [192]. For these five housekeeping genes, on average, about 16% of nucleotides and 5% of amino acids are polymorphic. With the exception of gnd, the level of sequence diversity is greater in S. enterica than in E. coli, which is attributable to the unusually high divergence of subspecies V {S. bongori) from the other subspecies. Comparisons of the individual trees based on the nucleotide sequences have revealed several cases in which the branching orders of lineages are not congruent. These disparities are due both to intragenic recombination events, which can involve regions ranging from six basepairs to more than 1 kb, and to the exchange of entire genes [63]. Notwithstanding low levels of recombination at some loci, the relationships among strains based on nucleotide sequences matched those established by DNA hybridization and MLEE. Based on these phylogenetic relationships, serovars that are exclusively or predominantly diphasic (subspecies I, II, Illb, and VI) cluster apart from the monophasic subspecies (Ilia, IV, V, and VII). This suggests that, following the divergence of S. enterica and E. coli from a common ancestor, E. coli evolved as an commensal of mammals while Salmonella remained associated with reptiles, which are still the primary hosts of the monophasic subspecies. Salmonella serovars are typically classified as either monophasic or diphasic based on their ability to produce one or two forms of flagellin. Subsequently, Salmonella evolved as an intracellular pathogen through the acquisition of several pathogenicity islands, which conferred the ability to invade host epithelial cells and circumvent host defenses. The diphasic condition originated in the lineage ancestral to subspecies I, II, Illb, and VI is a mechanism to further evade the host

1.


25

immune system and is likely to have assisted in the exploitation of birds and mammals of potential hosts [181]. This scenario is supported by the phylogenetic distribution of pathogenicity islands in Salmonella [193]. The SPI-1 island, which confers the ability to invade nonphagocytic host cells, was acquired very early in the evolution of Salmonella and is present in all subspecies, whereas the SPI-2 island, which is necessary for intracellular proliferation, is absent from Salmonella bongori, which were originally recovered from nonmammalian hosts. This suggests that the evolution of Salmonella as a pathogen has been marked by the acquisition and/or generation of several genes that facilitate interactions with the host [9, 194].

J.

Staphylococcus

The are two notable cases where MLEE has been applied to uncover the evolutionary history of infective strains of Staphylococcus aureus. The first involves strains of S. aureus causing toxic shock syndrome (TSS) in young, healthy menstruating women. Almost all strains of S. aureus recovered from TSS patients express a toxin (designated TSST-1) [195, 196], which is now known to be encoded as part of a 15-kb pathogenicity island present only in TSST-1 -positive strains [17]. The analysis of genetic variation in 315 isolates of S. aureus expressing TSST-1 revealed that toxin production occurs in association with chromosomal backgrounds representing the full breadth of genotypic diversity in the species as a whole, as might be expected for genes encoded on a mobile element [197]. But despite the diversity among strains expressing TSST-1, a single distinctive clone causes the majority of cases of toxic shock syndrome. It is not known whether the present-day distribution of the TSST-1 gene in S. aureus reflects an evolutionarily old association or the independent acquisition of the TSST-1 pathogenicity island by multiple strains. However, these results suggest that the particular clone causing TSS has properties conferring strong affinity for human cervicovaginal surfaces [197]. A second case where MLEE has provided insights into the nature of staphylococcal infections involved strains of S. aureus resistant to the antimicrobial agent methicillin. Soon after methicillin entered clinical use, there were several hospital outbreaks caused by methicillin-resistant S. aureus (MRSA), and these organisms have now achieved global distribution [198]. Methicillin resistance is conferred by the expression of a modified penicillin-binding protein encoded by the mec gene. MLEE was employed to determine the extent of genetic diversity of MRSA strains and the relationships among strains from temporally and geographically separated outbreaks [199, 200]. The mec gene is harbored by many divergent lineages of S. aureus and has spread through multiple episodes of horizontal transfer. Many of the MRSAs recovered from Europe and Africa soon after methicillin was introduced into clinical use in the 1960s were identified as the

26

HOWARD OCHMAN

same ET, suggesting that most early outbreaks were caused by dissemination of a single clone that had acquired the mec gene. However, the association of mec with genetically divergent strains contradicts the idea that all of the extant MRS As descended from a single methicillin-resistant clone [201].

K.

Streptococcus

A study on the diversity and population structure of Streptococcus pyogenes (Group A Streptococcus) was originally undertaken to determine the relationships among strains causing toxic-shock-like syndrome and other invasive diseases [202] (see Chapter 15 herein, by Caparon). Of the 108 strains analyzed by MLEE, there were 33 ETs; but nearly half of disease episodes (including more than 70% of the cases of toxic-shock-like syndrome) were caused by two related clones, designated ETl and ET2. The genetic structure of Streptococcus pyogenes has also been characterized with respect to the nucleotide sequence variation in several genes, including those encoding the scarlet fever exotoxin {speA) [202, 203], serotype M protein (emm) [204, 205] and complement inhibitor (sic) [206], streptokinase (ska) [205], pyrogenic enterotoxin B (speB) [205], C5a peptidase (scp) [205], superantigen SSA(ssa) [207], and hyaluronidase (hyl) [208]. In many cases, horizontal transfer and genetic exchange have contributed to the allelic diversity at these loci, and the phylogenies of strains produced from sequences of individual genes often do not match those based on MLEE. The level of polymorphism in the sic gene in M1 strains of Streptococcus greatly exceeds that observed at any other locus in this species. Moreover, virtually all nucleotide substitutions alter the amino acid sequence of the encoded protein, and all insertions and deletion are in frame [206]. Because the Sic protein functions to inhibit complement, this variability has been ascribed to selection acting to adapt this protein to host factors. Although it has long been known that certain genes of Streptococcus pneumoniae are subject to horizontal transfer and recombinational exchange [30, 209, 211], the population structure and genetic variation of this species has only recendy been assessed. Using MLST, the sequence diversity of strains of S. pneumoniae associated with disease was assessed by examining ~450-bp portions of seven housekeeping genes in nearly 300 isolates [212]. Despite relatively low levels of sequence variation, the numbers of alleles per locus ranged from 18 in aroE to 34 in xpt. Among the 143 STs, 34 contained more than one isolate and 12 included at least five invasive isolates. In 26 of the 34 STs containing multiple isolates, there was a perfect congruence between ST and serotype; however, strains of the same serotype could differ by as many as six of seven loci [212]. Moreover, there was evidence of recombination among loci, and the repeated recovery of identical isolates causing invasive disease in geographically distant regions suggests that certain STs define strains with increased capacity to cause disease.

1.

L.


27

Vibrio

Vibrio cholerae is a natural inhabitant of aquatic environments; however, some clones can cause severe diarrheal disease in humans (see Chapter 10 herein, by DiRita). Seven pandemics of cholera have been recorded since 1817, with the sixth pandemic subsiding in 1925 and the seventh beginning in Indonesia in 1961. Although V. cholerae, the causative agent of cholera, has been subdivided into nearly 200 serovars, the number of pathogenic serotypes is small, and all recorded pandemic and epidemic cases have been associated with the 01 serotype on a worldwide scale. Non-01 strains are more frequently isolated from environmental sources but have been implicated in cases of gastroenteritis, septicemia, and meningitis in humans. An outbreak of cholera in India that began in 1991 was caused by an 0139 strain. Toxigenic isolates of V. cholerae are also separated into Classical and El Tor types, which are differentiated on the basis of polymixin sensitivity, hemolysis and hemagglutination activity, phage resistance patterns, and the Voges-Proskauer reaction. The first six pandemics were thought to have originated from Classical strains, and the seventh from an El Tor strain. Based on MLEE, the seventh-pandemic clone is relatively homogenous worldwide. A large group of strains isolated over 30 years from several countries affected by the seventh pandemic constitute a single ET (ET3) [213-215], with isolates from Australia (ETl), the U.S. Gulf Coast (ET2), and Latin America (ET4) each representing a unique type. Among more recently isolated samples of V. cholerae from Latin America, ET4 was still prevalent, but about 10% of strains were of ET3, showing that the recent outbreaks of cholera in this region have originated from introduction of new toxigenic 01 strains. All seventh-pandemic clones have the same sequences for the asd (aspartatesemialdehyde dehydrogenase) gene and the ctxB (toxin) gene [216, 217], suggesting a single origin of this clone. The toxigenic 0139 strain is genetically closely related to the seventh-pandemic strains and presumably evolved from an early seventh-pandemic isolate [218]. But based on ribotyping, it was concluded that several clones, rather than one, caused the seventh pandemic, with one group of representing clones arising in 1961 and found only in Asia, and the other arising in 1966 and spreading worldwide [219]. However, the frequency of genetic exchange between rrn operons is high in seventh-pandemic clones—nine new ribotypes have been detected among 47 isolates sampled over a 33-year period [220]—and this degree of recombination could obscure the relationships among clones that are identified by this method. The original sequencing of asd in isolates of V. cholerae detected three homogeneous, but distantly related, clones representing the sixth-pandemic, seventh-pandemic, and U.S. Gulf Coast clones [218]. The subsequent sequencing of other loci [221] revealed virtually no variation in either the mdh (malate dehydrogenase) and hlyA (hemolysin) genes among sixth-pandemic, seventh-pandemic. Gulf Coast, and 0139 isolates, and these sequences were distinct from

28

HOWARD OCHMAN

environmental, nontoxigenic non-Ol strains. The frequent occurrence of identical gene sequences among pathogenic isolates suggests that these clones are indeed closely related, and that the variation observed at other loci, and in the nontoxigenic isolates, has been generated, in part, by recombination.

IX. Conclusions Infectious diseases caused by microbial pathogens—ranging from plague to tuberculosis—are among the world's leading causes of death. And while investigations into the molecular bases of virulence are underway for numerous bacterial pathogens, all but the most rudimentary information pertaining to the extent and organization of genetic variation, the factors contributing to allelic diversity, and the genetic structure of their populations is available for the majority of these microorganisms. Aside from identifying the forces shaping the extent and organization of genetic diversity in those microorganisms, these findings are also applicable to the choice of procedures used for identification, epidemiology, and control of bacterial pathogens. The identification and classification of a pathogenic species could potentially influence analysis of its genetic diversity and population structure (e.g., if serovars of Salmonella were each considered a separate species, as originally determined on the basis of serotyping, the amount of variation in each would be very different than that presently detected in populations of Salmonella enterica). In addition, the wide variety of mechanisms that generate differences in pathogenic potential among related bacteria will affect the origin of virulence and, hence, the apportionment of genetic diversity in natural populations. Despite these factors, several generalizations have emerged from the population genetic analysis of bacteria. Foremost is that most species of bacteria are highly variable but clonal and consist of relatively small numbers of genotypes. Moreover, only a small proportion of clones cause most of the disease worldwide, implying that there are large differences in virulence among strains. Among the most productive approaches for future research will be the analysis of variation in virulence-associated loci within populations of pathogens since these genes are likely to respond to selective pressures presented by the host. Not only do such studies provide information about the genetic structure of the populations and the factors that contribute to genetic diversity, but also they can identify polymorphisms associated with particular clinical symptoms or disease pathologies. Such analyses will be essential for understanding the endemic and epidemic spread of pathogenic organisms, and for the development of vaccines to abate their effects on human populations.

I.


29

References 1. Groisman, E. A., and Ochman, H. (1996). Pathogenicity islands: Bacterial evolution in quantum leaps. Ce//87, 791-794. 2. Lee, C. A. (1996). Pathogenicity islands and the evolution of bacterial pathogens. Infect. Agents Dis. 5, 1-7. 3. Hacker, J., Blum-Oehler, G., Muhldorfer. 1., and Tschape, H. (1997). Pathogenicity islands of virulent bacteria: Structure, function and impact on microbial evolution. Mol. Microbiol. 23, 1089-1097. 4. Blum, G., Ott, M., Lischewski, A.. Ritter. A.. Imrich. H., Tschape, H., and Hacker, J. (1994). Excision of large DNA regions termed pathogenicity islands from tRNA-specific loci in the chromosome of an Escherichia coli wild-type pathogen. Infect. Immiin. 62, 606-614. 5. McDaniel, T K., Jarvis, K. G.. Donnenberg, M. S.. and Kaper. J. B. (1995). A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens. Proc. Natl. Acad. Sci. U.S.A. 28, 1664-1668. 6. Guyer, D. M., Kao, J.-S., and Mobley. L. T (1998). Genomic analysis of a pathogenicity island in uropathogenic Escherichia coli CFT073: Distribution of homologous sequences among isolates from patients with pyelonephritis, cystitis, and catheter associated bacteriuria and from fecal samples. Infect. Immiin. 66. 4411—4417. 7. Rajakamur, K. (1997). Use of a novel approach, termed island probing, identifies the Shigella flexneri she pathogenicity island which encodes a homolog of the immunoglobin A protease-like family of proteins. Infect. Inmmn. 65. 4606-4614. 8. Yokes, S. A., Reeves, S. A., ToiTes, A. G., and Payne. S. M. (1999). The aerobactin iron transport system genes in Shigella flexneri are present within a pathogenicity island. Mol. Microbiol. 33, 63-73. 9. Groisman, E. A., and Ochman, H. (1997). How Salmonella became a pathogen. Trends Microbiol. 5, 343-349. 10. Buchreiser, C., Brosch, R., Bach, S., Guiyoule. A., and Carniel. E. (1998). The high-pathogenicity island of Yersinia pseudotuberculosis can be inserted into any of the three chromosomal asn tRNA genes. Mol. Microbiol. 30, 965-978. 11. Buchreiser, C., Prentice, M., and Carniel. E. (1998). The 102-kilobase unstable region of F^rvm/c/ pestis comprises a high-pathogenicity island linked to a pigmentation segment which undergoes internal rearrangement. J. Bacteriol. 180. 2321-2329. 12. Schubert, S., Rakin, A., Karch, H.. Carniel E.. and Heesemann, J. (1998). Prevalence of the "high-pathogenicity island" of Yersinia species among Escherichia coli strains that are pathogenic to humans. Infect. Immun. 66. 480-485. 13. Karaolis, D. K., Johnson, J. A., Bailey, C. C. Boedeker, E. C. Kaper, J. B., and Reeves, P R. (1998). A Vibrio cholerae pathogenicity island associated with epidemic and pandemic strains. Proc. Natl. Acad. Sci. U.S.A. 95, 3134-3139. 14. Martin, K., Morlin, G., Smith, A.. Nordyke. A.. Eisenstark, A., and Golomb, M. (1998). The tryptophanase gene cluster of Haemophilus influenzae type b: Evidence for horizontal gene transfer. /. Bacteriol. 180, 107-118. 15. Mhlanga-Mutangadura, T, Morlin, G.. Smith. A. L., Eisenstark, A., and Golomb, M. (1998). Evolution of the major pilus gene cluster of Haemophilus influenzae. J. Bacteriol. 180, 4693-4703. 16. Censini, S., Lange, C, Xiang, Z., Crabtree. J. E., Ghiara, P.. Borodovsky, M., Rappuoli, R., and Covacci, A. (1996). cag, a pathogenicity island of Helicobacter pylori, encodes type 1-specific and disease associated virulence factors. Proc. Natl. Acad. Sci. U.S.A. 93, 14648-14653.

30

HOWARD OCHMAN

17. Lindsay, J. A., Ruzin, A., Ross, H. F., Kurepina, N., and Novick, R. P. (1998). The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus. Mol. Microbiol. 29, 527-543. 18. Hacker, J., Bender, L., Ott, M., Wingender, J., Lund, B., Marre, R., and Goebel, W. (1990). Deletions of chromosomal regions coding for fimbriae and hemolysins occur in vitro or in vivo in various extraintestinal Escherichia coli isolates. Microb. Pathogen. 8, 213-225. 19. Campbell, A. M. (1992). Chromosomal insertion sites for phages and plasmids. J. Bacteriol. 174, 7495-7499. 20. Inouye, S., Sunshine, M. G., Six, E. W., and Inouye, M. (1991). Retronphage phi R73: An E. coli phage that contains a retroelement and integrates into a tRNA gene. Science 252, 969-971. 21. Pierson III, L. S., and Kahn, M. L. (1987). Integration of satellite bacteriophage P4 in Escherichia coli. DNA sequences of the phage and host regions involved in site-specific recombination. / Mol. Biol. 196, 487-496. 22. Cheetham, B. F., and Katz, M. E. (1995). A role for bacteriophages in the evolution and transfer of bacterial virulence determinants. Mol. Microbiol. 18, 201-208. 23. McDaniel, T. K., and Kaper, J. B. (1997). A cloned pathogenicity island from enteropathogenic Escherichia coli confers the attaching and effacing phenotype on E. coli K-12. Mol. Microbiol. 23, 399-407. 24. Moss, J. E., Cardozo, T. J., Zychlinsky, A., and Groisman, E. A. (1999). The ^^/C-associated SHI-2 pathogenicity island of Shigella flexneri. Mol. Microbiol. 33, 74-83. 25. Blanc-Potard, A. B., and Groisman, E. A. (1997). The Salmonella selC locus contains a pathogenicity island mediating intramacrophage survival. EMBO. J. 16, 5376-5385. 26. Maurelli, A. T, Baudry, B., d'Hauteville, H., Hale, T. L., and Sansonetti, P J. (1985). Cloning of plasmid DNA sequences involved in invasion of HeLa cells by Shigella flexneri. Infect. Immun. 49, 164-171. 27. Sasakawa, C , Kamata, K., Sakai, T, Makino, S., Yamada, M., Okada, N., and Yoshikawa, M. (1988). Virulence-associated genetic regions comprising 31 kilobases of the 230-kilobase plasmid in Shigella flexneri 2a. J. Bacteriol. 170, 2480-2484. 28. Gemski, P., Lazere, J. R., Casey, T, and Wohlhieter, J. A. (1980). Presence of virulence-associated plasmid in Yersinia pseudotuberculosis. Infect. Immun. 28, 1044-1047. 29. Portnoy, D. A., Moseley, S. L., and Falkow, S. (1981). Characterization of plasmids and plasmid-associated determinants of Yersinia enterocolitica pathogenesis. Infect. Immun. 31, 775-782. 30. Jackson, M. P, Neill, R. J., O'Brien, A. D., Holmes, R. K., and Newland, J. W. (1987). Nucleotide sequence analysis and comparison of the structural genes for Shiga-like toxin-I and Shiga-like toxin II encoded by bacteriophages from Escherichia coli. FEMS Microbiol. Lett. 44, 109-114. 31. Waldor, M. K., and Mekalanos, J. J. (1996). Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272, 1910-1914. 32. Karaolis, D. K., Somara, S., Maneval Jr., D. R., Johnson, J. A., and Kaper, J. B. (1999). A bacteriophage encoding a pathogenicity island, a type-IV pilus and a phage receptor in cholera bacteria. Nature 399, 375-379. 33. Nakata, N., Tobe, T, Fukuda, I., Suzuki, T, Komatsu, K., Yoshikawa, M., and Sasakawa, C. (1993). The absence of a surface protease, OmpT, determines the intercellular spreading ability of Shigella: The relationship between the ompT and kcpA loci. Mol. Microbiol. 9, 459^68. 34. Maurelli, A. T., Fernandez, R. E., Bloch, C. A., Rode, C. K., and Fasano, A. (1998). "Black holes" and bacterial pathogenicity: A large genomic deletion that enhances the virulence of Shigella spp., and enteroinvasive Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 95, 3943-3948.

1.


31

35. Sokurenko, E. V., Chesnokova, V., Dykhuizen, D. E., Ofek. I., Wu. X. R., Krogfelt, K. A., Struve, C , Schembri, M. A., and Hasty, D. L. (1998). Pathogenic adaptation of Escherichia coli by natural variation of FimH adhesin. Pwc Natl. Acad. Sci. U.S.A. 95, 8922-8926. 36. Sokurenko, E. V., Hasty, D. L., and Dykhuizen, D. E. (1999). Pathoadaptive mutations: Gene loss and variation in bacterial pathogens. Trends Microbiol. 7, 191-195. 37. Lee, C. A., Jones, B. D., and Falkow, S. (1992). Identification of a Salmonella typhimurium invasion by selection for hyperinvasive mutants. Proc. Natl. Acad. Sci. U.S.A. 89, 1847-1851. 38. Hromockyj, A. E., and Maurelli, A. T. (1989). Identification of Shigella invasion genes by isolation of temperature-regulated inv::lacZ operon fusions. Infect. Immun. 57, 2963-2970. 39. Lawrence, J. G., and Ochman, H. (1998). Molecular archaeology of the Escherichia coli genome. Proc. Natl. Acad. Sci. U.S.A. 95, 9413-9417. 40. Lawrence, J. G. (1997). Selfish operons and speciation by gene transfer. Trends Microbiol. 5, 355-359. 41. Groisman, E. A., and Ochman, H. (1994). How to become a pathogen. Trends Microbiol. 2, 289-294. 42. Fields, R I., Swanson. R. V., Haidaris, C. G.. and Heffron, R (1986). Mutants of Salmonella typhimurium that cannot survive within the macrophage are avirulent. Proc. Natl. Acad. Sci. U.S.A. 83,5189-5193. 43. Isberg, R. R., and Falkow, S. (1985). A single genetic locus encoded by Yersinia pseudotuberculosis permits invasion of cultured animal cells by Escherichia coli K-12. Nature 317, 262-264. 44. Darasse, A., Kotoujansky, A., and Berthau, Y. (1994). Isolation by genomic subtraction of DNA probes specific for Erwinia carotovora subsp. atroseptica. Appl. Environ. Microbiol. 60, 298-306. 45. Wu, F. M., and Muriana, P. M. (1995). Genomic subtraction in combination with PCR for enrichment of L/.v/^nV/ monocytogenes-î^Q^c\f'\c sequences. Int. J. Food Microbiol. 11, 161-174. 46. Hughes, M. S., Beck, L. A., Skuce, R. A., and Neill S. D. (1997). Development of mycobacterial species-specific DNA probes by subtraction hybridization. EEMS Microbiol. Lett. 156, 31-36. 47. Kuflu, K. M., and Cuppels, D. A. (1997). Development of a diagnostic DNA probe for xanthomonads causing bacterial spot of peppers and tomatoes. Appl. Environ. Microbiol. 63, 4462^470. 48. Calia, K. E., Waldor, M. K., and Calderwood. S. B. (1998). Use of representational difference analysis to identify genomic differences between pathogenic strains of Vibrio cholerae. Infect. Immun. 66, 849-852. 49. Townsend, K. M., Frost, A. J., Lee, C. W., Papadimitriou, J. M., and Dawkins, H. J. (1998). Development of PCR assays for species- and type-specific identification of Pasteurella multocida isolates. / Clin. Microbiol. 36. 1096-1100. 50. Straus, D., and Ausubel, F. M. (1990). Genomic subtraction for cloning DNA corresponding to deletion mutations. Proc. Natl. Acad. Sci. U.S.A. 87, 1889-1893. 51. Brown, P. K., and Curtiss, R. (1996). Unique chromosomal regions associated with virulence of an avian pathogenic Escherichia coli strain. Proc. Natl. Acad. Sci. U.S.A. 93, 11149-11154. 52. Akopyants, N. S., Fradkov, A., Diatchenko, L., Hill J. E.. Siebert, P D., Lukyanov, S. A., Sverdlov, E. D., and Berg, D. E. (1998). PCR-based subtractive hybridization and differences in gene content among strains of Helicobacter pylori. Proc. Nad. Acad. Sci. U.S.A. 95, 13108-13113. 53. Le Minor, L., and M. Y. Popoff. (1987). Designation of Salmonella enterica sp. Nov., nom. Rev., as the type and only species of the genus Salmonella. Int. J. Syst. Bacteriol. 37, 465-468.

32

HOWARD OCHMAN

54. Reeves, M. W., Evins, G. M., Heiba, A. A., Plikaytis, B. D., and Farmer III, J. J. (1989). Clonal nature of Salmonella typhi and its genetic relatedness to other salmonellae as shown by multilocus enzyme electrophoresis, and proposal of Salmonella bongori comb. nov. J. Clin. Microbiol 27,313-320. 55. Ochman, H., Whittam, T. S., Caugant, D. A., and Selander, R. K. (1983). Enzyme polymorphism and genetic population structure in Escherichia coli and Shigella. J. Gen. Microbiol. 129, 2715-2726. 56. Pupo, G. M., Karaolis, D. K., Lan, R., and Reeves, R R. (1997). Evolutionary relationships among pathogenic and nonpathogenic Escherichia coli strains inferred from multilocus enzyme electrophoresis and mdh sequence studies. Infect. Immun. 65, 2685-2692. 57. Wang, R. F., Cao, W. W., and Cemiglia, C. E. (1997). Phylogenetic analysis and identification of Shigella spp. Mol. Cell. Probes 11, 427-432. 58. Rolland, K., Lambert-Zechovsky, N., Picard, B., and Denamur, E. (1998). Shigella and enteroinvasive Escherichia coli strains are derived from distinct ancestral strains of E. coli. Microbiology 144, 2667-2672. 59. Milkman, R. (1973). Electrophoretic variation in Escherichia coli from natural sources. Science 182, 1024-1026. 60. Selander, R. K., and Levin, B. R. (1980). Genetic diversity and structure in Escherichia coli populations. Science 210, 545-547. 61. Selander, R.K., and Musser, J. M. (1990). Population genetics of bacterial pathogenesis. In "Molecular Basis of Bacterial Pathogenesis," Vol. 11 (B. H. Iglewski and V. L. Clark, eds.), pp. 11-36. Academic Press, San Diego. 62. Musser, J. M. (1996). Molecular population genetic analysis of emerged bacterial pathogens: Selected insights. Emerg. Infect. Dis. 2, 1-17. 63. Selander, R. K., Li, J., Boyd, E. R, Wang, F-S., and Nelson, K. (1994). DNA sequence analysis of the genetic structure of populations of Salmonella enterica and Escherichia coli. In "Bacterial Diversity and Systematics" (F. G. Priest, A. Ramos-Cormenzana, and R. Tindall, eds.), pp. 17^9. Plenum, New York. 64. Selander, R. K., Caugant, D. A., Ochman, H., Musser, J. M., Gilmour, M. N., and Whittam, T. S. (1986). Methods of multilocus enzyme electrophoresis for bacterial population genetics and systematics. Appl. Environ. Microbiol. 51, 873-884. 65. Lewontin, R. C. (1974). "The Genetic Basis of Evolutionary Change." Columbia University Press, New York. 66. Kimura, M. (1983). "The Neutral Theory of Molecular Evolution." Cambridge University Press, London. 67. Whittam, T. S., Ochman, H., and Selander, R. K. (1983). Geographic components of linkage disequilibrium in natural populations of Escherichia coli. Mol. Biol. Evol. 1, 67-83. 68. Dykhuizen, D. E., de Framond, J., and Hartl, D. L. (1984). Selective neutrality of glucose-6phosphate dehydrogenase allozymes in Escherichia coli. Mol. Biol. Evol. 1, 162-170. 69. Maiden, M. C , Bygraves, J. A., Fell, E., Morelli, G., Russell, J. E., Urwin, R., Zhang, Q., Zurth, K., Caugant, D. A., Feavers, I. M., Achtman, M., and Spratt, B. G. (1998). Multilocus sequence typing: A portable approach to the identification of clones within populations of pathogenic microorganisms. Proc. Natl. Acad. Sci. U.S.A. 95, 3140-3145. 70. Spratt, B. G. (1999). Multilocus sequence typing: Molecular typing of bacterial pathogens in an era of rapid DNA sequencing and the Internet. Curr Opin. Microbiol. 2, 312-316. 71. 0rskov, F, 0rskov, I., Evans Jr., D. J., Sack, R. B., Sack, D. A., and Wadstrom, T. (1976). Special Escherichia coli serotypes among enterotoxigenic strains from diarrhoea in adults and children. Med. Microbiol. Immunol. (Berlin) 162, 73-80.

I.


33

72. 0rskov, L, 0rskov, K, Jann, B., and Jann, K. (1977). Serology, chemistry, and genetics of O and K antigens of Escherichia coli. Bacterial. Rev. 41, 667-710 73. Achtman, M., and Pluschke, G. (1986). Clonal analysis of descent and virulence among selected Escherichia coli. Annu. Rev. Microbiol. 40, 185-210. 74. Musser, J. M., Barenkamp, S. J., Granoff. D. M., and Selander, R. K. (1986). Genetic relationships of serologically nontypable and serotype b strains of Haemophilus influenzae. Infect. Immun. 52, 183-191. 75. van der Zee, A., Mooi, F, Van Embden, J., and Musser, J. (1997). Molecular evolution and host adaptation of Bordetella spp.: Phylogenetic analysis using multilocus enzyme electrophoresis and typing with three insertion sequences. J. Bacterial. 179, 6609-6617. 76. Steere, A. C., Grodzicki, R. L.. Kornblatt, A. N., Craft. J. E., Barbour, A. G., Burgdorfer, W., Schmid, G. P., Johnson, E., and Malawista. S. E. (1983). The spirochetal etiology of Lyme disease. New Engl. J. Med. 308, 733-740. 77. Boerlin, R, Peter, O., Bretz, A. G.. Postic. D.. Baranton. G., and Piffaretti, J. C. (1992). Population genetic analysis of Borrelia hurgdaiferi isolates by multilocus enzyme electrophoresis. Infect. Immun. 60, 1677-1683. 78. Balmelli, T, and Piffaretti, J. C. (1996). Analysis of the genetic polymorphism of Borrelia burgdorferi sensu lata by multilocus enzyme electrophoresis. Int. J. Syst. Bacterial. 46, 167-172. 79. Valsangiacomo, C, Balmelli, T, and Piffaretti. J. C. (1997). A phylogenetic analysis of Borrelia burgdorferi sensu lata based on sequence information from the hbb gene, coding for a histone-like protein. Int. J. Syst. Bacterial. 47. 1-10. 80. Dykhuizen, D. E., Polin, D. S., Dunn, J. J., Wilske. B.. Preac-Mursic, V., Dattwyler, R. J., and Luft, B. J. (1993). Borrelia burgdorferi is clonal: Implications for taxonomy and vaccine development. Prac. Natl. Acad. Sci. U.S.A. 90. 10163-10167. 81. Wang, I. N., Dykhuizen, D. E., Qiu, W., Dunn. J. J.. Bosler. E. M.. and Luft, B. J. (1999). Genetic diversity of aspC in a local population of Borrelia burgdaiferi sensu stricto. Genetics 151, 15-30. 82. Whittam, T. S. (1996). Genetic variation and evolutionary processes in natural populations of Escherichia coli. In ''Escherichia call and Salmonella^ Vol. 2 (F. C. Neidhardt, ed.), pp. 2708-2720. ASM, Washington, DC. 83. Selander, R. K., Musser, J. M., Caugant, D. A.. Gilmour. M. N., and Whittam, T S. (1987). Population genetics of pathogenic bacteria. Micrab. Pathogen. 3, 1-7. 84. Bettelheim, K. A. (1978). The source of "OH"" serotypes of Escherichia coli. J. Hyg. 80, 83-113. 85. 0rskov, F, and 0rskov, I. (1983). The clone concept in epidemiology, taxonomy, and evolution of the Enterobacteriaceae and other bacteria. J. Infect. Dis. 148, 346-357. 86. 0rskov, F, and 0rskov, I. (1992). Escherichia call serotyping and disease in man and animals. Can. J. Microbial 38, 699-704. 87. Levine, M. M., and Edelman, R. (1984). Enteropathogenic Escherichia coli of classic serotypes associated with infant diarrhea: Epidemiology and pathogenesis. Epidemiol. Rev. 6, 31-51. 88. 0rskov, F, Whittam, T. S., Cravioto, A., and 0rskov. I. (1990). Clonal relationships among classic enteropathogenic Escherichia call (EPEC) belonging to different O groups. J. Infect. Dis. 162,76-81. 89. Whittam, T. S., Wolfe, M. L., Wachsmuth, I. K.. 0rskov, F, 0rskov, I., and Wilson, R. A. (1993). Clonal relationships among Escherichia coli strains that cause hemorrhagic colitis and infantile diarrhea. Infect. Immun. 61, 1619-1629. 90. Riley, L. W., Remis, R. S., Helgerson, S. D.. McGee. H. B., Wells, J. G., Davis, B. R., Hebert, R. J., Olcott, E. S., Johnson, L. M.. Hargrett. N. T. Blake. P A., and Cohen, M. L. (1983). Hemorrhagic colitis associated with a rare Escherichia call serotype. A^^vv Engl. J. Med. 308, 681-685.

34

HOWARD OCHMAN

91. Boyce, T. G., Swerdlow, D. L., and Griffin, P. M. (1995). E. coli 0157:H7 and the hemolyticuremic syndrome. New Engl. J. Med. 333, 364-368. 92. Karmali, M. (1989). Infection by verocytotoxin-producing Escherichia coli. Clin. Microbiol. Rev. 2, 15-38. 93. Besser, R. E., Lett, S. M., Weber, J. T., Doyle, M. P., Barrett, T. J., Wells, J. G., and Griffin, P. M. (1993). An outbreak of diarrhea and hemolytic uremic syndrome from Escherichia coli 0157:H7 in fresh-pressed apple cider. JAMA 269, 2217-2220. 94. Watanabe, Y., Ozasa, K., Mermin, J. H., Griffm, P. M., Masuda, K., Imashuku, S., and Sawada, T. (1999). Factory outbreak of Escherichia coliO\51:Hl Infection in Japan. Emerg. Infect. Dis. 5, 424-^28. 95. Burland, V., Shao, Y., Perna, N. T., Plunkett, G., Sofia, H. J., and Blattner, F. R. (1998). The complete DNA sequence and analysis of the large virulence plasmid of Escherichia coli 0157:H7. Nucleic Acids Res. 26, 4196-4204. 96. Makino, K., Ishii, K., Yasunaga, T., Hattori, M., Yokoyama, K., Yutsudo, C. H., Kubota, Y, Yamaichi, Y, lida, T, Yamamoto, K., Honda, T., Han, C. G., Ohtsubo, E., Kasamatsu, M., Hayashi, T., Kuhara, S., and Shinagawa, H. (1998). Complete nucleotide sequences of 93-kb and 3.3-kb plasmids of an enterohemorrhagic Escherichia coli 0157:H7 derived from Sakai outbreak. DNA Res. 5, 1-9. 97. Perna, N. T, Mayhew, G. F, Posfai, G., Elliott, S., Donnenberg, M. S., Kaper, J. B., and Blattner, F. R. (1998). Molecular evolution of a pathogenicity island from enterohemorrhagic Escherichia coli 0157:H7. Infect. Immiin. 66, 3810-3817. 98. Plunkett, G., Rose, D. J., Durfee, T. J., and Blattner, F R. (1999). Sequence of Shiga Toxin 2 Phage 933W from Escherichia coli 0157:H7: Shiga Toxin as a phage late-gene product. J. Bacteriol. 181, 1767-1778. 99. Whittam, T. S., Wachsmuth, I. K., and Wilson, R. A. (1988). Genetic evidence of clonal descent of Escherichia coli 0157:H7 associated with hemorrhagic colitis and hemolytic uremic syndrome. J. Infect. Dis. 157, 1124-1133. 100. Ochman, H., and Selander, R. K. (1984). Evidence for clonal population structure in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 81, 198-201. 101. Herzer, P. J., Inouye, M., and Whittam, T. S. (1990). Phylogenetic distribution of branched RNA-linked multicopy single-stranded DNA among natural isolates of Escherichia coli. J. Bacteriol. 172,6175-6181. 102. Lecointre, G., Rachdi, L., Darlu, P., and Denamur, E. (1998). Escherichia coli molecular phylogeny using the incongruence length difference test. Mol. Biol. Evol. 15, 1685-1695. 103. Arnold, C., Metherell, L., Willshaw, G., Maggs, A., and Stanley, J. (1999). Predictive fluorescent amplified-fragment length polymorphism analysis of Escherichia coli: High-resolution typing method with phylogenetic significance. J. Clin. Microbiol. 37, 1274-1279. 104. Wieler, L. H., McDaniel, T. K., Whittam, T. S., and Kaper, J. B. (1997). Insertion site of the locus of enterocyte effacement in enteropathogenic and enterohemorrhagic Escherichia coli differs in relation to the clonal phylogeny of the strains. FEMS Microbiol. Lett. 156, 49-53. 105. Bingen, E., Picard, B., Brahimi, N., Mathy, S., Desjardins, P, Elion, J., and Denamur, E. (1998). Phylogenetic analysis oiEscherichia coli strains causing neonatal meningitis suggests horizontal gene transfer from a predominant pool of highly virulent B2 group strains. J. Infect. Dis. Ill, 642-650. 106. Boyd, E. F, and Hartl, D. L. (1998). Chromosomal regions specific to pathogenic isolates of Escherichia coli have a phylogenetically clustered distribution. J. Bacteriol. 180, 1159-1165. 107. Picard, B., Garcia, J. S., Gouriou, S., Duriez, P., Brahimi, N., Bingen, E., Elion, J., and Denamur, E. (1999). The link between phylogeny and virulence in Escherichia coli extraintestinal infection. Infect. Immun. 67, 546-553.

1.


35

108. Karaolis, D. K., Lan, R., and Reeves, R R. (1994). Sequence variation in Shigella sonnet (Sonnei), a pathogenic clone of Escherichia coli. over four continents and 41 years. J. Clin. Microbiol. 32, 796-802. 109. Musser, J. M., Granoff, D. M., Pattison, P. E., and Selander, R. K. (1985). A population genetic framework for the study of invasive diseases caused by serotype b strains of Haemophilus influenzae. Proc. Natl. Acad. Sci. U.S.A. 82, 5078-5082. 110. Porras, O., Caugant, D. A., Lagergard, T., and Svanborg-Eden, C. (1986). Application of multilocus enzyme gel electrophoresis to Haemophilus influenzae. Infect. Immun. 53, 71-78. 111. Caugant, D. A., Selander, R. K., and Olsen, I. (1990). Differentiation between Actinobacillus {Haemophilus) actinomycetemcomitans, Haemophilus aphrophilus and Haemophilus paraphrophilus by multilocus enzyme electrophoresis. J. Gen. Microbiol. 136, 2135-2141. 112. Musser, J. M., Kroll, J. S., Moxon, E. R., and Selander, R. K. (1988). Clonal population structure of encapsulated Haemophilus influenzae. Infect. Immun. 56, 1837-1845. 113. Musser, J. M., Kroll, J. S., Moxon, E. R., and Selander, R. K. (1988). Evolutionary genetics of encapsulated strains of Haemophilus influenz.ae. Proc. Natl. Acad. Sci. U.S.A. 85, 7758-7762. 114. Urwin, G., Musser, J. M., and Yuan, M. F. (1995). Clonal analysis of Haemophilus influenzae type b isolates in the United Kingdom. J. Med. Microbiol. 43, 45-49. 115. Kroll, J. S., Moxon, E. R., and Loynds, B. M. (1993). An ancestral mutation enhancing the fitness and increasing the virulence of Haemoplnlus influenzae type b. J. Infect. Dis. 168, 172-176. 116. Allan, I., Loeb, M. R., and Moxon, E. R. (1987). Limited genetic diversity of Haemophilus influenzae (type b). Microb. Pathogen. 2, 139-145. 117. Smith-Vaughan, H. C , Sriprakash, K. S., Leach, A. J.. Mathews, J. D., and Kemp, D. J. (1998). Low genetic diversity of Haemophilus influenz.ae type b compared to nonencapsulated H. influenzae in a population in which H. influenz.ae is highly endemic. Infect. Immun. 66, 3403-3409. 118. van Alphen, L., Caugant, D. A., Duim, B., O'Rourke, M., and Bowler, L. D. (1997). Differences in genetic diversity of nonencapsulated Haemophilus influenz.ae from various diseases. Microbiology U2>, 1423-1431. 119. Cover, T. L., and Blaser, M. J. (1996). Helicobacter pylori infection, a paradigm for chronic mucosal inflammation: Pathogenesis and implications for eradication and prevention. Adv. Intern. Med. 41, 85-117. 120. Cover, T. L., Glupczynski, Y., Lage, A. P., Burette, A., Tummuru, M. K., Perez-Perez, G. L, and Blaser, M. J. (1995). Serologic detection of infection with cagA-i- Helicobacter pylori strains. J. Clin. Microbiol. 33, 1496-1500. 121. Atherton, J. C , Peek Jr., R. M., Tham, K. T., Cover, T. L., and Blaser, M. J. (1997). Clinical and pathological importance of heterogeneity in vacA, the vacuolating cytotoxin gene of Helicobacter pylori. Gastroenterology 112, 92-99. 122. Covacci, A., Falkow, S., Berg, D. E., and Rappuoli, R. (1997). Did the inheritance of a pathogenicity island modify the virulence of Helicobacter pyloril Trends Microbiol. 5, 205-208. 123. Cover, T. L. (1996). The vacuolating cytotoxin of Helicobacter pylori. Mol. Microbiol. 20, 241-246. 124. Covacci, A., Censini, S., Bugnoli, M., Petracca, R., Burroni, D., Macchia, G., Massone, A., Papini, E., Xiang, Z., Figura, N., and Rappuoli, R. (1993). Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. U.S.A. 90. 5791-5795.

36

HOV/ARD OCHMAN

125. Xiang, Z., Censini, S., Bayeli, P. F, Telford, J. L., Figura, N., Rappuoli, R., and Covacci, A. (1995). Analysis of expression of CagA and Vac A virulence factors in 43 strains of Helicobacter pylori reveals that clinical isolates can be divided into two major types and that CagA is not necessary for expression of the vacuolating cytotoxin. Infect. Immun. 63, 94-98. 126. Weel, J. R, van der Hulst, R. W., Gerrits, Y., Roorda, R, Feller, M., Dankert, J., Tytgat, G. N., and van der Ende, A. (1996). The interrelationship between cytotoxin-associated gene A, vacuolating cytotoxin, Sind Helicobacter pylori-related diseases. J. Infect. Dis. 173, 1171-1175. 127. Marshall, D. G., Dundon, W. G., Beesley, S. M., and Smyth, C. J. (1998). Helicobacter pylori—a conundrum of genetic diversity. Microbiology 144, 2925-2939. 128. Marshall, D. G., Chua, A., Keeling, R W., Sullivan, D. J., Coleman, D. C , and Smyth, C. J. (1995). Molecular analysis oi Helicobacter pylori populations in antral biopsies from individual patients using randomly amplified polymorphic DNA (RAPD) fingerprinting. FEMS Immunol. Med. Microbiol. 10, 317-323. 129. Go, M. R, Kapur, V., Graham, D. Y., and Musser, J. M. (1996). Population genetic analysis of Helicobacter pylori by multilocus enzyme electrophoresis: Extensive allelic diversity and recombinational population structure. J. Bacteriol. 178, 3934-3938. 130. Maynard-Smith, J. N. H., O'Rourke, M., and Spratt, B. G. (1993). How clonal are bacteria? Proc. Natl. Acad. Sci. U.S.A. 90, 4384-4388. 131. Cover, T. L., Tummuru, M. K., Cao, R, Thompson, S. A., and Blaser, M. J. (1994). Divergence of genetic sequences for the vacuolating cytotoxin among Helicobacter pylori strains. / Biol. Chem. 269, 10566-10573. 132. Atherton, J. C , Cao, R, Reek Jr., R. M., Tummuru, M. K., Blaser, M. J., and Cover, T. L. (1995). Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori. Association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270, \111\-\1111. 133. Yamaoka, Y, Kodama, T, Kita, M., Imanishi, J., Kashima, K., and Graham, D. Y. (1998). Relationship of vac A genotypes of Helicobacter pylori to cagA status, cytotoxin production, and clinical outcome. Helicobacter 'i, 241-253. 134. Gamer, J. A., and Cover, T. L. (1995). Analysis of genetic diversity in cytotoxin-producing and non-cytotoxin-producing Helicobacter pylori strains. J. Infect. Dis. 172, 290-293. 135. Forbes, K. J., Fang, Z., and Pennington, T. H. (1995). Allelic variation in the Helicobacter pylori flagellin genes flaA and flaB: Its consequences for strain typing schemes and population structure. Epidemiol. Infect. 114, 257-266. 136. van der Ende, A., Rauws, E. A., Feller, M., Mulder, C. J., Tytgat, G. N., and Dankert, J. (1996). Heterogeneous Helicobacter pylori isolates from members of a family with a history of peptic ulcer disease. Gastroenterology 111, 638-647. 137. Cao, P., and Cover, T. L. (1997). High-level genetic diversity in the vapD chromosomal region of Helicobacter pylori. J. Bacteriol. 179, 2852-2856. 138. Suerbaum, S., Smith, J. M., Bapumia, K., Morelli, G., Smith, N. H., Kunstmann, E., Dyrek, I., and Achtman, M. (1998). Free recombination within Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 95, 12619-12624. 139. Bibb, W. R, GelHn, B. G., Weaver, R., Schwartz, B., Plikaytis, B. D., Reeves, M. W., Pinner, R. W., and Broome, C. V. (1990). Analysis of clinical and food-borne isolates of Listeria monocytogenes in the United States by multilocus enzyme electrophoresis and application of the method to epidemiologic investigations. Appl. Environ. Microbiol. 56, 2133-2141. 140. Boerlin, P., and Piffaretti, J. C. (1991). Typing of human, animal, food, and environmental isolates of Listeria monocytogenes by multilocus enzyme electrophoresis. Appl. Environ. Microbiol. 57, 1624-1629.

1.


37

141. Baloga, A. O., and Harlander, S. K. (1991). Comparison of methods for discrimination between strains of Listeria monocytogenes from epidemiological surveys. Appl. Environ. Microbiol. 57, 2324-2331. 142. Norrung, B., and Skovgaard, N. (1993). Application of multilocus enzyme electrophoresis in studies of the epidemiology of Listeria monocytogenes in Denmark. Apj?!. Environ. Microbiol. 59,2817-2822. 143. Avery, S. M., Hudson, J. A., and Buncic. S. (1996). Multilocus enzyme electrophoresis typing of New Zealand Listeria monocytogenes isolates. Int. J. Food Microbiol. 28, 351-359. 144. Caugant, D. A., Ashton, F. E., Bibb, W. P., Boerlin, P., Donachie, W., Low, C, Gilmour, A., Harvey, J., and Norrung, B. (1996). Multilocus enzyme electrophoresis for characterization of Listeria monocytogenes isolates: Results of an international comparative study. Int. J. Food Microbiol. 32, 301-311. 145. Louie, M., Jayaratne, P., Luchsinger, I., Devenish, J., Yao. J., Schlech, W., and Simor, A. (1996). Comparison of ribotyping, arbitrarily primed PCR, and pulsed-field gel electrophoresis for molecular typing of Listeria monocytogenes. J. Clin. Microbiol. 34, 15-19. 146. Bibb, W. F, Schwartz, B., Gellin, B. G., Plikaytis, B. D., and Weaver, R. E. (1989). Analysis of Listeria monocytogenes by multilocus enzyme electrophoresis and application of the method to epidemiologic investigations. Int. J. Food Microbiol. 8, 233-239. 147. Piffaretti, J. C , Kressebuch, H., Aeschbacher, M.. Bille, J., Bannerman, E., Musser, J. M., Selander, R. K., and Rocourt, J. (1989). Genetic characterization of clones of the bacterium Listeria monocytogenes causing epidemic disease. Proc. Natl. Acad. Sci. U.S.A. 86, 3818-3822. 148. Harvey, J., and Gilmour, A. (1994). Application of multilocus enzyme electrophoresis and restriction fragment length polymorphism analysis to the typing of Listeria monocytogenes strains isolated from raw milk, nondairy foods, and clinical and veterinary sources. Appl. Environ. Microbiol. 60, 1547-1553. 149. Graves, L. M., Swaminathan, B., Reeves. M. W.. Hunter, S. B.. Weaver, R. E., Plikaytis, B. D., and Schuchat, A. (1994). Comparison of ribotyping and multilocus enzyme electrophoresis for subtyping of Listeria monocytogenes isolates. J. Clin. Microbiol. 32, 2936-2943. 150. Vines, A., Reeves, M. W., Hunter, S., and Swaminathan, B. (1992). Restriction fragment length polymorphism in four virulence-associated genes of Listeria monocytogenes. Res. Microbiol. 143, 281-294. 151. Rasmussen, O. F., Beck, T., Olsen, J. E.. Dons, L., and Rossen, L. (1991). Listeria monocytogenes isolates can be classified into two major types according to the sequence of the listeriolysin gene. Infect. Immun. 59, 3945-3951. 152. Rasmussen, O. E, Skouboe, P., Dons. L.. Rossen, L.. and Olsen, J. E. (1995). Listeria monocytogenes exists in at least three evolutionary lines: Evidence from flagellin, invasive associated protein and listeriolysin O genes. Microbiology 141, 2053-2061. 153. Wiedmann, M., Bruce, J. L., Keating, C , Johnson, A. E., McDonough, P. L., and Batt, C. A. (1997). Ribotypes and virulence gene polymorphisms suggest three distinct Listeria monocytogenes lineages with differences in pathogenic potential. Infect. Immun. 65, 2707-2716. 154. Bloom, B. R., and Murray, C. J. L. (1992). Tuberculosis: Commentary on a reemergent killer. Science 251, 1055-1064. 155. Kapur, V., Whittam, T. S., and Musser, J. M. (1994). Is Mycobacterium tuberculosis 15,000 years old?y. Infect. Dis. 170, 1348-1349. 156. Cave, M. D., Eisenach, K. D., McDermott, P E, Bates, J. H., and Crawford, J. T. (1991). IS6110: Conservation of sequence in the Mycobacterium tuberculosis complex and its utilization in DNA fingerprinting. Mol. Cell. Probes 5, 73-80.

38

HOWARD OCHMAN

157. Yang, Z., Barnes, P. F, Chaves, R, Eisenach, K. D., Weis, S. E., Bates, J. H., and Cave, M. D. (1998). Diversity of DNA fingerprints of Mycobacterium tuberculosis isolates in the United States, y. Clin. Microbiol. 36, 1003-1007. 158. Sreevatsan, S., Pan, X., Stockbauer, K. E., Connell, N. D., Kreiswirth, B. N., Whittam, T. S., and Musser, J. M. (1997). Restricted structural gene polymorphism in the Mycobacterium tuberculosis complex indicates evolutionary recent global dissemination. Proc. Natl. Acad. Sci. U.S.A. 94, 9869-9874. 159. Yeh, R. W., Ponce de Leon, A., Agasino, C. B., Hahn, J. A., Daley, C. L., Hopewell, R C , and Small, P. M. (1998). Stability of Mycobacterium tuberculosis DNA genotypes. J. Infect. Dis. 177,1107-1111. 160. Achtman, M. (1995). Epidemic spread and antigenic variability of Neisseria meningitidis. Trends Microbiol. 3, 186-192. 161. Achtman, M. (1997). Microevolution and epidemic spread of serogroup A Neisseria meningitidis—a review. Gene 192, 135-140. 162. Achtman, M. (1998). Microevolution and epidemic spread of Neisseria meningitidis. Electrophoresis 19, 593-596. 163. Caugant, D. A. (1998). Population genetics and molecular epidemiology of Neisseria meningitidis. Apmis 106, 505-525. 164. Peltola, H. (1983). Meningococcal disease: Still with us. Rev. Infect. Dis. 5, 71-91. 165. Caugant, D. A., Froholm, L. O., Bovre, K., Holten, E., Frasch, C. E., Mocca, L. P., Zollinger, W. D., and Selander, R. K. (1986). Intercontinental spread of a genetically distinctive complex of clones of Neisseria meningitidis causing epidemic disease. Proc. Nat. Acad. Sci. U.S.A. 83, 4927^931. 166. Caugant, D. A., Kristiansen, B. E., Froholm, L. O., Bovre, K., and Selander, R. K. (1988). Clonal diversity of Neisseria meningitidis from a population of asymptomatic carriers. Infect. Immun. 56, 2060-2068. 167. Olyhoek, T., Crowe, B. A., and Achtman, M. (1987). Clonal population structure of Neisseria meningitidis serogroup A isolated from epidemics and pandemics between 1915 and 1983. Rev. Infect. Dis. 9, 665-692. 168. Caugant, D. A., Mocca, L. F, Frasch, C. E., Froholm, L. O., Zollinger, W. D., and Selander, R. K. (1987). Genetic structure of Neisseria meningitidis populations in relation to serogroup, serotype, and outer membrane protein pattern. J. Bacteriol. 169, 2781-2792. 169. Morelli, G., Malomy, B., Muller, K., Seller, A., Wang, J. F, del Valle, J., and Achtman, M. (1997). Clonal descent and microevolution of Neisseria meningitidis during 30 years of epidemic spread. Mol. Microbiol. 25, 1047-1064. 170. Feil, E., Carpenter, G., and Spratt, B. G. (1995). Electrophoretic variation in adenylate kinase of Neisseria meningitidis is due to inter- and intraspecies recombination. Proc. Natl. Acad. Sci. U.S.A. 92, 10535-10539. 171. Feil, E., Zhou, J., Maynard Smith, J., and Spratt, B. G. (1996). A comparison of the nucleotide sequences of the adk and recA genes of pathogenic and commensal Neisseria species: Evidence for extensive interspecies recombination within adk. J. Mol. Evol. 43, 631-640. 172. Zhou, J., Bowler, L. D., and Spratt, B. G. (1997). Interspecies recombination, and phylogenetic distortions, within the glutamine synthetase and shikimate dehydrogenase genes of Neisseria meningitidis and commensal Neisseria species. Mol. Microbiol. 23, 799-812. 173. Maynard-Smith, J., and Smith, N. H. (1998). Detecting recombination from gene trees. Mol. Biol. Evol. 15, 590-599. 174. White, P. B. (1926). Further studies on the Salmonella group. Med. Res. Spec. Rep. Counc. (Great Brittain) Sen 103, 3-160.

1.


39

175. Kauffman, F. (1966). "The Bacteriology of Enterobacteriaceae." Williams & Wilkins, Baltimore. 176. Crosa, J. H., Brenner, D. J., Ewing, W. H.. and Falkow, S. (1973). Molecular relationships among the salmonellae. J. Bactehol. 115. 307-315. 177. Le Minor, L., Veron, M., and Popoff, M. Y. (1982). Taxonomie de Salmonellae. Ann. Inst. Pasteur Microbiol. 133B, 223-243. 178. Nelson, K., and Selander, R. K. (1992). Evolutionary genetics of the proline permease gene iputP) and the control region of the proline utilization operon in populations of Salmonellae and Escherichia coli. J. Bacteriol. 174, 6886-6895. 179. Nelson, K., and Selander, R. K. (1994). Intergeneric transfer and recombination of the 6-phosphogluconate dehydrogenase gene {gnd) in enteric bacteria. Proc. Natl. Acad. Sci. U.S.A. 91, 10227-10231. 180. Boyd, E. F, Nelson, K., Wang, F-S., Whittam. T S.. and Selander. R. K. (1994). Molecular genetic basis of allelic polymorphism in malate dehydrogenase (nulh) in natural populations of Escherichia coli and Salmonella enterica. Proc. Natl. Acad. Sci. U.S.A. 91, 1280-1284. 181. Li, J., Ochman, H., Groisman, E. A., Boyd, E. F, Solomon, F, Nelson, K., and Selander, R. K. (1995). Relationship between evolutionary rate and cellular location among the Inv/Spa invasion proteins of Salmonella enterica. Proc. Natl. Acad. Sci. U.S.A. 92. 7252-7256. 182. Boyd, E. F, Wang, F-S., Whittam, T. S., and Selander. R. K. (1996). Molecular genetic relationships of the salmonellae. Appl. Environ. Microbiol. 62, 804-808. 183. Beltran, R, Musser, J. M., Helmuth, R.. Farmer III J. J., Frerichs, W. M., Wachsmuth, I. K., Ferris, K., McWhorter, A. C, Wells, J. G.. Cravioto, A., and Selander, R. K. (1988). Toward a population genetic analysis of Salmonella: Genetic diversity and relationships among strains of serotypes S. choleraesuis, S. derby, S. dublin, S. enteritidis, S. heidelberg, S. infantis, S. newport, and S. typhimiirium. Proc. Natl. Acad. Sci. U.S.A. 85, 7153-7151. 184. Beltran, R, Plock, S. A., Smith, N. H., Whittam. T S., Old. D. C., and Selander, R. K. (1991). Reference collection of strains of the Salmonella typhinuirium complex from natural populations. J. Gen. Microbiol. 137, 601-606. 185. Boyd, E. R, Wang, F-S., Beltran, P, Plock. S. A., Nelson, K., and Selander, R. K. (1993). Salmonella reference collection B (SARB): Strains of 37 serovars of subspecies I. J. Gen. Microbiol. 139, 1125-1132. 186. Li, J., Smith, N. H., Nelson, K., Crichton, P B.. Old, D. C., Whittam, T S., and Selander, R. K. (1993). Evolutionary origin and radiation of the avian-adapted non-motile salmonellae. J. Med. Microbiol. 38, 129-139. 187. Selander, R. K., Beltran, P, Smith, N. H., Barker, R. M., Crichton, P B., Old, D. C., Musser, J. M., and Whittam, T. S. (1990). Genetic population structure, clonal phylogeny, and pathogenicity oi Salmonella paratyphi B. Infect. Immim. 58, 1891-1901. 188. Selander, R. K., Beltran, P, Smith, N. H., Helmuth, R., Rubin, F A., Kopecko, D. J., Ferris, K., Tall, B. D., Cravioto, A., and Musser, J. M. (1990). Evolutionary genetic relationships of clones of Salmonella serovars that cause human typhoid and other enteric fevers. Infect. Immun. 58, 2262-2275. 189. Selander, R. K., Smith, N. H., Li, J., Beltran, P, Ferris, K., Kopecko, D. J., and Rubin, F A. (1992). Molecular evolutionary genetics of the cattle-adapted serovar Salmonella dublin. J. Bacteriol 174, 3587-3592. 190. Nelson, K., Whittam, T S., and Selander, R. K. (1991). Nucleotide polymorphism and evolution in the glyceraldehyde-3-phosphate dehydrogenase gene {gapA) in natural populations of Salmonella and Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 88, 6667-6671. 191. Wang, F. S., Whittam, T. S., and Selander, R. K. (1997). Evolutionary genetics of the isocitrate dehydrogenase gene (icd) in Escherichia coli and Salmonella enterica. J. Bacteriol. 179, 6551-6559.

40

HOWARD OCHMAN

192. Nelson, K., Wang, F. S., Boyd, E. R, and Selander, R. K. (1997). Size and sequence polymorphism in isocitrate dehydrogenase kinase/phosphatase gene (acek) and flanking regions in Salmonella enterica and Escherichia coli. Genetics 147, 1509-1520. 193. Ochman, H., and Groisman, E. A. (1996). Distribution of pathogenicity islands in Salmonella spp. Infect. Immun. 64, 5410-5412. 194. Baumler, A. J., Tsolis, R. M., Ficht, T. A., and Adams, L. G. (1998). Evolution of host adaptation in Salmonella enterica. Infect. Immun. 66, 4579-4587. 195. Bergdoll, M. S., Crass, B. A., Reiser, R. R, Robbins, R. N., and Davis, J. R (1981). A new staphylococcal enterotoxin, enterotoxin F, associated with toxic-shock-syndrome Staphylococcus aureus isolates. Lancet 1, 1017-1021. 196. Schlievert, R M., Shands, K. N., Dan, B. B., Schmid, G. R, and Nishimura, R. D. (1981). Identification and characterization of an exotoxin from Staphylococcus aureus associated with toxic-shock syndrome. J. Infect. Dis. 143, 509-516. 197. Musser, J. M., Schlievert, R M., Chow, A. W., Ewan, R, Kreiswirth, B. N., Rosdahl, V. T., Naidu, A. S., Witte, W., and Selander, R. K. (1990). A single clone of Staphylococcus aureus causes the majority of cases of toxic shock syndrome. Proc. Natl. Acad. Sci. U.S.A. 87, 225-229. 198. Grubb, W. B. (1990). Molecular epidemiology of methicillin-resistant Staphylococcus aureus. In "Molecular Biology of the Staphylococci" (R. Novick and R. A. Skurray, eds.), pp. 595-606. VCH, New York. 199. Musser, J. M., and Selander, R. K. (1990). Brazilian purpuric fever: Evolutionary genetic relationships of the case clone of Haemophilus influenzae biogroup aegyptius to encapsulated strains of Haemophilus influenzae. J. Infect. Dis. 161, 130-133. 200. Musser, J. M., and Kapur, V. (1992). Clonal analysis of methicillin-resistant Staphylococcus aureus strains from intercontinental sources: Association of the mec gene with divergent phylogenetic lineages implies dissemination by horizontal transfer and recombination. / Clin. Microbiol. 30, 2058-2063. 201. Kreiswirth, B., Komblum, J., Arbeit, R. D., Eisner, W., Maslow, J. N., McGeer, A., Low, D. E., and Novick, R. P. (1993). Evidence for a clonal origin of methicillin resistance in Staphylococcus aureus. Science 259, 227-230. 202. Musser, J. M., Hauser, A. R., Kim, M. H., Schliever, R M., Nelson, K., and Selander, R. K. (1991). Streptococcus pyogenes causing toxic-shock-like syndrome and other invasive diseases: Clonal diversity and pyrogenic exotoxin expression. Proc. Natl. Acad. Sci. U.S.A. 88, 26682672. 203. Nelson, K., Schlievert, R M., Selander, R. K., and Musser, J. M. (1991). Characterization and clonal distribution of four alleles of the speA gene encoding pyrogenic exotoxin A (scarlet fever toxin) in Streptococcus pyogenes. J. Exp. Med. 174, 1271-1274. 204. Whatmore, A. M., Kapur, V., Sullivan, D. J., Musser, J. M., and Kehoe, M. A. (1994). Non-congruent relationships between variation in emm gene sequences and the population genetic structure of group A streptococci. Molec. Microbiol. 14, 619-631. 205. Musser, J. M., Kapur, V., Szeto, J., Pan, X., Swanson, D. S., and Martin, D. R. (1995). Genetic diversity and relationships among Streptococcus pyogenes strains expressing serotype Ml protein: Recent intercontinental spread of a subclone causing episodes of invasive disease. Infect. Immun. 63, 994-1003. 206. Stockbauer, K. E., Grigsby, D., Pan, X., Fu, Y.-X., Perea Mejia, L. M., Cravioto, A., and Musser, J. M. (1998). Hypervariability generated by natural selection in an extracellular complementinhibiting protein of serotype Ml strains of gvou^ \ Streptococcus. Proc. Natl. Acad. Sci. U.S.A. 95,3128-3133.

1.


41

207. Reda, K. B., Kapur, V., Goela, D., Lamphear, J. G., Musser, J. M., and Rich, R. R. (1996). Phylogenetic distribution of streptococcal superantigen SSA allelic variants provides evidence for horizontal transfer of ssa within Streptococcus pyogenes. Infect. Immun. 64, 1161-1165. 208. Marciel, A. M., Kapur, V., and Musser, J. M. (1997). Molecular population genetic analysis of a Streptococcus pyogenes bacteriophage-encoded hyaluronidase gene: Recombination contributes to allelic variation. Microb. Pathogen. 22. 209-217. 209. Dowson, C. G., Hutchison, A., Brannigan, J. A., George, R. C , Hansman, D., Linares, J., Tomasz, A., Maynard Smith, J., and Spratt, B. G. (1989). Horizontal transfer of penicillin-binding protein genes in penicillin-resistant clinical isolates of Streptococcus pneumoniae. Proc. Natl. Acad. Sci. U.S.A. 86, 8842-8846. 210. Dowson, C. G., Hutchison, A., Woodford, N., Johnson, A. R, George, R. C., and Spratt, B. G. (1990). Penicillin-resistant viridans streptococci have obtained altered penicillin-binding protein genes from penicillin-resistant strains of Streptococcus pneumoniae. Proc. Natl. Acad. Sci. U.S.A. 87, 5858-5862. 211. Coffey, T. J., Enright, M. C., Daniels, M., Morona, J. K., Morona, R., Hryniewicz, W., Paton, J. C., and Spratt, B. G. (1998). Recombinational exchanges at the capsular polysaccharide biosynthetic locus lead to frequent serotype changes among natural isolates of Streptococcus pneumoniae. Molec. Microbiol. 27, 73-83. 212. Enright, M. C., and Spratt, B. G. (1998). Amultilocus sequence typing scheme for Streptococcus pneumoniae: Identification of clones associated with serious invasive disease. Microbiology 144, 3049-3060. 213. Wachsmuth, I. K., Evins, G. M., Fields, R I., Olsvik, O., Popovic, T., Bopp, C. A., Wells, J. G., Carrillo, C, and Blake, P. A. (1993). The molecular epidemiology of cholera in Latin America. y. Infect. Dis. 161, 621-626. 214. Evins, G. M., Cameron, D. N., Wells, J. G., Greene, K. D., Popovic, T, Giono-Cerezo, S., Wachsmuth, L K., and Tauxe, R. V. (1995). The emerging diversity of the electrophoretic types of Vibrio cholerae in the Western Hemisphere. J. Infect. Dis. 172, 173-179. 215. Beltran, P, Delgado, G., Navarro, A., Trujillo, F., Selander, R. K., and Cravioto, A. (1999). Genetic diversity and population structure of Vibrio cholerae. J. Clin. Microbiol. 37, 581-590. 216. Olsvik, O., Wahlberg, J., Petterson, B., Uhlen, M., Popovic, T, Wachsmuth, L K., and Fields, P. L (1993). Use of automated sequencing of polymerase chain reaction-generated amplicons to identify three types of cholera toxin subunit B in Vibrio cholerae 01 strains. J. Clin. Microbiol. 31, 22-25. 217. KaraoHs, D. K., Lan, R., and Reeves, P. R. (1995). The sixth and seventh cholera pandemics are due to independent clones separately derived from environmental, nontoxigenic, non-01 Vibrio cholerae. J. Bacteriol. Ill, 3191-3198. 218. Karaolis, D. K., Lan, R., and Reeves, P. R. (1994). Molecular evolution of the seventh-pandemic clone of Vibrio cholerae and its relationship to other pandemic and epidemic V. cholerae isolates. /. Bacteriol 176, 6199-6206. 219. Koblavi, S., Grimont, F, and Grimont, P A. (1990). Clonal diversity of Vibrio cholerae 01 evidenced by rRNA gene restriction patterns. Res. Microbiol. 141, 645-657. 220. Lan, R., and Reeves, P. R. (1998). Recombination between rRNA operons created most of the ribotype variation observed in the seventh pandemic clone of Vibrio cholerae. Microbiology 144, 1213-1221. 221. Byun, R., Elbourne, L. D., Lan, R., and Reeves, P. R. (1999). Evolutionary relationships of pathogenic clones of Vibrio cholerae by sequence analysis of four housekeeping genes. Infect. Immun. 67, 1116-1124.


CHAPTER 2

Germ Warfare: The Mechanisms of Virulence Factor Delivery JILL REISS HARPER THOMAS J. SILHAVY

I. II. III. IV. V. VI. VII. VIII.

Introduction The General Secretory Pathway Autotransporters: Type V Two-Step Secretion: Type II ABC Transporters: Type I Conjugal Transfer Systems: Type IV Contact-Dependent Secretion: Type III Concluding Remarks References

43 45 47 49 52 55 57 61 61

/. Introduction Pathogenic bacteria synthesize a diverse array of virulence determinants. These virulence proteins, which comprise the arsenal of bacterial weapons, have a wide variety of activities that require them to be targeted to specific locations. For example, proteins involved in attachment of the bacterium to the host cell must be localized to the bacterial surface, some bacterial toxins are secreted into surrounding fluids, and others are injected directly into the cytoplasm of the eukaryotic host cell. Thus, in pathogenic bacteria secretion itself is a virulence determinant; without the means to selectively target proteins, these bacteria are harmless.


A^

44

JILL REISS HARPER AND THOMAS J. SILHAVY

The problem of targeting proteins to their correct location is not unique to pathogenic bacteria. All cells have subcellular compartments that are bound by lipid bilayers, and it is essential that each of these compartments contain a characteristic set of proteins. To maintain this organization, cells have specific mechanisms to target proteins from their site of synthesis in the cytoplasm to each noncytoplasmic location. This targeting, of course, requires the movement of proteins into and across the lipid bilayers. In most cases, proteins destined to leave the cytoplasm are tagged with a signal sequence at the amino terminus that serves to target them to the cellular secretion (Sec) machinery that includes a heterotrimeric complex of integral membrane proteins. Signal sequences and the components of this heterotrimeric complex have been conserved in all three domains of life [1]. While bacterial cells are not as complex as their eukaryotic counterparts, they do exhibit some compartmentalization. If we consider lipid bilayers to be compartments, then Gram-positive bacteria have three destinations to which proteins can be targeted: the cytoplasm, the membrane, and the extracellular environment. Gram-negative bacteria, on the other hand, can target proteins to five distinct locations: the cytoplasm, the inner membrane, the periplasm, the outer membrane, and the extracellular environment. In Gram-positive bacteria, the Sec machinery is sufficient for targeting to the extracellular environment since the secreted proteins must pass through only one lipid bilayer. The outer membrane of Gram-negative bacteria complicates protein secretion to the extracellular environment. At least five different mechanisms, termed Types I through V, appear to be conserved among the Gram-negative bacteria. (Unfortunately, this nomenclature is confusing and incomplete because it does not include the mechanisms responsible for secretion of certain colicins [2] and the process of pilus assembly [3], for which there is currendy no name. Nonetheless, this nomenclature system has caught on and will be used in this chapter.) The various Gram-negative secretion systems can be divided into those that utilize the Sec machinery for translocation across the inner membrane (Sec-dependent) and those that do not (Sec-independent). The Sec-dependent secretion systems include Types II and V. The Type II system involves a two-step mechanism in which proteins are first targeted to the periplasm by the Sec machinery and then secreted from the cell by a complex reaction requiring a dozen or so additional proteins. The Type V systems, also called the autotransporters, utilize the normal pathway for outer membrane targeting. Autoproteolysis releases a secreted protein domain into the environment. The Sec-independent secretion systems include Types I, III, and IV. The Type I systems utilize a complex of three proteins that span both the inner and outer membranes, and they secrete proteins directly into the media. The Type IV systems are not well characterized at this time; they appear to be closely related to systems used for the conjugal transfer of DNA from one bacterium to another. The Type III systems are fascinating

2.

GERM WARFARE

45

devices that actually allow injection of bacterial proteins directly into the cytoplasm of the host cell. Sometimes they are called "contact-dependent" to reflect this capacity for injection. Altogether these systems provide a wide range of options for bacterial protein secretion and will be the focus of this chapter.

//. The General Secretory Pathway Protein export from the cytoplasm is a conserved pathway that was discovered and characterized in the Gram-negative bacterium Escherichia coli. As noted above, many proteins destined for noncytoplasmic locations are synthesized with a signal sequence that targets them for translocation. This amino-terminal signal is later cleaved during the export process [4]. The function of signal sequences was first demonstrated using gene fusions in which the amino-terminal end of an exported protein such as LamB, the receptor for bacteriophage X, is fused to the cytoplasmic enzyme (3-galactosidase, or LacZ (reviewed in [5]). Strains carrying gene fusions of this type exhibit several novel phenotypes that can be exploited to obtain mutations that alter the secretion process. First, LacZ is meant to fold in the cytoplasm; if it is located somewhere else, it folds improperly and exhibits no activity. Second, the attempted secretion of large amounts of LacZ lethally jams the secretion machinery. The original LamB signal sequence mutations were isolated as suppressors of this deadly jamming. Furthermore, these signal sequence mutations increased LacZ activity because they prevented enzyme export from the cytoplasm [6]. Further analysis of many signal sequences from both prokaryotic and eukaryotic proteins has shown that they consist of about 15-26 amino acids. There is no conserved sequence, but they do possess common features [7]. They contain a stretch of 10-12 hydrophobic amino acids preceded by one or two positively charged residues and followed at the carboxy-terminal end by a cleavage site for leader peptidase, which removes the signal sequence from the preprotein to yield the mature protein. This run of hydrophobic residues resembles a transmembrane domain. Mutations that block signal sequence processing do not prevent translocation, but leave the mutant precursor protein tethered to the inner membrane with the amino terminus of the signal sequence facing the cytoplasm [8]. To catalyze insertion into and translocation across the inner membrane, E. coli has a set of proteins—the Sec proteins—that were identified using two complementary genetic approaches. In one approach, suppressors of signal sequence mutations were selected based on their ability to restore localization of a mutant precursor protein. The genes identified by this approach were called/?r/ for protein localization [9]. In the complementary approach, mutations were sought that increased the LacZ activity of an exported fusion protein and simultaneously

46


conferred a conditional lethal phenotype. Genes identified in this way were termed sec for secretion [10]. Subsequent molecular analysis revealed that three of the sec genes were also/7r/ genes: secA/prlD [10, 11], secE/prlG [12, 13], and secY/prlA [9, 12, 14]. It is these three proteins that form the essential core of the secretion machinery; protein translocation can be reconstituted in vitro with only Sec A, SecE, and SecY [15, 16]. Later genetic experiments identified three additional nonessential sec genes—secB [17], secD [18], and secF [19]—and biochemical studies added two others—secG [16, 20] and jo/C [21]. Interestingly, prl alleles of secG were identified in 1997 [22]. Biochemical approaches have elucidated the role of the Sec proteins in protein export from the cytoplasm [23]. The Sec machinery is comprised of soluble cytoplasmic proteins and peripheral and integral cytoplasmic membrane proteins (Fig. 1, see color plate). In order to be exported from the cytoplasm, a protein must first be recognized by the secretion machinery. Multiple, partially redundant mechanisms ensure the accuracy of precursor recognition. In the cytoplasm, SecA binds signal sequences directly [24]. The secretion-specific chaperone SecB binds to the mature portions of exported proteins such as LamB or MalE [25, 26] and maintains them in an unfolded, export-competent state [27-31]. This contributes to recognition because SecB also binds SecA [32, 33]. SecA functions as a dimer to direct the precursor to the membrane and energize the translocation reaction [34]. The complex of SecA, precursor protein, and SecB interacts on the cytoplasmic face of the heterotrimeric SecYEG complex in the inner membrane. The SecYEG complex, based on studies with its eukaryotic counterparts, likely forms a protein-conducting channel through the inner membrane [35-37]. Binding of ATP then induces a change in the structure of SecA that allows it to insert into the inner membrane [38]. Once SecA enters the membrane, SecB is released into the cytoplasm and the precursor is partially translocated through the SecYEG complex [39]. This translocation allows access of signal peptidase to the precursor, and signal peptidase then cleaves the signal peptide. SecD, SecF, and YajC seem to stabilize the membrane-bound form of SecA [40]. ATP hydrolysis causes SecA to release the partially translocated precursor and deinsert from the membrane. Additional ATP binding and hydrolysis by SecA will repeat the insertion/deinsertion process to catalyze the progressive threading of the translocating protein through the inner membrane [40-42]. Translocation may also continue using energy from the proton-motive force in a step that is poorly characterized. E. coli possesses another mechanism for targeting proteins to the secretion apparatus in the inner membrane—SRP, or Signal Recognition Particle. This targeting factor is a diminutive version of its eukaryotic counterpart. Prokaryotic SRP contains a protein, Ffh, that is homologous to eukaryotic SRP54, the subunit that recognizes the signal peptide, and an RNA, 4.5S RNA, that resembles eukaryotic 7S RNA. E. coli also has a protein, FtsY, that is similar to the a subunit of the mammalian SRP receptor. These proteins are functional in targeting; the

2. GERM WARFARE

47

bacterial proteins can replace their eukaryotic counterparts in vitro [43]. Genetic analysis of secB mutants and ffh-dtp\eied strains argues that there may be two separate subsets of exported proteins in E. coli—those targeted by SecB and those targeted by Ffh [44, 45]. Alternatively, or in addition, prokaryotic SRP may function in the insertion of inner membrane proteins [46-48]. As noted above, components of the secretion machinery such as signal sequences, SecYEG, and SRP are conserved throughout biology. SecA is found only in eubacteria and chloroplasts [1]. Apparently, eukaryotes use a different mechanism to energize translocation. SecB has only been found in the family Enterobacteriaccae [ 1 ].

///. Autotransporters: Type V The simplest of the Sec-dependent secretion systems are the Type V systems, also known as the autotransporters [49]. The prototype of these is the IgA protease from M gonorrhoeae [50], which has been studied extensively for quite some time because the proteolysis of IgA by IgA protease is a major contributor to virulence. Other examples of autotransporters include the immunoglobulin A (IgA) protease from Haemophilus influenzae [51], the serine protease from Serratia marcescens [52], and the vacuolating cytotoxin VacA from Helicobacter pylori [53]. Autotransporters are first exported to the outer membrane, where a proteolytic event releases the final product into the medium (Fig. 2, see color plate) [54]. The secreted protein is first expressed in the cytoplasm as a large multidomain protein consisting of an amino-terminal Sec-dependent signal sequence, the 106-kD mature portion of the protein, a 30-amino-acid y-protein, a 15-kD secreted a-protein, and a 45-kD carboxy-terminal (J-protein, which remains inserted in the outer membrane [55]. The amino-terminal signal sequence targets the protein for translocation from the cytoplasm by the Sec machinery. Concomitant with translocation across this membrane, the signal peptide is cleaved and the remainder of the protein is released into the periplasm. The carboxy-terminal p-domain is targeted to the outer membrane, where it forms a P-barrel pore or channel through which the rest of the protein can pass in its unfolded state through the outer membrane onto the cell surface. The protein then undergoes autoproteolysis to cleave the (J-domain from the rest of the protein. The (J-protein remains in the outer membrane, while the mature, a-, and y-domains are released into the extracellular environment. Subsequent autoproteolytic cleavages release the mature protein from the small a- and y-proteins [56]. Their function is unknown; however, they could play a role as chaperones.

48


The signals that direct the secretion of IgA protease are twofold: first, the presence of an amino-terminal signal sequence to direct export from the cytoplasm via the Sec machinery and, second, the presence of the (i-domain for secretion from the periplasm across the outer membrane [57]. The P-domain contains the information necessary for insertion into the outer membrane, and this insertion is sufficient for the rest of the protein to pass onto the surface. The information in the ^-domain responsible for targeting to the outer membrane has been reduced to a minimal "core" region that retains the ability to translocate a passenger protein across the outer membrane. This core is conserved among the carboxy-terminal regions of the IgA p-domains of several species. Based on structural predictions, the p-core resembles the p-barrel motif conserved among Gram-negative outer membrane proteins, and presumably all of these proteins share a common mechanism for outer membrane insertion [56]. However, this important process remains poorly characterized. Little was known about the secretion of autotransporters until the discovery that IgA protease itself is the only protein, aside from the Sec machinery, required for its secretion into the extracellular environment. The key experiment was the isolation of a DNA fragment from A^. gonorrhoeae that allowed secretion of active IgA protease when expressed in Escherichia coli [58]. Because E. coli does not usually secrete proteins into the extracellular environment, this indicated that all the components for secretion were present on the plasmid. Sequencing revealed that the cloned fragment contained a single gene coding for a protein significantly larger than the secreted, active form of IgA protease. Subsequent work demonstrated that the mature IgA protease, a-protein, and p-protein are the result of autoproteolytic processing at cleavage sites similar to those on host proteins that are the target of IgA protease [55, 57]. Other important work has shown that fusion proteins consisting of a signal sequence, the Vibrio cholerae toxin B subunit, and the IgA p-domain are efficiently secreted by E. coli, demonstrating that the signal sequence and the p-domain are sufficient for secretion. This work also shows that sequences translocated through the p-domain are generally in an unfolded state and that the translocation occurs in a linear fashion [59]. That the passenger proteins are translocated through the p-core in their unfolded state is further supported by the lack of cysteines, and therefore disulfide bonds, in the translocated portions of the autotransporters. Disulfide bonds would presumably introduce structures that cannot pass through the P-core pore [60]. Work with S. marcescens serine protease has demonstrated that, at least for this IgA protease-like protein, part of the carboxy-terminal p-domain plays the role of a chaperone and assists in folding of the mature enzyme once located on the surface [61]. Other work involves the development of systems to facilitate kinetic and structural studies of autotransporter secretion [62].

2. GERM WARFARE

49

The simplicity of these systems is striking, especially in contrast to other systems we will discuss. Accordingly, these autotransporters have implications for biotechnology, and many are being examined for their potential ability to secrete heterologous proteins from E. coli.

IV. Two-step Secretion: Type II The Type II secretion systems provide a commonly used mechanism for the extracellular secretion of proteins from Gram-negative bacteria. The pal system of Klebsiella oxytoca is the founding member of this diverse family. This system directs the secretion of pullulanase, a starch-debranching enzyme. Pullulanase is a lipoprotein that is exported to the periplasm by Sec-dependent means, targeted to the outer membrane, and then secreted by a complex mechanism requiring more than a dozen other proteins [63]. Examples of Type II secretion systems include the out systems of Envinia dvysanthemi and Erwinia cawtovora [64, 65], the xcp system of Pseudomonas aeruginosa [66], exe from Aeromonas hydrophila [67], xps from Xanthomonas campestris [68], and eps from Vibrio cholerae [69]. Type II systems also show similarities to systems that catalyze biogenesis of type IV pili, assembly of filamentous phages, and competence for DNA uptake [70-72]. The Type II secretion systems consist of about 12-14 proteins that are encoded by a cluster of genes [72, 73] (Fig. 3, see color plate). Although there is no direct evidence to support it, one model suggests that some of the components of the Type II systems form a pilus-like structure in the envelope [71, 74, 75]. The presence of several proteins—PulG, H, I, and J—that have homology to type IV pilin subunits led to this suggestion. These proteins, the pseudopilins, contain consensus prepilin peptidase cleavage sites and are processed by the PulO protein, which is homologous to the PilD/XcpA prepilin peptidase involved in the biogenesis of type IV pili [76]. PilD/XcpA is a multifunctional protein from P. aeruginosa that processes and A^-methylates the subunits of the type IV pilus and also the components of the apparatus that secretes alkaline phosphatase, phospholipase C, elastase, and exotoxin A [77-83]. The purpose of this processing is not yet known; it does not seem to be involved in pseudopilin complex formation or in relocalization of the pseudopilins [73]. Furthermore, the pseudopilins in K. oxytoca have been shown to localize to the cytoplasmic membrane, which may not lend well to their role in a pilus-like complex [84]. Nonetheless, it would seem that the role of the O protein is to process the components of the pilus-like structure so that they can assemble properly. The pilus-like structure is thought to guide secreted proteins in their folded conformation to the secretin PulD, located in the outer membrane. The secretins

50


are a conserved family of proteins that form pores in the outer membrane large enough to allow secreted proteins to pass into the surrounding medium [85-87]. Support for the PulD protein forming a pore in the outer membrane comes from its sequence similarity to gpIV, a filamentous phage protein, and PilQ proteins, involved in type IV pilus assembly [88]. It has been suggested that this similarity reflects similar mechanisms at work in the filamentous phage type IV pili, and Type II secretion systems. Indeed, PulD protein forms large stable complexes in the outer membrane that are similar to complexes formed by the gpIV and PilQ proteins [87, 89, 90]. The gpIV protein forms pores [91], and one of the PulD-like proteins, XcpQ from P. aeruginosa, has been shown to form multimeric ring complexes in the outer membrane that could also serve as pores [92]. PulD targeting requires the activity of PulS, a lipoprotein that is also found in the outer membrane [93]. PulS also acts as a chaperone to protect PulD from degradation [87, 94]. Additional components required for secretion include PulC, F, K, L, M, and N, all of which are located in the inner membrane and have unknown functions [84, 93, 95, 96]. The final component required for secretion is PulE, a cytoplasmic protein with a putative nucleotide-binding motif [96, 97]. PulE associates with the inner membrane only in the presence of the other Pul proteins [97]. The specific protein with which PulE interacts is unknown, but in V. cholerae the PulE homolog EpsE interacts with the PulL homolog EpsL [98]. The obvious role for PulE, because of its nucleotide-binding site, would be to act, along with the proton-motive force, as the energizer for the process of secretion. PulE could also play a role in signaling [98]. In any event, ATP binding by PulE is essential for secretion [97]. Perhaps the most surprising feature of the current model is that, with the exception of PulD, PulS, and PulE, all of the Type II components are found in the inner membrane [72]. Why a system that targets proteins from the periplasm through the outer membrane needs so many cytoplasmic membrane proteins and one cytoplasmic protein is not understood. Like Type IV autotransporters, secretion via Type II systems requires the presence of a classical signal peptide recognized by the Sec machinery and cleaved by signal peptidase I in most cases, or signal peptidase II in the case of lipoproteins like K. oxytoca pullulanase. This signal, of course, is necessary and sufficient to get the secreted protein into the periplasm. What is the signal that distinguishes a protein to be secreted via the Type II system from any other periplasmic protein? One approach to this question has been deletion analysis of secreted proteins to analyze the regions necessary and sufficient for secretion. From this work in P. aeruginosa, it appears that the first 30 amino acids of exotoxin A play an important role in extracellular secretion [99, 100]. There also seems to be a second secretion signal located in the carboxy terminus of the protein [100]. Another strategy has implemented the use of

2. GERM WARFARE

51

reporter fusions to proteins such as p-lactamase. The secretion of these fusion proteins is surprisingly resistant to large internal deletions [101, 102]. Two regions, each of about 80 amino acids and separated by about 600 amino acids, have been identified as putative extracellular targeting signals in pullulanase [102]. In exotoxin A of f! aeruginosa, the signal has been localized to amino acids 60-120 in the amino-terminal end of the protein [101]. Based on the ambiguity of these results, it seems unlikely that all Type II secretion systems will recognize a common stretch of amino acid sequence like the Sec machinery does. It is more likely that a combination of specific sequences and conformational information targets secreted proteins such as pullulanase to the Type II secretion machinery [103]. Some evidence already supports the idea that these proteins must be folded into a certain conformation in order to be secreted [104-109]. It seems likely that proper folding of the protein allows regions that are far apart in the primary sequence to come together and form a signal patch that is then recognized by the secretion machinery [103]. The history of concept development with the Type II secretion systems is interesting and informative (reviewed in [73]). It had been established that a large number of proteins secreted by Gram-negative bacteria were synthesized with classical signal peptides at their amino termini. However, viable mutants had been isolated in several species that were defective in the secretion of proteins to the extracellular surroundings. Presumably, the genes identified were specifically involved in the process of extracellular protein secretion (e.g., [110, 111-114]); this secretion system could not be essential. Based on the simplicity of certain systems, such as the recendy identified autotransporters, it was originally thought that few cellular components would be required for export across the outer membrane [73]. The discovery of the extremely complicated Type II systems was, therefore, quite surprising. The breakthrough occurred with the examination of pullulanase secretion. Previous attempts to express pullulanase in E. coli resulted in accumulation of enzyme in the membranes [115]. Finally, a large 23-kb DNA fragment from Klebsiella oxytoca was cloned that supported pullulanase secretion in E. coli [116]. Subsequent work showed that the cloned fragment contained 14 genes—pulC, D, E, F, G, //, /, y, K, L, M, A^, O, and S—that are required for the secretion of pullulanase [84, 117, 95, 96]. In the meantime, genes involved in the secretion of proteins in several pathogenic species were shown to be homologous to several of the pul genes [64, 66, 81]. Surprisingly, a functional secretion system that is homologous to the Pul system has been identified in E. coli K-12, which historically had not been thought to actively secrete proteins into the environment [118-120]. Understanding how all of these components interact to catalyze protein secretion will not be easy. It seems likely that all of these components make up

52


a macromolecular machine. The isolation or visualization of such a machine would be an important advance.

V. ABC Transporters: Type I At least in terms of number of components, the Type I, or ABC, transporter systems are quite simple. In these systems, only three proteins are required for Sec-independent secretion. Examples of the proteins secreted by the Type I secretion systems include a-hemolysin from Escherichia coli, Proteus vulgaris, and Morganella morganii [121]; leukotoxin from Pasteurella haemolytica [122]; metalloprotease from Serratia marcescens [123]; alkaline protease from Pseudomonas aeruginosa [124]; and proteases A, B, and C from Erwinia chrysanthemi [125]. As their name implies, these systems include a component that is a member of the ATP-binding cassette, or ABC, family of transporter proteins (reviewed in [2]). ABC transporters are found in all organisms from bacteria to mammals and function in the import and export of a diverse array of molecules. This family of proteins can be subdivided into three groups—the bacterial importers, the eukaryotic transporters, and the bacterial exporters [2]—with the bacterial exporters comprising the largest subfamily. The first bacterial exporter to be identified was the transporter for the a-hemolysin secreted by uropathogenic E. coli (reviewed in [2]; see also Wandersman, 1996), and this system remains the paradigm. The Type I systems secrete proteins and other substrates via a mechanism in which transfer across both the inner and outer membranes occurs in a single step [126, 127] (Fig. 4, see color plate). The most surprising feature of these systems is that secretion across both the inner and outer membranes requires only three proteins. These components include an ABC transporter, a membrane fusion protein, or accessory factor, and an outer membrane component. The secreted protein, while still in the cytoplasm, is recognized by and binds to the ABC protein via a carboxy-terminal signal. Substrate binding inhibits the ATPase activity of the ABC protein [128] and promotes interaction of the ABC protein with the membrane fusion protein. This interaction then stimulates interaction of the membrane fusion protein with the outer membrane component, so that the secretion apparatus is assembled in an ordered fashion [129,130]. The mechanism by which the substrate passes from the cytoplasm to the environment is not clear. The most likely scenario is that the complex of the ABC protein, membrane fusion protein, and outer membrane factor forms a channel through which the proteins pass, in one step, from the cytoplasm to the external medium. This model, of course, depends on the pore-forming ability of the outer membrane factor, which is still under debate.

2. GERM WARFARE

53

The ABC transporter—HlyB, in the case of E. coli a-hemolysin—contains an ATP-binding domain that is homologous to ATP-binding cassettes found in many types of proteins in organisms from bacteria to mammals [131, 132]. These ATP-binding cassettes are absolutely required for secretion; mutations that block ATPase activity also block hemolysin secretion [128, 133]. These transporter proteins reside in the inner membrane [134] with the ABC cassette in the cytoplasmic carboxy-terminal domain of the protein. The transmembrane domain of HlyB does not perform any other functions other than the anchoring of HlyB. Replacement of the HlyB transmembrane domain with another unrelated domain does not affect secretion of hemolysin [135]. Several lines of evidence demonstrate that the ABC transporter proteins are also responsible for substrate specificity. Hybrid ABC transporter systems comprised of components from different bacteria tend to secrete ABC cognates [129]. In addition, mutations in the signal sequence of a-hemolysin have been used to identify compensatory mutations in the cytoplasmic domain of HlyB that alter the substrate specificity of the system [136]. Indeed, it is the recognition and binding of the substrate by the ABC transporter that promotes interactions between the ABC transporter and the other two components [130]. The second component of the Type I secretion systems is known as the membrane fusion protein, or accessory factor [2, 137]. These proteins share a common predicted topology consisting of an amino-terminal domain anchored in the inner membrane, a periplasm-spanning domain, and a (3-barrel carboxy-terminal domain that is thought to span the outer membrane [138, 139]. The predicted topology of these proteins is supported by fractionation studies. HlyD fractionates to both the inner and outer membranes, while truncated HlyD, missing its carboxy-terminal 10 amino acids, fractionates only to the inner membrane [140]. The third component of the a-hemolysin-like systems is the outer membrane factor, which in the a-hemolysin system is TolC. TolC is unusual because the structural gene is unlinked, and the protein performs other important functions in E. coli [141, 142]. Demonstration of TolC involvement in a-hemolysin secretion was an important advance that clarified a very confusing situation [143]. TolC and other members of the membrane fusion protein family are thought to form a pore in the outer membrane through which the secreted substrates pass into the external milieu, but the evidence for their pore-forming abilities is limited [144-146]. Structural characterization of TolC has demonstrated that the protein exists as a trimer in the outer membrane with extensive (3-barrel structures [145, 146]. The carboxy-terminal domain of TolC protrudes into the periplasm and most likely contacts the other components, thus providing a bridge between the inner and outer membranes [146]. Proteins secreted by the ABC transporters, of course, lack signal sequences [147, 148]. The first insight into the nature of the signal was that the carboxy-terminal 23-kDa fragment of the 107-kDa HlyA is secreted [126, 149], Gene fusion

54


approaches further dehneated the portion of the carboxy terminus that is necessary and sufficient for hemolysin secretion to the last 48-60 amino acids [150-155]. Other members of this family show similar carboxy-terminal signal peptides [156-159]. Attempts have been made to identify individual amino acids within this signal, but none emerged [160,161]. In addition to possible "contact" residues [160], these carboxy-terminal sequences may have other structural features that are important in the recognition of the secretion machinery. An extended amphipathic helix as well as other structural features is found in the carboxy-terminal regions of many of the proteins secreted via Type I mechanisms [155]. Yet another region just upstream of the carboxy terminus may play a role in signal recognition as well. This region, which consists of a glycine-rich GGXGXD sequence that is repeated up to 36 times, is found in all types of cytotoxic proteins secreted by the ABC transporters [162]. This repeated sequence is clearly important for secretion, and may function as an internal chaperone to keep the secretion signal exposed [163, 164]. The cloning and sequencing of the chromosomal hemolysin gene cluster helped establish the novelty of the secretion system for a-hemolysin [165]. The hemolysin, HlyA, has no signal peptide, indicating that secretion is independent of the Sec machinery [147], and the sequence of HlyB revealed features of an ABC transporter [2, 138, 166]. As noted above, the third component of the a-hemolysin secretion system, the outer membrane protein TolC, was not identified until several years later [143]. In the meantime, other proteins from numerous bacteria were found that are secreted by similar systems. For example, the hemolysins from Proteus vulgaris, Proteus mirabilis, and Morganella morganii have gene clusters homologous to the E. coli hly cluster, and mutations in the secretion factors for these hemolysins can be complemented by the homologous E. coli genes [121]. Type I systems are also involved in the secretion of Pasteurella haemolytica leukotoxin [167], Erwinia chrysanthemi proteases [125], and Serratia marcescens Has A heme-binding protein [168]. The presence of a large number of Type I secretion systems that are highly homologous to one another has expedited the characterization of these systems. Characterization of other Type I systems such as the Has system in Serratia marcescens has provided important new information. HasA is an extracellular heme-binding protein. Perhaps the most surprising result is that interfering with SecB [169] can prevent HasA secretion. This provides the first example of SecB function in a different secretion system, and it suggests that proteins secreted by Type I systems must be maintained in an unfolded state. Further study of Type I systems may provide a means to attenuate the virulence of organisms that possess them. In addition, these systems have been demonstrated to be an efficient method of delivering antigens, such as the p67 sporozoite antigen of Theileria parva, for vaccination purposes [170-172].

2.

GERM WARFARE

55

W. Conjugal Transfer Systems: Type IV Type IV secretion systems are used by the plant pathogen Agwbacterium tumefaciens for the transfer of oncogenic T-DNA and proteins to plants, and certain Type IV components are used by Bordetella pertussis for the secretion of pertussis toxin (reviewed in [173]). Based on sequence homologies, these secretion systems appear to be novel adaptations of the conjugal transfer system used for the horizontal transfer of plasmids. Identification of a similar type of system in Legionella pneumophila [174, 175] and of homologs in the pathogenicity island of Helicobacter pylori (reviewed in [176-178]) demonstrate that the Type IV systems may be more widespread than originally thought. The conjugal transfer systems in Gram-negative bacteria, which include the F plasmids and the IncN plasmids, are well characterized. These plasmids carry the tra genes, which are required for plasmid transfer into another bacterium. Over the years, it has become clear that the mechanism of T-DNA transfer in A. tumefaciens is quite similar to the mechanism of conjugal plasmid transfer in Gram-negative bacteria (reviewed in [173]). The ways in which the DNA is processed prior to transfer are similar [179-181]. Finally, sequence analysis has demonstrated that the virB genes from A. tumefaciens show striking similarity to the tra genes from the IncN, IncP, and other conjugative plasmids [182-184] and that the tra genes most likely encode the proteins that form the DNA-channeling pore in the outer membrane [185, 186]. It is reasonable that a system for transferring DNA between bacteria might evolve into a system for delivering bacterial DNA into host cells. More surprising, however, was the discovery that these conjugal transfer systems have also been adapted to promote the secretion of proteins, such as the Bordetella pertussis toxin. This finding makes more sense, though, if we look at DNA transport systems as systems that are designed to transport proteins. The DNA merely "hitches" a ride with the protein as the protein gets transported into the recipient cell, regardless of whether the recipient is another bacterium or a plant or animal host cell. Evidence for this model comes from early experiments that demonstrated that some A. tumefaciens proteins can be transported by themselves into plant cells. At least two proteins, VirE2 and VirD2, are secreted along with the T-DNA as a complex, the T-complex, into the recipient plant cells. Mixed infections of plant cells with one bacterial strain deleted for T-DNA and one mutant for the virulence protein VirE2 result in tumorigenesis in the host plant, indicating that the protein can be transported into the plant in the absence of the T-DNA [187]. This tumorigenesis is, however, dependent on the presence of the virB operon, which encodes the transport apparatus, and VirD2, which binds the T-DNA. These results indicate that the VirE2 protein forms complexes with the T-DNA once both are in the plant. This is further supported by the observation that plants with a virEl transgene undergo tumorigenesis when infected with a

56


normally avirulent virE2 mutant strain [188]. In addition, VirF, which is also required for virulence, has been shown by similar experiments to be exported into plant cells independent of T-DNA [189]. The mechanism of protein transport from Agrobacterium tumefaciens is not yet clear. VirD4 and 11 different proteins encoded by the virB operon are required for transport of the T-complex and are proposed to form a pilus-like transport apparatus through which the proteins and DNA can pass (Fig. 5, see color plate). The VirB proteins have been fairly well characterized as to their subcellular locations (reviewed in [176]). The proteins appear to form a macromolecular complex that traverses the entire bacterial envelope, leading to the exterior of the cell. This huge complex is most likely a channel that allows secreted proteins to pass from the cytoplasm to the surface of the cell. Starting from the outside, this macromolecular complex is comprised of VirB9 and its companion lipoprotein VirB7. These two proteins form the bulk of the structure that spans the outer membrane and periplasm. VirB6, which is the main integral inner membrane protein, and several other proteins, VirB3, VirB8, and VirB 10, which are associated with the inner membrane, are proposed to form the channel through which the T-complex passes across the inner membrane. In addition, three ATPases, VirD4, VirB 11, and VirB4, are associated with the inner membrane and provide energy for both assembly of the transport complex and secretion of the T-complex. The final component, the VirBl protein, resides in at least two locations in the cell. First, it is located in the large complex between the two membranes. Second, it is cleaved and exported to the surface. VirBl has homology to transglycosylases and has been proposed to bore holes in the peptidoglycan layer and allow pilus assembly to occur. For quite some time, the idea of a pilus-like structure, much like the conjugative F pili encoded by a subset of the tra genes, has been favored as a mechanism of T-complex transport from the bacterial surface to the interior of the plant cell. Until recendy, however, there was no evidence to support this hypothesis. Lai and Kado have now shown that VirB2 actually does form pilus-like structures [190], with VirB5 suggested to be a minor pilin subunit. The signal that directs the VirE2, VirD2, and VirF proteins to the secretion apparatus is not currently known. These proteins do not contain the amino-terminal signal sequences that are recognized by the Sec machinery. Thus, the Type IV systems appear to be Sec-independent. However, pertussis toxin subunits, which may be secreted by a similar mechanism, are synthesized with classical signal sequences and are secreted to the periplasm by Sec-dependent mechanisms. Thus, while it is true that the B. pertussis Ptl proteins, which comprise the pertussis toxin secretion apparatus, are homologous to the Tra and Vir proteins of conjugal plasmid and T-DNA transfer systems, the actual mechanisms of toxin secretion and DNA secretion may be very different. Indeed, some of the proteins that are absolutely required for DNA transfer in the Tra and Vir systems are absent from

2. GERM WARFARE

57

the Ptl system. The Ptl proteins are probably responsible for secreting the pertussis toxin from the periplasm, while the Tra and Vir proteins are transferring molecules from the cytoplasm [173]. Perhaps characterization of the Legionella and Helicobacter Type IV systems will provide answers to this dilemma.

VIL Contact-Dependent Secretion: Type III Like the Type I systems, the Type III systems secrete proteins via a Sec-independent mechanism. However, these systems are much more complicated than the three-component ABC transporter systems. The Type III systems, which are sometimes called contact-dependent secretion systems, allow not only secretion but also injection of virulence factors directly into the cytosol of eukaryotic host cells. Type III systems are known in animal pathogens such as Yersinia pestis and other Yersiniae, Shigella flexneri, and Salmonella typhinnirium, as well as plant pathogens such as Pseudomonas syringae and Erwinia species, and they will likely be found elsewhere. Since the Yersinia Ysc system is the best characterized of these systems, it will be our primary focus. For more detailed discussions see [191]. Yersinia secretes a collection of virulence proteins that are inappropriately called Yops (yersinia outer proteins) for historical reasons. While these Yops were known to be essential for virulence for some time, their site of action has been elusive for several reasons (reviewed in [192]). For one, the Yops are not secreted under conditions in which Ca^^ is present at the millimolar levels; the absence of Ca^+ stimulates Yops secretion into the media. This observation was exploited to identify mutants that are secretion defective, and the genes identified were called ysc (Yop secretion) [193]. The Ysc proteins mediate Yop secretion in a Sec-independent manner despite the presence of YscC, a homolog of the Type II outer membrane secretin PulD [193]. Homologs of the Ysc proteins have been found in many pathogenic bacteria such as Salmonella typhimurium (the Inv/Spa proteins). Shigella flexneri (the Mxi/Spa proteins), and Pseudomonas solanacearum (the Hrp proteins) [194-199]. This is not the whole picture, though. Another set of observations complicates the story. The secreted Yops aggregate terribly in the extracellular medium, and adding them direcdy to eukaryotic cells has no cytotoxic effects. The real breakthrough in characterization of Yersinia secretion came with the observation that the cytotoxic effect of YopE is dependent on the presence of YopD protein. However, this requirement for YopD can be circumvented by direct microinjection of YopE into HeLa cells [200]. Yops must require both secretion out of the bacteria and injection or translocation into the host cell to exert their toxicity. This

58


translocation of Yops is mediated by a subset of the Yops, YopD and YopB. This was further supported by two key experiments. In the first, YopE was shown by microscopy to be injected into the host cell in a YopD-dependent manner [201]. Second, a reporter gene (Cya) was fused to YopE, and the enzymatic activity of this hybrid protein was used to demonstrate YopB- and YopD-dependent injection into host cells [202]. Thus, the process of Yop injection occurs in two steps: secretion out of the bacterium and translocation into the eukaryotic host. These two steps require greater than 20 proteins to transfer toxins direcdy from pathogen to host (Fig. 6, see color plate). This direct transfer requires that the cell coregulate secretion and translocation. The bacterial cell senses contact with the host cell via the YopN protein. YopN (LcrE) appears to remain on the surface of the bacterial cell, where it plugs the secretion pore [201, 203, 204]. The TyeA protein, identified in 1998, may also play a role similar to that of YopN; it also localizes to the bacterial cell surface and seems to function in the negative regulation of secretion [205]. When the bacterium contacts a host cell or Ca^"*^ is removed, the secretion channel is unplugged. In Yersiniae, this allows the secretion of factors that negatively regulate the expression of Yops and toxin synthesis commences. It seems that Yersinia might also have a plug on the cytoplasmic side of the secretion apparatus. LcrG may function as a cytoplasmic YopN, with LcrV acting as its inhibitor [206, 207]. In other systems, such as Shigella Ipa secretion, the secreted toxins accumulate in the bacterial cytoplasm prior to the stimulation of secretion by contact with host cells. In the absence of Ca^"^, the fully folded Yops are secreted from the bacterial cytoplasm to the exterior of the cell, where they are retained on the surface or released into the external environment. Secretion probably occurs through a channel-like complex that traverses the entire bacterial envelope. LcrD and its homologs in other organisms function as the channel through the inner membrane [191, 208]. As mentioned earlier, YscC and related proteins show homology to the outer membrane secretins (PulD) from the Type II systems [193]. They localize to the outer membrane [197, 209] and form ringlike multimers through which the Yops could pass [209]. YscJ is a lipoprotein [193] and may function to bridge between the inner and outer membranes. Passage through the inner and outer membranes is likely energized by the cytoplasmic ATPase YscN [210]. Despite this progress, there remain many essential Ysc proteins about which little is known. YscD, J, Q, R, S, T, and U are inner membrane proteins that show homology to the components of the flagellar export apparatus [191, 193, 211]. Accordingly, these proteins, as well as YscO and YscG, are probably part of a macromolecular machine that spans the bacterial envelope [193, 211, 212]. Little is known about the components YscF, I, K, L, and P, except that the F, I, K, and L proteins may be cytoplasmic [191]. The next step, translocation, is facilitated by the formation of a translocator through which the effector Yops are injected into the eukaryotic cell. YopB and

2. GERM WARFARE

59

YopD are thought to form a channel through which the effector Yops can enter the host cytosol. Indeed, these proteins look like transmembrane proteins in contrast to most Yops, which are soluble proteins [213]. Furthermore, YopB has been shown to exhibit a pore-forming hemolytic activity consistent with its role as the pore [214]. More recent evidence, however, suggests that YopD is translocated into the cytosol of the host, which seems inconsistent with its role in forming the translocation pore [215]; the role for YopD in the translocation is, therefore, unclear at this time. LcrV, mentioned earlier as having a role in regulation of secretion, may also have an important role in the translocation process. LcrV has been shown to facilitate the secretion of YopB and YopD and, therefore, may be important for assembhng the translocation machinery [207, 216]. Similarly, TyeA appears to have dual roles since it has been found to participate in translocation in a subset of the Yops [205]. Another factor that is important in translocation of proteins into the host cytosol is YopK. Mutants defective in yopK exhibit an increased cytotoxicity resulting from increased levels of translocated YopE, while overexpression of YopK causes decreased translocation [217]. YopK is not translocated but remains associated with the bacteria from which it is secreted, which supports the hypothesis that YopK acts as a modulator of translocation by negatively controlling the size of the pore through an association with the pore-forming proteins [217]. Another group of proteins important in translocation of Yops into host cells is the chaperones (reviewed in [218]). These chaperones, several of which have been identified, are bacterial cytoplasmic proteins that demonstrate specificity for one or two substrates and are usually named Syc for specific Yop chaperone. SycE, also called YerA, acts as a chaperone for YopE; it binds but does not target YopE to the secretion machinery in the inner membrane [219, 220]. Unlike general chaperones, SycE and the other Syc chaperones bind to a specific sequence on the target protein, and these sequences coincide with the domain required for efficient translocation into eukaryotic cells [221]. The implication is that by binding to these translocation domains the Syc proteins prevent premature interaction of their respective Yops with components of the translocator, YopB and YopD [221]. The Type III chaperones include SycE, SycH (for YopH), SycD (for YopD), and YscB (for YopN) [218, 222]. In Salmonella typhimurium, the Type III secretion system involved in eukaryotic cell invasion has been isolated and visualized [223]. These "needle complexes" clearly resemble the flagellar basal body, and this strengthens the idea that the two systems have a common ancestor [223]. Undoubtedly, the isolation of similar macromolecular structures from organisms with other Type III systems will be forthcoming. Yops were isolated from culture supernatants in the early 1980s [224-226], and later work revealed the genes encoding the Yops and hinted at the mechanism of their regulation [227, 228]. Further examination of the sequence information

60


available for the Yops showed that these proteins do not contain typical Sec-dependent amino-terminal signal sequences, nor do they contain the carboxyterminal signals required for secretion by the Sec-independent Type I secretion systems, which was the only known Sec-independent secretion mechanism at the time [229-232]. The use of gene fusion technology narrowed down the Type III signal sequence to the amino-terminal 48 amino acids of the YopH protein; this sequence, fused to a reporter protein, directed the secretion of the reporter by the Yersinia export system [232]. Later studies with other yo/?//and v6>/7£'gene fusions located the signal for secretion in the amino-terminal 11-17 amino acids of the Yops, while the signal for translocation into the eukaryotic cytoplasm requires at least the first 50-70 amino acids [233, 234]. This suggests that the Yops contain several domains—a secretion domain, a translocation domain, and an effector domain [233]. Surprisingly, the secretion domains show very little similarity at the amino acid level [231, 233, 234]. A gene fusion approach reported in 1997 provided insight into the signal recognized by the Ysc secretion machinery [235]. As expected, hybrid proteins containing the first 15 amino acids of YopN fused to neomycin phosphotransferase are efficiendy secreted. Single point mutations that alter any of the YopN amino acids cause no significant reduction in the secretion of the fusion protein. Very unexpected, however, was the observation that frameshift mutations, which alter the entire YopN sequence, have little effect either. These results imply that the signal for Yop secretion lies in the mRNA and that Yop secretion may be coupled with translation [235]. Future work is necessary to determine the nature of the mRNA signal for Yop secretion and to identify the factors that recognize this signal. However, these striking findings present a paradox. Remember that many of the Yops require a cytoplasmic Syc chaperone to be translocated. If, in fact, Yop secretion occurs cotranslationally, there should be no protein in the cytoplasm for the Syc proteins to bind. One possible explanation for these two very different signals is that perhaps Yersinia switches between the signals during different stages of pathogenicity. The mRNA signal facilitates YopE secretion under conditions of Ca^"^ deprivation, yet YopE, in addition to having its first 15 codons intact, must bind to its cognate chaperone, SycE, for translocation into eukaryotic cells [236]. Perhaps secretion into the medium requires different signals than translocation, and Yersinia uses different modes of substrate recognition for these processes. It will be interesting to see if knowledge gained from the Yersinia system applies generally to all Type III systems. For example, is the mRNA signal universal? Does this apply even to flagellar subunits? Certain mechanistic steps must be conserved because selected components from one system can be exchanged with the homologous components from another system [237-240]. Obviously, the characterization of these secretion systems have important clinical applications, but these systems, which so efficiendy inject proteins, also have important biotechnological implications for the delivery of proteins to eukaryodc

2. GERM WARFARE

61

cells, such as their utility as antigen delivery systems for vaccine development [241].

VIIL Concluding Remarks At first glance, the five mechanisms for protein secretion that are conserved among Gram-negative bacteria appear quite diverse. Some work independently of the Sec apparatus—Types 11 and V—and some do not—Types I, III, and IV. The various systems require anywhere from one special component in the case of the autotransporters, to 20 or more components, in the case of the contact-dependent secretion systems. Despite this diversity, however, at least two major themes are emerging. First, whether it is a secretin such as PulD or other proteins such as the P-protein domain of the autotransporters, these secretion systems all possess a putative pore-forming protein that allows passage of the secreted substrate through the outer membrane. Mounting evidence supports a channel for the signal-sequence directed passage of proteins through the bacterial cytoplasmic membrane (SecYEG) or the endoplasmic reticular membrane of eukaryotic cells (the Sec61p complex), and YopBD seems to form a channel that allows Yop passage into the eukaryotic cytosol. Mother Nature seems to have a favored solution to the barrier problem posed by lipid bilayers. Second, we predict that the multiple components of complex secretion systems form a macromolecular machine that spans the Gram-negative cell envelope, and that such structures will likely resemble surface structures such as pili or flagella. This is most strikingly demonstrated by the Type III secretion system in Salmonella. In this case "needle complexes" have been isolated, visualized, and shown to resemble the flagellar basal body. We anticipate that similar techniques applied to the other systems will be fruitful as well.

References 1. PohlSchroder, M., Prinz. W. A.. Hartmann. E.. and Beckwith. J. (1997). Protein translocation in the three domains of life: Variations on a theme. Cell 91, 563-566. 2. Path, M. J., and Kolter, R. (1993). ABC Transporters: Bacterial exporters. Microbiol. Rev. SI, 995-1017. 3. Thanassi, D. G., Saulino, E. T., and Hultgren, S. J. (1998). The chaperone/usher pathway: A major terminal branch of the general secretory pathway. Ciirr. Opiii. Microbiol. 1, 223-231. 4. Blobel, G., and Dobberstein, B. (1975). Transfer of proteins across membranes, I: Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membranebound ribosomes of murine myeloma. J. Cell. Biol. 67. 835-851. 5. Bieker, K. L., Phillips, G. J., and Silhavy, T. J. (1990). The sec and prl genes of Escherichia coli. J. Bioenerg. Biomemb. 22, 291-310. 6. Emr, S. D., Schwartz, M., and Silhavy. T. J. (1978). Mutations altering the cellular localization of the phage lambda receptor, an Escherichia coli outer membrane protein. Proc. Natl. Acad. Sci. U.S.A. 75, 5802-5806.

62


7. von Heijne, G. (1985). Signal sequences. The limits of variation. J. Mol Biol 184, 99-105. 8. Fikes, J. D., and Bassford, P. J. J. (1987). Export of unprocessed precursor maltose-binding protein to the periplasm of Escherichia coli cells. J. Bacteriol 169, 2352-2359. 9. Emr, S. D., Hanley-Way, S., and Silhavy, T. J. (1981). Suppressor mutations that restore export of a protein with a defective signal sequence. Cell 23, 79-88. 10. Oliver, D., and Beckwith, J. (1981). E. coli mutant pleiotropically defective in the export of secreted proteins. Cell 25, 2765-2772. 11. Fikes, J. D., and Bassford, P. J. J. (1989). Novel secA alleles improve export of maltose-binding protein synthesized with a defective signal peptide. J. Bacteriol. 171, 402^09. 12. Riggs, P. D., Derman, A. I., and Beckwith, J. (1988). A mutation affecting the regulation of a secA-lacZ fusion defines a new sec gene. Genetics 118, 571-579. 13. Stader, J., Gansheroff, L. J., and Silhavy, T. J. (1989). New suppressors of signal-sequence mutations, prlG, are linked tightly to the secE gene of Escherichia coli. Genes Dev. 3, 1045-1052. 14. Ito, K., Wittekind, M., Nomura, M., Shiba, K., Yura, T, Miura, A., and Nashimoto, H. (1983). A temperature-sensitive mutant of E. coli exhibiting slow processing of exported proteins. Cell 32, 789-797. 15. Akimaru, J., Matsuyama, S., Tokuda, H., and Mizushima, S. (1991). Reconstitution of a protein translocation system containing purified SecY, SecE, and SecA from Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 88, 6545-6549. 16. Brundage, L., Hendrick, J. P, Schiebel, E., Driessen, A. J. M., and Wickner, W. (1990). The purified E. coli integral membrane protein SecY/E is sufficient for reconstitution of SecA-dependent precursor protein translocation. Cell 62, 649-657. 17. Kumamoto, C. A., and Beckwith, J. (1983). Mutations in a new gene, secB, cause defective protein localization in Escherichia coli. J. Bacteriol. 154, 253-260. 18. Gardel, C., Benson, S., Hunt, J., Michaelis, S., and Beckwith, J. (1987). secD, a new gene involved in protein export in Escherichia coli. J. Bacteriol. 169, 1286-1290. 19. Gardel, C., Johnson, K., Jacq, A., and Beckwith, J. (1990). The secD locus of E. coli codes for two membrane proteins required for protein export. EMBO J. 9, 3209-3216. 20. Nishiyama, K. I., Hanada, M., and Tokuda, H. (1994). Disruption of the gene encoding pl2(SecG) reveals the direct involvement and important function of SecG in the protein translocation of Escherichia coli at low temperature. EMBO J. 13, 3272-3277. 21. Duong, P., and Wickner, W. (1997). Distinct catalytic roles of the Sec YE, SecG, and SecDFyajC subunits of the preprotein translocase. EMBO J. 16, 2756-2768. 22. Bost, S., and Belin, D. (1997). prl mutations in the Escherichia coli secG gene. J. Biol. Chem. 272, 4087^093. 23. Duong, R, Eichler, J., Price, A., Leonard, M. R., and Wickner, W. (1997). Biogenesis of the Gram-negative bacterial envelope. Cell 91, 567-573. 24. Cunningham, K., and Wickner, W. (1989). Specific recognition of the leader region of precursor proteins is required for the activation of translocation ATPase of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 86, 8630-8634. 25. Gannon, P. M., Li, P., and Kumamoto, C. A. (1989). The mature portion of Escherichia coli maltose-binding protein (MBP) determines the dependence of MBP on SecB for export. J. Bacteriol. 171, 813-818. 26. Randall, L. L., Topping, T. B., and Hardy, S. J. S. (1990). No specific recognition of leader peptide by SecB, a chaperone involved in protein export. Science 248, 860-863. 27. Collier, D. N., Bankaitis, V., Weiss, J. B., and Bassford, P J. (1988). The antifolding activity of SecB promotes the export of the E. coli maltose-binding protein. Cell 53, 273-283.

2.

GERM WARFARE

63

28. Kusters, R., deVrije, T., Breukink, E., and deKruijff, B. (1989). SecB stabilizes a translocationcompetent state of purified prePhoE protein. / Biol. Chem. 264, 20827-20830. 29. Lecker, S. H., Driessen, A. J. M., and Wickner. W. (1990). ProOmpA contains secondary and tertiary structure prior to translocation and is shielded from aggregation by association with SecB protein. EMBO J. 9, 2309-2314. 30. Randall, L. L., Topping, T. B., and Hardy. S. J. S. (1994). The basis of recognition of non-native structure by the chaperone SecB. In "The Biology of Heat Shock Proteins and Molecular Chaperones" (R. I. Morimoto, A. Tissieres. and C. Georgopoulos, eds.), pp. 285-298. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 31. Smith, V. R, Hardy, S. J., and Randall, L. L. (1997). Determination of the binding frame of the chaperone SecB within the physiological ligand oligopeptide-binding protein. Protein Sci. 6, 1746-1755. 32. Hartl, E-U., Lecker, S., Schiebel, E., Hendrick, J. P, and Wickner, W. (1990). The binding cascade of SecB to SecA to SecY/E mediates preprotein targeting to the E. coli membrane. Cell 63, 269-279. 33. den Blaauwen, T, Terpetschnig, E., Lakowicz, J. R., and Driessen, A. J. (1997). Interaction of SecB with soluble SecA. FEBS Lett. 416. 35-38. 34. Driessen, A. J. (1993). SecA, the peripheral subunit of the Escherichia coli presursor protein translocase, is functional as a dimer. Biochemistry 32. 13190-13197. 35. Hanein, D., Matlack, K. E., Jungnickel. B.. Plath, K., Kalies, K. U., Miller, K. R., Rapoport, T. A., and Akey, C. W. (1996). Oligomeric rings of the Sec61p complex induced by ligands required for protein translocation. Cell 87, 721-732. 36. Hamman, B. D., Chen, J. C, Johnson, E. E., and Johnson, A. E. (1997). The aqueous pore through the translocon has a diameter of 40-60 A during cotranslational protein translocation at the ER membrane. Cell 89, 535-544. 37. Beckmann, R., Bubeck, D., Grassucci, R.. Penczek, P., Verschoor, A., Blobel, G., and Frank, J. (1997). Alignment of conduits for the nascent polypeptide chains in the ribosome-Sec61 complex. Science 278, 2123-2126. 38. Economou, A., and Wickner, W. (1994). SecA promotes preprotein translocation by undergoing ATP-driven cycles of membrane insertion and deinsertion. Cell 78, 835-843. 39. Eekkes, P., van der Does, C , and Driessen. A. J. (1997). The molecular chaperone SecB is released from the carboxy terminus of SecA during initiation of precursor protein translocation. EMBO J. 16,6105-6113. 40. Duong, E, and Wickner, W. (1997). The SecDFyajC domain of preprotein translocase controls preprotein movement by regulating SecA membrane cycling. EMBO J. 16, 4871-4879. 41. Driessen, A. J., Eekkes, P, and van der Wolk. J. P W. (1998). The Sec system. Cum Opin. Microbiol. 1, 216-222. 42. Economou, A. (1998). Bacterial preprotein translocase: Mechanism and conformational dynamics of a processive enzyme. Mol. Microbiol. 27, 511-518. 43. Powers, T, and Walter, P. (1997). Co-translational protein targeting catalyzed by the Escherichia coli signal recognition particle and its receptor. EMBO J. 16, 4880^886. 44. Kumamoto, C. A., and Beckwith, J. (1985). Evidence for specificity at an early step in protein export in Escherichia coli. J. Bacteriol. 163, 267-274. 45. Phillips, G. J., and Silhavy, T. J. (1992). The E. coli ffh gene is necessary for viability and efficient protein export. Nature 359, 144-146. 46. de Gier, J. W., Mansournia, P., Valent, Q. A., Phillips, G. J., Luirink, J., and von Heijne, G. (1996). Assembly of a cytoplasmic membrane protein in Escherichia coli is dependent on the signal recognition particle. FEBS Lett. 399, 307-309.

64


47. Ulbrandt, N. D., Newitt, J. A., and Bernstein, H. D. (1997). The E. coli signal recognition particle is required for the insertion of a subset of inner membrane proteins. Cell 88, 187-196. 48. Valent, Q. A., Scotti, P. A., High, S., de Gier, J. W., von Heijne, G., Lentzen, G., Wintermeyer, W., Oudega, B., and Luirink, J. (1998). The Escherichia coli SRP and SecB targeting pathways converge at the translocon. EMBO J. 17, 2504-2512. 49. Klauser, T, Pohlner, J., and Meyer, T. F. (1993). The secretion pathway of IgA protease-type proteins in Gram-negative bacteria. Bioessays 15, 799-805. 50. Koomey, J. M., and Falkow, S. (1984). Nucleotide sequence homology between the immunoglobulin Al protease genes of Neisseria gonorrhoeae. Neisseria meningitidis, and Haemophilus influenzae. Infect. Immun. 43, 101-107. 51. Poulsen, K., Brandt, J., Hjorth, J. R, Thogersen, H. C , and Kilian, M. (1989). Cloning and sequencing of the immunoglobulin Al protease gene (iga) of Haemophilus influenzae serotype b. Infect. Immun. 57, 3097-3105. 52. Yanagida, N., Uozimi, T, and Beppu, T. (1986). Specific excretion of Serratia marcescens protease through the outer membrane of Escherichia coli. J. Bacteriol. 166, 937-944. 53. Schmitt, W., and Haas, R. (1994). Genetic analysis of the Helicobacter pylori vacuolating cytotoxin: Structural similarities with the IgA protease type of exported protein. Mol. Microbiol. 12,307-319. 54. Henderson, I. R., Navarro-Garcia, P., and Nataro, J. P. (1998). The great escape: Structure and function of the autotransporter proteins. Trends Microbiol. 6, 370-378. 55. Pohlner, J., Halter, R., and Meyer, T. F. (1987). Neisseria gonorrhoeae IgA protease. Secretion and implications for pathogenesis. Antonie Van Leeuwenhoek 53, 479^84. 56. Klauser, T., Kramer, J., Otzelberger, K., Pohlner, J., and Meyer, T. F. (1993). Characterization of the Neisseria IgA beta-core: The essential unit for outer membrane targeting and extracellular protein secretion. J. Mol. Biol. 234, 579-593. 57. Pohlner, J., Halter, R., Beyreuther, K., and Meyer, T F. (1987). Gene structure and extracellular secretion on Neisseria gonorrhoeae IgA protease. Nature 325, 458^62. 58. Halter, R., Pohlner, J., and Meyer, T F. (1984). IgA protease of Neisseria gonorrhoeae: Isolation and characterization of the gene and its extracellular product. EMBO J. 3, 1595-1601. 59. Klauser, T., Pohlner, J., and Meyer, T. F. (1992). Selective extracellular release of cholera toxin B subunit by Escherichia coli: Dissection of Neisseria IgA beta-mediated outer membrane transport. EMBO J. 11, 2327-2335. 60. Jose, J., Jahnig, F, and Meyer, T. F. (1995). Common structural features of IgAl protease-like outer membrane protein autotransporters. Mol. Microbiol. 18, 378-380. 61. Ohnishi, Y., and Horinouchi, S. (1996). Extracellular production of a Serratia marcescens serine protease in Escherichia coli. Biosci. Biotechnol. Biochem. 60, 1551-1558. 62. Strauss, A., Pohlner, J., Klauser, T, and Meyer, T. F. (1995). C-terminal glycine-histidine tagging of the outer membrane protein IgA beta of Neisseria gonorrhoeae. FEMS Microbiol. Lett. Ill, 249-254. 63. Pugsley, A. P., Chapon, C , and Schwartz, M. (1986). Extracellular puUulanase of Klebsiella pneumoniae is a lipoprotein. J. Bacteriol. 166, 1083-1088. 64. He, S. Y, Lindeberg, M., Chatterjee, A. K., and Collmer, A. (1991). Cloned Erwinia chrysanthemi out genes enable Escherichia coli to selectively secrete a diverse family of heterologous proteins to its milieu. Proc. Natl. Acad. Sci. U.S.A. 88, 1079-1083. 65. Reeves, P. J., Whitcombe, D., Wharam, S., Gibson, M., Allison, G., Bunce, N., Barallon, R., Douglas, P, Mulholland, V., Stevens, S., Walker, D., and Salmond, G. P C. (1993). Molecular cloning and characterization of 13 out genes from Erwinia carotovora subspecies carotovora: Genes encoding members of a general secretion pathway (GSP) widespread in gram-negative bacteria. Mol. Microbiol. 8, 443-456.

2.

GERM WARFARE

65

66. Filloux, A., Bally, M., Ball. G., Akrim. M.. Tommassen. J., and Lazdunski, A. (1990). Protein secretion in Gram-negative bacteria: Transport across the outer membrane involves common mechanisms in different bacteria. EMBO J. 9, 4323^329. 67. Howard, S. P., Critch, J., and Bedi. A. (1993). Isolation and analysis of eight exe genes and their involvement in extracellular protein secretion and outer membrane assembly in Aewmonas hydrophila. J. Bacterial. 175, 6695-6703. 68. Dums, F., Dow, J. M., and Daniels, M. J. (1991). Structural characterization of protein secretion genes of the bacterial phytopathogenX

I

ddifN^ «ii€ or titort C nîdtte^ traiiscrt|itta«i tmUw binding site

promoter etement -polyC-

J

non^fttticlloiid promoter

Fig. 7/ Phase variation based on homopolymeric DNA tracts. Top: the spacer common to the promoters of the hi/A and hifB genes of H. influenzae contains a poly-TA tract. Loss of TA dinucleotide repeats results in a reduction in promoter efficiency or complete inactivation. Center: the coding region of the pilC gene of A^. gonorrhoeae includes a poly-G tract. Loss of a G residue from this tract results in a shift in the reading frame of the gene and premature termination of translation. Bottom: the spacer between the promoter and binding site for a positive regulator in iht fimS gene of Bordetella pertussis contains a poly-C tract. Loss of C residues from this tract alters the spacing between these elements and makes the promoter nonfunctional.

3.

REGULATION OF VIRULENCE GENE EXPRESSION IN BACTERIAL PATHOGENS

103

stretch of invertible chromosomal DNA. The promoter is required for transcription of the//mA gene, encoding the type 1 fimbrial subunit. The invertible element is flanked by 9-bp inverted repeats, and these form part of the binding sites for the site-specific recombinases that operate the switch. These proteins, FimB and FimE, are encoded by nearby genes {finiB and fimE, respectively), and are members of the integrase family of recombinases [179, 180]. They are very closely related, and show 48% identity in amino acid sequence. The FimB protein can invert the switch in both directions with approximately equal efficiency, whereas the FimE protein has a marked tendency to turn the switch off. Therefore, the switch is subject to directional bias, depending on which recombinase is acting at the inverted repeats at any time. The FimB and FimE recombinases do not act alone. They have requirements for two accessory proteins, IHF and LRP, both of which have been described earlier herein. The IHF protein binds to two sites, one in the switch and one just outside [181] (Fig. 8). The site within the switch is bifunctional since it is also required for operation of the fimA promoter. Without occupancy of this internal site by IHF, the switch ceases to function and the power of the promoter is reduced sevenfold [179]. The contribution of IHF to operation of the switch seems to be architectural; it probably promotes interactions between the inverted repeats and their associated recombinases that are favorable for recombination. A similar role has been described for LRP, which has three binding sites within the switch [182]. LRP interaction with the switch is enhanced by leucine, isoleucine, valine, and alanine, linking switch operation to nonpolar amino acid pools in the cell. The switch is also subject to modulation by H-NS. This protein binds to and represses the promoters of the recombinase genes and may also interact with the switch direcdy, retarding the rate of inversion. The latter possibility is consistent with the preference of H-NS for curved DNA since the fim switch includes DNA elements displaying a strong potential for curvature. In addition to the influences of two recombinases and three accessory proteins, the fim switch is also highly sensitive to the degree of supercoiling of its DNA. Departures from optimal levels of supercoiling can either inactivate the switch (too supercoiled), or cause it to become heavily biased toward one outcome (too relaxed) [183]. All of these influences serve to constrain the freedom of the switch to operate randomly. Instead, it seems to display several dependencies simultaneously. These may serve to direct the switch in ways that benefit the cell during its interacdons with a host, or while it is in a free-living state.

XIII. Pap Pilus Gene Transcription Pap pili assist uropathogenic E. coli in attachment to uroepithelial cells [184]. The pili are expressed at 37°C but not at 23°C, and are subject to phase variadon at the higher temperatures [185]. It has been suggested that low temperature is a cue that the bacteria are outside the host, and so it is unnecessary to express the pili [186]. The control of transcription and phase variation in the pap system is complex and multifactorial. Similar circuitry governs expression of the E. coli F1845 and S pili.

104

CHARLES J. DORMAN AND STEPHEN G. J. SMITH

encoded by the daa and sfa operons, respectively [187]. Phase variation is a function of the differential methylation by the Dam methylase of two GATC sites in the pap regulatory region called GATC-I and GATC-II [188] (Fig. 12). This methylation, involving just two basepairs, represents the modification to the genome that underlies phase variation in the pap system. In Phase-ON cells, GATC-I is unmethylated while GATC-II is methylated; the converse is true in Phase-OFF cells [ 189]. Methylation at these sites is blocked by the LRP protein. In Phase-OFF cells LRP binds cooperatively to three sites (LRP sites 1, 2, and 3) in the pap regulatory region and so blocks methylation of GATC-II (which is part of LRP site 2). In Phase-OFF cells, GATC-I remains available for methylation by Dam. Methylation of GATC-I reduces the affinity of LRP for sites 4 and 5. Failure of LRP to bind at sites 4 and 5 prevents/7(3/?^/transcription, locking cells in the OFF phase until after the next round of DNA replication [188, 190]. The Papl protein collaborates with LRP to block methylation of GATC-I. The proteins form a complex that has enhanced LRP affinity for sites 4 and 5 at GATC-I. Shifting LRP binding to GATC-I frees GATC-II for methylation by the Dam methylase. The LRP protein is a transcription activator when bound to GATC-I but is a repressor when bound to GATC-II [191, 192] (Fig. 12). Transcription of papl is activated by the PapB protein, and the cAMP-CRP complex activates both papB and papl, with cAMP-CRP playing the role of an

^ papB

C^ P papi

1

papA

pupH

binding siie

I •

4

LRP binding sites 5 6 1 2

GATC4

r impBA

3

I

GATC*II

Fig. 12 Phase-variable control oipap gene expression in E. coli. The pap regulatory sequences lying between the divergently transcribed papl and papB genes are shown. In the enlargement, the locations of the six binding sites for LRP, together with the cAMP-CRP-binding sites are shown. The two sites subject to Dam-mediated DNA methylation, GATC-I and GATC-II, are also illustrated, and these are associated with LRP sites 5 and 2, respectively. The promoters for the papl and papBA transcription units are at the borders of the enlarged area.

3.


105

antirepressor [193-195]. The nucleoid-associated protein H-NS is involved in thermoregulation of pap gene expression, where it acts as a repressor of transcription at low temperatures [196]. The binding of H-NS to the pap regulatory sequences containing the GATC-I and GATC-II sites prevents methylation, although the mechanism by which this occurs is not understood [186]. Thus, the regulation of pap gene expression occurs at many levels. On top of phase variation, expression of pap responds to the metabolic state of the cell (via cAMP-CRP) and to temperature via a mechanism that involves H-NS.

XIV. Contact'Dependent Gene Regulation The pap genes of uropathogenic E. coli have been implicated in a form of gene activation that is triggered by contact with the host. Induction requires the presence of the PapG adhesin protein that recognizes a specific carbohydrate receptor on uroepithelial cells, and the presence of this receptor. Among the targets of the induction is barA (or airS), a gene required for production of siderophores (iron-chelators) and their receptors, and without which E. coli cannot grow in urine. BarA/AirS shows homology to histidine protein kinases, raising the possibility that it senses Pap-mediated contact and transmits the signal to virulence genes via an unidentified response regulator [197]. In the human and rodent pathogens Yersinia pseudotuberculosis (which causes adenitis and septicaemia), Y. enterocolitica (gastrointestinal infections), and Y. pestis (bubonic plague), contact with host cells results in export of a bacterial transcriptional repressor, allowing previously repressed virulence genes on a plasmid to become activated [198, 199]. The repressor is the LcrQ protein (equivalent to the 57% identical YscMl and YscM2 proteins in K enterocolitica, encoded by closely related genes [200]), and is exported from the bacterial cytosol into the external medium via a plasmid-encoded type III secretion system that is triggered by host-microbe interaction. Plasmid-linked yop virulence genes that are subject to LcrQ repression are then transcribed. LcrQ is more accurately described as an anti-activator, since it inhibits the ability of the LcrF protein to activate yop gene transcription. (LcrF, called VirF in Y. enterocolitica, is an AraC-like transcription factor that responds to temperature.) The LcrQ protein has not been shown to bind to yop DNA, and it may exert its negative regulatory effects indirecdy [198, 201]. This regulatory mechanism is reminiscent of one employed in the control of flagellar biosynthesis in S. typhimurium. There, the counterpart of LcrQ is an anti-sigma factor called FlgM, which is secreted via the type III secretion system used to assemble flagellae and an integral component of the flagellum [202].

106


XV. The Virulence Gene Regulatory Cascade of S. flexneri S. flexneri causes bacillary dysentery in humans, and is a facultative intracellular pathogen. The bacteria enter and replicate within the cells of the colonic epithelium, and move between cells. The process involves M-cell entry, basolateral invasion of epithelia, actin-dependent intra- and intercellular movement, and macrophage killing. The genes required for expression of the pathogenic phenotype are located on a high-molecular-weight plasmid. They occupy 31 kb and consist of ipa genes coding for secreted proteins needed for invasion of host cells, the mxi and spa genes encoding components of a type III secretion system required for Ipa protein secretion [203, 204], and regulatory elements that govern the expression of the virulence factors (see below). The S. flexneri virulence gene regulatory system illustrates many points that are of general relevance in bacterial pathogens (Fig. 13). The key genes are grouped in operons under the collective control of a regulatory protein. The specific regulators and all of the structural genes are located on a plasmid, but chromosomally encoded proteins with wide-ranging effects in the cell also influence expression of the virulence genes. Gene expression is under tight control and responds to particular environmental influences. In the S. flexneri system, a considerable amount of molecular detail is available [60], and a very similar system is found in enteroinvasive E. coli (EIEC) [29, 205]. The control regime takes the form of a cascade, at the top of which is an AraC-like protein called VirF. This plasmid-encoded protein binds to and activates transcription of a gene coding for a subordinate regulator called VirB [206]. This protein is presumed to have DNA-binding activity (it shows homology to DNA-binding proteins involved in plasmid maintenance) and is required for transcriptional activation of the structural genes. These are organized in large, divergendy transcribed operons (Fig. 13). The virB regulatory region is the key control element in the cascade. It seems to be occupied by the VirF protein under both activated and nonactivated conditions. Activation requires a temperature of 37°C and physiological levels of osmolarity, together with a pH of 7.4 [207-209]. The pH control is exerted positively via the CpxA/CpxR two-component regulatory system, and CpxR has been shown to bind directly to the upstream region of the v/rFgene [210]. The organization of the genes into a regulatory cascade allows for very tight control of expression coupled with rapid activation of the system once appropriate environmental signals are received. The cascade displays a gearing effect in which the gene at the top, virF, is regulated over only a twofold range, while the next gene, virB, is controlled over a 10-fold range, and the structural genes at the bottom of the cascade experience a 100-fold range of regulation. Control of virF expression is loose at the level of transcription, while virB is more tightly

3.

107


i

.^

\ii r II %s

\irl

WW

\)\\-

-'ihni:

i 17^/^

w w Fig. 13 The virulence gene regulatory cascade of Shiiêlla flexueri. The virulence regulon occupies approximately 33 kb of a 230-kb virulence plasmid. Large gaps between genes have been deleted for convenience, and are represented by short diagonal lines through the genetic map shown in the center of the figure. The virF and virB genes are illustrated in more detail at the top and bottom of the figure, respectively. Positive regulatory inputs are represented by downward vertical arrows above, and negative influences by T-shaped symbols below the promoter regions. See text for details.

controlled, with the structural genes being under the tightest control of all [211]. This regime seems to ensure that energetically wasteful expression of the structural genes under inappropriate conditions is avoided while allowing for sufficient expression of the regulatory proteins under nonpermissive conditions to facilitate a rapid activation of the structural genes when inducing conditions

108


arise. The system also has the capacity to permit fine-tuning of individual structural operon promoters once activation has occurred [211]. The virB promoter provides one of the clearest examples of a role for DNA supercoiling in the environmental activation of a virulence gene [212, 213]. Although the VirF protein is required absolutely for virB promoter activation, no activation occurs at 30°C (i.e., in the absence of a thermal signal) unless the level of supercoiling of the virB promoter is increased by artificial means [213]. Since increases in temperature cause a general increase in negative supercoiling in bacterial DNA, the simplest interpretation of these data is that the change in DNA structure resulting from a transition to growth at 37°C helps either to drive open complex formation or to improve the interaction between bound VirF and RNA polymerase, or both. It is also a possibility that the multimerization of VirF on DNA may be driven by DNA structural changes at the binding site. Other AraC-like proteins, such as the E. coli MelR protein, display DNA-dependent multimer formation that results in a change in their biological activities [214]. The dependence of the virB promoter on DNA structure for activation is also consistent with a requirement for a normal complement of DNA topoisomerases in the cell [212, 215, 216]. In addition, the promoter is repressed by the H-NS nucleoid-associated protein and requires the IHF protein for full activity in stationary phase [118, 130, 212]. Both H-NS and IHF contact the virB regulatory sequences directly. H-NS levels in the cell are maintained at a near-constant value of about 20,000 copies per cell [217], except in cold-shock, when the level increases by about three- to fourfold [218, 219]. One of the tasks of VirF may be to overcome the repression imposed by H-NS at the virB promoter, a task it carries out in collaboration with thermally induced increases in the negative supercoiling of the DNA [60, 220]. As cells enter stationary phase, their DNA becomes relaxed, a situation that would inhibit supercoiling-dependent promoters such as that of the virB gene [141]. The function of IHF in stationary phase may be to offset this negative effect by maintaining the virB promoter in an active conformation [130]. The virF gene is under posttranscriptional control and provides a good example of the power of translational regulation to influence the expression of a virulence gene regulon. Mutations in the miaA gene, coding for the tRNA A^-isopentyladenosine (i^ A37) synthetase, which is required for the synthesis of the modified nucleoside 2-methyl-A^-isopentyladenosine, a component of tRNA, have a negative effect on the translation of virF mRNA, and hence on the expression of the other genes in the regulatory cascade [221]. Normal expression of the virF gene also requires a functional tgt gene (called vacC in S. flexneri) that codes for tRNA-guanine transglycosylase, required for the synthesis of modified bases in tRNA. Mutations in tgt {vacC) interfere negatively with translation of virF mRNA, and hence with expression of the virulence genes that depend on VirF [222]. The location of the S. flexneri virulence genes on a plasmid is also significant. In EIEC, this plasmid has been shown to integrate at a specific location on the chromosome. This integration event is accompanied by shutting down of vim-

3.


109

lence gene expression, caused by an H-NS-dependent repression of virB transcription [223]. The plasmid also displays significant instability, being lost from the cell when expression of the virulence genes is induced. The instability depends on the functioning of the virF and virB genes and suggests that expression of the structural gene operons makes the plasmid difficult to maintain [224]. The virulence genes are also prone to inactivation by insertion sequences and to DNA rearrangements. Inactivation of virF and virB prevents virulence gene expression and stabilizes the plasmid, although the significance of this in vitro observation is not understood fully. In the case of insertion sequence mutations, these may reverse precisely in the future, restoring virulence gene expression. Thus, the inactivating mutations may serve as a means of ensuring that the plasmids are maintained in the absence of selective pressure [224]. The S. flexneri virulence gene control network also illustrates the point that both specific and general regulators of gene expression can contribute to control of expression of a particular phenotype. In this case, the phenotype is invasiveness, and the specific control elements (VirF and VirB) are encoded by the virulence plasmid, while the general control elements (H-NS, IHF, the DNA topoisomerases, and the tRNA modification systems) are encoded by the chromosome. The general regulatory elements key the virulence genes into the global gene expression program of the cell, while the specific elements operate at a local level to optimize and fine-tune virulence gene expression.

XVL A Thermometer Protein from ttie Salmonella Virulence Plasmid The tlpA gene on the large virulence plasmid of S. typhimiiriiim encodes an autoregulatory repressor protein called TlpA [225]. This protein responds to changes in temperature by shifting ixom an inactive unfolded monomer to an active folded coiled-coil dimeric protein. Its ability to bind to the tlpA gene promoter and repress it is a function of the intracellular concentration of TlpA (which influences protein-protein interactions) and the thermally determined stmcture of the protein. At 22°C, the promoter is repressed, and derepression occurs as the temperature increases. The shift in the structure of the protein is reversible; in other words, TlpA does not become denatured permanently by increases in temperature. No accessory proteins are required for TlpA action at the tlpA promoter, and thermally induced changes in DNA conformation do not appear to be an issue. It will be interesting to see if more examples of this type of regulatory protein emerge in the future. It has been pointed out that a DNA sequence motif characteristic of the tipA promoter is found in several virulence genes, including spvA on the plasmid, and the PhoP/PhoQ-regulated genes prgH

110


and pagC on the chromosome [225]. It will be important to determine which genes, if any, in addition to tlpA are under the control of the TlpA protein.

XVII. CelI'Density'Dependent Regulation Although bacteria are unicellular, they can collaborate to perform certain functions, some of which are of benefit to pathogens during infection of a host. At the heart of these processes Hes a mechanism by which the bacteria sense their own population density. Various forms of bacterial group behavior are regulated by small diffusible self-produced regulatory molecules. Individually, bacteria are incapable of producing the signal in sufficient quantities to elicit a response, but as a population increases in density a critical threshold is crossed and signaling results in a response, usually at the level of transcription, that alters bacterial behavior [226-228]. Early research on cell-density-dependent regulation focused on the bioluminescent phenotype in certain marine bacteria, such as Photobacterium fischeri (formerly Vibrio fischeri). When free-living, and at a low culture density, these bacteria do not emit light. When the population density increases, as during growth in the specialized light organ of the fish with which they have a symbiotic relationship, the bacteria become bioluminescent and emit a blue-green light. This is caused by accumulation of a low-molecular-weight A^-acyl-L-homoserine lactone (AHL) that is synthesized by the bacteria, and sensed by the population. The particular AHL made by P. fischeri is A^-(3-oxo)-hexanoyl-L-homoserine lactone (OHHL). The sensor/regulator is the autoregulatory membrane-associated LuxR protein, and, on binding OHHL, this activates transcription of the bioluminescence genes, which are organized as an operon on the bacterial chromosome (Fig. 14). Also in this operon is the gene for Luxl, which is involved in production of OHHL. Thus, upregulation of the lux operon involves increased transcription of the gene for signal production. It is proposed that downregulation occurs because the LuxR protein is unstable and is subject to rapid turnover by the Lon protease [228]. The population-density-dependent system used by another marine bacterium, V harveyii, has already been discussed in the context of the modular nature of bacterial regulatory proteins (see above). Here, the signal-reception and -transmission system is of the type used by the "two-component" regulators and is based on a phosphotransfer mechanism [91]. Among Gram-negative pathogens, cell-density-dependent regulation based on AHL signaling has been characterized in P aeruginosa, where two systems operate to control expression of a wide array of virulence genes. One system is composed of LasI, a signal generator, and LasR, a response regulator, and is sensitive to A^-(3-oxo)-dodecanoyl-L-homoserine lactone (OdDHL). This controls

3.

11]


An\ iteiisih iiiliiiiv

±J

LELJ.

•J^ : .... \

'iih

''oi

;uM

Hiuli dcriMU ciilliire

/

• • • hixi^

'uî s a . H ; ..

HIOI.UMINESCENCE \

"\J-^V \ .

Fig. 14 Cell-density-dependent gene regulation in Photohactehiim {Vibrio) fischeri. Expression of lux genes is controlled by a diffusible autoinducer (OHHL, see text for details). This is in low concentrations at low culture densities, and the luxICDABEG operon is expressed at a basal level. Autoinducer is synthesized by the Luxl protein, and no single cell can produce enough to induce bioluminescence. At high culture densities, autoinducer levels are sufficient to activate luxICDABEG transcription. This occurs via the LuxR sensor protein, which also acts a repressor of its own gene.

112


the expression of alkaline protease, elastase, exotoxin A, exoenzyme S, neuraminidase, and hemolysin. The second system is composed of VsmI (or Rhll), a signal generator, and VsmR (or RhlR), a response regulator. It is sensitive to A^-butanoyl-L-homoserine lactone (BHL), and controls expression of chitinase, elastase, hydrogen cyanide, pyocyanin, rhamnolipid, lectins, staphylolytic activity, hemolysin, and alkaline protease [229, 230]. Cell-density-dependent regulation also contributes to the ability of P. aeruginosa to form biofilms, an important feature of the pathogen in colonization of catheters, and in growth in the lungs of cystic fibrosis patients [231]. Interestingly, the RhlR protein and the BHL signal are required for activation of the rpoS gene, encoding the sigma-38 component of RNA polymerase [232]. This points to a very wide network of effects exerted by this cell-density-dependent regulatory system in the cell, extending beyond dedicated virulence genes to include genes involved in general adaptation by the cell to changes in growth phase and environmental composition. Detailed information on cell-density-dependent regulation and bacterial virulence is also available for the Gram-negative plant pathogens A. tumefaciens, Erwinia carotovora, and Erwinia stewartii [233]. In A. tumefaciens, the cause of crown gall tumors in plants, A^-(3-oxooctonoyl)-L-homoserine lactone (OOHL) regulates transfer of the tumor-inducing plasmid from one bacterial cell to another by conjugation. In E. carotovora, OHHL regulates production of carbapenem antibiotics. It also regulates the expression of the virulence factors protease, cellulase, and pectinase in the same organism, and exopolysaccharide production in E. stewartii [233]. In the case of E. carotovora, OHHL coordinates the expression of an antimicrobial agent (the carbapenem antibiotic) and the expression of three virulence factors required for successful aggression against a plant host. This may be a way in which the pathogen can eliminate bacterial competitors while ensuring that it undertakes an attack on the host only when the bacterial population is of sufficient size. Among Gram-positive bacteria, similar population density regulatory strategies are in use, albeit with different types of signaling molecules. Typically, a complex phenotype is controlled by a two-component regulatory system that responds to a peptide signal. In the case of Staphylococcus aureus, several virulence traits are controlled by the accessory gene regulatory (agr) locus. In the laboratory, as the culture enters the late exponential phase of growth, agr coordinates repression of cell-surface-associated factors (e.g., protein A, fibronectin binding protein, and coagulase) and induction of secreted components (such as alpha toxin, betahemolysin, toxic-shock syndrome toxin, enterotoxin, lipases, and proteases). The signal is an 8-amino acid peptide called AgrD, which is derived from a 45-amino acid precursor. AgrD production requires AgrB, with which it is translationally coupled. These factors are translated from RNAII (Fig. 15). Transcribed divergently from RNAII is RNAIII, and this is under positive transcriptional control. The AgrD pheromone is one of three signals that activates RNAIII production; the others are the genetic loci sar, xpr, and the RNAIII-activating protein, or RAP [234-237].

3.


113

Fig. 15 The Agr regulatory system of Staphylococcus aureus. The effector molecule of the agr locus is RNAIII. It is under the control of the AgrD pheromone. whose synthesis requires the agrB and agrD genes. The agrA and agrC genes code for a two-component signal transduction system that responds to AgrD levels. AgrA-P probably operates on the '/?-independent regulation of hydroperoxidase I in Escherichia coli. Mol. Microbiol. 12, 571-578. 82. Zheng, M., Aslund, F., and Storz, G. (1998). Activation of the OxyR transcription factor by reversible disulfide bond formation. Science 279, 1718-1721. 83. Hausladen, A., Privalle, C. T., Keng, T., DeAngelo, J., and Stamler, J. S. (1996). Nitrosative stress: Activation of the transcription factor OxyR. Cell 86, 719-729. 84. Gaston, B., Reilly, J., Drazen, J. M., Fackler, J., Ramdev, R, Arnelle, D., Mullins, M., Sugarbaker, D. J., Chee, C, Singel, D. J., Loscalzo, J., and Stamler, J. S. (1993). Endogenous nitrogen oxides and bronchodilator 5-nitrosothiols in human airways. Proc. Natl. Acad. Sci. U.S.A. 90, 10957-10961. 85. Groisman, E. A. personal communication. 86. Altuvia, S., Weinstein-Fischer, D., Zhang, A., Postow, L., and Storz, G. (1997). A small, stable RNA induced by oxidative stress: Role as a pleiotropic regulator and antimutator. Cell 90,43-53, 87. Utsumi, R., Brisette, R. E., Rampersaud, A., Forst, S. A., Oosawa, K., and Inouye, M. (1989). Activation of bacterial porin gene expression by a chimeric signal transducer in response to aspartate. Science 245, 1246-1249. 88. Chang, C.-H., and Winans, S. C. (1992). Functional roles assigned to the periplasmic, linker, and receiver domains of the Agrobacterium tumefaciens VirA protein. J. Bacteriol. 174, 7033-7039. 89. Uhl, M. A., and Miller, J. F. (1994). Autophosphorylation and phosphotransfer in the Bordetella pertussis BvgAS signal transduction cascade. Proc. Natl. Acad. Sci. U.S.A. 91, 1163-1167. 90. Uhl, M. A., and Miller, J. F. (1995). Bordetella pertussis BvgAS virulence control system. In "Two-Component Signal Transduction" (J. A. Hoch and T. J. Silhavy, eds.), pp. 333-349. American Society for Microbiology Press, Washington, DC. 91. Bassler, B., Wright, M., and Silverman, M. (1994). Sequence and function of luxO, a negative regulator of luminescence in Vibrio harx'eyii. Mol. Microbiol. 12, 403^12. 92. Miller, V. L., Taylor, R. K., and Mekalanos, J. J. (1987). Cholera toxin transcriptional activator ToxR is a transmembrane DNA binding protein. Cell 48, 271-279. 93. Martinez-Hackert, E., and Stock, A. M. (1997). Structural relationships in the OmpR family of winged-helix transcription factors. J. Mol. Biol. 269, 301-312. 94. Ottemann, K. M., and Mekalanos, J. J. (1996). The ToxR protein of Vibrio cholerae forms homodimers and heterodimers. / Bacteriol. 178, 156-162. 95. Bajaj, V., Hwang, C , and Lee, C. A. (1995). hilA is a novel ompR/toxR family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715-727. 96. Johnson, C , Pegues, D. A., Hueck, C. J., Lee, A., and Miller, S. I. (1996). Transcriptional activation of Salmonella typhimurium invasion genes by a member of the phosphorylated response-regulator superfamily. Mol. Microbiol. 22, 715-727. 97. Dell, C. L., Neely, M. N., and Olsen, E. R. (1994). Altered pH and lysine signalling mutants of cadC, a gene encoding a membrane-bound transcriptional activator of the Escherichia coli cadBA operon. Mol. Microbiol. 14, 7-16.

3.


123

98. Benjamin Jr., W. H., Wu, X., and Swords. W. E. (1996). The predicted amino acid sequence of the Salmonella typhimiihwn virulence gene mviA^ strongly indicates that MviA is a regulator protein of a previously unknown S. typlniuuriwn response regulator family. Infect. Immun. 64, 2365-2367. 99. Muffler, A., Fischer, D., Altuvia, S., Storz. G., and Hengge-Aronis, R. (1996). The response regulator RssB controls stability of the a' subunit of RN A polymerase in Escherichia coli. EMBO J. 15, 1333-1339. 100. Pratt, L. A., and Silhavy, T. J. (1996). The response regulator SprE controls the stability of RpoS. Pwc. Natl. Acad. Sci. U.S.A. 93, 2488-2492. 101. Zhou, Y., and Gottesman, S. (1998). Regulation of proteolysis of the stationary phase sigma factor RpoS. J. Bacteriol. 180, 1154-1158. 102. Pettijohn, D. E. (1996). The nucleoid. In "Escherichia coli and Salmonella: Cellular and Molecular Biology"" (F. C. Neidhardt, R. Curtiss III J. L. Ingraham, E. C. C. Lin, K. Brooks Low, B. Magasanik, W. S. Reznikoff, M. Riley. M. Schaechter, and H. E. Umbarger, eds.), 2nd ed., pp. 158-166. American Society for Microbiology Press. Washington, DC. 103. Robbe-Saule, V., Schaeffer, E, Kowarz. L.. and Norel. E (1997). Relationships between H-NS, sigma S, SpvR and growth phase in the control of spvR, the regulatory gene of the Salmonella cliiblin virulence plasmid. Mol. Gen. Genet. 256, 333-347. 104. Atlung, T., and Ingmer, H. (1997). H-NS: A modulator of environmentally regulated gene expression. Mol. Microbiol. 24. 7-17. 105. Ussery, D. W., Hinton, J. C. D.. Jordi. B. J. A. M., Granum. R E., Seirafi, A., Stephen, R. J., Tupper, A. E., Berridge, G., Sidebotham, J. M.. and Higgins. C. E. (1994). The chromatin-associated protein H-NS. Biochimie 76, 968-980. 106. Cusick, M. E., and Belfort, M. (1998). Domain structure and RNA annealing activity of the Escherichia coli regulatory protein StpA. Mol. Microbiol. 28, 847-857. 107. Zhang, A., and Belfort, M. (1992). Nucleotide sequence of a newly identified Escherichia coli gene, stpA, encoding an H-NS-like protein. Nucleic Acids Res. 20. 6735. 108. Zhang, A., Derbyshire, V., Galloway Salvo. J. L.. and Belfort, M. (1995). E. coli protein StpA stimulates self-splicing by promoting RNA assembly in vitro. RNA 1, 783-793. 109. Zhang, A., Rimsky, S., Reaban, M. E.. Buc, H., and Belfort. M. (1996). Escherichia coli protein analogs StpA and H-NS: Regulatory loops, similar and disparate effects on nucleic acid dynamics. EMBO J. 15, 1340-1349. 110. Eree, A., and Dorman, C. J. (1997). The Escherichia coli stpA gene is transiently expressed during growth in rich medium and is induced in minimal medium and by stress conditions. J. Bacteriol. 179,909-918. 111. Eree, A., Williams, R. M., and Dorman. C. J. (1998). The StpA protein functions as a molecular adapter to mediate repression of the bgl operon by truncated H-NS in Escherichia coli. J. Bacteriol. 180, 994-997. 112. Spurio, R., Ealconi, M., Brandi, A., Pon. C. L.. and Gualerzi. C. O. (1997). The oligomeric structure of nucleoid protein H-NS is necessary for recognition of intrinsically curved DNA and for DNA bending. EMBO J. 16, 795-805. 113. Ueguchi, C , Suzuki, T., Yoshida. T., Tanaka. K., and Mizuno, T. (1996). Systematic mutational analysis revealing the functional domain organization of Escherichia coli nucleoid protein H-NS. J. Mol. Biol. 263. 149-162. 114. Ueguchi, C, Seto, C, Suzuki. T. and Mizuno. T. (1997). Clarification of the dimerization domain and its functional significance for the Escherichia coli nucleoid protein H-NS. J. Mol. Biol. 274. 145-151.

124


115. Williams, R. M., Rimsky, S., and Buc, H. (1996). Probing the structure, function, and interactions of the Escherichia coli H-NS and StpA proteins by using dominant negative derivatives. J. Bacterioi 178, 4335-4343. 116. Elliot, S. J., Wainwright, L. A., McDaniel, T. K., Jarvis, K. G., Deng, Y., Lai, L.-C., MacNamara, B., Donnenberg, M. S., and Kapen J. B. (1998). The complete sequence of the locus of enterocyte effacement (LEE) from enteropathogenic Escherichia coli E2348/69. Mol. Microbiol. 28, 1 ^ . 117. Goyard, S., and Bertin, R (1997). Characterization of BpH3, an H-NS-like protein in Bordetella pertussis. Mol. Microbiol. 24, 815-823. 118. Hromockyj, A. E., Tucker, S. C, and Maurelli, A. T (1992). Temperature regulation oi Shigella virulence: Identification of the repressor gene virR, an analogue of hus, and partial complementation by tyrosyl transfer RNA (tRNA,'>^). Mol. Microbiol. 6, 2113-2124. 119. Hulton, C. S. J., Seirafi, A., Hinton, J. C. D., Sidebotham, J. M., Waddell, L., Pavitt, G. D., Owen-Hughes, T, Spasskey, A., Buc, H., and Higgins, C. F. (1990). Histone-like protein HI (H-NS), DNA supercoiling and gene expression in bacteria. Cell 63, 631-642. 120. La Teana, A., Falconi, M., Scarlato, V., Lammi, M., and Ron, C. L. (1989). Characterization of the structural genes for the DNA-binding protein H-NS in Enterobacteriaceae. EEBS Letts. 244, 34-38. 121. Marsh, M., and Hillyard, D. R. (1990). Nucleotide sequence oi hns encoding the DNA-binding protein H-NS of Salmonella typhimiiriiim. Nucleic Acids Res. 18, 3397. 122. Ron, C. L., Calogero, R. A., and Gualerzi. C. O. (1988). Identification, cloning, nucleotide sequence and chromosomal map location of /z/?.v, the structural gene for Escherichia coli DNA-binding protein H-NS. Mol. Gen. Genet. 111. 199-202. 123. Sonnenfield, J. M., Raupach, B., Falkow. S., Higgins, C. R, and Hinton, J. C. D. (1997). Genbank Database Accession No. AF009363. 124. Free, A., and Dorman, C. J. Unpublished data. 125. Fleischmann, R. D., Adams, M. D., White, O., Clayton, R. A., Kirkness, E. F, Kerlavage, A. R., Bult, C. J., Tomb, J.-F., Dougherty, B. A., Merrick, J. M., McKenney, K., Sutton, G., FitzHugh, W, Fields, C. A., Gocayne, J. D., Scott, J. D., Shirley, R., Liu, L.-I., Glodek, A., Kelley, J. M., Weidman, J. R, Phillips, C. A., Spriggs, T, Hedblom, E., Cotton, M. D., Utterback, T. R., Hanna, M. C , Nguyen, D. T, Saudek, D. M., Brandon, R. C , Fine, L. D., Fritchman, J. L., Fuhrmann, J. L., Geoghagen, N. S. M., Gnehm, C. L., McDonald, L. A., Small, K. V., Eraser, C. M., Smith, H. O., and Venter, J. C. (1995). Whole-genome random sequencing and assembly oi Haemophilus influenzae Rd. Science 269, 496-512. 126. Drlica, K., and Rouviere-Yaniv, J. (1987). Histonelike proteins of bacteria. Microbiol. Rev. 51, 301-319. 127. Rice, P A., Yang, S.-W., Mizuuchi, K., and Nash, H. A. (1996). Cry.stal structure of an IHF-DNA complex: A protein-induced DNA U-turn. Cell 87, 1295-1306. 128. Goosen, N., and van de Putte, P. (1995). The regulation of transcription initiation by integration host factor. Mol. Microbiol. 16, 1-7. 129. Hill, S. A., Samuels, D. S., Carlson, J. H., Wilson, J., Hogan, D., Lubke, L., Belland, R. J. (1997). Integration host factor is a transcriptional cofactor of pilE in Neisseria gonorrhoeae. Mol. Microbiol. 23, 649-656. 130. Porter, M. E., and Dorman, C. J. (1997). Positive regulation of Shigella flexneri virulence genes by integration host factor. J. Bacterioi. 179, 6537-6550. 131. Appleman, J. A., Ross, W., Salomon, J., and Course, R. (1998). Activation of Escherichia coli rRNA transcription by FIS during a growth cycle. J. Bacterioi. 180, 1525-1532.

3.

REGULATION OF VIRULENCE GENE EXF'RESSION IN BACTERIAL PATHOGENS

125

132. Koch, C, Vanderkerck. J., and Kahmann. R. (1988). E. coli host factor for site-specific DNA inversion. Cloning and characterization of the fis gene. Proc. Natl. Acad. Sci. U.S.A. 85, 4237-4241. 133. Nilsson, L., Vanet, A.. Vijgenboom. E.. and Bosch. L. (1990). The role of FIS in trans activation of stable RNA operons of E. coll. EMBO J. 9. 727-734. 134. Wu, H., Tyson, K. L., Cole. J. A., and Busby. S. J. (1998). Regulation of transcription initiation at the Escherichia coll ulr operon promoter: A new mechanism to account for co-dependence on two transcription factors. Mol. Microbiol. 27. 493-505. 135. Xu. J., and Johnson, R. C. (1995). Identification of genes negatively regulated by FIS: FIS and RpoS comodulate growth-phase-dependent gene expression in Escherichia coll. J. Bacterlol. 17, 938-947. 136. Schneider, R.. Travers. A., and Muskhelishvili. G. (1997). FIS modulates growth phase-dependent topological transitions of DNA in Escherichia coll. Mol. Microbiol. 26, 519-530. 137. Calvo, J. M., and Matthews. R. G. (1994). The leucine-responsive regulatory protein, a global regulator of metabolism in Escherichia coll. Microbiol. Rev. 58. 466-490. 138. Drlica, K. (1992). Bacterial topoisomerases and the control of DNA supercoiling. Trends Genet. 6:433-437. 139. Luttinger, A. (1995). The twisted life of DNA in the cell: Bacterial topoisomerases. Mol. Microbiol. 15, 601-606. 140. Schmid, M. B., and Sawitzke. J. A. (1993). Multiple bacterial topoisomerases: Specialization or redundancy? BloEssaxs 15. 445-449. 141. Dorman, C. J., Barr, G. C. Ni Bhriain. N.. and Higgins. C. F (1988). DNA supercoiling and the anaerobic and growth phase regulation of tonB expression. J. Bacterlol. 170. 2816-2826. 142. Goldstein. E., and Drlica. K. (1984). Regulation of bacterial DNA supercoiling: Plasmid linking numbers vary with growth temperature. Proc. Natl. Acad. Scl. U.S.A. 81, 4046-4050. 143. Higgins, C. E, Dorman, C. J.. Stirling. D. A.. Waddell. L.. Booth. I. R.. May, G., and Bremer, E. (1988). A physiological role for DNA supercoiling in the osmotic regulation of gene expression in S. typhlmurlwn and E. coll. Cell 52. 569-584. 144. Hsieh, L. S.. Burger. R. M.. and Drlica. K. (1991). Bacterial DNA supercoiling and [ATP]/[ADP]: Changes associated with a transition to anaerobic growth. J. Mol. Biol. 219, 443-450. 145. Hsieh, L. S.. Rouviere-Yaniv. J., and Drlica. K. (1991). Bacterial DNA supercoiling and [ATP]/[ADP] ratio: Changes associated with salt shock. J. Bacterlol. 173. 3914-3917. 146. Karem, K., and Foster. J. W. (1993). The influence of DNA topology on the environmental regulation of a pH-regulated locus in Salmonella txphunurlwn. Mol. Microbiol. 10, 75-86. 147. van Workum, M., van Dooren. S. J. M.. Oldenburg. N.. Molenaar. D.. Jensen. P. R., Snoep, J. L., and Westerhoff, H. V. (1996). DNA supercoiling depends on the phosphorylation potential in Escherichia coll. Mol. Microbiol. 20. 351-360. 148. Mojica, F. J. M.. and Higgins, C. F. (1997). In vivo supercoiling of plasmid and chromosomal DNA in an Escherichia coll hns mutant. ./. Bacterlol. 179. 3528-3533. 149. Fang, M., and Wu. H.-Y. (1998). A promoter relay mechanism for sequential gene activation. / Bacterlol. 180, 626-633. 150. Liu, L. E, and Wang, J. C. (1987). Supercoiling of the DNA template during transcription. Proc. Natl. Acad. Scl. U.S.A. 84. 7024-7027. 151. Pruss, G. J., and Drlica. K. (1989). DNA supercoiling and prokaryotic transcription. Cell 56, 521-523. 152. Lilley, D. M. J., and Higgins. C. F. (1991). Local DNA topology and gene expression: The case of the leiiSOO promoter. Mol. Microbiol. 5. 779-783.

126


153. Cashel, M., Gentry, D. R., Hernandez, V. J., and Vinella, D. (1996). The stringent response. In ''Escherichia coli and Salmonella: Cellular and Molecular Biology" (F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. Brooks Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger, eds.), 2nd ed., pp. 1458-1496. American Society for Microbiology Press, Washington, DC. 154. Figueroa-Bossi, N., Guerin, M., Rahmouni, R., Leng, M., and Bossi, L. (1998). The supercoiling sensitivity of a bacterial tRNA promoter parallels its responsiveness to stringent control. EMBO J. 17, 2359-2367. 155. Wang, J.-Y, and Syvanen, M. (1992). DNA twist as a transcriptional sensor for environmental changes. Mol Microbiol. 6, 1861-1866. 156. Dorman, C. J. (1996). Flexible response: DNA supercoiling, transcription and bacterial adaptation to environmental stress. Trends Microbiol. 4, 214-216. 157. Jordi, B. J. A. M., Owen-Hughes, T A., Hulton, C. S., and Higgins, C. F. (1995). DNA twist, flexibility, and transcription of the osmoregulated proU promoter of Salmonella typhimurium. EMBO J. 14, 5690-5700. 158. Mukerji, M., and Mahadevan, S. (1997). Characterization of the negative elements involved in silencing the bgl operon of Escherichia coli: Possible roles for DNA gyrase, H-NS, and CRP-cAMP in regulation. Mol. Microbiol. 24, 617-627. 159. Matthews, K. S. (1992). DNA looping. Microbiol. Rev. 56, 123-136. 160. Schleif, R. (1992). DNA looping. Anna. Rev. Biochem. 61, 199-223. 161. Palecek, E. (1991). Local supercoil-stabilized structures. CRC Crit. Rev. Biochem. 26, 151-226. 162. Trifonov, E. N. (1985). Curved DNA. CRC Crit. Rev Biochem. 19, 89-106. 163. Nash, H. A. (1990). Bending and supercoiling of DNA at the attachment site of bacteriophage X. Trends Biochem. Sci. 15, 111-111. 164. Dybvig, K. (1993). DNA rearrangements and phenotypic switching in prokaryotes. Mol. Microbiol. 10,465-471. 165. Robertson, B. D., and Meyer, T. F. (1992). Genetic variation in pathogenic bacteria. Trends Genet. 8, 422^27. 166. Abraham, J. M., Freitag, C. S., Clements, J. R., and Eisenstein, B. L (1985). An invertible element of DNA controls phase variation of type I fimbriae of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 82, 5724-5727. 167. Marrs, C. F, Rozsa, F W., Hackel, M., Stevens, S. P, and Glasgow, A. C. (1990). Identification, cloning and sequencing of piv, a new gene involved in inverting the pilin genes of Moraxella lacunata. J. Bacteriol. 172, 4370-4377. 168. Haas, R., and Meyer, T. F. (1986). The repertoire of silent pilus genes in Neisseria gonorrhoeae: Evidence for gene conversion. Cell 44, 107-115. 169. Murphy, G. L., Connell, T. D., Barritt, D. S., Koomey, M., and Cannon, J. G. (1989). Phase variation of gonococcal protein II: Regulation of gene expression by slipped strand mispairing of a repetitive DNA sequence. Cell 56, 539-547. 170. Stem, A., and Meyer, T. F. (1987). Common mechanism controlling phase variation and antigenic variation in pathogenic Neiserriae. Mol. Microbiol. 1, 5-12. 171. Weiser, J. N., Love, J. M., and Moxon, E. R. (1989). The molecular mechanism of phase variation of H. influenzae lipopolysaccharide. Cell 59, 657-665. 172. van Ham, S. M., van Alphen, L., Mooi, F. R., and Putten, J. P. (1993). Phase variation of H. influenzae fimbriae: Transcriptional control of two divergent genes through a variable combined promoter region. Cell 12^, 1187-1196.

3.


127

173. Willems, R., Paul, A., van der Heide, H. G. J., ter Avest. A. R., and Mooi, F. R. (1990). Fimbrial phase variation in Bordetella pertussis: A novel mechanism for transcriptional regulation. EMBO J. 9, 2803-2809. 174. Jonsson, A.-B., Nyberg, G., and Normark, S. (1991). Phase variation of gonococcal pili by frameshift mutation at pilC, a novel gene for pilus assembly. EMBO J. 10, 477-488. 175. Olyhoek, A. J., Sarkari, J., Bopp, M., Morelli, G., and Achtman, M. (1991). Cloning and expression in Escherichia coli of ope, the gene for an unusual class 5 outer membrane protein from Neisseria meningitidis. Microb. Pathogen. 11, 249-257. 176. Stibitz, S., Aaronson, W., Monack, D., and Falkow, S. (1989). Phase variation in Bordetella pertussis by frameshift mutation in a gene for a novel two-component system. Nature 338, 266-269. 177. Belland, R. J. (1991). H-DNA formation by the coding repeat elements of neisserial opa gens. Mol. Microbiol. 5, 2351-2360. 178. Belland, R. J., Morrison, S. G., Carlson, J. H.. and Hogan, D. M. (1997). Promoter strength influences phase variation of neisserial opa genes. Mol. Microbiol. 23, 123-135. 179. Dorman, C. J., and Higgins, C. F. (1987). Fimbrial phase variation in Escherichia coli: Dependence on integration host factor and homologies with other site-specific recombinases. J. Bacteriol. 169, 3840-3843. 180. Eisenstein, B. L, Sweet, D. S., Vaughn, V., and Friedman, D. I. (1987). Integration host factor is required for the DNA inversion that controls phase variation in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 84, 6506-6510. 181. Blomfield, I. C, Kulasekara, D. H., and Eisenstein, B. I. (1997). Integration host factor stimulates both FimB- and FimE-mediated site-specific inversion that controls phase variation of type 1 fimbriae expression in Escherichia coli. Mol. Microbiol. 23, 705-717. 182. Roesch, P. L., and Blomfield, I. C. (1998). Leucine alters the interaction of the leucine-responsive regulatory protein (LRP) with the fim switch to stimulate site-specific recombination in Escherichia coli. Mol. Microbiol. 27, 751-761. 183. Dove, S. L., and Dorman, C. J. (1994). The site-specific recombination system regulating expression of the type 1 fimbrial subunit gene of Escherichia coli is sensitive to changes in DNA supercoiling. Mol. Microbiol. 14, 975-988. 184. Lindberg, F. P., Lund, B., and Normark, S. (1984). Genes of pyelonephritogenic E. coli required for digalactoside-specific agglutination of human cells. EMBO J. 3, 1167-1173. 185. Blyn, L. B., Braaten, B. A., White-Ziegler, C. A., Rolfson, D. H., and Low, D. A. (1989). Phase-variation of pyelonephritis-associated pili in Escherichia coli: Evidence for transcriptional regulation. EMBO J. 8, 613-620. 186. White-Ziegler, C. A., Angus Hill M. L., Braaten, B. A., van der Woude, M., and Low, D. A. (1998). Thermoregulation of Escherichia coli pap transcription: H-NS is a temperature-dependent DNA methylation blocking factor. Mol. Microbiol. 28, 1121-1137. 187. van der Woude, M. W., and Low, D. A. (1994). Leucine-responsive regulatory protein and deoxyadenosine methylase control the phase variation and expression of the sfa and daa pili operons in Escherichia coli. Mol. Microbiol. 11, 605-618. 188. Braaten, B. A., Nou, X., Kaltenbach, L. S., and Low. D. A. (1994). Methylation patterns in pap regulatory DNA control pyelonephritis-associated pili phase variation in E. coli. Cell 74, 577-588. 189. Blyn, L. B., Braaten, B. A., and Low, D. A. (1990). Reguladon of pap pilin phase variadon by a mechanism involving differential Dam methylation states. EMBO J. 9, 4045-^054. 190. Kaltenbach, L. S., Braaten, B. A., and Low, D. A. (1995). Specific binding of Papl to Lrp-pap DNA complexes. / Bacteriol. Ill, 6447-6455.

128


191. Nou, X., Skinner, B., Braaten, B., Blyn, B., Hirsh, L., and Low, D. (1993). Regulation of pyelonephritis-associated pili phase variation in Escherichia coli: Binding of the Papl and LRP regulatory proteins is controlled by DNA methylation. Mol. Microbiol. 7, 545-553. 192. van der Woude, M., Kaltenbach, L. S., and Low, D. A. (1995). Leucine-responsive regulatory protein plays dual roles as both an activator and a repressor of the Escherichia coli pap operon. Mol. Microbiol. 17, 303-312. 193. Baga, M., Goransson, M., Nomiark, S., and Uhlin, B. E. (1985). Transcriptional activation of a pap pilus virulence operon from uropathogenic Escherichia coli. EMBO J. 4, 3887-3893. 194. Forsman, K., Goransson, M., and Uhlin, B. E. (1989). Autoregulation and multiple DNA interactions by a transcriptional regulatory protein in E. coli pili biogenesis. EMBO J. 8, 1271-1277. 195. Forsman, K., Sonden, B., Goransson. M., and Uhlin, B. E. (1992). Antirepression function in Escherichia coli for the cAMP-cAMP receptor protein transcriptional activator. Proc. Natl. Acad. Sci. U.S.A. 89, 9880-9884. 196. Goransson, M., Sonden, B., Nilsson, P., Dagberg, B., Forsman, K., Emanuelsson, K., and Uhlin, B. E. (1990). Transcriptional silencing and thermoregulation of gene expression in Escherichia coli. Nature 344, 682-685. 197. Zhang, J. P., and Normark, S. (1996). Induction of gene expression in Escherichia coli after pilus-mediated adherence. Science 273, 1234-1236. 198. Cornelis, G. R., and Wolf-Watz, H. (1997). The Yersinia yop virulon: A bacterial system for subverting eukaryotic cells. Mol. Microbiol. 23, 861-867. 199. Pettersson, J., Nordfelth, R., Dubinina, E., Bergman, T, Gustafsson, M., Magnusson, K. E., and Wolf-Watz, H. (1996). Modulation of virulence factor expression by pathogen target cell contact. Science 273, 1231-1233. 200. Stainier, L, Iriarte, M., and Cornelis, G. R. (1997). YscMl and YscM2, two Yersinia enterocolitica proteins causing downregulation oi yop transcription. Mol. Microbiol. 26, 833-843. 201. Williams, A. W., and Staley, S. C. (1998). YopD of Yersinia pestis plays a role in negative regulation of the low-calcium response in addition to its role in translocation of yops. J. Bacteriol. 180, 350-358. 202. Hughes, K. T., Gillen, K. L., Semon, M. J., and Karlinsey, J. E. (1993). Sensing structural intermediates in bacterial flagellar assembly by export of a negative regulator. Science 161, 1277-1280. 203. Menard, R., Dehio, C , and Sansonetti, P. J. (1996). Bacterial entry into epithelial cells: The paradigm of Shigella. Trends Microbiol. 4, 220-226. 204. Parsot, C , Menard, R., Gounon. P., and Sansonetti, P. J. (1995). Enhanced secretion through the Shigella flexneri Mxi-Spa translocation leads to assembly of extracellular proteins into macromolecular structures. Mol. Microbiol. 16, 291-300. 205. Hale, T. L. (1991). Genetic basis of virulence in Shigella species. Microbiol. Rev. 55, 206-224. 206. Tobe, T., Nagai, S., Okada, N., Adler, B., Yoshikawa, M., and Sasakawa, C. (1991). Temperature-regulated expression of invasion genes in Shigella flexneri is controlled through the transcriptional activation of the virB gene on the large plasmid. Mol. Microbiol. 5, 887-893. 207. Maurelli, A. T, Blackmon, B., and Curtiss III, R. (1984). Temperature-dependent expression of virulence genes in Shigella flexneri. Infect. Immun. 43, 195-201. 208. Nakayama, S.-I., and Watanabe, H. (1995). Involvement of cpxA, a sensor of a two-component regulatory system, in the pH-dependent regulation of expression of Shigella sonnei virE gene. J. Bacteriol. Ill, 5062-5069. 209. Porter, M. E., and Dorman, C. J. (1994). A role for H-NS in the thermo-osmotic regulation of virulence gene expression in Shigella flexneri. J. Bacteriol. 176, 4187^191.

3.


129

210. Nakayama, S.-I., and Watanabe, H. (1998). Identification of cp.xR as a positive regulator essential for expression of the Shigella sonnei virF gene. J. Bacterial. 180. 3522-3528. 211. Porter, M. E., and Dorman, C. J. (1997). Differential regulation of the plasmid-encoded genes in the Shigella flexneri virulence regulon. Mol. Gen. Genet. 256. 93-103. 212. Dorman, C. J., Ni Bhriain, N.. and Higgins. C. F. {1990). DNA supercoiling and environmental regulation of virulence gene expression in Shigella flexneri. Nature 344, 789-792. 213. Tobe, T., Yoshikawa, M., and Sasakawa. C. (1995). Thermoregulation of virB transcription in Shigella fle.xneri by sensing of changes in local DNA superhelicity. J. Bactehol. 177, 1094—1097. 214. Bourgerie, S. J., Michan. C. M.. Thomas. M. S.. Busby, S. J. W.. and Hyde, E. I. (1997). DNA binding and DNA bending by the MelR transcription activator protein from Escherichia coli. Nucleic Acids Res. 25, 1685-1693. 215. McNairn, E., Ni Bhriain, N., and Dorman. C. J. (1995). Over-expression of the Shigella fle.xneri genes coding for DNA topoisomerase IV compensates for loss of DNA topoisomerase I: Effect on virulence gene expression. Mol. Microbiol. 15. 507-517. 216. Nf Bhriain, N., and Dorman, C. J. (1993). Isolation and characterization of a topA mutant of Shigella fle.xneri. Mol. Microbiol. 7, 351-358. 217. Free, A., and Dorman, C. J. (1995). Coupling of Escherichia coli hns mRNA levels to DNA synthesis by autoregulation: Implications for growth phase control. Mol. Microbiol. 18, 101-113. 218. Brandi, A., Pon, C. L., and Gualerzi. C. O. (1994). Interaction of the main cold shock protein CS7.4 (CspA) of Escherichia coli with the promoter region of hns. Biochimie 76, 1090-1098. 219. La Teana, A., Brandi, A., Falconi, M., Spurio, R.. Pon, C. L., and Gualerzi, C. O. (1991). Identification of a cold shock transcriptional enhancer of the Escherichia coli gene encoding nucleoid protein H-NS. Proc. Natl. Acad. Sci. U.S.A. 88, 10907-10911. 220. Tobe, T, Yoshikawa, M., Mizuno, T, and Sasakawa, C. (1993). Transcriptional control of the invasion regulatory gene virB of Shigella flexneri: Activation by VirF and repression by H-NS. J. Bacterial. 175, 6142-6149. 221. Durand, J. M. B., Bjork, G. R., Kuwae, A., Yoshikawa, M., and Sasakawa, C. (1997). The modified nucleoside 2-methylthio-A^'-isopentyladenosine in tRNA of Shigella flexneri is required for expression of virulence genes. J. Bacterial. 179, 5777-5782. 222. Durand, J. M., Okada, M., Tobe, T, Watarai. M.. Fukuda, I., Suzuki, T., Nakata, N., Komatsu, K., Yoshikawa, M., and Sasakawa, C. (1994). vacC, a virulence-associated chromosomal locus of Shigella flexneri, is homologous to tgt. a gene encoding tRNA-guanine transglycosylase (Tgt) of Escherichia call K-12. J. Bacterial. 176. 4627-4634. 223. Colonna, B., Casalino, M., Fradiani, P. A.. Zagaglia. C., Naitza. S., Leoni, L., Prosseda, G., Coppo, A., Ghelardini, P., and Nicoletti, M. (1995). H-NS regulation of virulence gene expression in enteroinvasive Escherichia coli harboring the virulence plasmid integrated into the host chromosome. J. Bacterial. 111. 4703^712. 224. Schuch, R., and Maurelli, A. T (1997). Virulence plasmid instability in Shigella fle.xneri 2a is induced by virulence gene expression. Infect. Inunun. 65, 3686-3692. 225. Hurme, R., Berndt, K. D., Normark, S. J., and Rhen. M. (1997). A proteinaceous gene regulatory thermometer in Salmonella. Cell 90, 55-64. 226. Dunny, G. M., and Leonard, B. A. B. (1997). Cell-cell communication in Gram-positive bacteria. Annu. Rev. Microbiol. 51, 527-564. 227. Gray, K. M. (1997). Intercellular communication and group behavior in bacteria. Trends Microbial. 5, 184-188. 228. Swift, S., Throup, J. P, Williams, P, Salmond, G. P C . and Stewart, G. S. A. B. (1996). Quorum sensing: A population-density component in the determination of bacterial phenotype. Trends Biatechnol. 21, 214-219.

130


229. Pearson, J. P., Pesci, E. C , and Iglewski, B. H. (1997). Roles of Pseudomonas aeruginosa las and rhl quorum-sensing systems in control of elastase and rhamnolipid biosynthesis genes, /. Bacteriol 179, 5756-5767. 230. Winson, M. K., Camara, M., Latifi, A., Foglino, M., Chhabra, S. R., Daykin, M., Bally, M., Chapon, V., Salmond, G. P C., Bycroft, B. W., Lazdunski, A., Stewart, G. S. A. B., and Williams, P. (1995). Multiple A^-acyl-L-homoserine lactone signal molecules regulate production of virulence determinants and secondary metabolites in Pseudomonas aeruginosa. Proc. Natl. Acad Sci. U.S.A. 92, 9427-9431. 231. Davies, D. G., Parsek, M. R., Pearson, J. P., Iglewski, B. H., Costerton, J. W., and Greenberg, E. P. (1998). The involvement of cell-to-cell signals in the development of a bacterial biofilm. Science 280, 295-298. 232. Latiffi, A., Foglino, M., Tanaka, K., Williams, P., and Lazdunski, A. (1996). A hierarchical quorum-sensing cascade in Pseudomonas aeruginosa links the transcriptional activators LasR and RhlR (VsmR) to expression of the stationary-phase sigma factor RpoS. Mol. Microbiol. 21, 1137-1146. 233. Robson, N. D., Cox, A. R. J., McGowan, J., Bycroft, B. W., and Salmond, G. P C. (1997). Bacterial A^-acyl-homoserine-lactone-dependent signalling and its potential biotechnological applications. Trends Biotechnol. 15, 458-464. 234. Balaban, N., Goldhom, T, Nhan, R. T., Dang, L. B., Scott, S., Ridgley, R. M., Rasooly, A., Wright, S. C , Larrick, J. W., Rasooly, R., and Carlson, J. R. (1998). Autoinducer of virulence as a target for vaccine and therapy against Staphylococcus aureus. Science 280, 438^40. 235. Cheung, A. L., Koomey, J. M., Butler, C. A., Projan, S. J., and Fischetti, V. A. (1992). Regulation of exoprotein expression in Staphylococcus aureus by a locus {sar) distinct from agr. Proc. Natl. Acad Sci. U.S.A. 89, 6462-6466. 236. Cheung, A. L., and Projan, S. J. (1996). Cloning and sequencing of sar of Staphylococcus aureus, a gene required for the expression of agr. J. Bacteriol. 176, 4168^172. 237. Hart, M. E., Smeltzer, M. S., and landolo, J. J. (1993). The extracellular protein regulator {xpr) affects exoprotein and agr levels in Staphylococcus aureus. J. Bacteriol. 175, 7875-7879. 238. Ji, G. Y, Beavis, R. C , and Novick, R. P (1995). Cell density control of staphylococcal virulence mediated by an octapeptide pheromone. Proc. Natl. Acad. Sci. U.S.A. 92, 1205512059. 239. Morfeldt, E., Tegmark, K., and Arvidson, S. (1996). Transcriptional control of the «gr-dependent gene regulator, RNAIII, in Staphylococcus aureus. Mol. Microbiol. 21, 1227-1237. 240. Morfeldt, E., Taylor, D., von Gabain, A., and Arvidson, S. (1995), Activation of alpha-toxin translation in Staphylococcus aureus by the rraAi^-encoded antisense RNA, RNAIII. EMBO J. 14, 4569-4577. 241. Bridges, B. A. (1997). Hypermutation under stress. Nature 387, 557-558. 242. Foster, P. L. (1993). Adaptive mutation: The uses of adversity. Annu. Rev. Microbiol. 47, 467-504. 243. Torkelson, J., Harris, R. S., Lombardo, M.-J., Nagendran, J., Thulin, C , and Rosenberg, S. M. (1997). Genome-wide hypermutation in a subpopulation of stationary-phase cells underlies recombination-dependent adaptive mutation. EMBO J. 16, 3303-3311. 244. Cairns, J. (1993). Directed mutation. Science 260, 1221-1222. 245. Foster, P. L. (1994). Population dynamics of a Lac- strain of Escherichia coli during selection for lactose utilization. Genetics 138, 253-261. 246. Hall, B. G. (1998). Activation of the bgl operon by adaptive mutation. Mol. Biol. Evol. 15, 1-5. 247. Moxon, E. R., and Thaler, D. S. (1997). The tinkerer's evolving toolbox. Nature 387, 659-662.

3.


131

248. LeClerc, J. E., Li, B., Payne, W. L., Cebula, T. A. (1996). High mutation frequencies among Escherichia coli and Scihnonella pathogens. Science 274, 1208-1211. 249. Konigsberg, W., and Godson, G. N. (1983). Evidence for use of rare codons in the dnaG gene and other regulatory genes of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 80, 687-691. 250. Sharp, P. M., and Li, W.-H. (1986). Codon usage in regulatory genes in Escherichia coli does not reflect selection for rare codons. Nucleic Acids Res. 14, 7737-7749. 251. Kurland, C. G. (1993). Major codon preference in theme and variations. Biochem. Soc. Trans. 21, 841-846. 252. Chen, G. T., and Inouye, M. (1994). Role of AGA/AGG codons, the rarest codons in global gene regulation in Escherichia coli. Genes Dev. 8, 2641-2652. 253. Chen, G. T., and Inouye, M. (1990). Suppression of the negative effect of minor arginine codons on gene expression: Preferential usage of minor codons within the first 25 codons of the Escherichia coli genes. Nucleic Acids Res. 18, 1465-1473. 254. Goldman, E., Rosenberg, A. H., Zubay, G., and Studier, F. W. (1995). Consecutive low-usage leucine codons block translation when near the 5' end of a message in Escherichia coli. J. Mol. Biol. 245, 467-473. 255. Bonekamp, R, and Jensen, K. F. (1988). The AGG codon is translated slowly in E. coli even at very low expression levels. Nucleic Acids Res. 16, 3013-3024. 256. Misra, R., and Reeves, P. (1985). Intermediates in the synthesis of TolC protein include an incomplete peptide stalled at a rare Arg codon. Eur J. Biochem. 152, 151-155. 257. Robinson, M., Lilley, R., Little, S., Emtage, J. S., Yarranton, G., Stephens, R, Millican, A., Eaton, M., and Humphreys, G. (1984). Codon usage can affect translation rate in Escherichia coli. Nucleic Acids Res. 12, 6663-6671. 258. Rosenberg, A. H., Goldman, E., Dunn, J. J., Studier, F W., and Zubay, G. (1993). Effects of consecutive AGG codons on translation in Escherichia coli demonstrated with a versatile codon test system. /. Bacteriol. 175, 716-722. 259. Spanjaard, R. A., and van Duin, J. (1988). Translation of the sequence AGG-AGG yields 50% ribosomal frameshift. Proc. Nad. Acad. Sci. U.S.A. 85, 1961-191 \. 260. Zahn, K., and Landy, A. (1996). Modulation of lambda integrase synthesis by rare arginine tRNA. Mol. Microbiol. 21, 69-76. 261. Ritter, A., Gaily, D., Olsen, P B., Dobrindt, U., Friedrich, A., Klemm, P, and Hacker, J. (1997). The Pai-associated leuX specific tRNA5Leu affects type 1 fimbriation in pathogenic Escherichia coli by control of FimB recombinase expression. Mol. Microbiol. 25, 871-882. 262. Blum, G., Ott, M., Lischewski, A., Ritter, A., Imrich, H., Tschiipe, H., and Hacker, J. (1994). Excision of large DNA regions termed pathogenicity islands from tRNA-specific loci in the chromosome of an Escherichia coli wild-type pathogen. Infect. Immun. 62, 606-614. 263. Colston, M. J., and Davis, E. O. (1994). The ins and outs of protein splicing elements. Mol. Microbiol 12, 359-363. 264. Fsihi, H., Vincent, V, and Cole, S. T. (1996). Homing events in the gyrA gene of some mycobacteria. Proc. Nad. Acad. Sci. U.S.A. 93, 3410-3415. 265. Cooper, A. A., and Steven, T. H. (1995). Protein splicing: Self-splicing of genetically mobile elements at the protein level. Trends Biochem. Sci. 20, 351-356. 266. Derbyshire, V, Wood, D. W., Wu, W., Dansereau, J. T, Dalgaard, J. Z., and Belfort, M. (1997). Genetic definition of a protein-splicing domain: Functional mini-inteins support structure predictions and a model for intein evolution. Proc. Natl. Acad. Sci. U.S.A. 94, 11466-11471. 267. Delihas, N. (1995). Regulation of gene expression by r/T//?.v-encoded antisense RNAs. Mol. Microbiol. 15.411-414.

132


268. Wagner, E. G. H., and Simons, R. W. (1994). Antisense RNA control in bacteria, phages, and plasmids. Annu. Rev. Microbiol. 48, 713-742. 269. Okamoto, K., and Freundlich, M. (1986). Mechanism for the autogenous control of the crp operon: Transcriptional inhibition by a divergent RNA transcript. Proc. Natl. Acad. Sci. U.S.A. 83, 5000-5004. 270. Andersen, J., Forst, A., Zhao, K., Inouye, M., and Delihas, N. (1989). The function of micF RNA. J. Biol. Chem. 264, 17961-17970. 271. Strauss, E. J., and Falkow, S. (1997). Microbial pathogenesis: Genomics and beyond. Science 276,707-712. 272. Valdivia, R. H., and Falkow, S. (1997). Fluoresence-based isolation of bacterial genes expressed within host cells. Science 211, 2007-2011. 273. Hoheisel, J. D. (1997). Oligomer-chip technology. Trends Biotechnol. 15, 465-469. 274. Massey, R. C , Rainey, R B., Sheehan, B. J., Keane, O. M., and Dorman, C. J. (1999). Environmentally constrained mutation and adaptive evolution in Salmonella. Cum Biol. 9, 1477-1480.

CHAPTER 4

Strategies to Identify Bacterial Pathogenicity Factors ANDREW CAMILLI D. SCOTT MERRELL JOHN J. MEKALANOS

I. Introduction II. Biochemical Strategies A. Classical Approaches B. Chemical Modification Screens C. Zymography D. Receptor/Ligand Affinity Screens E. Immunological Methods F Subtractive Hybridization G. Differential Display H. Reverse Genetics III. Genetic Screens A. In Vitro Screens B. In VMY; Screens IV. Genetic Selections A. Direct Selections B. Complementation Approaches C. Selection for Nongrowing Bacterial Mutants D. //? V/IY; Expression Technology V. Genomic Approaches A. Genome Walking B. Genomic Analysis and Mapping by /// Vitro Transposition C. Computational Screens D. Transcriptional Profiling and the Use of Microarrays VI. Concluding Remarks References

Principles of Bacterial Pathogenesis Copyright © 2001 by Academic Press AH rights of reproduction in any form reserxed. ISBN 0-12-304220-8

i oo

134 135 135 139 141 143 144 145 147 149 149 150 156 159 160 161 162 163 165 164 166 167 168 169 170

134

ANDREW CAMILLI, D . SCOTT MERRELL, AND JOHN J. MEKALANOS

/. Introduction Since Pasteur, Lister, Koch, and their contemporaries first proved the germ theory of disease in the second half of the nineteenth century, humans have been striving to understand the pathogenic mechanisms of the many reprehensible microbes that plague humans. This desire to understand the microbial causation of many human diseases has led to many epidemiological studies and to increased knowledge of the diverse environmental reservoirs occupied by pathogenic bacteria. The information garnered has been extremely helpful in combating many diseases and has shed light on the persistence of pathogenic bacteria in the absence of a human host. For example, an understanding of the ability of Vibrio cholerae to survive within an aqueous environment led to the development of water treatment strategies that have proven successful in the prevention of cholera and similar diarrheal diseases. Two additional lines of predominantly empirical research—vaccine development and the discovery of antimicrobial compounds—have also had tremendous impact on the prevention and treatment of many bacterial diseases. However, there is now considerable evidence that science has begun to exhaust these empirical lines of research, placing the scientific community at the dawn of an era when detailed knowledge of the cellular and molecular aspects of host-pathogen interactions will seed the future development of novel vaccines and novel antimicrobial compounds. Essential to the establishment of a complete understanding of the host-pathogen interactions that are necessary for the manifestation of disease is the identification and detailed characterization of virulence factors produced by pathogens at each stage of the infection process. The other chapters of this book detail the progression of the understanding of many host-pathogen interactions that have been elucidated during the past century. In contrast, the present chapter focuses on the various biochemical, immunological, and genetic strategies that have been successfully employed by microbiologists to identify these bacterial pathogenicity factors. In addition, speculation on the utilization of several postgenomic strategies that are currently being developed is presented in an attempt to illuminate future methodologies, which should prove useful in further discoveries. A listing of each technique/strategy along with a brief description of some of their advantages and disadvantages is shown in Table I. Bacterial pathogenicity is a multifactorial and dynamic process that requires coordinate production and repression of many bacterial gene products. Though each bacterial pathogen interacts with the human host in a unique way, common events occur in many infections. These common events involve: (1) primary attachment to host cells or tissue, (2) invasion into host cells or tissues (in the case of invasive pathogens) or increased adherence by extracellular pathogens, (3) avoidance of and/or resistance to host immune defenses, (4) acquisition of nutrients, (5) multiplication, and, lasdy, (6) evacuation from the host to either a new host or an environmental reservoir. Invariably, at one or more of these stages.

4.

STRATEGIES TO IDENTIFY BACTERIAL PATHOGENICITY FACTORS

135

bacterial-produced toxins and/or the host immune response result in damage to host tissues or organs and give rise to pathology. Although a tremendous amount of knowledge has been amassed on the identity and function of toxins and other secreted or surface factors produced by bacterial pathogens, it is becoming more and more evident that the investigation of bacterial pathogenicity must be broadened to include an understanding of cytosolic factors that are involved in processes as diverse as production of metabolites, multiplication, and adaptation to dynamic and stressful environments. The reason for this is simple: any factor produced by a pathogen that plays a role in survival or multiplication in the host environment is, a priori, suitable as a target for inclusion in vaccine formulations or for development of antimicrobial drugs. Accordingly, in some sections of this chapter the reader may note the use of a broader definition of "pathogenicity factor" that includes any factor produced by a pathogen that plays a role in its survival, multiplication, or spread to new hosts.

//. Biochemical Strategies A.

Classical Approaches

Prior to the dawn of the recombinant DNA era, microbiologists seeking to identify bacterial pathogenicity factors primarily focused their efforts on the biochemical purification of macromolecules that could be associated with disease symptoms and could be produced by the pathogen during laboratory growth. This classical approach is still utilized today, although usually in conjunction with immunological and/or genetic methods that increase the breadth of the approach. For example. Western blot analysis using human convalescent serum could be used to identify proteins secreted by a pathogen during in vitro growth. Once the protein was identified, it could be purified by two-dimensional gel electrophoresis, an amino-terminal portion of the protein sequenced, degenerate polymerase-chain reaction (PCR) primers synthesized, a portion of the gene amplified by PCR, and the gene encoding the protein identified in the pathogen genome using the PCR product as a probe. This series combines techniques from several sections of this chapter and highlights the value of understanding and utilizing multiple techniques to identify pathogenicity factors. An essential element of the initial, purely biochemical purifications was that an activity had to be associated with the pathogenicity factor to allow its purification. The range of activities of the pathogenicity factors that could be isolated were therefore limited to the presence of enzymatic activity that could be assayed in vitro, or toxic activity that could be assayed using cultured host cells or whole animals. For this reason, many of the first pathogenicity factors

136

Table I

ANDREW CAMILLI, D. SCOTT MERRELL, AND JOHN J. MEKALANOS

Techniques and Strategies for Identifying Pathogenicity Factors"

Technique or strategy

Advantages

Disadvantages

Classical biochemical approaches

Doesn't require genetic manipulation of pathogen. Can identify factors essential for viability.

Requires knowledge of enzymatic activity or function. Can be laborious.

Chemical modification screens

Doesn't require genetic manipulation. Can identify essential factors. Not necessary to know enzymatic activity or function.

Targets only a subset of pathogenicity factors (e.g., those exposed on the outer surface).

Zymography

Doesn't require genetic manipulation. Can identify essential factors. Direct detection of purified or partially purified factor.

Requires knowledge of enzymatic activity for which a substrate is available. Enzymatic activity must be stable and in most cases, renaturable. Can be expensive.

Receptor/ligand affinity screens

Doesn't require genetic manipulation. Direct purification of factor.

Requires knowledge of either ligand or receptor.

Immunological methods

Doesn't require genetic manipulation. Targets factors expressed during infection.

Targets only factors that are immunogenic, many of which may not be pathogenicity factors.

Subtractive hybridization

Doesn't require genetic manipulation. Can identify essential factors. Doe.sn't require knowledge of function.

Lacks sensitivity and can be laborious. Only a subset may encode pathogenicity factors.

Differential display

Doesn't require genetic manipulation. Can identify essential factors.

Lacks sensitivity, and can be laborious. Only a subset may encode pathogenicity factors.

Reverse genetics

Doesn't require genetic manipulation of pathogen. Can identify essential pathogenicity proteins.

Requires some amino acid sequence of protein factor.

Large-scale screenings

Can be comprehensive.

Is usually laborious. Usually requires an in vitro model of infection (e.g., tissue culture). Cannot identify redundant pathogenicity factors.

Targeting exported proteins

Many pathogenicity factors are exported. Simultaneously generates a null mutation in pathogenicity gene.

Only a subset of pathogenicity factors are exported proteins. Some protein-reporter fusions will be toxic.

Coordinate regulation screens

Directly identifies pathogenicity genes coregulated with known ones.

Requires knowledge of in vitro growth conditions that induce expression of pathogenicity genes.

Host mimicry

Directly identifies pathogenicity genes and may point to a role during infection.

Requires knowledge of in vitro conditions that mimic a host parameter(s).

continued

4.

137

STRATEGIES TO IDENTIFY BACTERIAL PATH(XIENICITY FAC TORS

Technique or strategy

Advantages

Disadvantages

Recombinationbased in vivo expression technology

Identifies pathogenicity genes that are transiently induced or induced at a low level in host tissues.

Not all pathogenicity genes are transcriptionally induced in host. Only a subset of infection-induced genes encode pathogenicity factors.

Signature-tagged mutagenesis

Targets pathogenicity genes that play essential roles during infection. Allows multiple mutant strains to be screened per host animal.

Can be laborious. Requires large infectious dose in some disease models. Cannot identify redundant pathogenicity factors.

Direct selections

Directly identifies strains with mutations in pathogenicity genes.

Requires knowledge of a phenotype associated with the pathogenicity factor. For some direct selections only a subset of mutations will be in pathogenicity genes.

Complementation approaches

Directly identifies genes necessary and in some cases sufficient to confer a pathogenic property.

Depending on strategy, requires efficient means of transformation into pathogen, or a simple in vitro model of disease (e.g., adherence to tissue culture cells).

Selection for nongrowing bacterial mutants

Directly identifies genes necessary for multiplication in vitro under host-like conditions (e.g.. within tissue culture cells).

Only a subset of mutations will be in genes specifically required for in vivo growth.

In vivo expression technology

Directly identifies pathogenicity genes that are expressed in host tissues.

Pathogenicity genes that are transiently expressed during infection or are expressed at low levels may not be identified.

Genome walking

Rapid and simple method of identifying pathogenicity genes.

Requires mutagenesis studies to confirm roles in pathogenicity.

Genomic analysis Targets pathogenicity genes that play and mapping essential roles during infection. Allows by in vitro multiple mutant strains to be screened transposition per host animal.

Laborious. Requires prior genome sequence information. Cannot identify redundant pathogenicity factors.

Computational screens

Rapid and simple method of identifying pathogenicity genes.

Requires prior knowledge of pathogenicity factors to serve as query sequences. Requires mutagenesis studies to confirm roles in pathogenicity.

Transcriptional profiling and the use of microarrays

Rapid and comprehensive method of identifying pathogenicity genes.

Requires prior genome sequence information. Expensive. Requires mutagenesis studies to confirm roles in pathogenicity.

"This table is meant to list most of the techniques and strategies discussed in the text and is not all-inclusive. Only a brief description of some of the major advantages and disadvantages of each is given. Please read the text for more detailed discussion, it is up to the reader to thoroughly investigate and carefully weigh the advantages and disadvantages of each technique or strategy before embarking on a study to identify virulence factors.

138


identified were bacterial exotoxins that possess dramatic activities whose effects can be easily monitored. For example, the Streptococcus pneumoniae pore-forming protein pneumolysin (Ply) lyses mammalian red blood cells (RBCs), producing clear, hemolytic zones around colonies grown on media containing RBCs. This activity was noted early in this century [72], and it provided a convenient assay to monitor purification of Ply to homogeneity as lysis of RBCs in solution could be easily monitored. A purification strategy that included acetic acid precipitation, ammonium sulfate precipitation, ion-exchange chromatography, gel filtration, and preparative acrylamide gel electrophoresis led to purification of Ply [108]. With purified Ply in hand, researchers were then able to convincingly show that Ply was sufficient to reproduce the toxic lethal effect of 5*. pneumoniae culture filtrates [108]. A second example of a purely biochemical identification of a pathogenicity factor is the isolation of cholera toxin, which built on the prior observations of De [27] and Dutta and colleagues [33] that the profuse watery diarrhea of cholera could be reproduced by enteral administration of sterile culture filtrates to animals. This observation, along with the ability to monitor purification by bioassay in infant rabbits and by immunoassay using polyclonal antiserum, led to the isolation of cholera toxin in a pure form using a purification strategy similar to that described above for Ply [38]. This purification precipitated many elegant studies (reviewed in [37]) that revealed cholera toxin's A-B5 subunit structure, its ganglioside GMj host cell receptor, and its primary mode of action—activation of host cell adenylate cyclase. The application of these purely biochemical approaches to other pathogens has been particularly successful in the identification of factors such as diphtheria toxin, pertussis toxin, and shiga toxin, and unquestionably validates the usefulness of such approaches. The potential for application of biochemical isolations is infinite and is Umited only by the ability of the researcher to measure the specific activity of a factor during purification. Potential applications could utilize prior understanding of enzymatic differences between pathogen and host. For instance, phosphorylated protein tyrosine residues are rare in bacteria, yet prevalent in signaling pathway components in human cells. Therefore, the Yersinia YopH protein, which was in actuality identified by other methods [12], could instead have been identified by a search for protein tyrosine phosphatase activity in whole cell bacterial extracts. The presence of such activity in a bacterial pathogen that is known to possess few examples of protein tyrosine phosphorylation could indicate a potential mode of interaction with the host by modification of host protein phosphorylation patterns. Understanding differences between host and pathogen enzymatic functions has many implications for application of biochemical strategies to isolate bacterial factors that specifically affect the host cell. For instance, though they themselves do not produce actin, many bacterial pathogens utilize host cell F-actin for attachment and intracellular spread. Therefore, one could conceivably design screens for bacterial factors that possess the ability to nucleate host cell G-actin

4.


139

or modify F-actin in vitro. Additionally, targeted biochemical screens could be designed to identify factors that proteolytically cleave key host proteins, phosphorylate or dephosphorylate protein serine and/or threonine residues, or covalently modify in other ways certain host proteins. Knowledge of differences between host and pathogen activities can also be used to screen for bacterial factors that alter normal host responses. For example, a recent study found that the Mycobacterium tuberculosis pathogenicity factor, lipoarabinomannan, stimulates the activity of a macrophage tyrosine phosphatase both in vivo and in vitro. This activity possibly inhibits proper macrophage function to provide a means of avoidance of this component of the host immune response [63]. Similar searches for factors that affect host enzymatic functions could also be conducted and could prove beneficial in the elucidation of heretofore undiscovered bacterial pathogenicity factors. Purely biochemical strategies for the discovery of bacterial pathogenicity factors have several advantages: (1) Many pathogenic bacteria are not easily genetically manipulatable and therefore are not easy targets for screens that employ large-scale mutagenesis strategies. Biochemical strategies to identify pathogenicity factors are therefore much more advantageous in these organisms. (2) Some pathogenicity factors are essential for bacterial growth in vitro and thus would never be isolated using genetic strategies that rely on mutagenesis. Therefore, biochemical isolations that do not involve disruption of normal bacterial processes allow isolation of these essential factors. (3) The activities of some pathogenicity factors are known, and this knowledge can facilitate design of strategies specific for the factor responsible for the activity. Despite these strengths, purely biochemical strategies for pathogenicity factor discovery also possess some disadvantages: (1) The advantage presented above as #3 can also serve as a major disadvantage when employing biochemical strategies to look for pathogenicity factors. The isolation and purification of factors that possess unknown functions is impossible since the ability to assay for function is critical to all of the classical biochemical strategies. (2) Biochemical purifications are often laborious as they rely on multiple purification steps that need to be designed or empirically determined to preserve specific activity of the factor one is attempting to isolate. For these reasons, genetic strategies are often more amenable for identification of pathogenicity factors having novel or unknown functions.

B.

Chemical Modification Screens

The first step in bacterial pathogenicity commonly involves adherence of the infectious agent to host cells or tissue. Because most of the host-pathogen interactions that allow this attachment to occur involve intimate contact between the bacteria and host cells, the outer surface of the bacterial cell has served as a

140

ANDREW CAMILLI. D. SCOTT MERRELL, AND JOHN J. MEKALANOS

prime hunting ground for factors involved in this aspect of pathogenicity. This section focuses on a few of the numerous biochemical strategies that have been used to dissect apart the intimate interactions that occur on the outer surface of bacterial pathogens. Although not a biochemical technique per se, direct visualization of bacterial surfaces by microscopy, in conjunction with fixation or biochemical staining procedures, can be used to identify surface-exposed pathogenicity factors. For example, light microscopy of bacterial cells stained with india ink and electron microscopy of polycationic ferritin-stained thin sections of bacterial cells have allowed visual detection of polysaccharide capsules surrounding some pathogens. Although much is already known about the roles of bacterial polysaccharide capsules in pathogenicity [6, 32], application of microscopy techniques and the ability to directly visualize capsule around newly isolated pathogens is useful in that it can provide an immediate advance in one's understanding of the potential pathogenicity of that organism. For example, two reports in 1994 demonstrated that a new epidemic strain of V cholerae, serotype 0139, expressed a polysaccharide capsule. This finding was novel in that the presence of a polysaccharide capsule had never before been seen in epidemic strains [57, 125]. Although the precise role of the capsule in the pathogenicity of this intestinal pathogen is not readily clear, it is possible that the capsule might be involved in adherence to the mucosal layer of the small bowel, or involved in resistance to certain host bactericidal mechanisms such as complement-mediated lysis. To this end, mutants that have lost the ability to express capsule have been shown to be attenuated for pathogenicity in an animal model of cholera [120]. This finding underscores the usefulness of visualization of bacterial surfaces as a means of detection of pathogenic factors. Electron microscopy has many orders of magnitude greater resolving power than light microscopy, and has been used to visualize the very long (0.5-15 |im) and thin (1-11 nm) adhesive fibers, called pili, that extend from the surfaces of many Gram-negative bacteria. Although it should be noted that not all pili expressed by pathogens play a role in human disease (e.g., [112]), the discovery of pili on the surface of a suspected or known pathogen provides a good starting point for further investigation, as some pili have been shown to be essential factors for establishment of infections. Direct application of both electron microscopy and light microscopy are therefore attractive in that they can provide immediate knowledge about surface structures such as pili or capsules. It should be noted, however, that other methods are required to assess the putative roles of these surface structures in pathogenicity. In addition to direct visualization of surface structures, electron microscopy and light microscopy have also proven themselves valuable tools for indirectly monitoring the presence of many bacterial surface adhesins—for example, surface factors involved in such diverse activities as in vitro autoagglutination [113], attachment to cultured host cells [117], and invasion into cultured cells [56]. Once

4.

STRATEGIES TO IDENTIFY BACTERIAL PATH(3GEMCITY FACTORS

141

again, though, other methods are ultimately needed to identify these pathogenicity factors, highlighting the importance of combining multiple techniques to gamer a greater understanding of host-pathogen interactions. Direct chemical modification of bacterial surface macromolecules using reagents that are unable to cross the bacterial outer membrane can facilitate identification of pathogenicity factors. The inability of these reagents to be internalized not only facilitates specific tagging of surface-exposed molecules but also often facilitates the purification and subsequent characterization of surfaceexposed pathogenicity factors. Two frequently used methods of surface modification involve radioiodination and biotinylation of intact cells followed by purification and electrophoretic separation of membrane proteins. Surface molecules that have been radioiodinated can be directly visualized by autoradiography following separation by gel electrophoresis, hi contrast, biotinylated molecules can only be visualized after binding of a secondary molecule such as streptavidin that possesses a high binding affinity for biotin. The secondary molecule is usually visualizable due to the fact that it has been either radiolabeled or conjugated to a colorimetric moiety such as alkaline phosphatase. Though these two surface modification techniques differ chemically, both have proven themselves useful for identification of surface factors involved in pathogenicity. Most recendy, these two labeling methods have been utilized to identify Bartonella henselae surface proteins, and to then further identify the subset of these surface proteins that function as adhesins for human endothelial cells [16]. Biochemical screens for pathogenicity factors that utilize chemical modification have advantages and disadvantages that force the investigator to consider the usefulness of the procedures. A primary advantage of chemical modification is that litde or no information need be known about the pathogenicity factors present on the cell surface other than the fact that they reside there. This allows the investigator to construct broadly based screens, which increases the chances of discovery of important pathogenicity factors. The primary limitations of these types of screens include the fact that many bacterial pathogenicity factors are indeed not expressed on the outer membrane and thus would be missed in screens that exclusively looked at this subset of molecules. In addition, isolation of surface-exposed molecules on a gel generally only provides a starting point for identification of putative pathogenicity factors, and must be followed up by other methodologies. For example, in order to assess a protein's role in pathogenicity, one might wish to mutate the gene encoding the surface protein by using reverse genetics.

C.

Zymography

Since some bacterial pathogenicity factors are enzymes, identification strategies that directly assay for enzymatic activity are useful in that they provide a more

142


direct method of identification. For example, many bacterial proteases are involved in cleavage of host cell extracellular matrix proteins or host cellular proteins. Cleavage of these host proteins often assists the pathogen in the establishment of an infection and sometimes aids the bacterium in avoidance of the host cell immune response. One strategy for identification of bacterial enzymatic pathogenicity factors is zymography. This strategy combines electrophoretic separation of bacterial proteins with the subsequent visual detection of enzymatic activity in situ, that is, in the gel itself. The most common zymographs target identification of proteases, as the construction of the substrate matrix is simple. For example, a denaturing acrylamide gel can be embedded with a substrate polypeptide such as casein. This gel would then be loaded with a protein preparation from a pathogen of interest, and electrophoresed to separate the secreted or membrane proteins. The separated proteins would then be renatured by subsequent removal of denaturants from the gel. Next, an incubation period would allow enzymatic proteolysis to occur. Finally, the gel can be stained with coomasie blue, and the presence of a small clearing zone within the casein matrix indicates that the protein species present at that location possesses proteolytic activity. This methodology relies on the fact that the embedded casein molecules do not migrate appreciably during electrophoresis due to their lack of substantial net charge. Candidate proteases can then be purified and further studied biochemically, or purified and subjected to reverse genetics to identify the encoding gene. Various zymography modifications that involve utilization of secondary gels impregnated with additional polypeptide substrates have been made (reviewed in [67]) and then utilized for the study and/or identification of several putative pathogenicity factors from a variety of different pathogens [22, 34, 41]. The usefulness of zymography in identification of pathogenic factors is not limited to proteases. In fact, many other enzymatic activities have also been identified by utilization of zymographic techniques. Examples of the diverse activities and factors that have been identified include a secreted S. sobrinus dextranase inhibitor believed to be involved in colonization of the oral cavity [111], a Staphylococcus aureus murein hydrolase that is under the regulation of the agr and sar pathogenicity gene regulators [44], and a Prevotella melanginogenica hemolysin that is produced by this periodontal pathogen [4]. Indeed, zymography is a method that can be adapted for identification of additional types of enzymes, and is advantageous in that the potential applications for identification of enzymatic factors are as limidess as the imagination of the investigator. It should be noted, however, that limitations to zymography do exist. First, embedding a large gel matrix with some substrates can be very costly. Second, many bacterial enzymes that are involved in pathogenicity may not be produced during normal laboratory growth or may not be present in an active form. Third, bacterial proteases often exhibit high substrate specificity, and the identity of the substrate may not be known. Fourth, detection of an enzymatic

4.


143

activity requires that the enzyme be not only stable but renaturable after electrophoretic separation. The latter problem may be solved by running a nondenaturing gel, but then information concerning the molecular weight of the enzyme may be lost. And, fifth, biochemical methods (such as zymography) assume that the activity being pursued is important for pathogenesis; however, this can only be assessed by side-by-side comparison of the corresponding mutant and wild-type strains, or by reproduction of the disease symptoms by the purified protein in the case of toxins.

D. Receptor/Ligand Affinity Screens Intimate interaction between host and pathogen occurs on many levels. Not only do most bacterial pathogens have to adhere to host cells or tissues, but often they produce polypeptides that subsequendy interact with host cell components. These interactions are often key steps in the development and progression of pathogenicity. Because of this, many biochemical screens specifically target pathogenicity factors that show the ability to bind to host molecules with high affinity. Receptor/ligand screens are one such type of screen and have proven beneficial in identification of a number of interacting components. Receptor/ligand screens in the most straightforward form involve the following steps. First, total protein from the pathogen of interest is prepared and then separated on a denaturing acrylamide gel. Following separation, protein constituents are renatured and subsequently transferred and bound to a solid membrane support. This membrane can then be exposed to any radiolabeled host ligand of interest. The immobilized bacterial polypeptide bands that interact with and are bound by the labeled host ligand can then be identified by autoradiography. This type of receptor/ligand screen has been used successfully numerous times. For example, a S. pyogenes adhesin was identified using radiolabeled collagen as a ligand [118]; and in a more recent example, the Enterobacter cloacae mannosesensitive pilus adhesin for human cells was identified from purified pili subunit polypeptides after exposure to mannose-containing biotinylated albumin [92]. In addition to the basic receptor/ligand screen, variations of this technique have been used to study the binding of bacterial proteins to immobilized host ligands (e.g., [129]), and the binding of entire host cells to immobilized bacteria expressing an adhesin (e.g., [70]). Yet another variation of this method combines the expression of individual pathogen proteins on the surface of a surrogate host bacterium (that lacks the binding activity being studied) with affinity binding selection of surrogate host cells expressing the receptor to a ligand matrix. This methodology has been used to identify a Neisseria gonorrhoeae adhesin for the host cell ligand, gangliotetraosylceramide [93]. Affinity binding can often be used as a direct method for the identification of bacterial pathogenicity factors. For instance, both panning and affinity chroma-

144


tography utilize immobilized host ligand as a means of purifying bacterial factors that interact with specific host substrates. A more recent innovation in these techniques has been developed, which is called Receptor activity-directed affinity Tagging (ReTagging) [55]. ReTagging is beneficial in that it utilizes a UV light-activatible crosslinker to covalently attach a biotin molecule onto a bacterial receptor when it comes into close contact with the host ligand (Fig. 1, see color plate). The attachment of the biotin molecule then allows direct identification and purification of the bacterial factor to which it is bound using streptavidin-coated magnetic beads (streptavidin possesses high binding affinity for biotinylated compounds). This method was recently utilized to identify the Helicobacter pylori receptor for the Lewis B histo-blood group antigen [55]. This simultaneous identification and purification of the bacterial receptor is a major advantage of the utilization of ReTagging. In addition, the affinity of the receptor-ligand pair enables identification and purification of low-abundance receptors.

E. Immunological Methods Many bacteria disrupt or circumvent the host immune system in order to produce a productive infection. Thus, it is fitting that researchers have developed strategies for discovery of pathogenicity factors that exploit the immunological system. Such strategies are usually based on the ability of antibodies to protect a host or surrogate host from disease. Before application of this strategy, a bank of B-lymphocyte hybridomas that produce monoclonal antibodies (MAbs) directed against antigens exposed on the bacterial surface must first be generated. Subsequently, each individual MAb can then be screened for the ability to protect the host from disease. The inability to cause disease represents a positive result and might indicate that the MAb is binding to and blocking the function of a pathogenicity factor residing on the surface of the pathogen. It should be noted, however, that mere binding of the MAb to the bacterial cell surface can be sufficient to protect the host from disease in that antibodies can act as opsonins for complement deposition and phagocytosis. For this reason, it is necessary to show protection using purified Fab fragments of the MAb, as Fab fragments lack opsonizing activity yet can block the function of the bacterial factor. Of course, studies in surrogate host systems that lack complement and/or phagocytic cells (e.g., cultured mammalian cells grown in medium supplemented with heat-inactivated serum) do not require the Fab control experiment. Once an MAb has been identified that provides protection of the host or surrogate host from disease, the MAb then becomes a powerful tool that can be used to identify the surface molecule to which it binds. Solubilized membrane protein extracts can be used in conjunction with the MAb of interest to direcdy purify the bacterial factor. This methodology has been used to identify many diverse bacterial factors (e.g., [78, 99]). Recently, this method was also used to

4.


145

look for factors involved in adherence of Mycobacterium tuberculosis to host cells. In this study an MAb that blocked adherence of M. tuberculosis to cultured mammalian nonprofessional phagocytic cells was identified from a bank of MAb generated against intact M. tuberculosis cells. This MAb was then used to purify and identify a phosphatidylinositol mannoside as the adhesin [52]. Since the production of hybridomas and subsequent screening of many individual MAbs can be very laborious, an often used variation on the immunological approach described above has been developed that uses the host immune system to select for immunogenic bacterial surface antigens during a sublethal infection. In this way, convalescence serum antibodies can then be gathered and used to identify bacterial surface proteins that were expressed during infection. In the most common implementation of this approach, bacterial outer membrane proteins are then electrophoretically separated or prepared as part of a bacteriophage X expression library. These polypeptide species can then be screened using the convalescence serum for the subset that were immunogenic during the course of the infection. Detection of protein species bound by serum antibodies requires binding of a secondary antibody conjugated to a colorimetric moiety. A further subset of these antigenic factors may indeed be pathogenicity factors, and their identification as such requires additional experimentation. This modified immunological approach has been beneficial as sera from lepromatous leprosy patients used to screen a X cosmid expression library of M. leprae genomic DNA [106], and resulted in the identification of a 15-kDa protein that is immunogenic to both B- and T-lymphocytes. The exact role this protein plays in pathogenicity remains as yet to be demonstrated. The application of immunologic strategies is advantageous in that it allows a broad approach to target for identification factors exposed on the bacterial surface. However, three limitations include the fact that these screens can often be quite laborious, many pathogenicity factors are either not exposed on the bacterial surface or are poorly immunogenic and thus would never be found (e.g., shiga toxin), and many highly immunogenic surface antigens are not pathogenicity factors (e.g., S. typhimurium flagella [76]).

F. Subtractive Hybridization The facts that related bacterial strains often have quite different pathogenic abilities and that gene expression patterns of pathogenic bacteria often change dramatically when going from an in vitro to an in vivo environment have led to the development of techniques that seek to identify differential expressed genes. One such technique is nucleic acid subtractive hybridization. In the first case, nucleic acids from two related bacterial strains, one pathogenic and the other nonpathogenic, can be physically compared to identify differentially expressed genes or differences in gene content that may be responsible for the

146


pathogenicity difference. In the second situation, the nucleic acids from a single strain grown under a nonhost condition, and a host-like condition can be physically compared to identify differentially expressed genes, with the understanding that some of these may encode pathogenicity factors. The foundation of the subtractive hybridization approach relies on the fact that DNA sequences common to both (denatured) DNA preparations can be subtracted out as doublestranded DNA hybrid duplexes while unique single-stranded sequences remain. The "nonpathogenic" DNA is supplied in molar excess to ensure that complete hybridization of common sequences occur. In addition, the nonpathogenic DNA is typically modified by biotinylation so that subsequent to the annealing reaction the common DNA hybrids can be removed by utilization of a streptavidin matrix. Thus, only the single-stranded fragments of interest are left behind. It should be noted that subtractive hybridization can be performed using either cDNA or genomic DNA derived from the strains of interest. However, differences in gene expression in a single strain grown under in vitro and host-like conditions can only be detected by this method using cDNA. The traditional subtractive hybridization protocols rely simply on subtractive hybridization followed by either cloning of the end products or utilization of the products as probes to identify the genes from genomic libraries (e.g., [15, 95]). A more recent subtractive hybridization protocol has been designed to incorporate modifications that vastly improve the efficiency of the technique. This modified procedure has been called Representational Difference Analysis (RDA) [75]. RDA has been applied to discover bacterial pathogenicity factors. Some examples are the identification of genetic differences between the two closely related pathogens Neisseria meningitidis and A^. gonorrhoeae, which might be responsible for their different etiologies [114], and the recent identification of genetic differences between an emergent epidemic strain of V. cholerae, serogroup 0139, and its progenitor serogroup Ol strain [17]. The RDA method involves initial restriction enzyme digestion of DNA from the virulent strain (called the tester) followed by ligation of a PCR adapter/primer onto the ssDNA overhangs formed by digestion of the tester DNA. Next, the modified tester DNA is amplified by PCR. This amplified tester DNA can then be mixed with a molar excess of digested, but otherwise unmodified, DNA from the nonvirulent strain (the driver), and the mixture is denatured and allowed to anneal. During the course of the annealing reaction, most ssDNA tester fragments anneal with complementary driver ssDNA fragments. These dsDNA hybrids represent sequences common to both the tester and the driver and are thus not of interest. In contrast, though, ssDNA tester fragments that are unique cannot hybridize with driver ssDNA, but instead can only anneal back to the complementary tester ssDNA fragment. This restores the original adapter/primer arms on both ends of the resulting dsDNA fragment. These dsDNA tester fragments can then be reamplified by PCR. The resulting fragments represent unique sequences and can be further enriched by additional rounds of subtractive hybridization and

4.


147

amplification. As a final step, the amplified unique tester DNA fragments can be cloned and further analyzed. Despite improvements such as RDA, subtractive hybridization methods are still somewhat laborious and are fairly insensitive, that is, many genetic differences between tester and driver are not identified due to a variety of technical difficulties. A more recent technical improvement to the RDA protocol, however, increases both the efficiency and sensitivity of the method [3]. In the first implementation of this method, which is a slighdy modified version of the Suppressive Subtractive Hybridization (SSH) technique originally developed for use in eukaryotes [28], more than a dozen genomic sequences present in a monkey-isolate of H. pylori (tester) and not present in a human isolate (driver) were identified. Many of the identified sequences were smaller than the limit of detection of tradidonal subtractive hybridization screens, thus validating the improved sensitivity of this technique [3]. The method for SSH is similar to RDA but contains one major procedural difference. Instead of ligating a single adapter/primer onto the ends of the tester DNA restriction fragments, two different adapter/primers are used (Fig. 2, see color plate). First, the digested tester dsDNA is divided into two pools, 1 and 2. Adapter/primer 1 is then ligated onto fragments contained in pool 1, and adapter/primer 2 is ligated to fragments contained in pool 2. Each pool is then subtractively hybridized, separately, to a molar excess of unmodified driver DNA, as described above. After annealing, pools 1 and 2 are mixed together and another molar excess of denatured driver DNA is added. This final mixture is then allowed to undergo additional annealing. During this second annealing reacdon, any remaining common ssDNA from the tester DNA from pools 1 or 2 can anneal to the additional driver ssDNA added to the second annealing reacdon. Likewise, any unique tester ssDNA fragments originally derived from pool 1 sdll remaining after the first annealing reaction can hybridize with a complementary ssDNA tester fragment from pool 2. These latter dsDNA tester fragments will thus contain adapter/primer 1 at one end and adapter/primer 2 at the other. Unique tester dsDNA fragments that contain both adapter/primers 1 and 2 at their two respective ends can then be amplified exponendally in a single PCR reacdon using forward and reverse primers complementary to adapter/primers 1 and 2, respecdvely (Fig. 2, see color plate).

G. Differential Display Since most bacterial pathogens must express different subsets of genes on infecdon of a host, many different methods have been developed that attempt to direcdy analyze these differentially expressed gene products. One approach that is related to, yet distinct from, subtractive hybridizadon, is Differential Display (DD). The DD methodology was originally designed to study differendal gene

148

ANDREW CAMILLI. D. SCOTT MERRELL, AND JOHN J. MEKALANOS

expression in eukaryotes [71] but has been modified to allow for screening of differentially expressed genes between two populations of bacteria [127]. Because of this, DD can be used to identify pathogenicity factors by comparing gene expression either between a pathogenic and a nonpathogenic strain of one species or, between a strain grown in vitro and in vivo. DD allows for differences in gene expression to be visually determined by direct comparison of two sets of randomly primed PCR products, which are generated from the two cDNA pools to be compared. Pools of cDNA are first generated from total RNA purified from each bacterial population. These cDNA pools are made using an arbitrary primer of approximately 18 bases in length. Utilization of an arbitrary primer and low-stringency hybridization allows priming to occur at pseudo-random sites during the reverse transcription reaction. These cDNAs are then PCR amplified in a second reaction, wherein the first amplification cycle uses a low-stringency annealing step that results in products that have the primer sequence at both ends. These double-end primed products are then amplified exponentially in subsequent rounds of the PCR using the arbitrary primer, a stringent annealing temperature, and radioactive deoxynucleotides. After the final PCR amplification, PCR products are separated on an acrylamide gel and visualized by autoradiography to screen for band differences between products derived from the two cell populations. Unique bands represent genes that were differentially expressed between the two cell populations. Finally, these differentially expressed bands can be cut out of the gel and sequenced. DD has been successfully used to identify several bacterial pathogenicity factors. For example, S. typhimurium genes that are stress-induced by exposure to hydrogen peroxide were identified in the study that first reported the DD method modified for use in prokaryotes [127]. More recently, a slightly modified version of this protocol was used to identify a Legionella pneumophila gene that was induced during intracellular infection of macrophages [1]. In this report, cDNAs were made using random hexamer primers, and PCR amplification was done using pairs of arbitrary primers that were 17-23 bases in length. The strengths of DD depend a great deal on a number of experimental parameters. In particular, the length of the arbitrary primer(s) and stringency conditions used directly correspond to the number of bands that one generates per reaction. Generally between 1 and 50 bands are produced per reaction, but since many of these bands are artifactual—for example, corresponding to ribosomal RNA-derived products—it is necessary to screen products generated in several different reactions each using different arbitrary primers. This multipronged screening strategy greatly increases the likelihood of identifying bands that correspond to bona fide differentially expressed genes. Because of these technical limitations, DD is not a comprehensive method by any means and can be quite laborious. However, there are some procedural and theoretical advantages of DD. For example, very little total RNA is needed as starting material, so that DD can be used on relatively small numbers of bacteria recovered after exposure to a wide

4.


149

variety of in vitro and in vivo conditions. In addition, DD is reported to be a sensitive method, able to identify genes that are transcribed at very low levels. H. Reverse Genetics Many of the methods described above (II.A-II.E) result in identification of a polypeptide with the potential to play a role in pathogenicity. Although this is a reasonable endpoint for some studies such as those whose goal it is to identify and biochemically characterize a toxin, it is usually of further benefit to proceed with identification of the gene encoding the polypeptide. Not only does this open the door to many additional studies, it can allow for precise determination of the actual role in pathogenicity of the discovered factor. Specifically, after the gene responsible for a putative pathogenicity factor has been identified, it can be mutated and the resulting strain determined to lack the polypeptide of interest. The pathogenicity of the mutant strain can then be compared in side-by-side experiments with that of the isogenic parental strain in, for example, an animal model of disease to assess the relative importance of the putative pathogenicity factor (see III.B.3). In addition to this validation, the gene can then be manipulated to allow purification of large quantities of the polypeptide for additional biochemical, immunological, or structural studies. Although there are several methods by which the gene encoding a polypeptide can be identified, utilization of reverse genetics provides a much more efficient method than, say, random mutagenesis followed by screening for loss of production of the polypeptide. Reverse genetics proceeds by first determining the amino acid sequence of a portion (at least five contiguous residues) of the purified polypeptide by protein microsequencing [65]. Second, the codon usage preference of the bacterial species being studied is used to design and synthesize degenerate DNA oligonucleotides that are predicted to code for the determined amino acid sequence. Third, the degenerate oligonucleotides are used to identify the gene from the bacterial genome. This can be done directly by Southern blot analysis using the oligonucleotides as a probe, or indirectly by PCR amplification of an internal portion of the gene using degenerate oligonucleotides that correspond to two separate parts of the gene, followed by use of the PCR product as a Southern blot probe. This latter methodology requires that the amino acid sequence of two separate parts of the polypeptide be known. Reverse genetics has been used numerous times in both of these manners to identify bacterial genes encoding pathogenicity factors (e.g., [30, 53, 73]).

///. Genetic Screens Coincident with gaining the knowledge that DNA was the genetic material came the realization that DNA also encoded the biosynthetic information for pathogenicity factors [7]. The five and a half decades that have followed this seminal discovery have seen the development of many experimental techniques that have

150


potentiated advances in understanding of bacterial pathogens and their ability to cause disease. Techniques such as sequence-specific DNA cleavage, ligation, and transformation, coupled with various mutagenesis protocols, allow scientists to manipulate DNA with amazing power. It is this ease of genetic manipulation that now allows scientists to identify and characterize pathogenicity factors from many organisms. It should be noted, however, that, even though genetic techniques have proven invaluable to culturable and transformable pathogens, many organisms are still unable to be grown in vitro and/or are untransformable. These organisms present a unique challenge to scientists wishing to study them and emphasize the importance of continued advances in molecular techniques. The remainder of this chapter attempts to describe many of the genetic strategies that have been employed throughout the last few decades of the twentieth century to facilitate identification of bacterial pathogenicity factors. Each technique, as described, can result in identification of a bacterial gene that is required for production of a putative pathogenicity factor. In practice, it is evident that the power of genetic techniques is only fully realized when used in combination with immunological or biochemical strategies. By implementation of combinations of strategies and techniques, scientists are able to gain a fuller understanding of the roles of pathogenicity factors in the disease process. Although a large number of genetic screens and selections have been designed to identify pathogenicity factors, only a sampling of these will be covered here for the sake of brevity. Emphasis will be placed on those strategies that have been particularly successful and/or those that introduce an interesting or novel concept.

A. In Vitro Screens 1.

LARGE-SCALE SCREENINGS

Most genetic strategies for identification of pathogenicity factors have in common the need to mutagenize the bacterial genome in a random fashion to generate a bank of mutant strains. Ideally, the mutagenesis is conducted in a manner such that each mutant strain contains a single, unique mutation, and the number of strains in the bank is comprehensive enough to ensure that most genes will have been mutated. The Poisson distribution equation, yV=[ln(l-P)]/[ln(l

-F)l

can be used to calculate the minimum number of strains (AO required to assure (with probability [P]) that any particular gene will be mutated, where F is equal to the average gene size divided by the genome size. For example, given an average gene size of 1 kb in a genome size of 4 Mb, the minimum number of strains in a mutant bank sufficient to assure that 95% of all genes were mutated would be 11,981. After construction of the mutant bank, a screen or a selection

4.


151

for strains with changes in pathogenicity is conducted. Intrinsically, a screen is more labor intensive than a selection, as it involves examination of individual strains to find those with the sought-after phenotype. In contrast, a selection is designed to enrich only those strains having the desired phenotype. In this way, the scientist is left with only the strains of interest at the completion of the enrichment. The most laborious yet comprehensive genetic screens are conducted on a large scale, whereby mutagenized strains are individually screened, one at a time, in some assay of pathogenicity chosen by the investigator. This method is powerful in that it allows the investigator the opportunity to assess each individual mutation's effect on the pathogenicity of the organism. In this manner, knowledge concerning each individual gene can, in theory, be gained. A classic example of a large-scale screening is the screening of 9,516 individual transposon insertion strains of 5'. typhimurium for loss of infectivity in microtiter plate wells containing monolayers of cultured murine peritoneal macrophages [36]. The infectivity of each strain was assessed by lysing each monolayer and titering viable bacteria on agar plates. From this screen, 83 strains exhibiting reduced pathogenicity were identified. Although laborious, large-scale screening is advantageous in that, if a large enough bank of mutant strains is screened, it is theoretically more comprehensive than any possible selection. The reason for this comprehensiveness is simple: in a large-scale screen each mutant strain is individually assayed for pathogenicity, whereas in a genetic selection strains that exhibit slightly attenuated pathogenicity are usually lost (or ignored). For example, if one were to search for gene mutations that adversely affected bacterial multiplication in the presence of hydrogen peroxide, a reactive oxygen species produced by host cell macrophages, a mutation that only slightly decreased (or increased) multiplication could be picked up in a screen. However, in a penicillin-based selection, where the penicillin selects for strains that are no longer able to replicate by killing all multiplying cells, a mutant strain that exhibited a slightly reduced multiplication rate would still be killed by the penicillin. Mutants that displayed an enhanced multiplication rate would also be killed in a selection of this type. All genetic selection strategies suffer this type of limitation to some degree. On the other hand, there are several major advantages to the selections discussed below. Because of the different strengths and weaknesses of different genetic strategies, investigators must carefully determine which strategy is most appropriate for isolation of the types of genes being sought after. When conducting genetic screens or selections, it is important that the researcher consider not only the type of screen or selection but also the method utilized for mutagenesis of the bacterial pathogen. As stated above, in most cases a bank of mutant strains containing highly random, single mutations is most desirable. Since the choice of mutagenesis strategy is an important consideration in all types of genetic screens and selections, a brief discussion of the advantages

152


and disadvantages of several commonly used methods is included within this section. Physical and chemical methods for generating point mutations, such as UV light and A^-methyl-A^-nitro-A^-nitrosoguanidine, are probably the most frequently utilized types of mutagenesis due to the fact that they are highly random and can generate a tremendous number of different mutant alleles for any gene. The list of possible types of mutant alleles obtainable by utilization of these methods is about as complete as possible in that one can obtain null, partial function, dominant negative, dominant positive, suppressor, promoter-up, promoter-down, polar, or nonpolar mutations. However, a major disadvantage of these methods can be found in that multiple unlinked mutations are commonly introduced into individual cells. The presence of multiple mutations within a single strain gready complicates subsequent identification of the relevant mutation. This problem, however, can be overcome if an efficient means of genetic complementation exists for the pathogen being studied. Another commonly used strategy for mutagenesis involves utilization of transposon insertions. This method is advantageous in that it can be used to make stable, single insertions, and subsequent linkage of the transposon with the mutation greatly facilitates identification of the disrupted gene. Disadvantages of transposon mutagenesis include the fact that even the best of transposons insert only pseudo-randomly, which provides a much more limited spectrum of mutant alleles (relative to chemical mutagenesis), and finally, some pathogens, such as Campylobacter jejuni [64], are not currendy amenable to transposon mutagenesis. An in vitro transposidon protocol using a mariner transposon has been developed [65, 102] that improves generalized transposon mutagenesis in three ways. First, mariner can insert with a high degree of randomness in that it requires only an 5'-AT-3' dinucleotide target site. Second, mariner mini-transposons have been recendy engineered that produce either polar or nonpolar inserdon mutations (B. Akerley, E. Rubin, J. Mekalanos, and A. Camilli, unpublished data). Third, the products of in vitro transposition can be moved into transformable bacterial species followed by double-crossover insertion events, thus obviating the need for transposidon in vivo [2]. Suicide plasmid insertion-duplication mutagenesis is an alternative approach that theoretically can be more random than transposon mutagenesis, provides a linked marker for gene recovery, and can generate both polar or nonpolar mutations in most genes. In this method, random small fragments of the pathogen genome are ligated into a suicide plasmid vector, and the recombinant plasmids are then introduced into the pathogen by transformation, whereupon they integrate into the chromosome by homologous recombination. However, generadng a null allele in small genes can be difficult or impossible given that an internal portion of the gene coding sequence usually needs to be inserted into the suicide vector in order to ensure disrupdon of the gene, and given that the likelihood of proper inserdon (by homologous recombinadon) is limited direcdy by the size of the

4.


153

coding sequence contained within the plasmid. Despite the limitations of transposon and suicide plasmid insertion-duplication mutagenesis, both of these methods have the major advantages of: (1) establishing linkage between the observed phenotype and inactivation of a particular locus, and (2) providing a selectable marker tighdy linked to the mutation to allow the mutation to be moved to different genetic backgrounds. 2.

TARGETING EXPORTED PROTEINS

As was discussed in section II.B, many pathogenicity factors are surface-exposed or secreted proteins. A common requirement for all of these proteins is that they be exported across the bacterial cytoplasmic membrane. Several genetic strategies have been designed to target exported proteins for identification and mutagenesis. The most successful and widely used of these is TnphoA mutagenesis. This strategy combines transposon mutagenesis with a series of screens. First, a prescreen to identify strains harboring insertions into genes encoding exported proteins is conducted, and subsequently this subset of strains is screened to identify nonpathogenic mutant strains in a model of disease [113]. This approach relies on the use of TnphoA, which contains at the 5' end of the transposon a truncated/?/zoA gene that lacks a promoter and the DNA region encoding its signal sequence [84]. Because PhoA requires export to the periplasm for disulfide bond formation and subsequent enzymatic activity (as a phosphatase), only those TnphoA insertions occurring into genes encoding extracytoplasmic proteins will result in export of the protein fusion (if the fusion is in the correct reading frame). These exported PhoA translational fusion products are then enzymatically active and can be assayed for colorimetrically by the formation of blue colonies on media containing 5-bromo-4-chloro-3-indolyl-phosphate. This TnphoA mutagenesis strategy has been used to identify many pathogenicity factors such as the V cholerae toxin-coregulated pilus (TCP) pilus [113], Escherichia coli (EPEC) chromosomal and plasmid genes required for adherence to host cells [29], and S. choleraesuis genes required for transcytosis across epithelial cell layers [39]. In addition, other phoA fusion vectors have been constructed and used to identify secreted proteins from other bacterial species in which TnphoA does not function properly [59]. Despite the fact that the TnphoA strategy has been employed so successfully, there are of course limitations to the strategy that must be considered when deciding the appropriateness of the screen. By definition, the techniques described in this section preclude the identification of nonexported pathogenicity factors. Secondly, the expression of some fusion products can be toxic to the host bacterium. Aside from excluding some genes from ever being detected, that is, those whose expression when fused to phoA is deleterious to growth in vitro, this phenomenon poses a problem when assessing the effect of loss of an infectioninduced gene on pathogenicity. Specifically, an attenuated level of pathogenicity may result partially, or totally, from slower /// vivo growth of the bacteria due to

154


toxic effects of the fusion product. Methods to alleviate this limitation have centered around the utilization of alternative reporter fusions that are perhaps not as deleterious to the cell. For instance, several other reporter systems have been developed that also identify exported proteins (e.g., p-lactamase [91, 101]), invasin [128], and nuclease [97]). However, as most of these alternative reporters allow for selection of strains expressing an exported fusion product, they will be discussed below under genetic selections. 3.

COORDINATE REGULATION SCREENS

Not only did the original V cholerae study using TnphoA take advantage of the knowledge that exported factors are often pathogenicity factors (III.A.2); it also took advantage of a pre-screen strategy to collect only strains harboring fusions that were coordinately regulated with a previously known pathogenicity factor [113]. Specifically, Tn/7/26>y4-mutagenized V. cholerae strains were grown under two separate in vitro conditions known to downregulate and upregulate, respectively, the expression of cholera toxin. Fusion strains expressing alkaline phosphatase activity only in the upregulatory growth condition could then be said to harbor TnphoA insertions in genes encoding exported polypeptides that are coordinately regulated with cholera toxin. The premise behind this strategy relies on the fact that many pathogenicity factors may often be coordinately regulated with a previously known pathogenicity factor. This premise has now been realized in several other pathogens aside from V cholerae. For instance, both the bvgAS regulon in Bordetella pertussis [40], and the phoPQ regulon in S. typhimurium have been shown to coordinately regulate expression of many pathogenicity factors [89, 110]. There are many other coordinate regulatory screens that have been designed that vary in the choice of reporter gene. Frequently used reporter genes include promoterless lacZ [126] and gfp [115] alleles. In contrast to the TnphoA translational fusion strategy, transcriptional fusions to cytosolic enzymes such as lacZ allow identification of virtually any gene that is coordinately regulated, that is, regardless of whether the gene encodes an exported or cytoplasmic protein. The two primary limitations of coordinate regulatory screens, though, are that, as discussed above, some gene fusions will be toxic to the bacteria, and the in vivo environmental conditions that induce expression of certain pathogenicity gene regulons remain unknown or are difficult to reproduce in the laboratory.

4.

HOST MIMICRY SCREENS

The requirement for bacterial pathogens to differentially regulate expression of genes depending on their environment continues to be a theme that can be exploited by the scientist to attempt to elucidate factors required for pathogenicity.

4.


155

In fact, this common theme runs through many of the strategies designed to elucidate pathogenicity factors that have been the most successful throughout the last few decades. Due to the facts that the environmental challenges that a pathogen encounters in its host are often different from one tissue to the next and that host environments are quite complex, screens designed to mimic certain host environmental parameters provide an attractive method for simplification of the process of pathogenicity factor identification. Although numerous mimicry screens have been designed, the most straightforward have proven to be those designed to look for gene fusions to a reporter such as lacZ, which are specifically induced during growth in a laboratory medium that mimics one particular host parameter. Examples of successfully employed screens include the identification of acid-inducible genes from various intestinal pathogens by exposing the bacteria to media having acidic pH values similar to those found in the human stomach or phagosomal microenvironment (e.g., [42]), and identification of high-temperature-inducible genes from facultative pathogens using an in vitro growth temperature equivalent to human body temperature (e.g., [5]). Host mimicry screens have not been limited to utilization of lacZ but have employed other reporter genes such as luciferase, gfp and phoA. In addition, a variety of host mimicking conditions have been employed. Due to the fact that many pathogenicity genes require more than a single signal to induce their expression, it is often advantageous for the investigator to be able to mimic several host signals at one time. It is for this reason that one commonly used in vitro model system is cultured mammalian cells. Cultured mammalian cells often provide many of the signals needed to simulate infection and have thus been used to identify pathogenicity factors produced by both adherent extracellular pathogens (e.g., [49, 94]) and intracellular pathogens (e.g., [61]). A recendy developed strategy combines the use of host macrophages as an inducing environment with Fluorescence-Activated Cell Sorting (FACS). This screen was utilized to search for S. typhimuriiim genes within the SPI-2 pathogenicity island locus, which are induced within the intracellular milieu of macrophages [24]. This study utilized gfp fusions to genes within the SPI-2 locus and resulted in identification of several genes that are transcriptionally induced within host cells and that are required for systemic spread in a mouse model of typhoid. One of the major advantages of host mimicry screens is the fact that there are coundess host-mimicking condidons that can be devised. In addition, most of these screens can be conducted with relative ease due to the fact that they take advantage of in vitro conditions to mimic the often more complex in vivo environment. In pardcular, the GFP/FACS screening strategy is attracdve due to the fact that it should allow near-saturating screens to be performed, as a thousand-fold enrichment of GFP-posidve bacterial cells can be achieved in a short period of dme [116]. However, it should be noted that host mimicry screens do suffer from the limitation that the host conditions required to trigger expression

156


of some bacterial factors remain unknown or may be too complex to reproduce in the laboratory. This is less true for tissue culture models of infection than for other in vitro systems, but it applies nonetheless. Because of this limitation, several in vivo screens have been recently developed that utilize the intact host animal to provide a more complete inducing environment (see below).

B. In Vivo Screens 1.

RECOMBINATION-BASED IN VIVO EXPRESSION TECHNOLOGY

Because utilization of large numbers of animals as disease models can be quite costly and trying due to ethical and logistical reasons, scientists have spent a considerable amount of time trying to minimize the necessity to conduct large screens that require intact animals. It is for such reasons that large-scale screens (III.A.l) using animals as hosts can be quite limiting since it is required that each individual strain of interest be screened in a separate animal. Efforts to develop screens that allow multiple mutant strains to be tested per animal have recendy been inspired by the advent of In Vivo Expression Technology (I VET). I VET is a series of genetic selection methods devised to identify bacterial genes that are transcriptionally induced during infection of a host animal (see IV.D). These efforts have resulted in the development of two new screens: RecombinationBased In Vivo Expression Technology (RIVET), and Signature-Tagged Mutagenesis (STM). Each of these screens addresses fundamentally different questions, but both are advantageous in that they allow for utilization of fewer numbers of intact animals to screen for genes of interest. With RIVET, bacterial genes that are transcriptionally induced during infection are identified, while with STM (described in III.B.2) genes that play an essential role in pathogenicity are identified. There are two major obstacles that must be overcome to comprehensively screen for bacterial genes induced during animal infection. First, because traditional gene reporters such as lacZ and phoA encode labile products, a positive signal can be transient and may be lost on recovery of bacterial cells from infected tissues. Second, in order to assay these reporters, a large number of cells is needed. This of course is not possible when screening a library of different strains in an animal. The technique of rapidly screening for GFP-expressing bacterial cells by FACS (described in III.A.4) may largely overcome both of these obstacles for some host-pathogen systems. However, a limitation that applies to all previously used gene reporters including gfp, is that gene fusions that are transiently expressed during infection or that are expressed at very low levels are difficult or impossible to detect. The RIVET strategy overcomes all three of these limitations to some degree. With RIVET, transcriptional gene fusions are made to a promoterless gene encoding a site-specific DNA recombinase, such as the tnpR gene (encoding

4.


157

resolvase) of Tn 1000. The resolvase enzyme is able to mediate recombination between two directly repeated copies of a specific target DNA sequence, called resl, which have been inserted flanking a genetic marker within the chromosome of the pathogen of interest. If a particular gtnt-tnpR fusion is transcriptionally induced during infection, even transiently and/or at a low level, the resolvase that is produced will catalyze a heritable change in the bacterium by excising the marker from the chromosome. The resolved strain can then be screened or selected for after recovering the bacteria from host tissues (Fig. 3, see color plate). For example, in the original study describing RIVET, induced resolvase fusions catalyzed the excision of a tetracycline-resistance gene, resulting in conversion of the fusion strain to a tetracycline-sensitive phenotype [18]. The strains harboring induced fusions were then identified by replica-plating colonies onto an agar medium supplemented with tetracycline. In the first implementation of RIVET, over a dozen V cholerae genes were identified whose levels of transcription increased during infection of the small bowel of infant mice, and a few of these genes were shown to play a role in pathogenicity by mutational analysis [19]. In addition to its application to V. cholerae, RIVET has also been used to identify S. aureus pathogenicity genes in a murine renal abscess model of disease [77]. RIVET-based screens suffer from two limitations. First, because the site-specific DNA recombinase acts on only one substrate sequence, a low level of expression of the recombinase gene fusion is sufficient to catalyze the excision event. Although this exquisite sensitivity allows for detection of transient and/or low-level gene induction events, it unfortunately prohibits identification of pathogenicity genes that have high basal levels of transcription during in vitro growth. This is due to the fact that a high basal level of transcription results in a strain that is unable to be constructed in the unrecombined state. However, more recent modifications have largely overcome this limitation. Specifically, a series of tnpR alleles containing "down" mutations in the Shine-Dalgarno sequence have been constructed. These mutations reduce the efficiency of translation over a wide range of transcriptional levels. This modification allows tnpR to be fused to genes that have high basal levels of transcription in vitro, without excising the substrate. If, though, these gene fusions are further transcriptionally induced during infection, the concomitant increase in translation product will result in excision of the marker from the strain [109]. The second limitation of RIVETbased screens, which holds for most other genetic screens and selections, is that only a subset of the genes identified will encode factors whose loss results in attenuation of pathogenicity in models of disease.

2.

SIGNATURE-TAGGED MUTAGENESIS

Signature-Tagged Mutagenesis (STM) is a genetic screen whereby transposon insertion mutant strains that are attenuated for survival and/or growth within the

158


host are identified [50]. Specifically, each host animal is infected with an input pool of limited complexity (usually 96 mutants), and the output of mutants recovered from host tissues is compared to the original input pool to identify strains that did not survive within the animal, that is, are missing from the output pool. In the most recent version of STM (Fig. 4, see color plate) [87], the strategy begins by marking each of 96 transposons with a small, unique DNA signature-tag which is inserted into a nonessential site within the transposons. Each of the resulting uniquely tagged transposons is then used to construct an insertion library in the bacterial pathogen. Strains from each of the 96 libraries are subsequently picked sequentially into microliter plate wells (see Fig. 4, see color plate), with each resulting microliter plate representing an input pool. A pooled sample of all 96 strains from each plate is then inoculated into an animal, and after a sufficient period of infection bacteria are recovered from one or more host tissues. The signature-tags are then PCR-amplified from both the input and output pools, and the representation of each strain in the input and output pools is measured by hybridizing the amplified tags to a master dot-blot filter that contains each of the original 96 DNA tags. Any strain present in the input pool, but is missing from the output pool, is attenuated for pathogenicity in the host animal. In the original implementation of STM, a large number of S. typhimurium pathogenicity factors were identified using a murine model of typhoid fever [50]. Since then, numerous STM screens have been successful in identifying pathogenicity factors from a variety of bacterial pathogens, such as V. cholerae [23] and S. pneumoniae [96]. In the first implementation of the modified STM protocol described above, several novel S. aureus pathogenicity factors were identified in a murine model of staphylococcal sepsis [87]. STM is a powerful genetic screen in that it allows many mutant strains to be screened per animal. In addition, the method is designed such that only genes that are essential for survival and pathogenicity in the animal are identified. Another advantage of STM, which is also shared with RIVET (III.B. 1), is that pathogenicity genes that are transiendy expressed, or expressed at low levels during infection, can in theory be identified. Perhaps the most notable limitation of STM, however, is that genes that play subtle roles in pathogenicity or have redundant functions are difficult or impossible to identify. Additional limitations of STM are that: (1) the presence of a colonization bottleneck in many disease models will necessarily reduce the allowable complexity of the input pool, and, relatedly, (2) a large inoculum, which can affect the natural course of the disease, is often needed to ensure adequate representation of all input strains. A significant property of both STM and large-scale screening that can make both methods powerful tools for analysis of potential pathogenicity factors, and which is not true for most other screening or selection strategies, is that a measure of the effect of a gene mutation on pathogenicity is built into the strategy. In contrast, the other strategies end with identification of bacterial factors whose expression levels are known to be increased during infection, or during in vitro

4.


159

growth in a host-mimicking condition. As these only lead to the possibility of a role in pathogenicity for such factors, a specific test of this possibility must follow. 3.

VERIFYING A ROLE IN PATHOGENICITY

As has been emphasized throughout this chapter, most genetic screens and selections result in identification of a gene whose role in pathogenicity is implicated, but not proven. Hence, it is usually desirable to fulfill what has been termed a molecular version of Koch's postulates in order to prove that the gene in question does play a role in pathogenicity [34]. The specific postulates are basically as follows: (1) The phenotype or property contributed by the gene should be associated with pathogenic strains. (2) A null mutation in the gene should attenuate pathogenicity in an appropriate model of disease. It should be noted at this point that the type of mutation that is least prone to artifacts is an in-frame deletion of virtually all of the gene coding sequence. For example, a deletion of only part of the gene could result in production of a truncated polypeptide that is toxic to the pathogen during infection. Alternatively, an out-of-frame deletion or insertional mutation might exert a polar effect on a downstream pathogenicity gene. (3) Pathogenicity should be fully restored after adding the gene back to the mutant strain. Such complementation is often best done by either inserting the gene back into the genome in a new location (e.g., [90]) or maintaining it on a low-copy plasmid. Insertion of the gene back into its original location, at the same time removing the mutant allele, does not rule out the possibility that a polar effect was responsible for the reduced level of pathogenicity originally observed. Optimally, the native promoter should be used to drive expression of the gene. However, this is often difficult and can be substituted for by engineering a heterologous promoter upstream of the gene. In practice, complementation of gene mutations usually ranges from easy to near impossible. For example, it is conceivable that some genes may require precisely regulated expression that can only be obtained in their native locations in the genome in order to mediate their effect on pathogenicity. A limitation of the second postulate is that mutations in some genes, which do indeed play roles in pathogenicity, may not result in a substantial decrease in pathogenicity when tested in the disease model. This might result from a limitation of the model of disease being used to measure pathogenicity, or it might result from the presence of another gene with redundant function that is also expressed by the pathogen. In the latter case, it would be ideal to identify and construct a null mutation in the redundant gene(s), and then attempt to fulfill Koch's molecular postulates for the first gene.

IV. Genetic Selections Since screens for pathogenicity factors can often be time consuming, requiring the isolation and analysis of many individual strains, it is of great benefit to the investigator if he/she can select (enrich) for the mutant strains that are being

160


sought. Such selection strategies greatly decrease the number of strains that must be analyzed to find a pathogenicity gene. The basis of most genetic selections is that strains harboring mutations that attenuate pathogenicity are selectively amplified to facilitate their isolation and identification. Alternatively, some genetic selections are based on identification of the gene mutations that enhance pathogenicity. The latter type of selection has a built-in enrichment scheme to amplify the strains of interest in that it searches for enhanced survival and/or multiplication in the host. Although the latter type of selection is straightforward, the former is less so, particularly in instances where the only measure of attenuated pathogenicity involves death of the mutant bacterial strain in the host. In this section, several examples of both of these types of genetic selections, which have been successfully employed, will be discussed.

A. Direct Selections There are a number of methods that can be employed for selection purposes, but perhaps the most straightforward enrichment strategy for gene mutations that attenuate pathogenicity is to selectively kill virulent strains, thus enriching avirulent strains. This could be done, for example, by using convalescence serum containing antibodies against a pathogenicity factor(s) exposed on the bacterial cell surface to select strains from a library of mutagenized bacteria that no longer express the pathogenicity factor(s). These strains would be enriched as a direct result of pathogenic bacteria being selected against due to the opsonizing activity of the antibodies. This method has been used, for example, to identify capsular polysaccharide-defective mutants of Klebsiella pneumoniae [9]. It should be noted that this particular method is not generally applicable, as convalescence sera to some pathogens may contain bactericidal antibodies recognizing multiple surface molecules, the losses of which cannot be obtained by the result of a single gene mutation. A second type of direct selection is built on the fact that some pathogenicity factors that reside on the surface of bacterial cells serve as receptors for lytic bacteriophages. This knowledge provides the basis for selection of bacterial mutant strains that have lost expression of the pathogenicity factor and thus are resistant to bacteriophage infection. For example, a major pathogenicity factor of many Gram-negative bacteria, LPS, serves as a receptor for many bacteriophages (e.g., [46]). Therefore, a strain that is deficient in LPS biosynthesis would be resistant to infection and lysis by bacteriophages. In addition to LPS, it has been found that other bacterial surface molecules or structures that are pathogenicity factors can serve as bacteriophage receptors (e.g., V. cholerae TCP [121]). A third type of direct selection takes advantage of growth phenotypes that are associated with the presence or absence of some pathogenicity factors. For example, wild-type L. pneumophila are growth sensitive to high salt concentra-

4.


161

tions, but a variety of nonpathogenic derivatives have been noted to be salt resistant. This correlation, which may or may not be relevant to the pathogenesis of this organism, was used as the basis for a selection of transposon mutagenized strains that were salt resistant. This selection was highly successful and resulted in the isolation of multiple strains that were subsequendy shown to be nonpathogenic due to disrupdons in several previously unidentified pathogenicity genes [119]. A limitation of this strategy is, of course, that the presence or absence of most pathogenicity factors cannot be associated with a growth phenotype in vitro, and thus this method is not generally applicable. However, it is likely that some pathogenicity factors, in particular those metabolic factors that assist growth of the pathogen during infecdon, can be identified by similar types of genetic selections. To take a hypothetical example, proton-pump inhibitors that are bactericidal for //. pylori in vitro [109] might be used to select for mutations in the proton-pump encoding genes that in turn encode an essential pathogenicity factor necessary for surviving the low-pH environment of the human stomach.

B.

Complementation Approaches

Many bacterial pathogenicity factors have been identified by complementation selecdon strategies. The most common of these strategies selects for a small fragment of genomic DNA carried on a plasmid, which complements the nonpathogenic phenotype of a strain containing a spontaneous or chemically induced mutadon in an unknown gene. In most cases, the complemendng fragment of DNA contains a functional version of the mutated gene or operon, and thus serves to identify the gene of interest. Three examples of the many pathogenicity factors idendfied by complementation analysis include the inv locus of S. typhimuriwn [46a], the dot/icm genes of L. pneumophila [10], and the phase-variable tcpH gene of V cholerae [21]. It should be noted, however, that it is somedmes possible that the complementing fragment contains a gene or set of genes disdnct from the mutated gene, but which nevertheless can phenotypically complement (i.e., bypass) the mutation. Because complementadon is the most suitable approach to idendfy genes disrupted by chemical mutagenesis, and because chemical mutagenesis produces the widest range of gene mutadons and functions in almost any bacterial species (see III.A.l), this combination of approaches comprises a powerful strategy for idendfying pathogenicity factors. However, it should be noted that complementation strategies are limited to bacterial species for which efficient transformation protocols exist. A second type of complementation strategy expands on the above strategy by searching for a gain of function. This strategy works by selecdng for a small genomic fragment of a pathogen, which when transferred into a nonpathogenic

162


Strain or heterologous species bestows some aspect of the pathogenic lifestyle. For example, the invasin protein of Y. pseudotuberculosis, which mediates invasion of nonprofessional phagocytic cells in the host, was identified by transforming a noninvasive E. coli strain with a library of Y. pseudotuberculosis genomic DNA, and then selecting for cells that had gained invasive ability [56]. The invasive strains of E. coli produced were selected for by the addition of an aminoglycoside antibiotic to the tissue culture medium to kill extracellular bacteria but not intracellular ones, as such antibiotics do not equilibrate with the interior of mammalian cells. A major attribute of this second type of complementation strategy is that an expressed genomic library of many pathogenic species, including some that are not genetically manipulatable, can be constructed in a nonpathogenic species that is much more easily manipulatable. However, this approach is limited by the need for relatively simple models of disease for which a single genetic locus can confer the sought-after phenotype on a heterologous bacterial species. As a result of this limitation, this type of approach has been used in the past usually to identify pathogenicity factors that either have enzymatic activities that the host bacterium lacks (e.g., hemolysins [31, 79]) or that complement a metabolic defect in the host bacterium (e.g., iron acquisition [8]). In addition, any factor identified by this method ultimately requires evaluation for a role, if any, in virulence in the original microorganism.

C. Selection for Nongrov^îng Bacterial Mutants Penicillin selection, which was originally developed to facilitate isolation of auxotrophs, functions by killing growing cells and sparing nongrowing cells [68]. Penicillin selections have been used as the basis for a strategy to isolate nonpathogenic bacterial mutants that fail to grow in host tissues. For example, in the original application of this strategy, transposon insertion mutants of Listeria monocytogenes that failed to multiply within cultured macrophages were selected for after the addition of methicillin, a (3-lactam that is able to freely penetrate into the interior of mammalian cells [20]. Amongst the mutant strains isolated were some that harbored transposon insertions in the gene encoding listeriolysin O, an essential pathogenicity factor that facilitates access to the host cell cytoplasm where multiplication and cell-to-cell spread occur [98]. For a review on the role of listeriolysin O, please refer to the Chapter 16 in this book. An analogous strategy to penicillin selection, termed "thymineless death," was developed to isolate L. pneumophila mutant strains unable to multiply within cultured macrophages [10]. The method is based on the observation that thymine auxotrophs of many bacterial species, including L. pneumophila, die when

4.


163

multiplying in the absence of exogenous thymine whereas nonmultiplying cells survive. This allowed an enrichment of nonpathogenic L. pneumophila mutant strains that failed to multiply intracellularly in cultured macrophages. This study identified a multigene locus, called dot (also referred to as icm ([14]), which is required for intracellular multiplication. One of the most important attributes of both penicillin selection and thymineless death strategies that is also shared with STM is that nongrowing mutant strains can be identified direcdy from host tissues or cells. However, in contrast to STM, the former two selection strategies are probably limited to in vitro and tissue culture host model systems, as recovery of a nongrowing mutant strain from an intact animal could be exceedingly difficult. For example, a thymineless death strategy to select for L. monocytogenes mutants that are no longer able to multiply inside macrophages in an animal host would probably fail due to clearance of bacteria by the host immune system. Moreover, an in vivo penicillin selection would require the difficult task of maintaining sufficient antibiotic concentrations at the site of infection in order to select for nongrowing mutant strains. However, the latter technical difficulty would not apply to the thymineless death strategy if used in an intact animal host.

D. In Vivo Expression Technology In Vivo Expression Technology (IVET) is a strategy designed to identify genes that are transcriptionally active only during infection. The premise behind IVET is that some, perhaps many, genes that are transcriptionally active during infection encode pathogenicity factors. The system relies on the enrichment of bacterial strains harboring gene fusions to a marker whose expression is or can be selected for within the intact host animal [80]. In the first implementation of IVET, an S. typhimurium transcriptional gene fusion library iopurA was constructed in a strain background in which the native purA gene had been deleted. Because purine auxotrophs cannot survive during systemic infection of mice, only strains from the library that harbor transcriptional gene fusions to purA that were expressed during infection were complemented and therefore survived and multiplied (Fig. 5, see color plate). This method has been used to identify several genetic loci encoding pathogenicity factors in S. typhimurium [48, 81]. Additionally, IVET has been successfully used to identify pathogenicity factors in other pathogens, such as P. aeruginosa [122, 123]. The above IVET strategy requires prior knowledge of selectable genes, such as purA, whose expression is required for multiplication within the host. It is essential that the chosen selectable gene be required for growth in the same host compartment that the pathogenicity gene (to be identified) is expressed in. Thus, a limitation of IVET strategies that employ complementing genes like purA is that, if the selectable gene is required throughout the entire infectious process.

164


then preferential enrichment will occur for strains harboring gene fusions that are constitutively expressed in the host animal. Thus, pathogenicity genes that are expressed at one stage of infection only, or which are transcribed at low levels, may not be identifiable via this approach. A second type of IVET selection is based on enrichment of bacterial strains harboring gene fusions to an antibiotic-resistance gene. In this method, a sufficient concentration of the corresponding antibiotic is maintained in the infected tissues to allow enrichment of strains containing active gene fusions. This method has been used to identify S. typhimurium pathogenicity factors [82]. Clearly, this method is more complicated than the auxotrophy complementation approach mentioned above since suitable antibiotics must be administered and proper concentrations maintained in infected host tissues. However, this property can provide some flexibility in designing the selection. For example, administering the antibiotic at only one stage of infection, or in one host organ, should facilitate identification of pathogenicity factors which may be expressed at only one stage of infection, or in only one host compartment, respectively. Recently, IVET has been combined with a direct selection scheme to identify pathogen genes that are transcriptionally induced during infection and which also encode exported proteins (J. J. Mekalanos and J. W. Tobias, unpublished data). The first component of this method proceeds by isolating pathogen DNA sequences that encode protein export signals, in a manner analogous to the TnphoA strategy (III.A.2). A plasmid library is constructed by ligating genomic fragments immediately upstream of a truncated p-lactamase gene that lacks the signal sequence coding portion required for protein export and subsequent enzymatic activity. The plasmids are moved into an E. coli host, and ampicillin is used to select for plasmids containing genomic fragments that encode protein export signals fused to the p-lactamase. Next, the selected pool of genomic fragments is excised from the plasmids and subcloned into pi VET 1 (see schematic diagram of pIVETl at the top of Fig. 5, see color plate). The resulting pool of pi VET 1 recombinants is then moved into the pathogen genome and subjected to the IVET selection to identify those strains containing infection-induced gene fusions, as described above (Fig. 5, see color plate). In the first implementation of this strategy, several S. typhimurium genes were identified. These included sti, a previously identified gene that encodes a glycosidase that interferes with IL-2 receptor function [85], and /v/-S2, a previously unknown gene whose putative product has a high degree of similarity to AmpD from E. coli, except that /v/-S2 encodes a putative signal sequence (the E. coli AmpD lacks a signal sequence). Roles for sti and ivi-Sl in pathogenicity have not been tested for, as yet. It is likely that this combination strategy will serve as a powerful method to identify additional infection-induced, exported factors from this and other pathogens. However, it should be noted that this combination strategy will be subject to the same limitations that apply to its two component methods.

4.


165

V. Genomic Approaches At the time of this writing, the complete genome sequences of 19 bacterial pathogens were publicly available, and another 40 or so were in progress (www.tigr.org). It is likely that within a decade the genomic sequences of almost all clinically important bacterial species will be known. This body of information, accompanied by the many computer algorithms already available for nucleotide and amino acid searching, have made possible a multitude of so-called, postgenomic strategies for identifying pathogenicity factors. In this chapter, several such strategies will be discussed, along with a few genomic strategies that do not require knowledge of a pathogen's genomic sequence. A.

Genome Walking

The past 30 years of molecular genetic research has shown conclusively that the genes encoding pathogenicity factors are more often than not physically linked on chromosomes or on extrachromosomal elements. Genetic linkage is particularly acute for genes that encode the factors necessary for production of complex factors such as pili (e.g., [83]) or type III secretion apparatuses (e.g., [25]). Although some genes encoding pathogenicity factors are not linked in this way (e.g., [47]), it nonetheless remains a fruitful strategy to target one's search for regions adjacent to previously identified pathogenicity genes. This is most easily accomplished by sequencing nearby genes, and analyzing the resulting nucleotide sequence for the presence of putative pathogenicity genes. Any such identified gene would then need to be mutated to assess its role, if any, in pathogenicity. One example of the numerous reports of genome walking is the recent completion of the Y. pestis Pmtl plasmid nucleotide sequence, which revealed the identity of at least seven genes encoding putative pathogenicity factors [54, 74]. A second genome walking strategy involves generating a series of mutations in specific DNA regions that surround a known pathogenicity gene. This strategy is usually accomplished by first cloning the region of interest into a plasmid, mutating the entire plasmid by transposon insertion mutagenesis or by site-directed means such as restriction enzyme-mediated frame-shift or deletion, and finally by moving the mutated plasmids back into the pathogen. The plasmids may then be tested for their ability to complement a complete deletion of the corresponding genomic locus or used to replace the genomic locus by allelic exchange. Strains that are isolated can then be tested for pathogenicity in a model of disease. This genome walking strategy has been used extensively. For example, a cluster of genes required for fimbrial expression in enteroaggregative E. coli, was identified by randomly inserting a transposon into a cloned DNA fragment known to contain genes essential for aggregation [103]. Genome walking has the advantage of discovering pathogenicity genes often with a minimal amount of labor, and it is also useful for identifying genes that function in concert to produce a complex virulence determinant. The only major

166


limitation of this strategy is that additional studies need to be done on the putative pathogenicity genes identified to assess their roles in pathogenicity in a model of disease.

B. Genomic Analysis and Mapping by in Vitro Transposition Genomic Analysis and Mapping By In vitro Transposition (GAMBIT) was recently developed primarily as a method to identify genes of bacteria that are essential for in vitro growth [2]. However, GAMBIT can also be used to identify genes that are essential for growth during infection in a host. In this respect, GAMBIT is similar to STM (III.B.2). An essential step in the GAMBIT protocol is the initial PCR-amplification of large segments of the pathogen genome (-15 kb in size). This step, of course, requires knowledge of some, or all, of the genomic sequence of a pathogen in order to allow the design of sets of PCR primer pairs. If the entire genome sequence is known, it would technically be possible to use GAMBIT to identify all essential genes by contiguous amplification and analysis (see below) of 15 kb regions. As shown in Figure 6 (see color plate), when using GAMBIT, each PCR product is subjected to in vitro transposon mutagenesis, which results in random insertions along the length of the fragment. The mutated DNA fragments are then transformed into the pathogen, and strains in which a fragment has integrated by allelic exchange are selected for by the use of the antibiotic resistance marker that is carried on the transposon. Transformant colonies are pooled and subsequently used both to infect a host and to prepare total DNA (the input DNA). After an appropriate period of infection, bacteria are recovered from host tissues, and total bacterial DNA (the output DNA) is prepared. During the course of the infection, strains that harbor transposon insertions in essential pathogenicity genes are selected against. Both the input and output DNAs are used as templates in PCR reactions using one of the primers originally used to amplify the genomic fragment, and a primer that is complementary to a sequence within the inverted repeat present at each end of the transposon. The PCR products are separated on an agarose gel, and their estimated sizes are used in conjunction with the genomic sequence to identify the sites of transposon insertion along the genomic fragment. The patterns of bands derived from the input and output DNA form two "ladders." The two ladders are visually compared to identify gaps (missing bands in one or more regions) in the output-derived ladder that correspond to gene(s) that are essential for survival and/or multiplication in the host (Fig. 6, see color plate). GAMBIT is a new methodology and has not been used to date in a general screen for pathogenicity factors. Nevertheless, a discussion of its advantages and limitations is warranted. Like STM, GAMBIT has the unique advantage of directly identifying genes that are essential for infectivity. However, because GAMBIT is predicted to be more labor intensive than STM when screening the

4.


167

entire genome of a pathogen, the former will probably find greater use in analyzing individual loci for the presence of pathogenicity genes. As with most genetic methods, GAMBIT can only be used in pathogens that are readily transformable. Although GAMBIT was originally implemented in two naturally transformable bacterial species, Haemophilus influenzae and S. pneumoniae, it should also be useful for examining small genomic regions in less efficiently transformable species. Finally, GAMBIT has the advantage that near-saturating mutagenesis can be achieved for a locus, thus allowing the essentiality of small regions in any locus to be examined in great detail. For example, in the original GAMBIT study [2], it was found that, although transposon insertions into most of the secA gene of H. influenzae were lethal for growth in vitro, those in the carboxy-terminal coding region were well tolerated. This finding was in accordance with a previous report [100], and it thus serves to demonstrate the ability of GAMBIT to allow detailed analysis of loci of interest.

C. Computational Screens Perhaps the most efficient means of identifying pathogenicity factors using genomic sequence is to search for predicted polypeptides that have amino acid sequence similarity to previously known pathogenicity factors, and which therefore are predicted to function similarly. For example, a recent report identified the prepilin peptidase, PilD, of V. cholerae from genomic sequence analysis. After identification of this gene, the authors then confirmed the requirement of the pilD product for secretion of a number of pathogenicity factors including cholera toxin [45]. Numerous additional similarity search strategies exist. For example, one could search for putative polypeptides with amino acid similarity, and thus possible functional similarity, to host encoded proteins. Such pathogen encoded proteins may function as pathogenicity factors by interfering with or otherwise affecting host cellular functions. Aside from searches based on amino acid similarity, a number of nucleotide search strategies also exist that can be used to identify pathogenicity factors. For example, two characteristics of pathogenicity islands—a different GC content than the overall GC content of the species and the presence of flanking tRNA sequences—have been used to identify previously unknown pathogenicity islands (e.g., [11]). Another strategy could be to search for promoter or coding-sequence elements that are known to be conserved in other pathogenicity genes. For example, most of the pathogenicity genes previously identified in L. monocytogenes are transcriptionally activated by the PrfA protein, which binds to a conserved 14-bp palindrome within their promoters [43]. Thus, after the L. monocytogenes genome sequence (currently in progress; www.pasteur.fr) is completed, it might be fruitful to identify and analyze the subset of genes that contain this palindrome in their promoter regions. Conserved virulence gene

168


promoter elements have been defined in a number of other pathogenic species as well, for example, B. pertussis [60, 104]), and thus this approach might be widely applicable. A third potential strategy could be to search for genes that contain small sequence repeats that might undergo slipped-strand mispairing during replication in order to phase-vary the expression of a bacterial factor [51]. For more information on phase variation, please refer to Chapter 14 in this book. Genes that phase-vary are likely to encode proteins that are seen by the host immune system, or that are involved in the biosynthesis or export of such factors, and thus are more likely than not to play a role in pathogenicity (e.g., [58]). As more pathogen genomes are sequenced, pathogenicity gene discovery by computer algorithm searching will become commonplace. Although computational screens are rapid and efficient means of identifying putative pathogenicity genes, it must be understood that these screens are only a starting point, as genes of interest must still be studied by traditional means, such as mutagenesis and complementation to demonstrate a role in pathogenicity.

D.

Transcriptional Profiling and the Use of Microarrays

The advent of extensive microbial and eukaryotic DNA sequence databases has prompted the development of DNA chip and microarray technologies in several commercial and academic laboratories [13, 26, 105, 107, 124]. These methods provide a powerful new way to measure the expression of virtually every gene in a genome by simultaneously measuring the concentration of individual messenger RNAs in an experimental sample. This methodology is referred to as "transcriptional profiling" and has applications in virulence gene identification. In brief, DNA samples that correspond to each individual gene of an organism are spotted by a high-resolution robot (the microarrayer) on the surface of, for example, an appropriately treated glass slide. Each gene has a known location on the microarray, and literally thousands of genes can be arrayed per glass slide. These DNA spots then serve as hybridization targets that will bind complementary nucleic acid labeled with appropriate fluorescent tags. In practice, mRNA is extracted from cell preparations, converted to fluorescently labeled cDNA, and hybridized to the microarray. The resulting hybridization intensities are measured by simply inserting the slide into a special fluorescence confocal microscope (the scanner) that can measure and record the individual fluorescent intensities of each spot in the microarray in a matter of minutes. There are clearly numerous applications for microarray technology. The conceptually easiest applications to envision involve comparisons of various sorts—for example, mutant strain versus parental strain, or bacteria grown in laboratory medium versus those grown in host-derived sample or host cells. In each of these experimental formats, bacteria are grown under a given condition or are allowed to interact with host cells for a defined period of time. Next, RNA

4.


169

is extracted and converted to fluorescently labeled DNA probes by reverse transcription. In each case, the comparison of RNA-derived probes from two different sources is made by differentially labeling (e.g., with red and green tags) and then hybridizing the two different probe preparations to the same DNA microarray. Comparison of DNA sequence distribution between different strains or even different culture preparations of a given strain provides an additional application for microarray technology that will fmd use in pathogenicity gene identification. In theory, hybridization to genomic arrays can be used to identify DNA segments that undergo amplification or rearrangement in response to the selective pressure of growth in a mammalian host. Because genetic elements encoding or controlling the expression of pathogenicity genes can undergo amplification [87], inversion [62, 130], or deletion [86], the mapping of any genomic segment that is rearranged in vivo provides a potentially powerful way of identifying genes involved in host-microbe interactions. Variations in sequence content between different clinical isolates might alert investigators to the presence of new pathogenicity-encoding plasmids, phages, and transposons, and pathogenicity islands that have moved horizontally into strains or have been recently lost. Thus, microarray analysis is expected to augment other established methods such as subtractive hybridization and differential display (discussed in II.F and II.G), which have been effectively used to identify genetic differences between closely related clinical isolates. The use of microarrays to define infection-induced genes, or genes coordinately regulated with other pathogenicity genes, will become more common as this technology is explored and improved. Other applications of microarrays will be developed. For example, it is easy to imagine that one could make a custom microarray carrying all oligonucleotide "tags" used to label individual transposons in the STM protocol (III.B.2). Such an array could be hybridized with PCR-derived probe pools derived from transposon mutants before and after animal passage, thus allowing quantification of the relative ratios of various mutants in a mixture without the need to grid mutants out spatially. Thus, like other IVET and STM techniques, microarray technology will eventually provide a new approach to defining which bacterial genes are induced during infection. The primary limitation of microarray technology, currently, is the high cost of manufacturing the microarrays. However, as the utilization of microarrays becomes more commonplace, and as the technology continues to develop, it is likely that the extreme costs associated with this technology will become more affordable.

VL Concluding Remarks As we settle into the postgenomic era, microbiologists will continue to use a combination of biochemical, genetic, and genomic techniques to identify patho-

170


genicity factors, although the latter will become a more frequent starting point. In particular, the biggest advances in pathogenicity gene discovery will likely come from large-scale genomic analyses, using either computational methods or DNA microarray technologies. Specifically, these two methods should facilitate identification of sets of pathogenicity genes with related function and/or that are coordinately regulated, many of which may be recalcitrant to discovery by current methods. For example, assessing the roles in pathogenicity of factors that have redundant function is very difficult, because knocking out the expression of one factor may not attenuate pathogenicity in the model of disease being used. However, knowing the identity of the redundant genes would allow a multiple gene knockout strategy to be used to assess their roles. In addition, many pathogenicity genes that are small in size may have eluded identification because they are difficult to mutate by transposon or insertion-duplication mutagenesis. Thus, knowing their locations would allow a site-directed mutagenesis strategy to be used to construct null mutations in them, in order to assess their collective roles in pathogenicity. It is likely that we are at the dawn of new discoveries of pathogenicity factors. With the technical advances of the last few years, in particular the ability to sequence large genomes, it is likely that many of the genes that are involved in pathogenicity will soon be discovered. This jackpot of discovery will more than likely lead to a general slowing of the pace of discovery of new pathogenicity genes within the more well-studied pathogens that are genetically manipulatable. In contrast to this slowing down of gene discovery, though, the scientific community will continue to be faced with a frontier of discovery that is perhaps even more challenging: a true understanding of the actual functions of bacterial pathogenicity factors during infection of the human host.

References Abu Kwaik, Y., and Pederson, L. L. (1996). The use of differential display-PCR to isolate and characterize a Legionella pneumophila locus induced during the intracellular infection of macrophages. Mol. Microbiol. 21(3), 543-556. Akerley, B. J., Rubin, E. J., Camilli, A., Lampe, D. J., Robertson, H. M., and Mekalanos, J. J. (1998). Systematic identification of essential genes by in vitro mariner mutagenesis. Proc. Natl. Acad. Sci. U.S.A. 95(15), 8927-8932. Akopyants, N. S., Fradkov, A., Diatchenko, L., Hill, J. E., Siebert, R D., Lukyanov, S. A., Sverdlov, E. D., and Berg, D. E. (1998). PCR-based subtractive hybridization and differences in gene content among strains of Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 95(22), 13108-13113. Allison, H. E., and Hillman, J. D. (1997). Cloning and characterization of a Prevotella melaninogenica hemolysin. Infect. Immiin. 65(7), 2765-2771. Andrews, G. P., Hromockyj, A. E., Coker, C , and Maurelli, A. T. (1991). Two novel virulence loci, mxiA and mxiB, in Shigella flexneri 2a facilitate excretion of invasion plasmid antigens. Infect. Immun. 59(6), 1997-2005.

4.


171

6. Avery, O. T., and Dubos, R. (1931). The protective action of a specific enzyme against type III pneumococcus infections in mice. J. Exp. Med. 54, 73-89. 7. Avery, O. T., MacLeod, C. M., and McCarty, M. (1944). Studies on the chemical nature of the substance inducing transformation of pneumococcal types: Induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus type III. Exp. Med. 79, 137-158. 8. Bearden, S. W., Staggs, T. M., and Perry, R. D. (1998). An ABC transporter system of Yersinia pestis allows utilization of chelated iron by Escherichia coli SABll. J. Bacterial. 180(5), 1135-1147. 9. Benedi, V. J., Ciurana, B., and Tomas, J. M. (1989). Isolation and characterization of Klebsiella pneumoniae unencapsulated mutants. J. Clin. Microbiol. 27(1), 82-87. 10. Berger, K. H., and Isberg, R. R. (1993). Two distinct defects in intracellular growth complemented by a single genetic locus in Legionella pneumophila. Mol. Microbiol. 7(1), 7-19. 11. Blanc-Potard, A. B., and Groisman, E. A. (1997). The Salmonella selC locus contains a pathogenicity island mediating intramacrophage survival. EMBO J. 16(17), 5376-5385. 12. Bliska, J. B., Guan, K. L., Dixon, J. E., and Falkow, S. (1991). Tyrosine phosphate hydrolysis of host proteins by an essential Yersinia virulence determinant. Proc. Natl. Acad. Sci. U.S.A. 88(1187), 1187-1191. 13. Bowtell, D. D. (1999). Options available—from start to finish—for obtaining expression data by microarray. Nature Genet. 21(1 Suppl.), 25-32. 14. Brand, B. C , Sadosky, A. B., and Shuman, H. A. (1994). The Legionella pneumophila icm locus: A set of genes required for intracellular multiplication in human macrophages. Mol. Microbiol. 14(4), 797-808. 15. Brown, P. K., and Curtiss III, R. (1996). Unique chromosomal regions associated with virulence of an avian pathogenic Escherichia coli strain. Proc. Natl. Acad. Sci. U.S.A. 93(20), 1114911154. 16. Burgess, A. W,, and Anderson, B. E, (1998). Outer membrane proteins of Bartonella henselae and their interaction with human endothelial cells. Microb. Pathogen. 25(3), 157-164. 17. Calia, K. E., Waldor, M. K., and Calderwood, S. B. (1998). Use of representational difference analysis to identify genomic differences between pathogenic strains of Vibrio cholerae. Infect. Immun. 66(2), 849-852. 18. Camilli, A., Beattie, D., and Mekalanos, J. (1994). Use of genetic recombination as a reporter of gene expression. Proc. Natl. Acad. Sci. U.S.A. 91, 2634-2638. 19. Camilli, A., and Mekalanos, J. J. (1995). Use of recombinase gene fusions to identify Vibrio cholerae genes induced during infection. Mol. Microbiol. 18, 671-683. 20. Camilli, A., Paynton, C. R., and Portnoy, D. A. (1989). Intracellular methicillin selection of Listeria monocytogenes mutants unable to replicate in a macrophage cell line. Proc. Natl. Acad. Sci. U.S.A. 86, 5522-5526. 21. Carroll, P A., Tashima, K. T, Rogers, M. B., DiRita, V. J., and Calderwood, S. B. (1997). Phase variation in tcpH modulates expression of the ToxR regulon in Vibrio cholerae. Mol. Microbiol. 25(6), 1099-1111. 22. Chhatwal, G. S., Walker, M. J., Yan, H., Timmis, K. N., and Guzman, C. A. (1997). Temperature dependent expression of an acid phosphatase by Bordetella bronchiseptica: Role in intracellular survival. Microb. Pathogen. 22(5), 257-264. 23. Chiang, S. L., and Mekalanos, J. J. (1998). Use of signature-tagged transposon mutagenesis to identify Vibrio cholerae genes critical for colonization. Mol. Microbiol. 27(4), 797-805. 24. Cirillo, D. M., Valdivia, R. H., Monack, D. M., and Falkow, S. (1998). Macrophage-dependent induction of the Salmonella pathogenicity island 2 type III secretion system and its role in intracellular survival. Mol. Microbiol. 30(1), 175-188.

172


25. Comelis, G. R., Boland, A., Boyd, A. P., Geuijen, C , Iriarte, M., Neyt, C , Sory, M. P., and Stainier, I. (1998). The virulence plasmid of Yersinia, an antihost genome. Microbiol. Mol Biol. Rev. 62(4), 1315-1352. 26. de Saizieu, A., Certa, U., Warrington, J., Gray, C , Keck, W., and Mous, J. (1998). Bacterial transcript imaging by hybridization of total RNA to oligonucleotide arrays [see comments]. Nature Biotechnol. 16( 1), 45-48. 27. De, S. N. (1959). Enterotoxicity of bacteria-free culture-filtrate of Vibrio cholerae. Nature 183, 1533-1534. 28. Diatchenko, L., Lau, Y. R, Campbell, A. P., Chenchik, A., Moqadam, F, Huang, B., Lukyanov, S., Lukyanov, K., and Gurskaya. N. (1996). Suppression subtractive hybridization: A method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc. Natl Acad. Sci. U.S.A. 93(12), 6025-6030. 29. Donnenberg, M. S., Calderwood, S. B., Donohue-Rolfe, A., Keusch, G. T., and Kaper, J. B. (1990). Construction and analysis of TnphoA mutants of enteropathogenic Escherichia coli unable to invade HEp-2 cells. Infect. Immun. 58(6), 1565-1571. 30. Dorrell, N., Li, S. R., Everest, P H., Dougan, G., and Wren, B. W. (1998). Construction and characterisation of a Yersinia enterocolitica 0:8 ompR mutant. FEMS Microbiol. Lett. 165(1), 145-151. 31. Drazek, E. S., Dubois, A., Holmes, R. K., Kersulyte, D., Akopyants, N. S., Berg, D. E., and Warren, R. L. (1995). Cloning and characterization of hemolytic genes from Helicobacter pylori. Infect. Immun. 63(11), 4345-4349. 32. Dubos, R., and Avery, O. T. (1931). Decomposition of the capsular polysaccharide of pneumococcus type III by a bacterial enzyme. J. Exp. Med. 54, 51-71. 33. Dutta, N. K., Pause, M. W., and Kulkarni, D. R. (1959). Role of cholera toxin in experimental cholera. / Bacteriol. 78, 594-595. 34. Engel, L. S., Hobden, J. A., Moreau, J. M., Callegan, M. C, Hill, J. M., and O'Callaghan, R. J. (1997). Pseudomonas deficient in protease IV has significantly reduced corneal virulence. Invest. Ophthalmol. Vis. Sci. 38(8), 1535-1542. 35. Falkow, S. (1988). Molecular Koch's postulates applied to microbial pathogenicity. Rev. Infect. Dis. 10(Suppl. 2), S274-S276. 36. Fields, P I., Swanson, R. V., Haidaris, C. G., and Heffron, F (1986). Mutants of Salmonella typhimurium that cannot survive within the macrophage are avirulent. Proc. Natl. Acad. Sci. U.S.A. 83(14), 5189-5193. 37. Finkelstein, R. A. (1992). History of cholera. In "Cholera" (D. Barua and W. B. Greenough III, eds.), pp. 155-187. Plenum, New York. 38. Finkelstein, R. A., and LoSpalluto, J. J. (1969). Pathogenesis of experimental cholera: Preparation and isolation of choleragen and choleragenoid. J. E.xp. Med. 130(185), 185-202. 39. Finlay, B. B., Starnbach, M. N., Francis, C. L., Stocker, B. A., Chatfield, S., Dougan, G., and Falkow, S. (1988). Identification and characterization of TnphoA mutants of Salmonella that are unable to pass through a polarized MDCK epithelial cell monolayer. Mol. Microbiol. 2(6), 757-766. 40. Finn, T. M., Shahin, R., and Mekalanos, J. J. (1991). Characterization of v/>-activated TnphoA gene fusions in Bordetella pertussis. Infect. Immun. 59(9), 3273-3279. 41. Fletcher, H. M., Schenkein, H. A., and Macrina, F L. (1994). Cloning and characterization of a new protease gene (prtH) from Porphyromonas gingivalis. Infect. Immun. 62(10), 4279-4286. [published erratum appears in Ifjfect. Immun. 1994 62(12), 5707]. 42. Foster, J. W., Park, Y. K., Bang, I. S., Karem, K., Belts, H., Hall, H. K., and Shaw, E. (1994). Regulatory circuits involved with pH-regulated gene expression in Salmonella typhimurium. Microbiology 140(Pt. 2), 341-352.

4.


173

43. Freitag, N. E., Youngman, P., and Portnoy, D. A. (1992). Transcriptional activation of the Listeria monocytogenes hemolysin gene in Bacillus siibtilis. J. Bacteriol. 174(4), 1293-1298. 44. Fujimoto, D. F., and Bayles, K. W. (1998). Opposing roles of the Staphylococcus aureus virulence regulators, Agr and Sar, in Triton X-100- and penicillin-induced autolysis. / Bacteriol. 180(14), 3724-3726. 44a. Galan, J. E., and Curtiss, R. (1989). Cloning and molecular characterization of genes whose products allow Salmonella typhimurium to penetrate tissue culture cells. Proc. Natl. Acad. Sci. U.S.A. 86, 6383-6387. 45. Fullner, K. J., and Mekalanos, J. J. (1999). Genetic characterization of a new type IV-A pilus gene cluster found in both classical and El Tor biotypes of Vibrio cholerae. Infect. Immun. 67(3), 1393-1404. 46. Guidolin, A., and Manning, P. A. (1985). Bacteriophage CP-Tl of Vibrio cholerae. Identification of the cell surface receptor. Eur J. Biochem. 153( 1), 89-94. 47. Hardt, W. D., Urlaub, H., and Galan, J. E. (1998). A substrate of the centisome 63 type III protein secretion system of Salmonella typhimurium is encoded by a cryptic bacteriophage. Proc. Natl. Acad. Sci. U.S.A. 95(5), 2574-2579. 48. Heithoff, D. M., Conner, C. P, Hanna, P C, Julio, S. M., Hentschel, U., and Mahan, M. J. (1997). Bacterial infection as assessed by in vivo gene expression. Proc. Natl. Acad. Sci. U.S.A. 94(3), 934-939. 49. Hendrixson, D. R., and St. Geme III, J. W. (1998). The Haemophilus influenzae Hap serine protease promotes adherence and microcolony formation, potentiated by a soluble host protein. Mol. Cell. Biol. 2(6), 841-850. 50. Hensel, M., Shea, J. E., Gleeson, C , Jones, M. D., Dalton, E., and Holden, D. W. (1995). Simultaneous identification of bacterial virulence genes by negative selection. Science 269(5222), 400-403. 51. Hood, D. W., Deadman, M. E., Jennings, M. P., Bisercic, M., Fleischmann, R. D., Venter, J. C, and Moxon, E. R. (1996). DNA repeats identify novel virulence genes in Haemophilus influenzae. Proc. Natl. Acad. Sci. U.S.A. 93(20), 11121-11125. 52. Hoppe, H. C , de Wet, B. J., Cywes, C , Daffe. M., and Ehlers, M. R. (1997). Identification of phosphatidylinositol mannoside as a mycobacterial adhesin mediating both direct and opsonic binding to nonphagocytic mammalian cells. Infect. Immun. 65(9), 3896-3905. 53. Hu, L. T, Pratt, S. D., Perides, G., Katz, L., Rogers, R. A., and Klempner, M. S. (1997). Isolation, cloning, and expression of a 70-kilodalton plasminogen binding protein of Borrelia burgdorferi. Infect. Immun. 65(12), 4989^995. 54. Hu, P, Elliott, J., McCready, P, Skowronski, E., Games, J., Kobayashi, A., Brubaker, R. R., and Garcia, E. (1998). Structural organization of virulence-associated plasmids of Yersinia pestis. J. Bacteriol. 180(19), 5192-5202. 55. liver, D., Arnqvist, A., Ogren, J., Frick, I. M., Kersulyte, D., Incecik, E. T, Berg, D. E., Covacci, A., and Engstrand, L. (1998). Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science 279(5349), 373-377. 56. Isberg, R. R., and Falcow, S. (1985). A single genetic locus encoded by Yersinia pseudotuberculosis permits invasion of cultured animal cells by Escherichia coli K-12. Nature 317, 262-264. 57. Johnson, J. A., Salles, C. A., Panigrahi, P, Albert, M. J., Wright, A. C , Johnson, R. J., and Morris, J. G. (1994). Vibrio cholerae 0\39 synonym Bengal is closely related to Vibrio cholerae El Tor but has important differences. Infect. Immun. 62, 2108-2110. 58. Jonsson, A.-B., liver, D., Falk, P., Pepose, J., and Normark, S. (1994). Sequence changes in the pilus subunit lead to tropism variation of Neisseria gonorrhoeae to human tissue. Mol. Microbiol. 13(3), 403-416.

174


59. Kahrs, A. F, Bihlmaier, A., Facius, D., and Meyer, T. F. (1994). Generalized transposon shuttle mutagenesis in Neisseria gonorrhoeae: A method for isolating epithelial cell invasion-defective mutants. Mol. Microbiol. 12(5), 819-831. 60. Karimova, G., and UUmann, A. (1997). Characterization of DNA binding sites for the BvgA protein of Bordetella pertussis. J. Bacteriol. 179(11), 3790-3792. 61. Klarsfeld, A. D., Goossens, P. L., and Cossart, P. (1994). Five Listeria monocytogenes genes preferentially expressed in infected mammalian cells: plcA, purH, purD, pyrE and an arginine ABC transporter gene, arpJ. Mol. Microbiol. 13(4), 585-597. 62. Klemm, P. (1986). Two regulatory fim genes, fimB and fimE, control the phase variation of type 1 fimbriae in Escherichia coli. EMBO J. 5(6), 1389-1393. 63. Knutson, K. L., Hmama, Z., Herrera-Velit, P, Rochford, R., and Reiner, N. E. (1998). Lipoarabinomannan of Mycobacterium tuberculosis promotes protein tyrosine dephosphorylation and inhibition of mitogen-activated protein kinase in human mononuclear phagocytes: Role of the Src homology 2 containing tyrosine phosphatase 1. J. Biol. Chem. 273(1), 645-652. 64. Labigne-Roussel, A., Courcoux, P., and Tompkins, L. (1988). Gene disruption and replacement as a feasible approach for mutagenesis of Campylobacter jejuni. J. Bacteriol. 170(4), 17041708. 65. Lampe, D. J., Churchill, M. E., and Robertson, H. M. (1996). A purified mariner transposase is sufficient to mediate transposition in vitro. EMBO J. 15(19), 5470-5479. [published erratum appears in EMBO J. 1997 16(13), 4153]. 66. Lane, W. S., Galat, A., Harding, M. W., and Schreiber, S. L. (1991). Complete amino acid sequence of the FK506 and rapamycin binding protein, FKBP, isolated from calf thymus. / Prot. Chem. 10(2), 151-160. 67. Lantz, M. S., and Ciborowski, P. (1994). Zymographic techniques for detection and characterization of microbial proteases. Methods Enzymol. 235, 563-594. 68. Lederberg, J., and Zinder, N. (1948). Concentration of biochemical mutants of bacteria with penicillin. Am. Chem. Soc. J. 70, 4267. 69. Lee, S. H., Hava, D. L., Waldor, M. K., and Camilli, A. (1999). Regulation and temporal expression patterns of Vibrio cholerae virulence genes during infection. Cell 99(6), 625-634. 70. Leong, J. M., Moitoso de Vargas, L., and Isberg, R. R. (1992). Binding of cultured mammalian cells to immobilized bacteria. Infect. Immun. 60(2), 683-686. 71. Liang, P., and Pardee, A. B. (1992). Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction [see comments]. Science 257(5072), 961-91 \. 72. Libman, E. (1905). A pneumococcus producing a peculiar form of hemolysis. Proc. N. Y. Pathol. Soc. 5, 168. 73. Lin, B., Hollingshead, S. K., Coligan, J. E., Egan, M. L., Baker, J. R., and Pritchard, D. G. (1994). Cloning and expression of the gene for group B streptococcal hyaluronate lyase. / Biol. Chem. 269(48), 30113-30116. 74. Lindler, L. E., Piano, G. V., Burland, V., Mayhew, G. F, and Blattner, F R. (1998). Complete DNA sequence and detailed analysis of the Yersinia pestis KIM5 plasmid encoding murine toxin and capsular antigen. Infect. Immun. 66(12), 5731-5742. 75. Lisitsyn, N., and Wigler, M. (1993). Cloning the differences between two complex genomes. Science 259(5097), 946-951. 76. Lockman, H. A., and Curtiss, R. (1990). Salmonella typhimurium mutants lacking flagella or motility remain virulent in BALB/c mice. Infect. Immun. 58(137), 137-143. 77. Lowe, A. M., Beattie, D. T, and Deresiewicz, R. L. (1998). Identification of novel staphylococcal virulence genes by in vivo expression technology. Mol. Microbiol. 27(5), 967-976.

4.


175

78. Lu, Y. S., Lai, W. C , Pakes, S. P, and Nie, L. C. (1991). A monoclonal antibody against a Pasteurella multocida outer membrane protein protects rabbits and mice against pasteurellosis. Infect. Immun. 59(1), 172-180. 79. Maclnnes, J. I., Kim, J. E., Lian, C. J., and Sokes, G. A. (1990). Actinobacillus pleuropneumoniae hlyX gene homology with the fnr gene oi Escherichia coli. J. Bacterial. 172(8), 4587^592. 80. Mahan, M. J., Slauch, J. M., Hanna, P C , Camilli, A., Tobias, J. W., Waldor, M. K., and Mekalanos, J. J. (1993). Selection for bacterial genes that are specifically induced in host tissues: The hunt for virulence factors (review). Infect. Agents Dis. 2(4), 263-268. 81. Mahan, M. J., Slauch, J. M., and Mekalanos, J. J. (1993). Selection of bacterial virulence genes that are specifically induced in host tissues. Science 259, 686-688. 82. Mahan, M. J., Tobias, J. W., Slauch, J. M., Hanna, P C , Collier, R. J., and Mekalanos, J. J. (1995). Antibiotic-based selection for bacterial genes that are specifically induced during infection of a host. Proc. Natl. Acad. Sci. U.S.A. 92(3), 669-673. 83. Manning, P A. (1997). The tcp gene cluster of Vibrio cholerae. Gene 192(1), 63-70. 84. Manoil, C , and Beckwith, J. (1985). TnphoA: A transposon probe for protein export signals. Proc. Natl. Acad. Sci. U.S.A. 82, 8129-8133. 85. Matsui, K., Nagano, K., Arai, T, Hirono, I., and Aoki, T. (1998). DNA sequencing of the gene encoding Salmonella typhimurium-dQr'wQd T-cell inhibitor (STI) and characterization of the gene product, cloned STI. FEMS Immunol. Med. Microbiol. 22(4), 341-349. 86. Maurelli, A. T, Fernandez, R. E., Bloch, C. A., Rode, C. K., and Fasano, A. (1998). "Black holes" and bacterial pathogenicity: A large genomic deletion that enhances the virulence of Shigella spp., and enteroinvasive Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 95(7), 3943-3948. 87. Mei, J. M., Nourbakhsh, F, Ford, C. W., and Holden, D. W. (1997). Identification of Staphylococcus aureus virulence genes in a murine model of bacteraemia using signature-tagged mutagenesis. Mol. Microbiol. 26(2), 399-^07. 88. Mekalanos, J. J. (1983). Duplication and amplification of toxin genes in Vibrio cholerae. Cell 35(0,253-263. 89. Miller, S. I., Kukral, A. M., and Mekalanos, J. J. (1989). A two-component regulatory system (phoP phoQ) controls Salmonella typhimurium virulence. Proc. Natl. Acad. Sci. U.S.A. 86(13), 5054-5058. 90. Nassif, X., and Sansonetti, P. J. (1986). Correlation of the virulence of Klebsiella pneumoniae Kl and K2 with the presence of a plasmid encoding aerobactin. Infect. Immun. 54(3), 603-608. 91. Odenbreit, S., Till, M., and Haas, R. (1996). Optimized BlaM-transposon shuttle mutagenesis of Helicobacter pylori allows the identification of novel genetic loci involved in bacterial virulence. Mol. Microbiol. 20(2), 361-373. 92. Pan, Y. T, Xu, B., Rice, K., Smith, S., Jackson, R., and Elbein, A. D. (1997). Specificity of the high-mannose recognition site between Enterobacter cloacae pili adhesin and HT-29 cell membranes. Infect. Immun. 65(10), 4199-4206. 93. Paruchuri, D. K., Seifert, H. S., Ajioka, R. S., Karlsson, K. A., and So, M. (1990). Identification and characterization of a Neisseria gonorrhoeae gene encoding a glycolipid-binding adhesin. Proc. Natl. Acad. Sci. U.S.A. 87(1), 333-337. 94. Pichichero, M. E., Loeb, M., Anderson, and Smith, D. H. (1982). Do pili play a role in pathogenicity oi Haemophilus influenzae type B? Lancet 2(8305), 960-962. 95. Plum, G., and Clark-Curtiss, J. E. (1994). Induction oi Mycobacterium avium gene expression following phagocytosis by human macrophages. Infect. Immun. 62(2), 476-483. 96. Polissi, A., Pontiggia, A., Feger, G., Altieri, M., Mottl, H., Ferrari, L., and Simon, D. (1998). Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect. Immun. 66(12), 5620-5629.

176


97. Poquet, I., Ehrlich, S. D., and Gruss, A. (1998). An export-specific reporter designed for gram-positive bacteria: Application to Lactococciis lactis. J. Bacterial. 180(7), 1904-1912. [published erratum appears in J. Bactehol. 1998 180(16), 4324]. 98. Portnoy, D. A., Chakraborty, T., Goebel, W., and Cossart, P. (1992). Molecular determinants of Listeria monocytogenes pathogenesis. Infect. Immun. 60(4), 1263-1267. 99. Purven, M., Frisk, A., Lonnroth, I., and Lagergard, T. (1997). Purification and identification of Haemophilus ducreyi cytotoxin by use of a neutralizing monoclonal antibody. Infect. Immun. 65(8), 3496-3499. 100. Rajapandi, T., and Oliver, D. (1994). Carboxy-terminal region of Escherichia coli SecA ATPase is important to promote its protein translocation activity in vivo. Biochem. Biophys. Res. Commun. 200(3), 1477-1483. 101. Reidl, J., and Mekalanos, J. J. (1995). Characterization of Vibrio cholerae bacteriophage K139 and use of a novel mini-transposon to identify a phage-encoded virulence factor. Mol. Microbiol. 18(4), 685-701. 102. Rubin, E. J., Akerley, B. J., Novik, V. N., Lampe, D. J., Husson, R. N., and Mekalanos, J. J. (1999). In vivo transposition of mariner-based elements in enteric bacteria and mycobacteria. Proc. Natl. Acad. Sci. U.S.A. 96(4), 1645-1650. 103. Savarino, S. J., Fox, P., Deng, Y., and Nataro, J. P. (1994). Identification and characterization of a gene cluster mediating enteroaggregative Escherichia coli aggregative adherence fimbria I biogenesis. /. Bacteriol. 176(16), 4949-4957. 104. Scarlato, V., Prugnola, A., Arico, B., and Rappuoli, R. (1991). The bvg-dependent promoters show similar behaviour in different Bordetella species and share sequence homologies. Mol. Microbiol. 5(10), 2493-2498. 105. Schena, M., Shalon, D., Davis, R. W., and Brown, P. O. (1995). Quantitative monitoring of gene expression patterns with a complementary DNA microarray [see comments]. Science 270(5235), 467-^70. 106. Sela, S., Thole, J. E., Ottenhoff, T H., and Clark-Curtiss, J. E. (1991). Identification of Mycobacterium leprae antigens from a cosmid library: Characterization of a 15-kilodalton antigen that is recognized by both the humoral and cellular immune systems in leprosy patients. Infect. Immun. 59(11), 4117-4124. 107. Shalon, D., Smith, S. J., and Brown, P. O, (1996). A DNA microarray system for analyzing complex DNA samples using two-color fluorescent probe hybridization. Genome Res. 6(7), 639-645. 108. Shumway, C. N., and Klebanoff, S. J. (1971). Purification of pneumolysin. Infect. Immun. 4(4), 388-392. 109. Sjostrom, J. E., Kuhler, T., and Larsson, H. (1997). Basis for the selective antibacterial activity in vitro of proton pump inhibitors against Helicobacter spp. Antimicrob. Agents Chemother 41(8), 1797-1801. 110. Soncini, F C , Garcia Vescovi, E., Solomon, F, and Groisman, E. A. (1996). Molecular basis of the magnesium deprivation response in Salmonella typhimurium: Identification of PhoP-regulated genes. J. Bacteriol. 178(17), 5092-5099. 111. Sun, J. W., Wanda, S. Y., Camilli, A., and Curtiss III, R. (1994). Cloning and DNA sequencing of the dextranase inhibitor gene (del) from Streptococcus sobrinus. J. Bacteriol. 176(23), 7213-7222. 112. Tacket, C. O., Taylor, R. K., Losonsky, G., Lim, Y, Nataro, J. P, Kaper, J. B., and Levine, M. M. (1998). Investigation of the roles of toxin-coregulated pili and mannose-sensitive hemagglutinin pili in the pathogenesis of Vibrio cholerae 0139 infection. Infect. Immun. 66(2), 692-695.

4.


177

113. Taylor, R. K., Miller, V. L., Furlong, D. B., and Mekalanos, J. J. (1987). Use of p hoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc. Natl Acad. Sci. U.S.A. 84(9), 2833-2837. 114. Tinsley, C. R., and Nassif, X. (1996). Analysis of the genetic differences between Neisseria meningitidis and Neisseria gonorrhoeae: Two closely related bacteria expressing two different pathogenicities. Proc. Natl. Acad. Sci. U.S.A. 93(20), 11109-11114. 115. Valdivia, R. H., and Falkow, S. (1996). Bacterial genetics by flow cytometry: Rapid isolation of Sahnonella typhimuriiim acid-inducible promoters by differential fluorescence induction. Mol. Microbiol. 22(2), 367-378. 116. Valdivia, R. H., Hromockyj, A. E., Monack, D., Ramakrishnan, L., and Falkow, S. (1996). Applications for green fluorescent protein (GFP) in the study of host-pathogen interactions. Gene 173(1), 47-52. 117. van den Berg, B. M., Beekhuizen, H., Willems, R. J., Mooi, F. R., and van Furth, R. (1999). Role of Bordetella pertussis virulence factors in adherence to epithelial cell lines derived from the human respiratory tract. Infect. Immiin. 67(3), 1056-1062. 118. Visai, L., Bozzini, S., Raucci, G., Toniolo, A., and Speziale, R (1995). Isolation and characterization of a novel collagen-binding protein from Streptococcus pyogenes strain 6414. J. Biol. Chem. 270(1), 347-353. 119. Vogel, J. P., Roy, C, and Isberg, R. R. (1996). Use of salt to isolate Legionella pneumophila mutants unable to replicate in macrophages. Ann. N.Y. Acad. Sci. 191, 271-272. 120. Waldor, M. K., Colwell, R., and Mekalanos, J. J. (1994). The Vibrio cholerae 0139 serogroup antigen includes an 0-antigen capsule and lipopolysaccharide virulence determinants. Proc. Natl. Acad. Sci. U.S.A. 91(24), 11388-11392. 121. Waldor, M. K., and Mekalanos, J. J. (1996). Lysogenic conversion by a filamentous phage encoding cholera toxin [see comments]. Science 272(5270), 1910-1914. 122. Wang, J., Lory, S., Ramphal, R., and Jin, S. (1996). Isolation and characterization of Pseudomonas aeruginosa genes inducible by respiratory mucus derived from cystic fibrosis patients. Mol. Microbiol. 22(5), 1005-1012. 123. Wang, J., Mushegian, A., Lory, S., and Jin, S. (1996). Large-scale isolation of candidate virulence genes of Pseudomonas aeruginosa by in vivo selection. Proc. Natl. Acad. Sci. U.S.A. 93(19), 10434-10439. 124. Wang, K., Gan, L., Jeffery, E., Gayle, M., Gown, A. M., Skelly, M., Nelson, P S., Ng, W. V., Schummer, M. (1999). Monitoring gene expression profile changes in ovarian carcinomas using cDNA microarray. Gene 229{\-2), 101-108. 125. Weintraub, A., Widmalm, G., Jansson, P E., Jansson, M., Hultenby, K., and Albert, M. J. (1994). Vibrio cholerae 0139 Bengal possesses a capsular polysaccharide which may confer increased virulence. Microb. Pathogen. 16(3), 235-241. 126. Weiss, A. A., Melton, A. R., Walker, K. E., Andraos, S. C., and Meidl, J. J. (1989). Use of the promoter fusion transposon Tn5 lac to identify mutations in Bordetella pertussis y/r-regulated genes. Infect. Immun. 57(2674), 2674-2682. 127. Wong, K. K., and McClelland, M. (1994). Stress-inducible gene of Salmonella typhimurium identified by arbitrarily primed PCR of RNA. Proc. Natl. Acad. Sci. U.S.A. 91(2), 639-643. 128. Worley, M. J., Stojiljkovic, I., and Heffron, F. (1998). The identification of exported proteins with gene fusions to invasin. Mol. Microbiol. 29(6), 1471-1480. 129. Yu, J. L., Mansson, R., Flock, J. I., and Ljungh, A. (1997). Fibronectin binding by Propionibacterium acnes. FEMS Immunol. Med. Microbiol. 19(3), 247-253. 130. Zhao, H., Li, X., Johnson, D. E., Blomfield, I., and Mobley, H. L. (1997). In vivo phase variation of MR/P fimbrial gene expression in Proteus mirabilis infecting the urinary tract. Mol. Microbiol. 23(5), 1009-1019.


CHAPTER 5

Mechanisms of Bacterial Pathogenesis in Plants: Familiar Foes in a Foreign Kingdom

JAMES R. ALFANO ALAN COLLMER

I. Introduction II. An Overview of Bacterial Plant Pathogens and Plant Diseases A. Three Types of Gram-Negative Bacterial Pathogens: Tumorigenic, Stealth Necrogenic, and Brute-Force Necrogenic B. The Fate of Nonpathogenic Bacteria in Plants C. The Overriding Importance of Protein Secretion Systems in Pathogenesis III. Tumorigenic Agrobacterium tumefaciens: Using the Type IV Secretion System to Transform the Host into a Factory for Bacterial Nutrients IV. Necrogenic, Stealth Pathogens: Parasites Strongly Dependent on the Hrp (Type III) Protein Secretion System A. The Hypersensitive Response (HR) Plant Defense Syndrome B. Gene-for-Gene (avr-R) Interactions and the Antiparasite Surveillance System of Plants C. hrp and hrc Genes and the Type III Protein Secretion System of Plant Pathogenic Bacteria D. Harpins and Pilins: Proteins Secreted in Culture in Abundance by the Hrp System E. Avr Proteins as Injected, Interchangeable Effectors of Parasitism V. Necrogenic, Brute-Force Pathogens: Soft-Rotters Dependent on Type II Secretion of Pectic Enzymes VI. Other Virulence Factors of Gram-Negative Plant Pathogens Compared with Those of Animal Pathogens A. Attachment Factors: Important in Agrobacterium but Role Unclear in Necrogenic Pathogens B. EPS: A Special Role in Wilt Diseases of Plants


i '7Q

180 181 181 184 185 186 189 189 190 191 194 196 200 201 201 202

180

JAMES R. ALFANO AND ALAN COLLMER

VII. VIII. IX. X.

C. Toxins: Some Similarities, but Differently Defined D. Iron Uptake Mechanisms: Common Strategies for Overcoming a General Eukaryote Defense E. Resistance to Host Antimicrobial Peptides: Related Mechanisms Underlying Bacterial Virulence in Mouse and Potato F. Regulation of Virulence: A Potpourri of Common Components, Strategies, and Signals Host Innate Immune Systems: Common Components in Pathogen Recognition and Defense Signaling The R Gene Surveillance System: An Innate Immune System with Elaborate Recognition Specificity Pseudomonas aeruginosa: Dual-Kingdom Pathogenesis Conclusions References

202 203 204 205 206 207 209 210 211

/. Introduction With vast areas of their surface bearing perforations large enough to admit most bacteria, and without an adaptive immune system, plants would seem easy targets for hungry prokaryotes. But, surprisingly, plants are relatively resistant to bacterial attack, and most of the bacterial diseases that do afflict them are caused by strains in a handful of species. When the interactions between these successful pathogens and plants are viewed from the perspective of animal pathogenic microbiology, we find a mix of the familiar and the alien. The plant pathogens use familiar, conserved regulatory components and secretion systems to deploy key virulence proteins, and the attacked plant cells often respond defensively with familiar factors like oxidative bursts, nitric oxide synthesis, and programmed cell death. However, the virulence proteins themselves and the antiparasite surveillance systems that trigger these defenses are significantly different. Both the familiar and alien features are becoming more relevant to students of animal pathogenesis. Aspects of conserved virulence systems are sometimes more experimentally accessible in the plant pathogens, and the prevalence of horizontally transferred pathogenicity islands highlights the universal nature and potential interchangeability of these systems. Food safety concerns encourage more attention to the association of many animal pathogenic bacteria with fruits and vegetables and the possible interactions of these bacteria with plant pathogens. And the development of a model system involving a Pseudomonas aeruginosa strain pathogenic to both plants and animals suggests that novel genes involved in animal pathogenesis may be efficiendy identified using a plant system. The purpose of this chapter is to provide an introduction to the world of bacterial plant pathogenesis that is accessible to students of animal pathogens. We will delineate the conserved and

5.

BACTERIAL PLANT PATHOGENS

181

unique mechanisms and emphasize those features that will contribute most to a broader understanding of mechanisms of bacterial pathogenesis.

//. An Overview of Bacterial Plant Pattiogens and Plant Diseases The most important prokaryotic pathogens of plants are Gram-negative bacteria in the genera Agrobacterium, Envinia, Pseudomonas, Xanthomonas, Ralstonia, and Xylella, Gram-positive bacteria in the genera Clavibacter and Streptomyces, and Mollicutes in the genus Spiroplasma and the Candidatus genus Phytoplasma. The Mollicutes and Xylella are fastidious parasites that are typically delivered by insects into the vascular system of plants. The other bacterial pathogens are generally facultative parasites that colonize the intercellular spaces between plant cells (or the nonliving xylem cells that conduct water up from the roots). Thus, most bacterial plant pathogens colonize a microniche in the host that is separated from the host cytoplasm not only by a plasma membrane, but also by a cell wall that is ca. 200 nm thick and comprised of a matrix of polysaccharide and protein thought to exclude macromolecules with a Stokes radius >4.6 nm [1].

A.

Three Types of Gram-Negative Bacterial Pathogens: Tumorigenic, Stealth Necrogenic, and Brute-Force Necrogenic

Most of what we know about pathogenic mechanisms is based on work with the Gram-negative bacteria, and in subsequent sections we will distinguish three types of pathogens within this group (Table I). Agrobacterium tumefaciens represents the first type, which is distinguished by its tumorigenicity. The pathogen transforms plant cells with bacterial DNA, producing tumors, but generally not necrosis, in the host (Fig. 1). In contrast, bacteria in the genera Erwinia, Pseudomonas, Xanthomonas, and Ralstonia are necrogenic. That is, they have a characteristic ability to elicit plant cell death. Pseudomonas syringae is representative of a subgroup of the necrogenic pathogens and of the second pathogen type we will discuss in that it is host specific and triggers host cell death only after prolonged "stealthy" parasitic multiplication. Eminia carotovora, on the other hand, is representative of another subgroup of the necrogenic pathogens and of the third pathogen type we will discuss in that it is relatively host promiscuous and kills host tissues rapidly during *'brute-force" pathogenesis. We will use the descriptive terms "stealth" and "brute-force" here instead of the formal terms

182 Table I


Model Gram-Negative Plant Pathogens

Pathogentype/ representative species

Model hosts

Typical disease symptoms

Major virulence factor secretion pathway"

Key virulence factors secreted

tobacco, carrot. and other dicots

crown gall tumors

Type IV

T-DNA

Pseudomonas syringae pvs.

tomato, Arabidopsis, legumes

necrotic lesions. often with chlorotic halos on leaves and fruit

Type III

Avr proteins^

Xanthomonas campestris pvs.

pepper, tomato. brassicas, rice

necrotic lesions on leaves and fruit

Type III

Avr proteins^

Erwinia amyiovora

apple and pear

fire blight

Type III

harpins, DspE (and other Avrlike proteins?)

Ralstonia solanacearum

tomato, tobacco. banana

wilt and necrosis of whole plant

Type III

Avr proteins?

potato, Saintpaulia, and many other plants with fleshy tissues

maceration "softrof' of fleshy plant organs

Type II

pectic enzymes

Tumorigenic Agrobacterium tumefaciens Necrogenic: host-specific ("stealth")

Necrogenic: host-promiscuous ("brute-force") Erwinia carotovora and E. chrysanthemi

^Note that plant cell-wall-degrading enzymes secreted by the type II pathway also contribute to the virulence of X. campestris and R. solanacearum and that the type III pathway contributes to the ability of E. chrysanthemi to initiate infections at low levels of inoculum [27, 77]. As discussed in the text, the known Avr proteins are thought to represent a larger class of effector proteins injected into host cells by the type III pathway.

biotroph and necrotroph, respectively, which denote the feeding relationship of the parasite with host cells [2]. The necrogenic pathogens produce a wide array of symptoms in plants (Fig. 1). For example, P. syringae typically causes small necrotic lesions, often surrounded by a toxin-induced yellow halo, on leaves. Ralstonia solanacearum colonizes the xylem cells in the water-supplying vascular system and causes plants to wilt. Erwinia amyiovora attacks blossoms and then vascular tissues in

5.


183

Fig. 1 Representative disease symptoms and plant responses caused by Gram-negative plant pathogens. (A) Crown gall of cherry caused by A. tumefaciens; note the massive gall exposed after removing soil. Reprinted through the courtesy of Wayne A. Sinclair. (B) Brown spot of bean caused by P. syringae pv. syringae; note the necrotic lesion (arrow) that has developed after several days of symptomless bacterial multiplication. Reprinted through the courtesy of Susan S. Hirano [269]. (C) Southern bacterial wilt of tomato caused by R. solanacearum; note the wilted plant on the right, which was grown in infested soil. Reprinted through the courtesy of H. David Thurston. (D) Fire blight of pear; note the blackened, collapsed stem and white strands of extruded bacterial ooze. Reprinted through the courtesy of Steven V. Beer. (E) Bacterial soft rot of potato caused by E. carotovora; note the sharply demarcated zone of macerated tissue in this close-up of a sectioned, infected potato tuber. (F) The hypersensitive response (HR) elicited in tobacco following infiltration with P. syringae pv. tomato at a concentration of 1 x 10^ cells/ml; note the infiltrated panel (arrow) has collapsed 24 hr after infiltration.

184


the shoots of apple trees, causing aptly named "fire blight" symptoms. And E. carotovora macerates fleshy plant tissues, producing "soft rots" in organs such as potato tubers. Beyond these major differences, there is another level of more subtle variation in plant diseases. For example, a field of tobacco attacked by P. syringae may have a dynamic mixture of strains that differ only in their ability to produce a toxin: those producing the toxin will cause "wildfire" symptoms with massive yellow halos around lesions, whereas the others will produce more limited "angular leaf spot" symptoms [3]. As will be discussed below, pathogenesis by the necrogenic pathogens is a highly multifactorial process involving much apparent redundancy and subtle variation in virulence factors. Because of this and the diversity of susceptible plants, a small handful of pathogen species using highly conserved virulence factor delivery systems can cause hundreds of different diseases. Host specificity is a key part of this pathogen variability and is well defined at the pathovar and race level. That is, P. syringae is divided into over 40 pathovars based primarily on specificity for different plant species. For example, many strains off! syringae pv. syringae are virulent on bean but "avirulent" on tobacco, whereas strains of P. syringae pv. tabaci have the opposite specificity. Over 100 pathovars similarly have been defined for Xanthomonas campestris (although taxonomic studies support the redefinition of some pathovar groups as new species [4]). Races can be distinguished within many pathovars off! syringae and X. campestris based on host range among cultivars of the host species. For example, P. syringae pv. glycinea is pathogenic to soybean, but race 6 is avirulent on Harosoy and a few other cultivars of soybean, where it triggers a strong defense syndrome known as the hypersensitive response (HR). The search for the basis of this specificity and the strong defenses elicited in resistant cultivars has driven major advances in our understanding of how these bacteria attack plants.

B. The Fate of Nonpathogenic Bacteria in Plants When nonpathogenic bacteria, such as Escherichia coli and Pseudomonas fluorescens, or disarmed pathogen mutants are introduced into the intercellular spaces of a plant leaf, they neither grow nor trigger the strong defense reactions associated with avirulent pathogens [5, 6]. However, they do trigger several weak, transient, localized defense responses, including an oxidative burst, plant cell wall papilla deposition, and increased expression of the phenylpropanoid pathway (which leads to phenolics involved in wall fortification, defense signaling, and pathogen inhibition) [7-9]. These responses may also be triggered nonspecifically by two common bacterial surface features, LPS and flagella. LPS isolated from a variety of bacteria can elicit weak defenses [10-12], and plants also respond to flagellins or a 20-amino-acid domain that is conserved in the N terminus of flagellins from bacteria as diverse as Escherichia and BaciUus spp. [13]. Virulent

5.


185

necrogenic stealth pathogens, such as P. syringae, suppress these defense responses [8]. What actually accounts for the stasis of nonpathogenic bacteria in plant leaves (or what enables pathogens to grow) remains a puzzle. It is likely that pathogens, in addition to suppressing defenses, also possess active mechanisms to trigger the release of nutrients to the surface of plant cells. This has been established for tumorigenic A. tumefaciens and for necrogenic brute-force pathogens (which lyse host cells) [14, 15]. However, the relative importance of defense suppression and nutrient acquisition in the success of the necrogenic stealth pathogens is unknown. Interestingly, mixed inoculations of these bacteria and nonpathogens result in growth of the nonpathogens [16], which indicates that pathogen-induced changes in host metabolism render plants broadly susceptible to bacteria. This issue has direct implications for human health given the contamination of many horticultural products with fecal coliforms and the finding that Salmonella grows better in vegetable tissues that are diseased [17]. In this regard, it is also noteworthy that in some situations (e.g., in the commercial production of alfalfa sprouts) human pathogens like Salmonella can grow significantly on the surface of plant tissues [18].

C. The Overriding Importance of Protein Secretion Systems in Pathogenesis The primary determinants of pathogenicity in the Gram-negative plant pathogens are T-DNA for tumorigenic Agrobacterium and secreted proteins for the necrogenic pathogens in Pseudomonas, Xanthomonas, Envinia, and Ralstonia. Other factors, such as peptide toxins and extracellular polysaccharides, appear to have a secondary role and are discussed later. Typically, necrogenic pathogen mutants altered in the production of a single virulence protein are only partially reduced in virulence. Redundancy among such proteins seems to be the rule. In contrast, mutants deficient in protein secretion pathways are generally nonpathogenic. This is true for the type IV pathway in Agrobacterium, the type III pathway in stealth necrogenic pathogens, and the type II pathway in brute-force necrogenic pathogens [15, 19]. Indeed, screens for random mutants that have lost pathogenicity generally yield deficiencies in the respective secretion systems [19]. The pathogenic personality of the three types of pathogens is largely determined by the unique capabilities of the dominant secretion pathway, which are to deliver a nucleoprotein complex into plant cells (type IV), to inject effector proteins into plant cells (type III), or to efficiently secrete massive quantities of degradative enzymes to the bacterial milieu and the surface of plant cells (type II). The type IV secretion system is also described as "conjugation-like" and is used by Bordetella pertussis to secrete pertussis toxin, as discussed below. The type III secretion system is also referred to as the "Hrp system" in plant pathogens and

186


the "contact-dependent" secretion system in animal pathogenic Yersinia, Salmonella, and Shigella because of its stimulation by contact with host cells [20]. And, the type II secretion system is also known as the "main terminal branch of the general secretory pathway" [21]. Aspects of the operation of these pathways that are particularly relevant to pathogenesis will be treated in subsequent sections, and an earlier chapter in this book is devoted to protein secretion systems (see Chapter 2 by Harper and Silhavy).

///. Tumorigenic Agrobacterium tumefaciens; Using the Type IV Secretion System to Transform ttie Host into a Factory for Bacteriai Nutrients Central to the pathogenesis of A. tumefaciens is the Ti (tumor inducing) plasmid and the transfer of a DNA segment (T-DNA) from the plasmid to the cytoplasm of the plant cell, where it is imported into the nucleus and integrated into the plant genome. The T-DNA encodes enzymes that synthesize plant growth hormones, thereby producing neoplastic growths, "crown galls," which develop at the crown of the plant just below the soil line. T-DNA also encodes opines and amino acid or sugar derivatives that provide a specialized food source for Ti-plasmid-bearing A. tumefaciens in the surrounding soil. This interkingdom DNA transfer system has aspects of both conjugation and protein secretion systems [22, 23]. For example, T-DNA borders are similar to the origin of transfer region oriT sites; in some plasmids, T-DNA is transferred in a single-stranded form, and components of the translocation apparatus are similar [22]. However, both T-DNA transfer and bacterial conjugation systems transfer both DNA and proteins, and, as discussed below, the VirE2 protein is probably transferred independently of T-DNA [23,24]. Thus, these conjugal transport systems can also be thought of as protein secretion systems. Further supporting this concept is the finding that these systems (encoded by virB genes in Agrobacterium and trb genes in conjugation systems) share extensive similarity with the Ptl secretion system present in the animal pathogen B. pertussis, which is dedicated to the secretion of the pertussis protein toxin [23, 25, 26]. These systems are presently being referred to as type IV protein secretion systems [27]. Agrobacterium pathogenesis and the associated type IV secretion system are summarized below and illustrated in Figure 2. A more detailed description can be found in several recent reviews that focus on different aspects of this phenomenon [15, 28, 29]. Agrobacteria in the soil are attracted to plant roots by a variety of signals, including monocyclic phenolics and monosaccharides associated with plant cell wall repair [15]. (The subsequent attachment to plant cells is discussed below.) The signals are thought to be released from plant wounds at the root-shoot

5.

187


Hormone prodyction

Fig. 2 Overview of the type IV secretion system of Agrobacterium and key events in the development of crown gall disease. Signals from wounded plant cells (white hexagons) induce the expression of vir genes via the VirA/VirG two-component regulatory system. The type IV secretion apparatus (encoded by the virB operon) delivers single-stranded (ss) T-DNA (squiggled line) across both the inner membrane (IM) and the outer membrane (OM) of the bacterial cell, and the cell wall (CW) and plasma membrane (PM) of the plant cell. T-DNA bound covalently to VirD2 (white trapezoids) at the 5'-phosphoryl end of the T-DNA may be transferred separately from the ssDNA-binding protein, VirE2 (hatched circles). Inside the plant cell, the T-DNA becomes coated with VirE2 and is targeted by the nuclear localization signals contained in both VirD2 and VirE2 to enter the nucleus through a nuclear pore (NP), whence it integrates into the plant genome. The transformed plant cell then produces tumor-promoting growth hormones and also opines, which support bacterial growth and induce tra genes directing conjugational spread of the Ti plasmid.

188


interface and are perceived by a two-component regulatory system comprised of the VirA and VirG proteins, which in turn activates virulence (vir) gene expression [30]. The vir genes are carried on the Ti plasmid within a 35-kb region that is made up of at least 6 operons involved in different aspects of tumorigenesis. With the exception of 11 chromosomal genes, some of which are known to be involved in bacterial attachment or sugar perception, the Ti plasmid carries all of the genetic information necessary to transfer T-DNA complexes from the bacterial cell to the plant nucleus. The T-DNA carries the information for tumor formation and opine production in the plant, whereas opine utilization by the bacterium is directed by genes carried on the Ti plasmid outside of the T-DNA region [28]. Since the ability to utilize these compounds is rare in nature, this trait gives plasmid-bearing strains a competitive advantage over other neighboring soil bacteria. Furthermore, opines stimulate a conjugational transfer system encoded by the Ti plasmid, resulting in transfer of the plasmid to Agwbacteriiim strains lacking it. Hence, the prime beneficiary of this curious disease is the ca. 200-kb Ti plasmid. The T-DNA complex translocation apparatus is encoded primarily by the virB operon. A current focus of research involves localization of different VirB proteins and determining specific protein-protein interactions occurring among components of the translocation apparatus [29]. Since most of these proteins are localized to the inner and outer bacterial membranes, the VirB proteins likely assemble to form a pore through both bacterial membranes, permitting one-step secretion of T-DNA complexes from the bacterial cell. It is important to note that the type IV (Ptl) system in B. pertussis secretes pertussis toxin subunits to the periplasmic space by using an export system similar to the E. coli Sec system. The subunits are assembled into the mature toxin in the periplasm and secreted to the extracellular milieu via the type IV secretion system. Thus, it appears that in different bacteria type IV secretion systems can secrete dissimilar macromolecules and initiate the secretion of these molecules from either the bacterial cytoplasm or the periplasmic space [23]. The substrate that travels the type IV pathway in Agrobacterium, the T-DNA complex, is a single-stranded DNA molecule containing the VirD2 protein attached at the 5'-end. The T-DNA molecule itself apparently provides no secretion signals targeting it for the type IV pathway; therefore, all of the information that enables DNA to be delivered into the plant cell apparently resides in the protein component of the complex. VirD2 (along with VirDl) nicks the T-DNA and attaches to the 5' end of the released single-stranded DNA [31, 32]. The binding of VirE2 to the T-DNA is not needed in the bacterium but is essential in the plant, and VirE2 is likely to be secreted independendy of the T-DNA. That is, the T-DNA complex can be translocated into the plant cell without VirE2 as long as VirE2 is supplied by another Agrobacterium strain (producing VirE2 but not T-DNA) or by expression of VirE2 in transgenic plants [33, 34]. VirD2 and VirE2 contain functional nuclear localization signals, implicating both proteins in transporting the T-DNA to the nucleus [28, 34, 35]. The T-DNA is randomly and

5. BACTERIAL PLANT PATHOGENS

189

Stably inserted into the plant genome. Plant molecular biologists have exploited the Agrobacterium system by replacing T-DNA genes with desired genes and a selectable marker, and then infecting germinal tissues or somatic cells capable of regenerating whole plants to construct transgenic plants expressing foreign genes. The recent discovery of type IV secretion systems in the animal pathogens Legionella pneumophila and Helicobacter pylori indicates that type IV secretion systems may be widely important in bacterial pathogenicity [36-38]. Because of the well-studied potential of the Agrobacterium type IV system to transfer both protein and DNA into host cells, it will be interesting to learn what macromolecules are transferred by these novel systems.

IV. Necrogenic, Stealth Pattiogens: Parasites Strongly Dependent on ttie Hrp (Type III) Protein Secretion System The stealth pathogens can trigger strong plant defenses in nonhost plants, or even in host plants in the latter stages of pathogenesis (e.g., in the common leaf spot diseases where lesions typically are limited). The capacity to trigger these defenses is intimately linked with the ability of the bacteria to be pathogenic, and the defense reactions (specifically the HR) have traditionally provided more convenient assays for this basic bacterial ability [39]. Hence, we will start with a general overview of plant defenses and a description of the antiparasite surveillance system of plants before moving on to the all-important bacterial type III (Hrp) protein secretion system and its parasite-promoting protein traffic.

A. The Hypersensitive Response (HR) Plant Defense Syndrome At the heart of plant resistance against most necrogenic, stealth pathogens is the HR, a defense-associated, rapid (usually within 24 hr), programmed death of plant cells in contact with an "incompatible" pathogen (i.e., an avirulent bacterium that is virulent on some other plant) [39, 39a]. Single bacterial cells can trigger the HR death of single plant cells in a one-to-one manner [40], and if bacteria are infiltrated into leaf intercellular spaces at a level higher than 5x10^ cells/ml, then usually enough plant cells die to produce a readily assayed macroscopic tissue collapse (Fig. 1). The HR is elicited in nature only by plant pathogens (including bacteria, fungi, viruses, and nematodes), and the gene-for-gene interactions and signals underlying this phenomenon are discussed in the next section. Paradoxically, plant cell death itself does not appear necessary for effective resistance

190


against P. syringae [41]. However, the HR syndrome includes a panoply of antimicrobial responses, and exploring these has revealed much about plant defense systems. The ability to respond with the HR is not restricted to any specialized cell types but is a general property of plant cells that enables defense anywhere in the plant to be immediate and localized. Many of the defense responses are further localized to the plant cell wall at the site of contact with the pathogen. These are early responses occurring within the first few hours of pathogen contact and include an oxidative burst [42, 43], dityrosine crosslinking of tyrosine-rich proteins in the wall [44], increased levels of phenolic compounds associated with wall fortification [45], callose synthesis and deposition of a callose-rich papilla between the cell wall and the plasma membrane [46], and redistribution of phospholipase D in the plasma membrane to the site of contact [47]. The affected plant cell and its neighbors then produce phytoalexins (low-molecular-weight antimicrobial compounds produced in response to microbes) [48], and a variety of "pathogenesisrelated" (PR) proteins that are secreted into the cell wall and intercellular fluids or into the vacuole (from which they are released if the cell dies) [49]. If enough plant cells in the region detect pathogen, then a signal for systemic acquired resistance (SAR) is sent throughout the plant, leading to induction of a subset of PR proteins, known as SAR proteins, in distal tissues. SAR development takes several days, and it confers increased (but often not complete) resistance against a broad array of pathogens for several weeks [49]. Salicylic acid is involved in SAR signaling, although it may not be the long-distance signal [50, 51]. (Serendipitously, because of the defense-inducing effects of salicylates, the standard medical prescription for two aspirins and a phone call in the morning is not entirely off the mark with plants.) There is litde evidence that these HR-associated defense responses are tailored to the potential invader. Many of them are also triggered (but more weakly and without host cell death) by nonpathogenic bacteria, and an SAR that is triggered by a bacterium, a fungus, or a virus is equally (usually partially) effective against members of all three groups of pathogens [49].

B.

Gene-for-Gene (avr-R) Interactions and the Antiparasite Surveillance System of Plants

We now know that the HR is triggered when a bacterial avirulence (avr) gene product interacts with the product of a cognate plant resistance (R) gene in a "gene-for-gene" manner [52]. Figure 3 depicts the dynamic nature of gene-forgene interactions. It is such interactions that control race-cultivar specificity within various pathovars off! syringae andX. campestris as well as similar highly host-specific interactions of plants with many fungi, viruses, and nematodes. It is also likely that avr-R gene interactions are important in host-range determination

5.


191

at the level of pathogen pathovar and host species. For example, R syringae pv. tomato carries several avr genes that interact with R genes in soybean, thereby contributing to the avirulence of this tomato pathogen on soybean [53]. Thus, the R gene products form the antennae of a surveillance system that is constitutively arrayed against a broad array of stealth parasites. We will further discuss the R gene system below in the section on innate immunity. Interestingly, most (but not all) of the known R genes recognizing bacteria encode cytoplasmic plant proteins [54]. Thus, the recognition event that betrays unsuccessful stealth parasites appears to occur most commonly inside of plant cells.

C. hrp and hrc Genes and the Type III Protein Secretion System of Plant Pathogenic Bacteria The hrp genes were discovered in P. syringae and so named because mutants lose their ability to elicit the HR in nonhosts or to be pathogenic in hosts [6, 55]. hrp gene clusters have now been at least partially characterized in several other Gram-negative plant pathogens, including X. campestris pv. campestris (black rot of crucifer), X. campestris pv. vesicatoria (bacterial spot of pepper), R. solanacearum (Southern bacterial wilt of tomato), E. amylovora (fire blight of apple), Pantoea (Erwinia) stewartii (Stewart's wilt of com), E. chrysanthemi (soft rot), E. carotovora (soft rot), and Erwinia herbicola pv. gypsophilae (gypsophila gall) [56]. Thus, they appear to be universal among the necrogenic pathogens, although they are not essential for the virulence of the brute-force pathogens. They have not been reported for A. tumefaciens. hrp genes are always clustered and appear to be flanked by genes encoding Avr proteins, harpins, and other potential virulence factors in apparent pathogenicity islands [57-60]. In P. syringae and E. amylovora, the Hrp pathogenicity islands have integrase and tRNA sequences at one border [60a] (J. F. Kim and S. V. Beer, unpublished results), which is typical of pathogenicity islands [61, 62]. hrp genes encode the type III protein secretion pathway, and nine of the hrp genes have been renamed ''hrc'' (HR and conserved) to indicate that they encode conserved components that are also present in the type III secretion systems of Yersinia, Shigella, and Salmonella [63]. The hrc genes were given the last letter designation of their Yersinia ysc homolog. With the availability of more DNA sequences for comparison, it now appears that there are hrp homologs of yscD and yscL, and these can be conceptually included among the hrc genes [56], thus bringing the total to 11. Ten of these widely conserved genes have homologs involved in flagellum biogenesis and flagellar-specific secretion, whereas hrcC encodes a "secretin," an outer membrane protein with homologs involved in type II and filamentous phage secretion [64]. Secretins are thought to multimerize and form a channel through the outer membrane [65]. Thus, the hrc genes appear to

192

JAMES R . ALFANO AND ALAN COLLMER

Round

Wild relatives

Host

J

Cuitivar 0

"i

"2

Pothogen

Outcome

Disease

Race 0

"3"' Resistance

Race 0

Disease

Cuitivar 1

Race 1

Resistance

Cuitivar 2

Race 1

Disease

Cuitivar 2

Race 2

Fig. 3 Gene-for-gene interactions of plants and stealth parasites. Panel A depicts the gene-for-gene concept portrayed in terms of a "game" played between plant breeders and the parasite. Round 1 depicts a highly inbred cuitivar of a crop species being attacked by a stealth parasite (e.g.. P. syringae, X. campestris, a rust fungus, etc.) that presumably has many adaptations for parasitism, but the only relevant features depicted are two molecules, encoded by the avrl and avr2 genes, that have the potential to be recognized by the host. In round 1, however, the plant lacks the factor(s) needed to detect race 0, and disease ensues. In round 2. recognition of the parasite is achieved by introgressing into the crop a major resistance gene {R\) derived from a wild relative. The interaction between the products of/?i and avr\ betrays the parasite, and strong. HR-associated defenses are triggered. Thus, the introduction of a single gene into the plant produces a new. resistant cuitivar. In round 3, the wide planting of cuitivar 1 has put strong selection pressure on the emergence of a new race of the parasite that lacks a functional avr\. There may be a fitness cost to this avr loss, but it is often slight or undetectable. The R\ gene in cuitivar 1 has now been ""defeated." and the crop is susceptible to this new race. In round 4, the plant breeder introgresses a second R gene, which is effective against both races 1 and 0. The resultant cuitivar 2 is widely planted, which leads to round 5. when a new race arises that lacks avn_ function. The game can go on for many more rounds because stealth parasites appear to carry a great number oi avr genes and wild relatives of the crop (typically in the geographic origin of the crop where complex host and parasite populations are in equilibrium) have a similarly large supply of/? genes.

5.

193


B.

HR Test cultivar

Avirulent donor race

D" [ «

Test cultivar

Virulent recipient race Cosmid library

\

Avirulent transconjugant

C.

Tomato {Pto)

m

Pepper (Bs3j

P. syringae pv. tomato

^\r

f< ovrBsS

^

X.campe8trls pv. voslcatorta

Fig. 3 (cont'd) Panel B depicts the universal strategy established by Staskawicz and Keen and their coworkers [53, 255, 270] for cloning bacterial ovr genes. The process requires a test cultivar that is resistant to the donor race and susceptible to the recipient race. A cosmid library of donor DNA is constructed in E. coli and then conjugated into the recipient race. Random transconjugants are inoculated into the test cultivar, and those that have acquired the avr gene are identified on the basis of their avirulence phenotype. That is, they elicit the HR instead of multiplying to high levels and causing disease. Note that the donor can also be another pathovar that is virulent in some other plant species and avirulent on all cultivars of the test species. Panel C clarifies the location of Avr-R protein interactions, which was schematically depicted at the spatial interface between the host and parasite in panels A and B, but for several bacterial Avr proteins appears to occur inside the plant cell. For example, as discussed in the text, there is convincing evidence that recognition of AvrPto and AvrBs3 occurs in the plant cytoplasm and nucleus, respectively.

194


have been recruited from two different protein secretion systems and to direct the translocation of proteins across the bacterial envelope. The Hrp systems of plant pathogens have been exploited to advance our understanding of some universal type III secretion system functions. For example, the Yersinia pestis LcrD and Bacillus subtilis FlhA proteins, which are homologs of HrcV, were initially thought to be regulatory proteins [66, 67]. However, the unambiguous secretion phenotype of an E. amylovora hrcV mutant provided evidence for the function of this protein family in secretion [68]. Similarly, the HrcC protein of Z. campestris pv. vesicatoria was the first type III secretin shown to induce the phage shock protein operon in E. coli, thus providing evidence that it forms a multimeric channel in the outer membrane [69]. Also, the observation that P. syringae pv. syringae hrcC mutants accumulate HrpZ (a type Ill-secreted harpin protein discussed below) in the periplasm, whereas mutants deficient in hrcU or several other hrc genes accumulated HrpZ only in the cytoplasm, provided the first direct evidence that the flagellar homologs specifically directed secretion across the inner membrane in a ^^c-independent manner [70]. The primary function of the type III protein secretion system is to deliver effector proteins into host cells (as will be discussed below). Thus, there are likely to be additional components of Hrp systems that function outside of the bacterial cell to enable effector proteins to cross the plant cell wall and plasma membrane. These components may be quite different between plant and animal pathogens because of the fundamental differences in the surfaces of the host cells. The hrp gene clusters of various plant pathogens can be divided into two groups based on protein similarities, gene arrangements, and regulatory components (discussed below) [71]. The hrp clusters oiR syringae and Erwinia spp. are in group I, those of X. campestris and R. solanacearum are in group II. Interestingly, the 11 /zrc-class genes are the only genes that are obviously conserved between the two groups. Although this may reflect the fact that extracellular components of the type III secretion system are more variable than those contained within the bacterial envelope [72], it is also possible that there are significant differences in the way these two groups of pathogens translocate proteins across the host cell wall.

D.

Harpins and Pilins: Proteins Secreted in Culture in Abundance by the Hrp System

It is important to note at the outset that there are conceptual discrepancies between the biological activities of isolated harpins and the phenotypes of harpin mutants, and this leaves the function of these proteins somewhat enigmatic. Harpins are defined as glycine-rich proteins that lack cysteine, are secreted by the Hrp system, and possess heat-stable HR elicitor activity when infiltrated into the intercellular spaces of leaves of tobacco and several other plants [19]. The first harpin was

5.


195

discovered in culture supematants of an E. coli transformant strongly expressing a functional E. amylovora hrp cluster [73]. Mutations in the encoding hrpN gene in E. amylovora essentially abolished the ability of the bacterium to elicit the HR in nonhost tobacco or to be pathogenic in susceptible immature pear fruit [73]. That is, hrpN mutants had a strong Hrp phenotype. These observations suggested that the HrpN harpin protein was the long sought bacterial elicitor of the HR. Genes encoding proteins with similar properties were subsequently found in R syringae [74, 75], R. solanacearum [76], E. chrysanthemi [77], and other plant pathogens, including the highly virulent strain CFBP1430 of £". amylovora [78]. No significant sequence similarity was found between the harpin genes in different bacterial genera, although the encoded proteins all had the properties that defined them as harpins (the PopA protein of R. solanacearum is unique in selectively triggering the HR in plants in which the bacterium is avirulent). Unexpectedly, with the exception of E. chrysanthemi, harpin gene mutations had little or no effect on the ability of the bacteria to elicit the HR. These observations suggested that harpins are not generally responsible for bacterial HR elicitation or that each bacterium has redundant harpin genes. Observations with the HrpZ harpin for P. syringae pv. syringae highlight the puzzling nature of harpins. The hrpZ gene is conserved in divergent P. syringae pathovars [75], and the isolated protein elicits an apparent programmed cell death in plants that is indistinguishable from the HR elicited by living bacteria [74]. Furthermore, a functional cluster of cloned R syringae pv. syringae hrp genes is greatly reduced in its ability to direct HR elicitation in E. coli when hrpZ is deleted [79]. However, mutation of hrmA [80, 81], which is in a variable region flanking the conserved hrp cluster in the cloned genes, has the paradoxical effect of abolishing the ability of this heterologous system to direct HR elicitation in tobacco without diminishing HrpZ synthesis or secretion [79]. Thus, isolated HrpZ is sufficient to elicit an HR in tobacco leaves, but HrpZ produced by bacteria in plants is not. hrmA has several properties of an avr gene (explained in Fig. 3 and below) that interacts with an unknown R gene in tobacco [82], and these observations suggest that the HrmA protein is the actual elicitor of the HR in this particular system and that HrpZ has a secondary or supporting activity in HR elicitation. Genes encoding HrpW, a second harpin, have recently been found near the hrp gene clusters of E. amylovora and P. syringae [59, 60]. In both bacteria, the N-terminal half of HrpW has general characteristics of harpins, whereas the C-terminal half is homologous to a newly defined class of pectate lyases found in fungal and bacterial plant pathogens. The harpin domain has elicitor activity, but neither the whole protein nor the pectate lyase domain possesses detectable pectolytic activity. As will be discussed in a later section, pectate lyases are major virulence factors in the brute-force pathogenesis of soft-rot Erwinia spp. because they efficiendy destroy plant cell wall structure. Furthermore, the a-l,4-linked galacturonide (pectic) polymers they attack are thought to control the porosity of plant cell walls [1]. Thus, it is intriguing that the R syringae pv. syringae HrpW pectate lyase domain binds specifically to calcium pectate [60]. The R syringae

196


HrpZ harpin binds to plant cell walls and has biological activity only with walled cells [83]. Thus, the current evidence points to plant cell walls as the site of harpin activity, but the function of these proteins remains a puzzle. One of the most abundant proteins secreted by the P. syringae Hrp system in culture is the HrpA pilin [84, 85]. HrpA is the subunit of a pilus that is formed on bacteria in an Hrp-dependent manner and is required for pathogenicity and elicitation of the HR. The Hrp pilus is 6-8 nm in diameter, and thus similar to the pilus required by A. tumefaciens for T-DNA transfer [85, 86]. E. Avr Proteins as Injected, Interchangeable Effectors of Parasitism Avr proteins are so named because of the avirulence phenotype they can confer to pathogenic bacteria, as depicted in Figure 3A. Over 30 bacterial avr genes have been cloned on the basis of this phenotype, as described in Figure 3B [87, 88]. Unlike harpins, no bacterial Avr protein has shown HR elicitor activity when infiltrated into the intercellular spaces of test plant leaves. There are several lines of evidence that typical Avr proteins act inside plant cells following injection by the Hrp system. These lines of evidence also highlight several experimental tools that are unique to plant pathosystems: (1) Cloned hrp gene clusters functioning heterologously in nonpathogenic bacteria, like E. coli, can direct elicitation of an 03

x;

o

O

C

3

cu

N

C3 C3 X

o

II

S

ID

03

_C

II

13 E E T3 O

0)

II

n

cn

D-

?3

fc 'E T-

>>

a.

"c O o C X O o

S

T3

-o" ^ o 5^

-? '^^ E

>

o

GXJ

o x:

•o C

c o OJ

O

7i

c c

^

C)

O

OH

C) X

E

CJ

13

s

o T3

'^

PQ -7^

II

5 tj

03

II

3

E

^ O

c

II S

g^ -^ I ^ E 13 -^ 2i ?i i I ^ ^ -7

X

>>

c

^

2 S S $> •-

'C

OJ

;-> c3

OJ

0)

j2

c o X

1J

E H

X

•>n

2 •- S 13 ^ y: 15 v: CJ

_"^ ~o

E

.,^

O

X

3

7i o 2P Q £

t^

274

CHRISTINA A. SCHERER AND SAMUEL I. MILLER

inflammatory response that might represent an appropriate model for human gastroenteritis [46]. Salmonellae enter normally nonphagocytic epithelial cells by a process known as bacterial-mediated endocytosis, or BME. Early EM studies of guinea pig enterocytes showed that Salmonella first destroy the enterocyte microvilli and then induce the formation of large membrane ruffles in the enterocytes [47]. Bacteria are internalized when the ruffles fuse to form phagosomes. After a period of time, the enterocytes eventually repair the brush border and return to normal. The ability to invade epithelial cells may enable Salmonella to colonize and cross epithelial barriers with greater efficiency, as noninvasive mutants are somewhat attenuated on oral inoculation [48, 49]. Once bacteria have crossed through the intestinal epithelium, they enter phagocytic cells (macrophages) in the underlying lymph tissue. Salmonellae are able to survive and replicate within macrophages, a feature that is correlated with pathogenesis in the mouse typhoid model of S. typhimurium infection [50]. Little is known about the molecular mechanisms whereby Salmonella serovars cause typhoid fever versus gastroenteritis in humans. Typhoid fever is a severe systemic illness characterized by dissemination of bacteria from the intestinal submucosal tissue through the lymphatic system to organs rich in reticuloendothelial tissues such as the liver, spleen, lymph nodes, and bone marrow. Nontyphoidal strain infection, on the other hand, is limited to the intestine. Infection with nontyphoidal Salmonellae results in an influx of neutrophils characteristic of acute inflammation followed by self-limited inflammatory diarrhea. This is in contrast to typhoidal infections, where gastrointestinal symptoms are unusual and lymphoid tissues throughout the body, including the Peyer's patches, contain organisms and a mononuclear infiltrate of lymphocytes and macrophages characteristic of chronic inflammation.

B. Inbred Mouse Enteric Fever Model Infection of BALB/c and C57B1/6 mice with Salmonella typhimurium is widely used as a model system for human enteric fever. Oral infection of these inbred strains with 10,000-100,000 colony forming units (CPUs) of S. typhimurium results in a systemic infection that results in death. The dose required to cause death in 50% of infected mice (LD50) is less than 10 when bacteria are administered intravenously (iv) or through the peritoneum (ip). On oral inoculation, bacteria quickly cross the epithelial layer to colonize the Peyer's patches by infection of macrophages in this lymphatic tissue. Bacteria survive and replicate within macrophages, and usually migrate (via the lymphatic system) to the spleen and liver within 3 to 4 days. Figure 2 (see color plate) shows two intracellular bacteria found in a macrophage within the liver of an infected BALB/c mouse [51]. Bacterial replication within reticuloendothelial tissues and the influx of inflammatory cells (e.g., macrophages and neutrophils) results in hepato-

7.

MOLECULAR PATHOGENESIS OF SALMONELLAE

275

splenomegaly, focal necrosis, and bacteremia, followed by death of the mouse within 5-8 days. Sublethal doses of attenuated bacteria can provide protection against subsequent infection. Such sublethal infections or "vaccinations" are commonly used to investigate the acquired immune response to Salmonella infection. Because inbred mice are so susceptible to Salmonella infection, this model is a very specific, but not necessarily sensitive, system for identification of virulence factors. The majority of strains highly attenuated for virulence in the mouse model contain mutations in regulatory genes or secretory genes that influence expression of multiple factors. In addition, it must be kept in mind that these mice have a genetic defect that alters macrophage function (see the next paragraph), and hence mutants attenuated in this model system are more likely to have mutations that affect survival after macrophage phagocytosis. This association has been noted in genetic studies performed by Fields, Heffron, and coworkers in which in vitro survival within macrophages correlated with mouse pathogenesis [50]. The susceptibility of BALB/c and other inbred mice (C57B1/6, DBA/1, and BIO) to Salmonella infections has been linked to a single locus, Bcg/Ity/Lsh, which mediates innate resistance to Mycobacteria, Salmonellae, and Leishmania, respectively [52, 53]. Bcg^lIty^lLsh^ phenotypes are all linked to the same gene, Nrampl [52, 54], which encodes a macrophage-specific phosphoglycoprotein that is recruited to the phagosomal membrane during phagocytosis [55, 56]. The allele linked to susceptibility to all three bacteria contains a single inactivating mutation at amino acid 169 (Nrampl^^^^^"^). Nrampl is hypothesized to have a transport function that promotes killing of intravacuolar microorganisms, either directly or indirecdy. Other loci that mediate resistance to Salmonella infection have also been described. CBA/N mice, which succumb to Salmonella infection about 3 weeks after infection, carry a mutation in the X-linked Xis locus, which results in a defect in antibody formation. This defect can be rescued by reconstituting the mice with immunologically normal bone marrow cells [57]. Genes within the major histocompatibility complex (MHC) have also been shown to influence the ability to clear Salmonella from infected mice [58, 59]. The Lps locus governs the ability to respond to bacterial lipopolysaccharide (LPS, or endotoxin) [60]. Mice homozygous for the recessive Lps^ (defective) allele (i.e., C3H/HeJ mice) are hyporesponsive to LPS, and are therefore completely resistant to the toxic effects of LPS. At the same time, they are extremely susceptible to Salmonella infection [60-62]. Although this effect is seemingly paradoxical, it presumably occurs because the mice are unable to activate an appropriate innate immune response when challenged with Salmonella. In 1998, at least two more independent loci that mediate innate resistance to infection with S. typhimurium have been identified in MOLF/Ei mice, a wild-derived inbred strain that is also extremely sensitive to Salmonella infection [63]. Thus, susceptibility of mice to lethal S. typhimurium infection is governed by multiple loci. Although these specific

276


defects have led to a better understanding of the immune responses necessary to combat Salmonella infections in mice, it is unclear to what extent these findings will apply to human disease susceptibility and host specificity.

C. Immunology of Salmonella Infections Little information is available about human and inbred murine antigen-specific immune responses to Salmonella infection. It is clear that protective immunity can be achieved against typhoidal Salmonellae infection in humans and a variety of experimental and farm animals. Immunity requires both humoral and cellular responses, but the most important protective antigens are not known. The role of cytotoxic T lymphocytes in human and mouse disease and immunity is also unknown. It is known that flagella, LPS, and Vi antigen are recognized by the immune system and that THl-type lymphocytes that recognize these antigens are promoted by Salmonellae infection. The role of innate immune responses in promoting specific immune responses as well as in disease pathogenesis is an important area for future research that may define important principles of bacterial interactions with immune cells.

1.

INNATE IMMUNE RESPONSE TO SALMONELLA INFECTION

Innate immunity to bacterial pathogens, including Salmonella, has received considerable attention in recent years. By definition, innate immune responses are host defense mechanisms that are not acquired on exposure to infectious agents. These mechanisms do not involve clonal lymphocyte proliferation in response to antigen and do not require a prolonged activation period. Innate immune mechanisms relevant to Salmonellae infection include gastric acidity, peristalsis, complement, opsonins, antimicrobial peptides, cilia, mucin, lysozyme, and the intestinal cell glycocalyx. Innate immune mechanisms relevant after invasion include phagocytosis, antimicrobial activity within phagosomes (antimicrobial peptides, nitrates, oxygen radicals, acidity), and secretion of chemokines and cytokines in response to signature bacterial molecules such as LPS. Chemokines and cytokines secreted by activated phagocytic cells elicit multiple responses, including recruitment of additional phagocytic cells to the site of infection (host inflammatory response) and modulation of the subsequent acquired immune response. A list of a number of innate immune mechanisms and corresponding bacterial properties that allow Salmonellae to evade or interact with the host immune system is provided in Table II. Innate immunity plays an important role in controlling Salmonella infections, both in mice and in humans. Macrophages are a crucial component of this response, as illustrated by the susceptibility of Ity^ mice to infection by S.

7.

277


Table II

Innate Immunity and Salmonella Infection Host immune mechanisms

Bacterial properties/factors

• Stomach acid • Extracellular matrix (glycocalyx) barrier • Antimicrobial peptides — lysozyme — defensins and other cAMP — complement

» Envelope modifications — outer membrane proteins, lipoproteins — lipid A modifications — transporters

•Phagocytosis — mannose binding protein — lectins — complement

»Vi antigen — inhibition of phagocytosis • Macropinocytosis/spacious phagosomes — unique phagosome trafficking • Intracellular survival

• Cytokine and chemokine production • Immune cell chemotaxis — PMN transmigration

• Stimulators of innate immunity — lipid A — flagellin — Type III secretion

typhimurium. Wild-type Salmonellae capable of surviving within macrophages will cause a lethal disease in Ity^ mice, but mutant strains defective in macrophage survival are avirulent in this model. A number of mouse studies have suggested that several inflammatory cytokines, including interleukin-12 (IL-12), interferongamma (IFN-y), and tumor necrosis factor-alpha (TNF-a), also play an important role in the early response to Salmonella infections. For instance, IFN-y-receptor null mice, which cannot respond to IFN-y, are hypersusceptible to infection with normally avirulent S. typhimurium [64]. Likewise, depletion of IL-12 with specific antibodies exacerbates Salmonella infections [65, 66]. Genetic evidence in humans also points to a role for IL-12 in the immune response to Salmonella, Individuals with a mutant form of the IL-12 (31-receptor (which is present on T cells and NK cells) suffer from severe and recurrent mycobacterial and Salmonella infections [67].

2.

ACTIVATION OF INNATE IMMUNE RESPONSES BY BACTERIAL FACTORS

The major bacterial factor responsible for stimulation of the innate immune system during Salmonella infection is lipopolysaccharide (LPS), or endotoxin (reviewed in [68]). Several different mechanisms for the recognition and response

278


to endotoxin exist. First, binding of LPS by macrophage membranes sensitizes them to the action of IFN-y, which results in macrophage activation independent of T-cell stimulation. Second, opsonization of bacteria with LPS-binding protein (LBP), a serum glycoprotein that binds to LPS with high affinity, mediates the adhesion of coated bacteria to macrophages. A third mechanism for endotoxin recognition takes place when LBP interacts with soluble LPS in the serum. This complex binds to the CD 14 receptor on macrophages and monocytes [69] and results in secretion of TNF-a by macrophages [70]. This process is thought to require a coreceptor to activate the signal transduction cascade necessary for TNF-a secretion. A candidate for this coreceptor, the Toll-like receptor-2, has been identified in humans [71]. LBP null mice exhibit the same phenotype as LPS^ mice, in that they are hyporesponsive to LPS [72, 73]. Localized secretion of TNF-a is an important mediator of the host inflammatory response, as it can induce expression of E-selectin on endothelial cells. E-selectin promotes the adherence of neutrophils to endothelial cells and their subsequent migration to sites of infection [74]. Although small amounts of LPS promote a beneficial host inflammatory response, large amounts of serum LPS (in bacteremic individuals, for instance) can induce overproduction of inflammatory cytokines, which results in serious tissue damage and septic (or endotoxic) shock [75]. Although LPS is the major inflammatory agent produced by Salmonella, other bacterial factors have been shown to induce expression of inflammatory molecules in vitro. Expression and production of the flagellar filament protein FliC (either monomers or filaments), for instance, correlates with the induction of TNF-a expression by infected human monocytes [76]. The alternate filament protein FljB induces TNF-a expression to a lesser extent. The significance of these findings has not yet been determined, but they indicate that unique protein structures, such as flagellin, may be recognized by specific innate immune receptors and contribute to inflammatory responses.

3.

ANTIMICROBIAL PEPTIDES

Antimicrobial peptides constitute an important component of innate immune defense against bacterial infection, including Salmonella (reviewed in [77]). Cationic antimicrobial peptides (CAMPs) have been found in a wide variety of animal tissues, including neutrophil granules, phagosomes, mucosal epithelia, and skin. These peptides are amphipathic, cationic molecules that bind to the bacterial surface, at least in part through electrostatic interactions with the negative charges of LPS. Once bound, the peptides permeabilize the bacterial membrane, leading to cell death. Inducible resistance to cationic antimicrobial peptides by Salmonella is required for virulence in the BALB/c mouse model, indicating that wild-type

7. MOLECULAR PATHOGENESIS OF SALMONELLAE

279

Salmonella can at least partially avoid the action of these potent bactericidal compounds [78-81]. Resistance to a variety of other membrane-active molecules, including complement, is likely important to Salmonellae pathogenesis. Such molecules are important both at mucosal surfaces as well as within phagosomes. Which of the myriad of known and unknown membrane-active compounds Salmonellae must resist within host tissues is unknown. It seems likely that the ability to regulate the permeability of the outer membrane to such compounds is an important virulence property, as discussed in more detail in section VILA.

4.

ACQUIRED IMMUNE RESPONSE

Although the innate immune response is crucial for containing Salmonella infections early in disease, protection against subsequent infection, and the ability to cure late or latent infection is related to the development of an antigen-specific (or adaptive) immune response. Both the humoral and cell-mediated arms of the immune response are necessary, as both immune serum and immune T cells are required for protection of susceptible mice [82]. The requirement of T cells for the development of acquired immunity has been further demonstrated in experiments with nude mice or mice artificially depleted of T cells. Infection of such mice results in increased colonization of the liver and spleen, the inability to clear Salmonella infections, and eventual death [83]. Like most intracellular pathogens. Salmonella induces primarily an inflammatory helper T-cell response, designated the THI response. Development of naive CD4-positive T cells into THI cells is induced by secretion of IL-12 by infected macrophages and IFN-y by NK cells during the early (or acute) phase of infection. THI cells help eradicate bacterial infections by activating the microbicidal properties of uninfected macrophages and by inducing production of circulating antibodies that can opsonize extracellular bacteria to maximize phagocytosis by activated macrophages. Vaccination of mice with attenuated Salmonella-Qxpressing heterologous antigens also can result in induction of cytotoxic (008"^) T cells (CTLs) that recognize these heterologous antigens. It is unknown whether CTLs can be induced to native Salmonella proteins. Litde is known about the antigen specificity of immune responses generated during Salmonella infections. However, CD4 helper T cells have been shown to recognize specific flagellin epitopes [84]. It has been known for many years that humans and animals produce antibodies to the major cell surface components: LPS, Vi polysaccharide, and flagella. Neither the diversity of antigens nor the role of specific antigens in immunity to salmonellosis has been defined. Although typhoidal Salmonellae are primarily intracellular pathogens, humoral immunity does play a role, as evidenced by the fact that CBA/N mice, which are unable to mount a normal antibody response to infection, eventually succumb to

280


Salmonella infections [57]. This indicates that extracellular Salmonella can be targeted by antibodies, perhaps preventing spread or initial invasion through the intestinal barrier [85]. The production of secretory IgA antibodies is an important component of the humoral response. These antibodies have been shown to play a protective role against Salmonella infection [86], perhaps by contributing to the development of mucosal immunity. The predominant antigen recognized by Salmonella-induced sera is LPS; however, many other antigens, including outer membrane proteins (Omps), are recognized by immune sera and appear to be important in developing protective immunity [87, 88].

VL In Vitro Models o/Salmonella Virulence A. Modeling Interactions with Macrophages Infection of cultured macrophages is widely used as a model pathogenic system. Such studies have been performed using both primary macrophages and immortalized cell lines from a variety of species (including mouse and human) and using a variety of Salmonella serotypes. Most studies use S. typhimuriiim to infect the susceptible BALB/c mouse-derived J774 and RAW264.7 cell lines, which contain an Nrampl mutation, or primary bone marrow and peritoneal macrophages from BALB/c mice. S. typhi infection has been modeled using primary human peripheral blood monocytes as well as cell lines from a variety of human cells. The standard assay to calculate the ability to survive within macrophages is the gentamicin protection assay. After incubation of cultured cells with bacteria, gentamicin sulfate is added to the extracellular medium. As the eukaryotic membrane is impermeable to gentamicin, internalized bacteria are protected from the antibiotic, whereas extracellular bacteria are killed. The number of internalized bacteria is calculated by releasing intracellular bacteria with a nonionic detergent such as Triton X-100 and plating dilutions of the lysate onto appropriate plates. Strains capable of surviving and replicating within macrophages will show a small increase in the number of intracellular bacteria over a period of 3-24 hours. These numbers must be interpreted with caution, as Salmonella can also exert a cytotoxic effect on infected cells, as will be discussed in more detail below. Despite this concern, this assay was used in a seminal study to demonstrate that survival within macrophages was a useful screening technique that predicted the mouse virulence phenotype [50]. The ability to survive within macrophages depends on the serotype of Salmonella and the source of macrophage, indicating that macrophages may be a

7.


281

component of Salmonellae host specificity. For example, S. typhimurium survives better in mouse splenic macrophages than in peritoneal macrophages [89]. In addition, S. typhi is able to survive in human monocytes, but does not survive as well within murine macrophages [90, 91 ]. Specific bacterial mutations that reduce survival within macrophages will be discussed in a later section.

1.

MACROPINOCYTOSIS AND SPACIOUS PHAGOSOME (SP) FORMATION

Interaction of cultured macrophages with S. typhimurium results in immediate formation of large membrane ruffles that are dependent on actin polymerization. Bacteria are taken up in vacuoles formed from these membrane ruffles. As these vacuoles resemble the macropinosomes formed on induction of membrane ruffling in cells stimulated with growth factors, the process of Salmonella uptake by macrophages has also been termed macropinocytosis [92]. Phagosomes containing wild-type S. typhimurium have been termed "spacious phagosomes" because of their unusually large size (approximately 2-6 |im) and because bacteria appear to be swimming freely within them. An example of spacious phagosomes formed by infected bone-marrow-derived macrophages is shown in Figure 3. Unlike phagosomes containing dead bacteria or non-Salmonellae, spacious phagosomes do not shrink immediately after internalization, but maintain their size for approximately 4-6 hours, after which they decrease in size. The ability to form and/or maintain spacious phagosomes in cultured macrophages correlates

Fig. 3 Spacious phagosomes formed by mouse bone-marrow-derived macrophages. Macrophages were infected with wild-type Salmonella typhimurium and filmed using a Metamorph video imaging system (lOOx phase objective). Several large (spacious) phagosomes are apparent, including one containing a bacterium (arrow). Arrowhead shows site of membrane ruffling. Image reprinted through the courtesy of C. Alpuche-Aranda.

282


with virulence in the BALB/c mouse model [90]. For instance, infection of BALB/c bone-marrow-derived macrophages with serotypes that are not pathogenic for mice, including S. typhi, S. arizonae, and S. pullorum, results in formation of fewer spacious phagosomes that appear to shrink faster than S. typhimurium-conimmng phagosomes. 2.

TRAFFICKING OF SALMONELLA-CONTAINING VACUOLES (SCVS)

Pathogenic microorganisms employ various strategies to survive within the macrophage phagosome, an intracellular membrane-bound compartment containing a variety of antimicrobial factors. Strategies utilized by bacteria to combat this host defense mechanism include: (1) escape from the phagosome to the host cell cytoplasm {Shigella and Listeria; reviewed in [93]); (2) inhibition of phagolysosomal fusion (Legionella); and (3) survival in a novel phagosome that fuses with the lysosomal compartment (Mycobacteria; reviewed in [94]). The strategies that Salmonellae use to survive within macrophages have not yet been fully elucidated, but results to date suggest two general mechanisms: (1) that Salmonellae traffic to the phagolysosome and synthesize factors, including those that remodel the bacterial surface, that promote resistance to microbicidal activities within the phagosome; or (2) that Salmonellae alter host cell processes that modify the SCVs and promote bacterial survival. Though both factors may contribute to survival of Salmonellae, conflicting reports exist in the literature on whether the trafficking of the SCVs to the lysosomal compartment is altered. At least two different phases of SCVs have been observed: initial spacious phagosomes, which appear to fuse with the lysosomal compartment within 15 minutes and which, at least in some cases, take a longer time to acidify due to their large volume, and a later (V2 to 4 hours) maturing acidified phagosome in which the phagosomal membrane is tightly adherent to the bacteria. Although some studies have addressed the composition of later phagosomes, the majority of studies have focused on the biochemical properties of early (spacious) phagosomes.

3.

ACIDIFICATION OF SALMONELLA-CONTAINING VACUOLES

Two studies have utilized modem methods to measure the pH of SCVs (see [95] for a review of methodology). Using fluorescein-isothiocyanate-conjugated dextran (FITC-Dx) as a probe, one group measured the pH of individual spacious phagosomes formed in bone-marrow-derived macrophages. They observed that within the first 2 hours of formation the mean pH of the SPs was 5.5. The mean pH of SPs that persisted for 4-6 hours subsequently decreased to approximately 4.9 [96]. In contrast, phagosomes containing killed bacteria (which were taken up

7.


283

by receptor-mediated phagocytosis) rapidly acidified to a pH of less than 4.9 within the first 2 hours. This study indicated that acidification of the SPs during the time of phagolysosomal fusion might be delayed due to the presence of a large volume of fluid of neutral pH. Another study of this type used a different pH-sensitive probe, DM-NERF dextran, to measure the pH of SCVs [97]. In this study, the pH of SCVs was observed to decrease more rapidly than previously observed. The pH of SCVs formed within bone-marrow-derived macrophages decreased to approximately 4.5 in 50 minutes, whereas phagosomes formed within RAW 267.4 macrophages dropped to an approximate pH of 4.0 within 10-20 minutes. Phagosomes containing latex beads or killed bacteria acidified to a pH of 4.0 within 15 min. Despite the conflicting data, the results of Alpuche-Aranda et al (see [96]) are consistent with delayed acidification of some of the larger SPs formed. The discrepancies described above could be the result of differences in experimental protocol. First, it is possible that the measurement of pH may depend on the population of phagosomes observed. In the first study, the pH of the most spacious phagosomes was reported. Because of their large volume, it is likely that acidification of this population of phagosomes is retarded and some vacuoles measured had neutral pH. As the number of spacious phagosomes drops significantly within the first 2 hours postinfection [92], continued observation of spacious phagosomes might bias the measurement toward a higher pH. Second, the composition of the bacterial inoculum may have played a role in determining the fate of the SCVs. The first group used opsonized stationary-phase bacteria to infect macrophages, whereas the second group used log-phase bacteria grown under limiting oxygen conditions. As bacterial growth state and oxygen limitation can affect infection of epithelial cells [98, 99] and expression of various virulence genes [100], it is reasonable to hypothesize that phagosome development in the macrophage might be similarly affected by alterations in virulence gene expression. In order to resolve the conflicting data, infection of macrophages and measurement of intravacuolar pH will have to be normalized so that these results can be direcdy compared.

4.

PHAGOLYSOSOMAL FUSION

Phagosome maturation is normally accomplished by fusion with endosomal and lysosomal compartments. This results in acidification of the vacuole and digestion of phagocytosed particles by lysosomal enzymes. Phagolysosomal fusion is usually monitored by either the fusion of phagosomes with fluid phase markers from the lysosomal compartment (discussed in [95]) or by colocalization of phagosomes with endosomal- or lysosomal-specific markers (reviewed in [94]). Fusion with the early endosome results in delivery of the GTPase Rab 5 to the phagosome, whereas fusion with the late endosome is characterized by the presence of the mannose 6 phosphate receptor and Rab 7. Lysosomal markers

284


include the lysosomal-associate membrane proteins (LAMPs), lysosomal-associated proteins (LAPs), and cathepsin L. These markers are not entirely restricted to lysosomes, however, as LAMPs can also be found in the late endosome compartment. At least three groups have reported that SCVs display incomplete fusion with lysosomes [97, 101, 102]. One of these studies utilized modem immunofluorescence techniques to determine the extent of phagolysosomal fusion [103]. On infection of both primary and cultured macrophages with wild-type S. typhimurium, the lysosomal markers LAMP-1 and LAP were detected in over 90% of SCVs 30 minutes after infection. The mannose-6-phosphate receptor and cathepsin L were only detected in a small proportion of phagosomes, even after a 10-hour incubation. In contrast, 80% of vacuoles containing latex beads colocalized with the latter two markers. In addition, SCVs appeared inaccessible to subsequently loaded endocytic markers, including rhodamine-labeled transferrin (which should traffic to early endosomes) and the fluid phase lysosomal marker fluorescein dextran. Conversely, another group, using similar techniques, found that, in addition to LAMP-1 and LAP, most SCVs also contained cathepsin L within 20 minutes of phagocytosis [104]. Observation of SCVs in this study was enhanced by pulse-chase labeling of phagosomes with fluorescent dextran during infection so the exact age of the vacuole could be determined. In addition, this group showed that SCVs readily fused with a fluid-phase lysosomal marker, Texas Red-ovalbumin, which was preloaded into infected macrophages. It is clear that SCVs have a novel morphology on formation. The delayed shrinkage of these vacuoles is also novel compared to growth-factor-induced macropinosomes. Given the above results, it seems likely that SCVs can undergo differential trafficking. Factors affecting trafficking might include bacterial growth conditions (which promote expression of different virulence factors) or the number of bacteria in contact with macrophages or within a vacuole, which may effect initiation of apoptosis in macrophages. Given the results of Oh et al. [104], it is clear that SCVs can rapidly fuse with the fluid phase of the lysosomal compartment and that bacteria can survive in this environment. However, since not all endocytic and lysosomal markers are known or have been tested, and given the morphologic novelty of the SPs, it would seem likely that SCVs have some biochemical novelty. The uniqueness of the more mature, acidified postspacious SCV remains to be better defined. Discovery of this novelty and the biochemical mechanism by which Salmonellae induce alterations of the phagosome are required to resolve these issues.

5.

CYTOTOXICITY OF SALMONELLAE ON MACROPHAGES

Several groups have observed that Salmonella infection of macrophages results in cytotoxicity, either in the form of necrosis (generalized cell death) or apoptosis (programmed cell death). Apoptotic death is characterized by membrane blebbing.


285

cell shrinkage, and chromatin condensation and fragmentation. Chromatin fragmentation is usually visualized using the fluorescent-based TUNEL assay, in which labeled dUTP is incorporated into nicked DNA by the enzyme terminal deoxynucleotide transferase (TdT). Necrotic death, on the other hand, is characterized by cytoplasmic swelling, membrane disruption, the disappearance of nuclear chromatin, and eventual lysis of the cell. Necrotic death usually produces an inflammatory response in the area, whereas ingestion of apoptotic cell fragments by neighboring cells prevents this response. Cytotoxicity of Salmonella on macrophages has been observed under several different situations. In one situation, infection of macrophages with relatively low numbers of bacteria- (a multiplicity of infection [moi] of 10-20 bacteria per cell) induced apoptosis within 45 minutes to 2 hours [105, 106]. Initiation of apoptosis required a functional type III secretion system (or TTSS) from Salmonella pathogenicity island 1, a specialized secretion apparatus that mediates the translocation of bacterial effector proteins into the host cell cytosol (described in §VII.B). Although both macrophage cell lines and bone-marrow-derived macrophages underwent apoptosis in these studies, primary macrophages appear to be much more sensitive to killing than the cell lines [106]. The requirement for bacterial entry is uncertain at this time, as there are conflicting results as to the effect of Cytochalasin D on induction of apoptosis. Apoptosis of cultured macrophages has also been observed when cells are infected with a high moi of stationary phase bacteria [107]. In this case, apoptosis is not apparent until 10-12 hours after infection. Induction of apoptosis in this system does not require the TTSS, but does require genes in the ompR/envZ regulon. Infection of macrophages with a high moi of 5". typhimurium has also been shown to induce necrotic death of these cells [106]. The differential induction of apoptosis and necrosis in culture may represent similar situations during in vivo infections. For instance, the type III secretion-dependent induction of apoptosis by small numbers of bacteria might mimic the situation during early stages of infection. Conversely, necrotic and apoptotic death in response to large numbers of bacteria might be indicative of late-stage infections, when the reticuloendothelial tissues contain high bacterial loads and undergo major inflammatory and necrotic responses. Nevertheless, it is unclear if activation of apoptotic pathways in cultured cells is physiologically significant. For instance, although induction of apoptosis might help bacteria escape from macrophages in the Peyer's patch, the TTSS associated with apoptosis in vitro is not required for pathogenesis in vivo. In one study, confocal microscopy of livers and spleens isolated from infected BALB/c mice has shown that, not only does S. typhimurium reside mostly within macrophages in vivo, but also that many of these macrophages display the hallmarks of apoptotic death [51]. Interestingly, a number of apoptotic macrophages did not actually contain bacteria, indicating that 5. typhimurium can induce apoptosis both direcdy and indirectly. Thus, it is possible that apoptotic macrophages are not responding to infection with Salmo-

286


nella per se, but rather to the secretion of cytokines and chemokines by infected (and dying) macrophages as a result of the acute inflammation of these tissues.

B. Modeling Salmonella Interactions with Epithelial Cells 1.

BACTERIAL-MEDIATED ENDOCYTOSIS

Although the main portal of entry for Salmonella into the gut epithelia is thought to be M cells, Salmonellae are also capable of entering normally nonphagocytic epithelial cells. A small percentage of these bacteria also transcytose from the apical to the basolateral surface, effectively crossing the epithelial barrier. A number of different cultured cell lines have been used to study this process, including nonpolarized epithelial cell lines such as Hela, Hep-2, and Henle-407 cells, and the polarized cell lines Caco2 and T84. As with macrophages, Salmonella induces membrane ruffling in epithelial cells that is morphologically similar to membrane ruffling induced by growth factors [108-110]. However, these ruffles are localized to the site of bacteria-cell interaction. Bacteria are taken up within vacuoles formed from the ruffles, a process known as bacterial-mediated endocytosis (BME). A scanning electron micrograph of Salmonella interacting with membrane ruffles is shown in Figure 4. Bacterial-mediated endocytosis is accompanied by major cytoskeletal rearrangements, in particular the accumulation of actin filaments around the bacteria [111]. Actin reorganization is required for bacterial entry, as the addition of cytochalasin B and cytochalasin D, actin filament inhibitors, greatly decreases the proportion of internalized bacteria. In addition to actin, a number of other cytoskeletal components, including a-actinin, tropomyosin, tubulin, and vincuHn, are observed around the bacterial vacuole. Interestingly, despite the presence of tubulin around the vacuole, microtubule inhibitors have no effect on bacterial invasion [HI, 112]. Invasion-defective mutants do not induce membrane ruffling or BME. Induction of membrane ruffling appears to be sufficient to allow bacterial entry, as passive entry of noninvasive bacteria has been induced by addition of either growth factors (such as EGF) or wild-type bacteria [108, 113]. A number of bacterial factors are required for epithelial cell invasion. Most invasion-defective strains contain mutations in the type III secretion system located in Salmonella pathogenicity island 1, which will be discussed in detail later in this chapter. Bacterial-mediated endocytosis in epithelial cells results in the formation of spacious phagosomes similar to macropinosomes in the host cell [114]. Salmonella are capable of replicating within these vacuoles, a feature which in some cases correlated with disease in the inbred mouse enteric fever model [115]. Replication within epithelial cells does not require the same gene products as survival within macrophages, as mutants that cannot replicate within epithelial

7.


287

Fig. 4 Scanning electron micrograph of polarized MDCK cells infected with wild-type Salmonella typhimurium. Two bacteria are attached to a large membrane ruffle; one bacterium appears to be in the process of internalization (white arrow). Note that the other adherent bacteria exhibit multiple surface appendages. Magnification 8400x. Photo reprinted through the courtesy of M. A. Clark, T. A. Booth, and M. A. Jepson.

cells can survive within macrophages. Like the phagosomes formed in macrophages, Salmonella-conidiXmng spacious vacuoles persist within epithelial cells and are still visible 2 hours after invasion [114]. The vacuoles appear to be mildly acidic [116], and contain some markers of the lysosomal compartment. However, fusion with lysosomes appears to be incomplete, as most vacuoles contain the lysosomal membrane glycoproteins LAMP-2 and LAP, but do not contain the mannose 6-phosphate receptor and associated lysosomal hydrolytic enzymes [117]. In addition, fusion with fluid lysosomal tracers appears to be minimal. Interestingly, BME of Salmonella also induces the formation of a filamentous network of lysosomal glycoprotein-containing tubular lysosomes that extend from Salmonella-conidiming vacuoles [118]. The formation of this network requires bacterial protein synthesis, vacuole acidification, and an intact microtubule network. Unlike Listeria and Yersinia, Salmonella enters cells by induction of macropinocytosis rather than receptor-mediated endocytosis. Therefore, inhibitors of phosphatidylinositol (PI)-3-kinase do not prevent Salmonella invasion [119]. The

288


molecular mechanism by which Salmonella induces membrane ruffling and BME in epithelial cells has not yet been fully elucidated. Induction of membrane ruffling is an extremely fast event (within seconds), and is likely to take advantage of cellular signal transduction pathways that can be immediately activated. Although early experiments suggested a role for tyrosine phosphorylation of the epidermal growth-factor receptor (EGF-R) during invasion of Henle-407 cells by Salmonella [113], subsequent experiments by other groups were unable to reproduce this result, and entrance does not require host cell tyrosine phosphorylation [120]. Investigation of the activation of membrane ruffling by Salmonella has focused on the small GTP-binding proteins that modulate actin rearrangements in eukaryotic cells, including Ras, Rac, Rho, and CDC42. Expression of constitutively active alleles of these proteins result in various actin-based cytoskeletal changes: expression of activated Rac induces the formation of lamellipodia, expression of activated Rho results in the formation of stress fibers, and activated CDC42 induces filopodia formation (reviewed in [121]). Although uptake of Shigella flexneri into epithelial cells has been shown to require Rho [122], Salmonella enters epithelial cells by a Rho-independent process [110] that requires CDC42 and, to a lesser extent, Rac [123]. Expression of a dominant negative CDC42 mutant interferes with bacterial invasion, whereas expression of a constitutively active CDC42 allele facilitates internalization of noninvasive strains. The mechanism by which Salmonella activates CDC42 and Rac is not yet known, but it is likely that they are activated by translocated bacterial effector proteins, as at least one bacterial protein (SopE) has been shown to interact with CDC42 and Rho and induce ruffling when expressed at high levels in mammalian cells [124].

2.

CELLULAR RESPONSES TO SALMONELLA INFECTION

Invasion of epithelial cells by Salmonella results in initiation of a complex cellular signal transduction cascade. One of the first documented cellular responses was an apparent increase in free intracellular calcium (Ca^"^) [125]. The ability to mobilize Ca^"*^ is necessary for the process of invasion, as Ca^"^ channel antagonists such as lanthanum or cadmium chloride block invasion. Interestingly, the Ca^"^ response is not blocked by cytochalasin D, and therefore does not appear to require bacterial entry. Ca^"^ fluxes have never been documented in individual infected cells, so it is unclear if the observed changes are a direct effect of bacteria-cell communication or if they are due to decreased membrane integrity as a result of bacterial protein translocation or cytotoxicity. In addition to Ca^"^ fluxes, increased levels of other second messenger molecules, including phospholipase A and arachidonic acid metabolites, appear to contribute to invasion [125].


289

Studies in the late 1990s have shown that three members of the mitogen-activated protein (MAP) kinase family—Jun kinase (JNK), p38, and ERK—are activated on internalization of wild type, but not invasion-defective, Salmonella [126]. These kinases appear to activate a signaling cascade that eventually results in the expression and secretion of IL-8, an inflammatory chemokine, as a specific inhibitor of p38 MAP kinase prevents IL-8 secretion. Activation of IL-8 expression is mediated by the transcription factors AP-1 and NF-KB, and is in part due to degradation of the NF-KB inhibitor, I-KB, and an increase in intranuclear c-Jun concentrations. These effects are not immediate; rather, they occur approximately 2 hours after infection. Activation of this signaling cascade by Salmonella requires a functional type III secretion apparatus and translocase, indicating that these effects might be mediated by translocated effector proteins. However, there is a high degree of crosstalk between mammalian signal transduction molecules, and paths of activation are quite difficult to elucidate. Cytokines other than IL-8 are also secreted by infected epithelial cells. In vitro studies have documented secretion of IL-6 by human small intestinal epithelial cells (lECs) infected with S. typhi [127]. The induction of IL-6 secretion is cytochalasin D independent, indicating that only bacterial adherence is required.

3.

INTERACTIONS WITH POLARIZED EPITHELIAL CELLS IN VITRO

Polarized tissue culture cells have been used to study the interaction of Salmonella with epithelial cells under slightly more physiological conditions. When grown on permeable filters, some cell lines will form polarized epithelial monolayers. Salmonellae can enter polarized epithelial cells from the apical side via formation of large membrane ruffles. Although bacterial invasion generally does not affect monolayer integrity [128], infection of polarized epithelial cells with high doses of bacteria can result in a decrease in transepithelial electrical resistance, which might be the result of major remodeling of the cytoskeleton at intercellular junctions at the apical (but not basolateral) pole [129]. In addition. Salmonella typhimurium does not stimulate chloride secretion at the apical pole of polarized T84 cells, indicating that it is unlikely that human gastroenteritis is a result of Salmonella secretion of a cholera-toxin-like activity [128]. Association of Salmonella with the apical side of T84 monolayers results in basolateral secretion of IL-8 and induction of neutrophil transmigration across the monolayer, both hallmarks of the host inflammatory response [128]. Basolateral secretion of IL-8 directs neutrophils to a subepithelial (basolateral) position, but transmigration of neutrophils across the monolayer appears to occur in response to a novel cytokine signal [130]. The induction of IL-8 secretion and transepithelial migration requires bacterial invasion and subsequent protein synthesis in both bacteria and cells. This in vitro system may be a model for human gastroenteritis, as the ability to induce an inflammatory response in vitro

290


correlates with the abihty of strains to cause human gastroenteritis [131]. Serotypes such as S. typhimurium that cause gastroenteritis in humans also induce transepithelial migration in cultured T84 cells. Conversely, host-restricted serotypes that do not cause gastroenteritis, including S. typhi, S. pullorum, and S. arizonae, cannot induce neutrophil migration in vitro.

VIL Virulence Factors In this section we will discuss a number of factors that contribute to pathogenicity in in vitro and/or in vivo assays. A list of many proposed virulence factors and their in vitro and in vivo phenotypes is also provided in Table III. The interaction of Salmonellae with their hosts is complex and requires many different bacterial factors. Approximately 4% of the Salmonella genome has been estimated to be required for virulence in the BALB/c mouse model system [132]. A number of virulence factors map to regions of the genome known as Salmonella pathogenicity islands (SPIs). Pathogenicity islands contain large segments of DNA that appear to have been acquired by horizontal transmission from an exogenous source, as the ratio of GC to AT basepairs in these regions differs from that of the rest of the Salmonella chromosome [133]. To date, five such islands have been described [134-139], at least two of which are specific for Salmonella species.

A. Major Transcriptional Regulators 1.

THE PHOP/PHOQ REGULON

One of the best-characterized transcriptional regulons required for Salmonella pathogenesis is made up of the two-component regulatory system [187] PhoPand PhoQ, which controls expression of more than 40 genes [159, 161]. PhoP/PhoQ is required for virulence in mice and humans, survival within macrophages, growth on succinate as a sole carbon source, and growth in the presence of magnesium limitation. PhoQ is a sensor histidine kinase [188] that phosphorylates PhoP, a response regulator, in response to environmental conditions. PhoQ activity is repressed by the divalent cations magnesium and calcium. PhoP-P04 activates expression of a set of genes arbitrarily designated pags (for PhoP-activated genes), which promote Salmonella survival within host tissues. Proteins encoded by ^pag include a nonspecific acid phosphatase [189], cation transporters, outer membrane

7.


291

proteins, and enzymes important for lipopolysaccharide modification. In addition, PhoP-P04 activates expression of the pmrCAB (also known as pagBpmrAB) operon encoding another two-component regulatory system that is activated in response to an acidic environment [190, 191]. PhoP-P04 represses transcription of another set of genes found on SPI1 (designated /7rgs, for PhoP-repressed genes) required for epithelial cell invasion and spacious phagosome formation, including hilA, which encodes a transcriptional regulator, and the prgHIJKorgA operon, which encodes components of a type III secretion system [48, 92, 171, 192]. Regulation and function of SPIl will be discussed in the following section and in section VLB. In addition to regulating a large number of genes, iht phoPQ operon is autoregulated, as full expression requires both PhoP and PhoQ [193]. A summary of the known components of the PhoP regulon is shown in Figure 5. Expression of PhoP-activated genes is maximally induced within nonspacious acidified phagosomes as much as several hours after phagocytosis [96, 194]. Although the in vivo signals for PhoQ activation are not fully defined, activation can be induced in vitro by low pH or growth in media containing low (micromolar) concentrations of the divalent cations Mg^"^ and Ca^"*" [96, 195, 196]. PhoQ contains distinct binding sites for Mg^+ and Ca^^, and is maximally repressed in the presence of both cations [197, 198]. Therefore, the best-defined signal in vitro is depletion of the divalent cations Mg^"*" and Ca^"^. This has led to the hypothesis that PhoQ is also activated by limiting concentration of divalent cations in vivo, and that expression of a subset of pags in response to low pH is mediated by PmrA and PmrB. Mild acidic growth conditions have been shown to promote transcription of the subset of PhoP-activated genes that are also PmrA-dependent [199]; in addition, transcriptional activation of psiD (also called pmrC or pagB) by mild acidification is independent of the PhoQ protein [196]. However, a recent report indicates that the expression of several proteins induced upon acid shock is dependent on PhoP/PhoQ but not PmrA [195]. Though posttranscriptional effects cannot be ruled out, this suggests that PhoQ or some unidentified target of PhoP can respond to low pH. In addition, as pH can affect the effective concentration of cations in solution, it is difficult to definitively rule out the possibility that PhoQ is affected by pH in vivo. Since the complex cationic milieu of the phagosome, other than pH, has not been directly measured, this question cannot be completely resolved. Two classes of mutations within phoPQ have been particularly useful in the study of this regulon. Mutations which inactivate PhoP (phenotype PhoP-null or PhoP") cannot repress prgs and are phenotypically similar to wild-type bacteria grown in high Mg^"^. Conversely, the phenotype of a phoP constitutive mutation (pho-24 or phenotype PhoP), in which a mutation in the periplasmic domain of PhoQ results in increased phosphorylation of PhoP, is similar to that of bacteria grown in micromolar concentrations of Mg^"^ [188]. The PhoP phenotype is also thought to mimic in part the environment within host phagosomes, where PhoP is activated. Both PhoP-null and PhoP mutants are avirulent in the mouse typhoid

292


Table III

Salmonella Virulence Factors

Protein/ locus

Function

Phenotype in vitro

Phenotype in vivo (mouse, unless specified)

Slightly attenuated on oral inoculation [141]

agf

Thin aggregative fimbriae (or curli)

aw

Aromatic amino acid synthesis

Avirulent in mice [142] and humans (5. typhi) [143]

Crp, Cya

Cyclic AMP receptor, adenylate cyclase

Avirulent in mice [144] and humans {S. typhi) [143]

fim

Type I fimbriae

Adherence to (and invasion of) HeLa cells [145]

Slight decrease in LD50 on oral inoculation [141]

GalE

UDP-galactose-4epimerase, LPS synthesis

Galactose sensitive and rough phenotype

Avirulent in mice [146]; S. typhi mutant still causes disease in humans [147]

HilA

Transcriptional regulator of SPIl

Epithelial cell invasion defect [148]

InvABCEFGHIJ

Type III secretion components (SPI1)

Secretion, translocation, and invasion defects [49, 149-152]

Attenuated by oral inoculation [49]

Ipf

Long polar fimbriae

Adherence to (and invasion oO Hep-2 cells [145]

Adherence to murine Peyer's patch [153]; slight decrease in LD50 on oral inoculation [154]

MetL

Methionine biosynthesis (homocysteine production)

Resistance to host nitric oxide [155]

Attenuated in mice [155]

MgtC

Unknown

Intracellular survival [137, 156]

Avirulent [137, 156]

OmpR/EnvZ

Transcriptional regulators (osmolarity)

Intracellular survival [157]

Avirulent [157]

pef

Plasmid-encoded fimbria

Adherence to murine small intestine [158]

Reduced fluid accumulation in infant mice [158]

PhoPQ

Transcriptional regulators (pH, Mg-^

Intracellular survival, CAMP resistance [78, 80, 159], stimulation of cytokine secretion [160]

Avirulent in mice [159, 161]; PhoP" mutants are immunostimulatory compared to wild type [162]; S. typhi PhoP null mutants are avirulent in humans [163, 164]

PmrAB

Transcriptional regulators (pH)

Polymyxin resistance [165]

PrgHIJK

Type III secretion components (SPI 1)

Invasion defect [48]

pur

Purine biosynthesis

Adherence to cultured mouse small intestinal epithelial cells [140]

Attenuated by oral inoculation [48, 90] Avirulent [166]

continued

7.

293


Protein/ locus

Function

Phenotype in vitro

Phenotype in vivo (mouse, unless specified)

Rck

Resistance to complement

RpoS

Transcriptional regulator (stationary phase, stress response)

Sip/SspBCD

Translocase (SPIl)

Invasion and translocation defect, decreased macrophage cytotoxicity [105, 171-173]

SlyA

Transcriptional regulator


SopB

Inositol phosphate phosphatase [176]

Fluid secretion and inflammation in cow ileal loops [177]

SopD

Unknown

Fluid secretion and inflammation in cow ileal loops [178]

SopE

Guanine exchange factor

Slight invasion defect iS.dublin, [179]), actin rearrangements when expressed exogenously [124]

SpaOPQRS

Type III secretion components (SPI1)

Secretion, translocation and invasion defects [180]

Attenuated by oral inoculation

SPI3

Unknown {mgtQ

Macrophage survival [137]

Avirulent [137]

SPI4

Type I secretion system?

Survival within macrophages [139]

SPI5

Unknown

SptP

Protein tyrosine phosphatase

SpvABCD

Unknown (sopB)

Growth within reticuloendothelial system [183]

SpvR

Transcriptional regulator of spvABCD

Growth within reticuloendothelial system [183]

SsaBCDE,G-V

Type III secretion apparatus of SPI2


Avirulent [184]

SseBCD

SPI2 translocase?

Intracellular survival [185]

Avirulent [185]

SsrAB

Transcriptional regulators of SPI2


Avirulent [184]

TolC

Unknown

Increased sensitivity to complement and detergents


Prevents complementmediated lysis [167] Avirulent in mice [168, 169] and humans (S. typhi) [170]


Effectors of gastroenteritis in mice and cows [138] Actin rearrangements when exogenously expressed [181]

Slight decrease in colonization of liver and spleen [182]

- / c

l\ aT

S

o v>

+"

^• sa -> '-».

0£

+ O «s

0£ 03

1

2C3

Ui

3

^

(/D

c«

fcJO

-:^ g- ^

{ ^ HH C8

o

BP''

p.

J|::,;:?:^ ....

I'-'-yPi-iiil

/••111 :::||il| . ^ I M i V§M:L2:M ^ ^^-^"J^M

^^•1 "fl

| \ ill

1

d

^'

ili^^^^^pife. /

J rY ^ l 1

\v

( y

^^^^

1 1

••

I^^^^B

ffl

Q =2

>% ^ Xi ""

Co -^^ c y.

^

dj X

^

3 i3

Il» t-ll o

3

^ .- c

>

x:

s^ ^ 00 "i> o = -^3 c^ C3 ^

's:; bx)

^ o 1 -DCD h

c/5

.?^

306


245], and are thus sometimes referred to as "translocase" proteins. Based on homologies to proteins in Yersinia and Shigella, it is thought that the translocase proteins interact with the TTSS and each other to faciUtate formation of a pore in the eukaryotic cell membrane. For instance, there is some evidence that IpaB and IpaD can associate in the bacterial membrane, while IpaB and IpaC can associate in bacterial supematants (this interaction is thought to be prevented by the IpgC chaperone in the bacterial cytosol) [246, 247]. IpaB, YopB (from Yersinia), and SipB all contain several potential membrane-spanning domains and exhibit some homology to the RTX family of pore-forming toxins. These proteins have been hypothesized to make pores in the eukaryotic plasma membrane [248]. SipC and SipD may help form the pore or a channel through which effector proteins are translocated or, alternatively, could help link SspB to the TTSS structure. Instability or shearing of the translocase once it is inserted into the host plasma membrane might explain the resultant presence of SipC and SipB inside the cell. A diagram of this model is shown in Figure 9 (see color plate). In addition to its role as a translocase, Sip/SspB is required for induction of TTSS-dependent apoptosis in macrophages [105]. Interestingly, IpaB has been shown to be sufficient to induce apoptosis in macrophages [249]. The current hypothesis for the role of IpaB is that it direcdy binds to interleukin 1 p-converting enzyme (ICE, also known as Caspase-1) to activate the apoptotic pathway [250]. ICE, which cleaves IL-lp to its mature form, is a natural inducer of apoptosis in cells. Recent data indicate that SipB plays a similar role in the induction of apoptosis by S. typhimurium [251].

6.

TRANSLOCATED PROTEINS (PUTATIVE EFFECTOR PROTEINS)

To date, six Salmonella proteins (excluding SipB and SipC) have been shown to be translocated into the eukaryotic cell cytosol in an SPIl-dependent manner. These include SptP [182, 245], AvrA [252], SopE [124, 179, 181], SopB [177], SopD [178], and Ssp/SipA [253]. The AvrA protein is homologous to the avirulence factor AvrRxv from Xanthomonas campestris pv. vesicatoria and the apoptosis-inducing protein YopJ from Yersinia pseudotuberculosis [252]. No virulence defects are associated with an avrA mutant, and no known function has been assigned to this protein at this time. The SopB protein, which is present on SPI5 in S. dublin [138], has been shown to promote fluid secretion and an inflammatory response in infected catde [177]. Recendy, it has been demonstrated to be an inositol phosphate phosphatase [176]. Mutations in the S. typhimurium sopB homolog sigD has been implicated in invasion of epithelial cells [254]; this phenotype is not observed in S. dublin sopB mutants. S. typhi and S. dublin infection of human polarized epithelial cells results in an increase in intracellular levels of inositol 1,4,5,6-tetrakisphophate

7.


307

[Ins(l,4,5,6)P4], a signaling molecule that induces secretion of chloride in Salmonella-inftcicd cells [255]. SopB inositol phosphate phosphatase activity has been implicated in this induction, which may contribute to fluid secretion and diarrhea [176]. Another translocated protein from Salmonella dublin, SopD, has also been shown to act in concert with SopB to promote an inflammatory response in bovine hgated ileal loops [178]. Since SopB and SopD have been shown to play a role in intestinal inflammation and fluid secretion in cattle, and since SPIl is required for inflammatory responses that result in neutrophil transmigration in a human model system for gastrointestinal inflammation [126, 131], it is attractive to hypothesize that SPIl-translocated effector molecules promote responses that lead to bovine and human gastroenteritis. Several translocated effector proteins have been associated with cytoskeletal rearrangements in eukaryotic cells. Recent evidence indicates that Sip A binds actin, and may be directly involved in Salmonella-induced membrane ruffling [253]. SptP and SopE have both been shown to induce actin rearrangements when expressed exogenously in eukaryotic cells [124, 245]. The Salmonella protein tyrosine phosphatase (SptP) is an interesting protein that appears to be a chimera between two Yersinia proteins, YopE and YopH [182]. The amino terminus of SptP (residues 107-290) is homologous to two bacterial cytotoxins: Exotoxin S from Pseudomonas aeruginosa and YopE from Yersinia spp. The carboxyl-terminal portion of SptP (residues 340-513) is homologous to the catalytic portion of the Yersinia protein tyrosine phosphatase YopH. SptP appears to be required for full virulence of Salmonella in mice, as sptP mutants display a slight defect in the ability to colonize spleens [182]. SopE was first described in S. dublin, where mutants displayed mild defects in epithelial cell invasion [179]. Recendy, SopE from S. typhimurium has been shown to bind to and induce GDP/GTP exchange by the small GTP-binding proteins CDC42 and Racl, resulting in their activation [124]. SopE also facilitates activation of downstream signal transduction pathways, as expression of SopE resulted in the increased phosphorylation of c-Jun. This is the first example of a Salmonella effector protein that directly modulates a cellular factor to induce actin rearrangements and nuclear responses. The role of SopE, SptP, SipA, and other translocated effector proteins in promoting invasion is unclear, as mutants are still able to enter epithelial cells (albeit with slightly lower efficiency, at least in the case of the sopE mutant). It seems most possible, given the identified phenotypes of expression in host cells, that multiple effectors are involved in redundant functions to induce actin rearrangements and membrane ruffling in infected cells; thus, individual mutants would only display mild or no defects in invasion. Alternatively, the effects observed are due to high nonphysiological expression of these proteins, and other unidentified factors with different functions promote membrane ruffling and BME.

308

C.


Factors Required for Systemic Infection

Initiation of systemic infections by Salmonellae requires that bacteria survive various severe environments, including the low pH of the stomach and phagosome, oxidative stress within the phagosome, and the relative unavailability of nutrients such as carbon and iron. As discussed before, the acid tolerance response is highly regulated and is probably necessary for Salmonella to survive in vivo. In this section, we will discuss other systems required for virulence in the mouse model, including metabolic pathways and the TTSS encoded in another Salmonella pathogenicity island.

1.

SALMONELLA PATHOGENICITY ISLAND 2

Salmonellae are unique among the Enterobacteriaceae in that they appear to utilize two separate type III secretion systems. Recent studies have defined a second type III secretion system in Salmonella that is required for growth within macrophages and virulence in the mouse model [135, 184]. Genes encoding this apparatus are located in a 25-kb region at centisome 30 in Salmonella pathogenicity island 2 (diagrammed in Fig. 7, see color plate). Sequences in SPI2 are conserved throughout the Salmonellae, with the exception of S. bongori [133, 256], but are not found in other enteric bacteria [136]. Thus, it is thought that this pathogenicity island contributes to Salmonella-specific aspects of infection. Like SPIl, SPI2 contains a number of genes predicted to encode a secretion apparatus (designated ssa by [185]), as well as chaperones (ssc), transcriptional regulators (ssr), and putative effector proteins (sse). Expression of genes within SPI2 is regulated by the two-component regulators SsrA (or SpiR) and SsrB. Regulation of these proteins has been difficult to study in vitro because optimal conditions for gene expression have not been determined. However, both apparatus genes and putative secreted effector genes are expressed when bacteria are internalized within cultured macrophages [194, 257]. Expression is dependent on SsrAB and acidification of vacuoles, as bafilomycin treatment prevents expression of SPI2 genes. 2.

SPI2 EFFECTOR

PROTEINS

Six potential secreted effector proteins have been identified in SPI2 [185]. These proteins are minimally expressed in vitro in an SsrAB-dependent manner. Secretion of these proteins has not yet been observed during in vitro growth, and confirmation that the putative effector proteins are secreted and/or translocated awaits further characterization of infected cells. Nevertheless, the Sse proteins do appear important for virulence, as sseA, sseB, and sseC mutants are avirulent, and sseF and sseG mutants are attenuated in the mouse model. It is likely that some


309

of these proteins assemble into a functional translocase associated with the TTSS from SPI2. Several Sse proteins contain homologies to proteins involved in translocation by other bacteria [185]. SseC, for instance, is homologous to YopB from Yersinia and EspD from enteropathogenic E. coli (EPEC), and contains three predicted membrane spanning alpha helices that might enable its insertion into host cell membranes. In addition, predicted transmembrane helices are present in all of the Sse proteins except SseA. SseB and SseD are predicted to encode proteins that are similar to EspA and EspB of EPEC. This is quite interesting, as EspA, B, and D have previously been shown to be required for induction of host cell signaling and translocation of the intimin receptor Tir [258-260]. In addition, EspA has been shown to form extracellular filaments that contact host epithelial cells and might represent a translocation channel [261]. As Salmonella have been shown to form many pili and fimbrial-like filaments that are not SPIl dependent [262], it is possible that one of the filaments is formed by proteins in SPI2, and the Esp-like proteins in particular.

3.

METABOLIC MUTANTS

A number of mutations within metabolic loci have been reported to attenuate Salmonellae in vivo, presumably because of the scarcity of required nutrients in the intracellular environment. Auxotrophic mutants with defects in the biosynthetic pathways of aromatic amino acids {aw) [142], purines (pur) [166], pyrimidines, histidine, and methionine [50], or defects in ethanolamine utilization, through which carbon and nitrogen might be acquired [263], are all attenuated in the mouse model system. In addition, the simultaneous prevention of glutamine synthesis and transport also results in bacterial attenuation [264]. Because of their inability to persist in vivo, auxotrophic strains have been investigated as potential vaccine strains [87, 142, 143, 166, 265-267], as will be discussed later.

4.

SURVIVAL WITHIN MACROPHAGES

Although little is known about the mechanisms utilized by Salmonella to survive within macrophage phagosomes, a few genes that affect this process have been identified. For instance, recombination-deficient mutants, which are hypothesized to be unable to repair DNA damaged by the macrophage oxidative burst, do not survive within macrophages and are avirulent in the mouse model [268]. Another gene product that appears to be required for resistance to oxidative stress in macrophages is the Sly A transcription factor [174], which is induced during stationary phase and required for survival within macrophages [175]. The ability to survive within macrophages may be due in part to resistance to host nitric oxide (NO). The production of nitric oxide is a host defense mechanism associated with broad-spectrum antimicrobial activity, especially important dur-

310


ing intracellular infections. Resistance to nitric oxide in macrophages appears to be mediated at least in part by the metL gene product, which controls the production of homocysteine in the methionine biosynthetic pathway. Salmonella with a mutation in metL are hypersensitive to 5-nitrosothiol NO donor compounds and are attenuated for virulence; this is apparently because homocysteine can act as an endogenous antagonist of NO, thus protecting the bacteria [155].

5.

Vi ANTIGEN

The major surface antigen of 5. typhi, Vi antigen, appears to protect these bacilli from some host innate immune mechanisms [269]. Expression of the Vi antigen correlates with prevention of antibody-mediated opsonization, increased resistance to host peroxide, and resistance to complement activation by the alternate pathway and complement-mediated lysis. Vi antigen thus may function to inhibit phagocytosis of typhoidal Salmonella by neutrophils while not interfering with the induction of phagocytosis by more permissive macrophages and epithelial cells.

D. Salmonella Toxins Several cholera-like toxins [270] and enterotoxins [271, 272] have been described in Salmonella, but none have been isolated or molecularly characterized. Interestingly, a 27-kb pathogenicity island (designated SPI4) that is predicted to encode proteins with homology to type I secretion systems has been identified [139]. Genes within this locus display considerable homology to proteins involved in the secretion of several members of the RTX toxin family, including the Bordetella pertussis CyaA protein, the Serratia marcescens 8000 lipase LipA, and the E. coli hemolysin HlyD. Preliminary evidence indicates that genes within this operon may encode another secretion system, perhaps utilized for secretion of a Salmonella toxin. A transposon insertion within SPI4 prevents secretion of at least one uncharacterized protein and is required for survival within macrophages.

E. Virulence Plasmids A number of Salmonella strains contain large (50-100 kb) virulence plasmids (reviewed in [221]). Although they are present in other host-adapted serotypes such as 5. dublin, S. choleraesuis, S. gallinarum/pullorum, and S. abortusovis.


311

they have not been detected in the human pathogen S. typhi. They are also present in the broad-host range serotypes S. typhimurium and S. enteritidis. In strains that do harbor virulence plasmids, their presence has been associated with full virulence in vivo. Phenotypes attributed to genes present on the virulence plasmid include increased intracellular growth, resistance to complement, and macrophage cytotoxicity [221, 273]. Plasmid-cured strains colonize the Peyer's patch and reticuloendothelial systems with wild-type efficiency; however, subsequent growth in the reticuloendothelial system is impaired, and bacteria are eventually cleared by the host. A single 8-kb region, which contains the Salmonella plasmid-virulence genes {spv), is sufficient to return full virulence to plasmid-cured strains of S. dublin [183]. The spv locus consists of five genes, the first of which, spvR, encodes a transcriptional regulator required for expression of the spvABCD operon. The spv genes are expressed during stationary phase in an RpoS-dependent manner [274]. In addition, expression of spvR has been reported to be autoregulated in a positive manner and negatively regulated by the spvA and spvB gene products [275]. Although one group has reported that expression of a promoter trap IVET {in vitro expression technology) vector-generated gene fusion to spvB in S. typhimurium was regulated by PhoP [276], another group has shown that spv expression in S. dublin is not dependent on the phoP, ompR, or cyalcrp regulatory loci [277]. Analysis of the predicted amino-acid sequence of SpvB revealed homology between the amino-terminal region and the CatM repressor of Acinetobacter calcoaceticus [111]. Mutations in either spvR or spvB result in the attenuated phenotype associated with plasmid loss. Mutations in spvC or spvD are partially virulent, but plasmid maintenance in vivo is impaired. Specific functions of the spv gene products have not yet been elucidated. Although plasmid-cured strains appear wild type in their ability to resist complement-mediated lysis, four loci on the virulence plasmid have been reported to affect complement lysis (reviewed in [278]). These include Rck (resistance to complement killing), an outer membrane protein that prevents polymerization of C9, the pore-forming component of the classical pathway of complement lysis. A surface-exposed loop appears to mediate this ability [279]. Rck is similar to two other virulence proteins, the Ail invasin from Yersinia enterocolitica and the S. typhimurium virulence protein PagC [167, 280]. Like Ail, Rck confers on E. coli the ability to adhere to and invade epithelial cells, but the molecular mechanisms for this are unknown. PagC, Ail, and Rck are members of an interesting family of outer membrane proteins that are present in all Gram-negative bacteria. Family members are present in bacteriophages, plasmids, and chromosomal regions that indicate their acquisition by horizontal transmission. All the phenotypes for these proteins, which include complement resistance, antibiotic resistance, and invasion, have been defined on the basis of multicopy expression. Since multicopy

312


expression of these proteins results in more global effects, including transcription and expression of unlinked genes, their function remains to be defined. Other virulence plasmid proteins that may modulate Salmonella virulence have also been described. A protein homologous to the LuxR family of quorum sensors that controls expression of genes encoded on the S. typhimurium virulence plasmid was discovered in 1998 [281]. SdiA is the first quorum-sensing protein to be identified in Salmonella. SdiA regulates at least one uncharacterized operon on the virulence plasmid, but its contribution to virulence is unknown at this time. Another interesting protein encoded on the virulence plasmid is TlpA, an autoregulatory coiled-coil transcriptional repressor. TlpA is unique in that its structure is modified by temperature, which in turn affects its transcriptional activity. Increased temperature results in a more unfolded (monomeric) state, which decreases repressor activity. It has been proposed that TlpA might act as a thermosensor to help regulate changes in gene expression on entry into hosts [282]. Although this is an intriguing hypothesis, the role of TlpA in host-induced gene expression is unknown at this time.

VllL Antibiotic-Resistant Salmonellae

A. Multi-Antibiotic Resistant Salmonellae The recent emergence of Salmonellae carrying stable resistance to multiple clinically relevant antibiotics is a significant health problem worldwide. Antibiotic resistance of S. typhi has been an issue since 1950, when strains resistant to chloramphenicol were isolated in Great Britain, only 2 years after the successful use of chloramphenicol in treatment of typhoid fever [283]. Currently, S. typhi isolates resistant to six different antimicrobial agents prevail in highly endemic typhoid areas, particularly China, Pakistan, and India. These strains of S. typhi carry a 120-kb plasmid that encodes resistance to ampicillin, chloramphenicol, streptomycin, sulfonamides, tetracycHne, and trimethoprim. In addition, S. typhi strains isolated from recent outbreaks in Tadjikistan and Pakistan have also acquired resistance to ciprofloxacin (a fluoroquinolone), one of the preferred antibiotics for treatment of typhoid fever [16]. As all of the above antibiotics, except streptomycin and tetracycline, are clinically relevant for the oral treatment of typhoid fever, the existence of a hepta-resistant agent for typhoid fever is a serious health problem. In addition, multi-antibiotic-resistant (MAR) typhoid has been a significant cause of death in children; the mortality rates of children

7.


313

infected with MAR S. typhi range from 7 to 16%, compared to a rate of 2% for children infected with susceptible strains of Salmonella [284]. Resistance of nontyphoidal Salmonellae is also a growing health problem. Particularly troubling is the penta-resistant strain of S. typhimurium known as DTI04 (definitive phage type 104), which emerged in Great Britain in 1984 and was reported in 1997 to have been isolated in the United States [23]. This strain has been isolated from numerous species of animals (wild and farm) and is resistant to ampicillin, chloramphenicol, streptomycin, sulfonamides, and tetracycline (R type ACSSuT). In addition, there have been reports of resistance to two other antibiotics, trimethoprim and fluoroquinolones, in Great Britain [285]. The veterinary use of antibiotics, such as ciprofloxacin and trimethoprim, to treat DTI04 infections in cattle has been proposed as a factor in the acquisition of resistance to these (and other) antibiotics by Salmonellae. Interestingly, resistance to ciprofloxacin has not been observed in the United States yet, possibly because fluoroquinolones are only licensed for use in poultry, where DTI04 may not yet be established as a pathogen [24]. Reports documenting the emerging resistance to fluoroquinolones by Salmonellae is worrisome, as ciprofloxacin is the antibiotic of choice for treating human salmonellosis; resistance to this drug will leave few available options. In addition to the problematic treatment of DTI04 infections, there are reports indicating that DTI04 may be more virulent for humans. In a study performed in the United Kingdom, 41% of patients infected with DTI04 were hospitalized, and 3% of culture-confirmed patients died (compared to an average death rate of 0.1%) [286].

B. Development of New Antibiotics The emergence of multidrug resistant Salmonella (and other bacterial pathogens) underlines the necessity for the development of new antibiotics. Salmonella is an ideal organism for the development of new drugs because of the readily available small animal model (mouse), in which antibiotics can be quickly and cheaply tested. In addition, as a number of the molecular mechanisms contributing to pathogenesis have been elucidated, the opportunity exists to develop antibiotics that target specific virulence mechanisms in Salmonella (i.e., pathogenesis-based antimicrobial therapy). As many of these mechanisms are likely to be utilized by other Gram-negative bacteria as well, the development of new drugs targeting Salmonella should also be beneficial for treatment of other pathogens. Such antibiotics might reduce the acquisition of antibiotic resistance while at the same time preserving normal bacterial flora. One active area of research is the development of new cationic antimicrobial peptides. Specific bacterial mechanisms that could be targeted include the PhoP-regulated modifications of the

314


bacterial outer membrane and LPS that mediate resistance to CAMPs and the type III secretion system through which bacterial effector proteins are translocated to the eukaryotic cell cytosol.

IX. Salmonella-Bosed Vaccines

A. Development of More Effective Typhoid Vaccines As new antibiotics are still in the developmental stage, it is also important to continue the effort to produce effective vaccines for typhoid fever. Three vaccines are currently available for typhoid fever, none of which are 100% effective, even when tested on endemic populations [287]. The live attenuated Ty21a strain (manufactured by the Swiss Serum and Vaccine Institute), which is orally administered, requires at least four doses to achieve 51-76% protective efficacy [288]. The heat- and phenol-inactivated typhoid vaccine (manufactured by Wyeth) has similar efficacy, and requires at least two doses (by injection). However, this vaccine has a variety of severe local and systemic side effects that limit its use. Finally, the cell-free, parenteral Vi-antigen vaccine produced by Pasteur Merieux (ViCPS) requires only a single dose, but most likely has a shorter duration of protection. A number of new live strains have been evaluated in human volunteers for their vaccine potential. Live attenuated vaccine strains may be preferable for typhoid fever because they can induce a wide spectrum of protective immunity, including mucosal, humoral, and cell-mediated immunity. Ideal vaccine strains should be genotypically stable (containing at least two nonreverting deletions) and should offer long-term protection after one or two initial doses. Data generated in the S. typhimurium inbred mouse model of typhoid fever are used to identify candidate vaccine strains. Unfortunately, although mutations in single pathways can effectively attenuate S. typhimurium in mice, they are not always adequate in humans. For instance, galE mutants, which are rough, galactose sensitive, and defective in LPS synthesis, are avirulent in mice, but S. typhi galE mutants can still cause typhoid fever [147]. Other more promising vaccine strains include aw deletion mutants [142, 143], crplcya deletion mutants [143, 144] and the phoPlphoQ deletion mutants [163, 164]. aw and crplcya mutants are equally attenuated in mice and humans, and produce vigorous mucosal, humoral, and cellular immune responses on oral inoculation. The PhoP/PhoQ null strain is also highly immunogenic after a single dose in human volunteers. This strain is also promising because it does not persist for long periods of time in vivo, produces very few


315

side effects, and has been shown to be more immunostimulatory compared to wild type [162]. Although there is little requirement for human vaccines for nontyphoidal Salmonellae, as the infections are usually self-limiting, various vaccines for use in animals have been developed. Such vaccines would be quite useful if they could eliminate the source of most Salmonella infections without increasing the potential for additional antibiotic resistance. Both aro-negative [289] and crp/cya mutant [290] S. typhimurium strains have been shown to be effective vaccine strains in chickens. These strains offer long-term protection against infection with both S. typhimurium and 5. enteritidis. In addition, immunization with the crp/cya strain in ovo (up to 7 days before hatching) protected chicks from infection [290], a fact that is particularly important in light of the ability of S. enteritidis to colonize intact shell eggs.

B. Salmonellae as Multivalent Vaccine Strains Induction of mucosal immunity is a hallmark of Salmonella infection, and has driven the development of multivalent Salmonella vaccines that can induce immunity not only to Salmonella but also to heterologous antigens. Antigen delivery to the gut-associated lymphoid tissue (GALT) by Salmonellae results in the eventual secretion of IgA antibodies at a number of sites, including the respiratory system, gastrointestinal tract, genitourinary tract, mammary glands, and salivary glands. The development of effective multivalent vaccine strains is dependent on several factors (reviewed in [291]). First, it is imprudent to introduce antibiotic resistance genes on plasmids carried by the vaccine strain. Therefore, heterologous antigens must either be integrated into the chromosome in single copy or they must be expressed on plasmids that do not utilize selectable antibiotic resistance markers. Expression of antigens from the chromosome is also problematic, as the level of expression may be too low for adequate immunogenicity; therefore, the use of in v/v(9-induced bacterial promoters to maximize expression has been investigated [292, 293]. An alternative approach is to express antigens on balanced lethal plasmids, which are required for bacterial survival in the host but do not require antibiotic selection for maintenance in vivo [294, 295]. Using the techniques described above, a number of studies have shown that immunity to foreign antigens, including tetanus toxin [292, 296], can be attained in mice. An interesting twist on the use of live attenuated Salmonella strains for delivery of heterologous antigens is the use of an aroA strain to transfer eukaryotic expression vectors to host cells [297]. Oral immunization of mice with S. typhimurium strains containing plasmids expressing portions of the Listeria monocytogenes protein listeriolysin resulted in transfer of these plasmids to host cells and subsequent protective immunity to infection with Listeria.

316


Investigators have utilized the type III secretion system encoded in SPIl to induce MHC Class I-restricted T-cell responses to heterologous antigens [298]. Fusion of an amino-terminal domain of the translocated protein SptP to various viral antigens allowed translocation of these antigens to the cytosol of infected cells and induced a protective Class I-restricted CTL response. Mice infected with the vaccine strains were completely protected from infection by murine lymphocytic choriomeningitis virus (LCMV), a normally lethal virus. C. Salmonella-Based Cancer Therapy Salmonella vaccine strains have also been investigated as potential cancer treatment vectors. Interestingly, some attenuated Salmonella strains have been shown to preferentially colonize tumors (rather than liver and spleen) and suppress tumor growth when inoculated into mice [299]. As Salmonella could be engineered to express and translocate (via the type III secretion system) anti-tumor prodrugs, such schemes appear to be worth investigating further.

Acknowledgments We would like to thank former and present members of our laboratory for helpful discussions, in particular C. Lesser, M. Hantman, and J. Gunn for critical reading of this manuscript. We also thank R. Valdivia, D. Holden, A. Zychlinsky and S. Falkow for communicating results prior to publication, and C. Alpuche-Aranda, S.-I. Aizawa, B. Finlay, and M. Jepson for providing photos. Work in our laboratory was supported by grants AI30479 (S.I.M.) and AI09312 (C.A.S.) from the National Institutes of Health.

References 1. Zinder, N. D. and Lederberg, J. (1952). Genetic exchange in Salmonella. J. Bacteriol. 64, 679-699. 2. Reeves, M. W., Evins, G. M., Heiba, A. A., Plikaytis, B. D., and Famer, J. J. I. (1989). Clonal nature of Salmonella typhi and its genetic relatedness to other salmonellae as shown by multilocus enzyme electrophoresis and proposal of Salmonella bongori comb. nov. J. Clin. Microbiol. 27,313-320. 3. WHO (1997). Fact Sheet no. 139: Multi-drug resistant Salmonella typhimuriiim. World Health Organization, http://www.who.ch/. 4. Vaillant, V., Haeghebaert, S., Desenclos, J. C., Bouvet, P., Grimont, F., Grimont, P. A., and Bumens, A. P. (1996). Epidemic d'infections a Salmonella dublin en France, novembre-decembre 1995. Euro Surveillance, http://www.b3e.jussieu.fr/ceses/EuroSurv/n2/en2-121.html.

7.


317

5. Hennessy, T. W., Hedberg, C. W., Slutsker, L., White, K. E., Besser-Wiek, J. M., Moen, M. E., Feldman, J., Coleman, W. W., Edmonson, L. M., MacDonald, K. L., and Osterholm, M. T. (1996), A national outbreak of Salmonella enteritidis infections from ice cream. New Engl. J. Med. 334, 1281-1286. 6. Svitlik, C , et al. (1992). Salmonella hadar associated with pet ducklings—Connecticut, Maryland, and Pennsylvania, 1991. MMWR 41, 185-187. 7. Cook, K. A., Dobbs, T. E., Hlady, W. G., Wells, J. G., Barrett, T. J., Puhr, N. D., Lancette, G. A., Bodager, D. W., Toth, B. L., Genese, C. A., Highsmith, A. K., Pilot, K. E., Finelli, L., and Swerdlow, D. L. (1998). Outbreak oi Salmonella serotype Hartford infections associated with unpasteurized orange juice. JAMA 280, 1504-1509. 8. Mermin, J., Hoar, B., and Angulo, F. J. (1997). Iguanas and Salmonella marina infection in children: A reflection of the increasing incidence of reptile-associated salmonellosis in the United States. Pediatrics 99, 399-402. 9. Kapil, A., Sood, S., Reddaiah, V. P, Das, B., and Seth, P (1997). Paratyphoid fever due to Salmonella enterica serotype paratyphi A [letter]. Emerg. Infect. Dis. 3, 407. 10. Mahon, B. E., Ponka, A., Hall, W. N., Komatsu, K., Dietrich, S. E., Siitonen, A., Cage, G., Hayes, P S., Lambert-Fair, M. A., Bean, N. H., Griffin, P M., and Slutsker, L. (1997). An international outbreak of Salmonella infections caused by alfalfa sprouts grown from contaminated seeds. J. Infect. Dis. 175, 876-882. 11. Pang, T., Levine, M. M., Ivanoff, B., Wain, J., and Finlay, B. B. (1998). Typhoid fever: Important issues still remain. Trends Microbiol. 6, 131-133. 12. CDC (1990). Epidemiologic notes and reports: Typhoid fever—Skagit County, Washington. MMW^/? 39, 749-751. 13. Hosek, G., Leschinksy, D., Irons, S., and Safranek, T. J. (1997). Multidrug-resistant Salmonella serotype typhimurium—United States, 1996. MMWR 46, 308. 14. WHO (1998). Typhoid fever. World Health Organization, http://www.who.ch/. 15. Mirza, S. H., Beeching, N. J., and Hart, C. A. (1996). Multi-drug resistant typhoid: A global problem. J. Med. Microbiol. 44, 317-319. 16. Hampton, M. D., Ward, L. R., Rowe, B., and Threlfal, E. J. (1998). Molecular fingerprinting of multidrug-resistant Salmonella enterica serotype Typhi. Emerg. Infect. Dis. 4, 317-320. 17. Altekruse, S. F, Cohen, M. L., and Swerdlow, D. L. (1997). Emerging foodbome diseases. Emerg. Infect. Dis. 3, 285-293. 18. Tauxe, R. V. (1997). Emerging foodbome diseases: An evolving public health challenge. Emerg. Infect. Dis. 3, 425-434. 19. Aserkoff, B., and Bennett, J. V. (1969). Effect of antibiotic therapy in acute salmonellosis on the fecal excretion of salmonellae. New Engl. J. Med. 281, 636-640. 20. CDC (1998). Salmonellosis: Frequently asked questions. Centers for Disease Control and Prevention, http://www.cdc.gov/ncidod/diseases/foodbom/salmon.htm. 21. Pang, T., Bhutta, Z. A., Finlay, B. B., and Altwegg, M. (1995). Typhoid fever and other salmonellosis: A continuing challenge. Trends Microbiol. 3, 253-255. 22. USDA (1998). Salmonella enteritidis risk assessment: Shell eggs and egg products. United States Department of Agriculture, Food Safety and Inspection Service, http://www.fsis.usda.gov/ ophs/risk/contents.htm. 23. CDC (1997). Multidmg-resistant Salmonella serotype Typhimurium—United States, 1996. MMW/? 46, 308-310. 24. Glynn, M. K., Bopp, C , Dewitt, W., Dabney, P, Mokhtar, M., and Angulo, F J. (1998). Emergence of multidrug-resistant Salmonella enterica serotype typhimurium DTI04 infections in the United States. New Engl. J. Med. 338, 1333-1338.

318


25. Ryan, C. A., Nickels, M. K., Hargrett-Bean, N. T., Potter, M. E., Endo, T., Mayer, L., Langkop, C. W., Gibson, C , McDonald, R. C , Kenney, R. T., et al. (1987). Massive outbreak of antimicrobial-resistant salmonellosis traced to pasteurized milk. JAMA 58, 3269-3274. 26. Sperber, S. J., and Schleupner, C. J. (1987). Salmonellosis during infection with the human immunodeficiency virus. Rev. Infect. Dis. 9, 925-934. 27. Pegues, D. A., and Miller, S. I. (2000). Salmonella including Salmonella typhi. In "Principles and Practice of Infectious Diseases" (G. L. Mandell, J. E. Bennett, and R. Dolin, eds.). Vol. 2, pp. 2344-2363. Churchill Livingstone, New York. 28. WHO (1997). Fact sheet no. 149: Typhoid fever. World Health Organization, http://www .who.ch/. 29. Kaye, D., Merselis Jr., J. G., Connolly, S., and Hook, E. W. (1967). Treatment of chronic enteric carriers oiSalmonella typhosa with ampicillin. Ann. N.Y. Acad. Set. 145, 429^35. 30. Goldberg, M. B., and Rubin, R. H. (1988). The spectrum oi Salmonella infection. Infect. Dis. Clin. N.Amer. 2,571-598. 31. Leavitt, J. W (1996). "Typhoid Mary: Captive to the Public's Health." Beacon, Boston. 32. Blaser, M. J., and Newman, L. S. (1982). A review of human salmonellosis, I: Infective dose. Rev. Infect. Dis. 4, 1096-1106. 33. Mintz, E. D., Cartter, M. L., Hadler, J. L., Wassell, J. T, Zingeser, J. A., and Tauxe, R. V. (1994). Dose-response effects in an outbreak of Salmonella enteritidis. Epidemiol. Infect. 112, 13-23. 34. (1971). A waterbome epidemic of salmonellosis in Riverside, California, 1965: Epidemiologic aspects. A collaborative report. Am. J. Epidemiol. 93, 3 3 ^ 8 . 35. Giannella, R. A., Broitman, S. A., and Zamcheck, N. (1972). Gastric acid barrier to ingested microorganisms in man: Studies in vivo and in vitro. Gut 13, 251-256. 36. Waddell, W. R., and Kunz, L. J. (1956). Association of Salmonella enteritis with operation of stomach. New Engl. J. Med. 255, 555-559. 37. Carter, P. B., and Collins, F. M, (1974). The route of enteric infection in normal mice. / Exp. Med. 139, 1189-1203. 38. Hohmann, A. W., Schmidt, G., and Rowley, D. (1978). Intestinal colonization and virulence of Salmonella in mice. Infect. Immun. 22, 163-110. 39. Kohbata, S., Yokobata, H., and Yabuuchi, E. (1986). Cytopathogenic effect of Salmonella typhi GIFU 10007 on M cells of murine ileal Peyer's patches in ligated ileal loops: An ultrastructural study. Microbiol. Immunol. 30, 1225-1237. 40. Brandtzaeg, P. (1989). Overview of the mucosal system. Curr Topics Microbiol. Immunol. 146, 13-25. 41. Neutra, M. R., Pringault, E., and Kraehenbuhl, J.-P. (1996). Antigen sampling across epithelial barriers and induction of mucosal immune responses. Ann. Rev. Immunol. 14, 275-300. 42. Clark, M. A., Jepson, M. A., Simmons, N. L., and Hirst, B. H. (1994). Preferential interaction of Salmonella typhimurium with mouse Peyer's patch M cells. Res. Microbiol. 145, 543-552. 43. Jones, B. D., Ghori, N., and Falkow, S. (1994). Salmonella typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer's Patches. J. Exp. Med. 180, 15-23. 44. Owen, R. L. (1994). M cells: Entry ways of opportunity for enteropathogens. J. Exp. Med. 180, 7-9. 45. Penheiter, K. L., Mathur, N., Giles, D., Fahlen, T, and Jones, B. D. (1997). Non-invasive Salmonella typhimurium mutants are avirulent because of an inability to enter and destroy M cells of ileal Peyer's Patches. Mol. Microbiol. 24, 697-709.

7.


319

46. Watson, P. R., Paulin, S. M., Bland, A. P, Jones, P W., and Wallis, T. S. (1995). Characterization of intestinal invasion by Salmonella typhimurium and Salmonella dublin and the effect of a mutation in the invH gene. Infect. Immun. 63, 2743-2754. 47. Takeuchi, A. (1967). Electron microscopic studies of experimental Salmonella infection I. Penetration into the intestinal epithelium by Salmonella typhimurium. Am. J. Pathol. 50, 109-136. 48. Behlau, I., and Miller, S. I. (1993). A PhoP-repressed gene promotes Salmonella typhimurium invasion of epithelial cells. J. Bacteriol. 175, 4475^M84. 49. Galan, J. E., and Curtiss III, R. (1989). Cloning and molecular characterization of genes whose products allow Salmonella typhimurium to penetrate tissue culture cells. Proc. Natl. Acad. Sci. U.S.A. 86, 6383-6387. 50. Fields, P I., Swanson, R. V., Haidaris, C. G., and Heffron, F. (1986). Mutants of Salmonella typhimurium that cannot survive within the macrophage are avirulent. Proc. Natl. Acad. Sci. U.S.A. S3, 5\S9-5\93. 51. Richter-Dahlfors, A., Buchan, A. M. J., and Finlay, B. B. (1997). Murine Salmonellosis studied by confocal microscopy: Salmonella typhimurium resides intracellularly inside macrophages and exerts a cytotoxic effect on phagocytes in vivo. J. Exp. Med. 186, 569-580. 52. Vidal, S. M., Malo, D., and Vogan, K. (1993). Natural resistance to infection with intracellular parasites: Isolation of a candidate for beg gene. Cell 73, 469-485. 53. Lissner, C. R., Swanson, R., and O'Brien, A. (1983). Genetic control of the innate resistance of mice to Salmonella typhimurium: Expression of the Ity gene in peritoneal and splenic macrophages isolated in vitro. J. Immunol. 131, 3006-3013. 54. Vidal, S., Tremblay, M. L., Govoni, G., Gauthier, S., Sebastiani, G., Malo, D., Skamene, E., Olivier, M., Jothy, S., and Gros, P. (1995). The ItylLshlBcg locus: Natural resistance to infection with intracellular parasites is abrogated by disruption of the Nrampl gene. / Exp. Med. 182, 655-666. 55. Vidal, S. M., Pinner, E., Lepage, P., Gauthier, S., and Gros, P. (1996). Natural resistance to intracellular infections: Nrampl encodes a membrane phosphoglycoprotein absent in macrophages from susceptible (NrampP'^^) mouse strains. J. Immunol. 157, 3559-3568. 56. Gruenheid, S., Pinner, E., Desjardins, M., and Gros, P. (1997). Natural resistance to infection with intracellular pathogens: The Nrampl protein is recruited to the membrane of the phagosome. J. Exp. Med. 185, 717-730. 57. O'Brien, A. D., Scher, I., and Metcalf, E. S. (1981). Genetically conferred defect in anti-Salmonella antibody formation renders CBA/N mice innately susceptible to Salmonella typhimurium infection. J. Immunol. 126, 1368-1372. 58. Hormaeche, C. E., Harrington, K. A., and Joysey, H. S. (1985). Natural resistance to Salmonellae in mice: Control by genes within the major histocompatibility complex. J. Infect. Dis. 152, 1050-1056. 59. Maskell, D. J., and Hormaeche, C. E. (1986). Genes within the major histocompatibility complex influence the response to ampicillin therapy and severity of relapse in H-2 congenic, susceptible /ry' mice infected with virulent Salmonella typhimurium. J. Immunogen. 13, 451-^57. 60. O'Brien, A. D., Rosenstreich, D. L., Scher, I., Campbell, G. H., MacDermott, R. P., and Formal, S. B. (1980). Genetic control of susceptibility to Salmonella typhimurium in mice: Role of the LPS gene. J. Immunol. 124, 20-24. 61. Qureshi, S. T, Lariviere, L., Sebastiani, G., Clermont, S., Skamene, E., Gros, P., and Malo, D. (1996). A high-resolution map in the chromosomal region surrounding the Lps locus. Genomics 31, 283-294.

320


62. Stultzer, B. M., Catagna, R., Bandekar, J., and Wong, P. (1993). Lipopolysaccharide nonresponder cells: The C3H/HeJ defect. Immunobiology 187, 257-271. 63. Sebastian!, G., Olien, L., Gauthier, S., Skamene, E., Morgan, K., Gros, P., and Malo, D. (1998). Mapping of genetic modulators of natural resistance to infection with Salmonella typhimurium in wild-derived mice. Genomics 47, 180-186. 64. Hess, J., Ladel, C , Midko, D., and Kaufmann, S. H. (1996). Salmonella typhimurium aroAinfection in gene-targeted immunodeficient mice: Major role of CD4^ TCR-alpha beta cells and IFN-gamma in bacterial clearance independent of intracellular location. J. Immunol. 156, 3321-3326. 65. Kincy-Cain, T., Clements, J. D., and Bost, K. L. (1996). Endogenous and exogenous interleukin12 augment the protective immune response in mice orally challenged with Salmonella dublin. Infect. Immun. 64, 1437-1440. 66. Mastroeni, P., Harrison, J. A., Chabalgoity, J. A., and Hormaeche, C. E. (1996). Effect of Interleukin 12 neutralization on host resistance and gamma interferon production in mouse typhoid. Infect. Immun. 64, 189-196. 67. de Jong, R., Altare, F, Haagen, I.-A., Elferink, D. G., de Boer, T., van Breda Vriesman, P. J. C, Kabel, P. J., Draaisma, J. M. T., van Dissel, J. T, Kroon, F. P, Casanova, J.-L., and Ottenhoff, T. H. M. (1998). Severe mycobacterial and Salmonella infections in Interleukin-12 receptor-deficient patients. Science 280, 1435-1438. 68. Wyckoff, T. J. O., Raetz, C. R. H., and Jackman, J. E. (1998). Antibacterial and anti-inflammatory agents that target endotoxin. Trends Microbiol. 6, 154-159. 69. Wright, S. D., Ramos, R. A., Tobias, P S., Ulevitch, R. J., and Mathison, J. C. (1990). CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPD binding protein. Science 249, 1431-1433. 70. Schumann, R. R., Leong, S. R., Flaggs, G. W., Gray, P W., Wright, S. D., Mathison, J. C, Tobias, P. S., and Ulevitch, R. J. (1990). Structure and function of lipopolysaccharide binding protein. Science 249, 1429-1431. 71. Yang, R. B., Mark, M. R., Gray, A., Huang, A., Xie, M. H., Zhang, M., Goddard, A., Wood, W. I., Gumey, A. L., and Godowski, P. J. (1998). Toll-like receptor-2 mediates lipopolysaccharide-induced cellular signalling. Nature 395, 284-288. 72. Jack, R. S., Fan, X., Bernheiden, M., Rune, G., Ehlers, M., Weber, A., Kirsh, G., Mentel, R., Furll, B., Freudenberg, M., Schmitz, G., Stelter, F, and Schutt, C. (1997). Lipopolysaccharidebinding protein is required to combat a murine Gram-negative bacterial infection. Nature 389, 742-745. 73. Wurfel, M. M., Monks, B. G., Ingalls, R. R., Dedrick, R. L., Delude, R., Zhou, D., Lamping, N., Schumann, R. R., Thieringer, R., Fenton, M. J., Wright, S. D., and Golenbock, D. (1997). Targeted deletion of the lipopolysaccharide (LPS)-binding protein gene leads to profound suppression of LPS responses ex vivo, whereas in vivo responses remain intact. J. Exp. Med. 186,2051-2056. 74. Raetz, C. R. (1993). Bacterial endotoxins: Extraordinary lipids that activate eucaryotic signal tranduction. J. Bacteriol. 175, 5745-5753. 75. Janeway, C. A., and Travers, P. (1997). "Immunobiology: The Immune System in Health and Disease," 3rd ed. Garland, New York and London. 76. Ciacci-Woolwine, F, Blomfield, J. C , Richardson, S. H., and Mizel, S. B. (1998). Salmonella flagellin induces tumor necrosis factor alpha in a human promonocytic cell line. Infect. Immun. 66, 1127-1134. 77. Ganz, T, and Weiss, J. (1997). Antimicrobial peptides of phagocytes and epithelia. Sem. Hematol. 34, 343-354.

.


321

78. Fields, P. I., Groisman, E. A., and Heffron, F. (1989). A Salmonella locus that controls resistance to microbicidal proteins from phagocytic cells. Science 243, 1059-1062. 79. Miller, S. I., Pulkkinen, W. S., Selsted, M. E., and Mekalanos, J. J. (1990). Characterization of defensin resistance phenotypes associated with mutations in the phoP virulence regulon of Salmonella typhimurium. Infect. Immun. 58, 3706-3710. 80. Groisman, E. A., Parra-Lopez, C , Salcedo, M., Lipps, C. J., and Heffron, F. (1992). Resistance to host antimicrobial peptides is necessary for Salmonella virulence. Proc. Natl. Acad. ScL U.S.A. 89, 11939-11943. 81. Martin, E., Ganz, T., and Lehrer, R. I. (1995). Defensins and other endogenous peptide antibiotics of vertebrates. J. Leukoc. Biol. 58, 128-136. 82. Mastroeni, P., Villarreal-Ramos, B., and Hormaeche, C. E. (1993). Adoptive transfer of immunity to oral challenge with virulent Salmonellae in innately susceptible BALB/c mice requires both immune serum and T cells. Infect. Immun. 61, 3981-3984. 83. Nauciel, C. (1990). Role of CD4"^ T cells and T-independent mechanisms in acquired resistance to Salmonella typhimurium infection. J. Immunol. 145, 1265-1269. 84. Cookson, B. T., and Bevan, M. J. (1997). Identification of a natural T-cell epitope presented by Salmonella-'mi^ciQd macrophages and recognized by T cells from orally immunized mice. J. Immunol. 158,4310^319. 85. Casadevall, A. (1998). Antibody-mediated protection against intracellular pathogens. Trends Microbiol. 6, 102-107. 86. Michetti, P, Mahan, M. J., Slauch, J. M., Mekalanos, J. J., and Neutra, M. R. (1992). Monoclonal secretory immunoglobulin A protects mice against oral challenge with the invasive pathogen Salmonella typhimurium. Infect. Immun. 60, 1786-1792. 87. Harrison, J. A., Villarreal-Ramos, B., Mastroeni, P., Demarco de Hormaeche, R., and Hormaeche, C. E. (1997). Correlates of protection induced by live Aro-Salmonella typhimurium vaccines in the murine typhoid model. Immunology 90, 618-625. 88. Hormaeche, C. E., Mastroeni, P., Harrison, J. A., Demarco de Hormaeche, R., Svenson, S., and Stocker, B. A. D. (1996). Protection against oral challenge three months after i.v. immunization of BALB/c mice with live Aro Salmonella typhimurium and Salmonella enteritidis vaccines is serotype (species)-dependent and only partially determined by the main LPS O antigen. Vaccine 14,251-259. 89. Buchmeier, N. A., and Heffron, F. (1989). Intracellular survival of wild-type Salmonella typhimurium and macrophage-sensitive mutants in diverse populations of macrophages. Infect. Immun. 51, 1-7. 90. Alpuche-Aranda, C. M., Berthiaume, E. P, Mock, B., Swanson, J. A., and Miller, S. I. (1995). Spacious phagosome formation within mouse macrophages correlates with Salmonella serotype pathogenicity and host susceptibility. Infect. Immun. 63, 4456-4462. 91. Ishibashi, Y, and Arai, T. (1996). A possible mechanism for host-specific pathogenesis of Salmonella serovars. Microb. Pathogen. 21, 435-446. 92. Alpuche-Aranda, C. M., Racoosin, E. L., Swanson, J. A., and Miller, S. I. (1994). Salmonella stimulates macrophage macropinocytosis and persists within spacious phagosomes. J. Exp. Med. 179, 601-608. 93. Theriot, J. A. (1995). The cell biology of infection by intracellular bacterial pathogens. Ann. Rev. Cell Dev. Biol. 154, 213-239. 94. Sinai, A. P., and Joiner, K. A. (1997). Safe haven: The cell biology of nonfusogenic pathogen vacuoles. Ann. Rev. Microbiol. 51, 415-462. 95. Steinberg, T. H., and Swansson, J. A. (1994). Measurement of phagosome-lysosome fusion and phagosomal pH. Methods Enzymol. 236, 147-160.

322


96. Alpuche-Aranda, C. M., Swanson, J. A., Loomis, W. P., and Miller, S. I. (1992). Salmonella typhimurium activates virulence gene transcription within acidified macrophage phagosomes. Pwc. Natl Acad. ScL U.S.A. 89, 10079-10083. 97. Rathman, M., Sjaastad, M. D., and Falkow, S. (1996). Acidification of phagosomes containing Salmonella typhimurium in murine macrophages. Infect. Immun. 64, 2765-2773. 98. Ernst, R. K., Dombroski, D. M., and Merrik, J. M. (1990). Anaerobiosis, type 1 fimbriae, and growth phase are factors that affect invasion of HEp-2 cells by Salmonella typhimurium. Infect. Immun. 58, 2014-2016. 99. Lee, C. A., and Falkow, S. (1990). The ability of Salmonella to enter mammalian cells is affected by bacterial growth state. Proc. Natl. Acad. Sci. U.S.A. 87, 4304-4308. 100. Bajaj, V., Lucas, R. L., Hwang, C , and Lee, C. A. (1996). Co-ordinate regulation of Salmonella typhimurium invasion genes by environmental and regulatory factors is mediated by control of hilA expression. Mol. Microbiol. 22, 703-714. 101. Buchmeier, N. A., and Heffron, F. (1991). Salmonella inhibits phagolysosomal fusion. Infect. Immun. 59, 2232-2238. 102. Ishibashi, Y., and Arai, T. (1990). Specific inhibition of phagosome-lysosome fusion in murine macrophages mediated by Salmonella typhimurium infection. FEMS Microbiol. Immunol. 64,

35^4. 103. Rathman, M., Barker, L. R, and Falkow, S. (1997). The unique trafficking pattern of Salmonella typhimurium-coniaining phagosomes in murine macrophages is independent of the mechanism of bacterial entry. Infect. Immun. 65, 1475-1485. 104. Oh, Y.-K., Alpuche-Aranda, C. M., Berthiaume, E., Jinks, T, Miller, S. I., and Swansson, J. A. (1996). Rapid and complete fusion of macrophage lysosomes with phagosomes containing Salmonella typhimurium. Infect. Immun. 64, 3877-3883. 105. Chen, L. M., Kaniga, K., and Galan, J. E. (1996). Salmonella spp. are cytotoxic for cultured macrophages. Mol. Microbiol. 21, 1101-1115. 106. Monack, D. M., Raupach, B., Hromockyj, A. E., and Falkow, S. (1996). Salmonella typhimurium invasion induces apoptosis in infected macrophages. Proc. Natl. Acad. Sci. U.S.A. 93, 9833-9838. 107. Lindgren, S. W., Stojilkovic, L, and Heffron, F. (1996). Macrophage killing is an essential virulence mechanism of Salmonella typhimurium. Proc. Natl. Acad. Sci. U.S.A. 93, 4197-4201. 108. Francis, C. L., Ryan, T. A., Jones, B. D., Smith, S. J., and Falkow, S. (1993). Ruffles induced by Salmonella and other stimuli direct macropinocytosis of bacteria. Nature 364, 639-642. 109. Ginnochio, C , Pace, J., and Galan, J. E. (1992). Identification and molecular characterization of a Salmonella typhimurium gene involved in triggering the internalization of Salmonellae into cultured epithelial cells. Proc. Natl. Acad Sci. U.S.A. 89, 5976-5980. 110. Jones, B. D., Paterson, H. F, Hall, A., and Falkow, S. (1993). Salmonella typhimurium induces membrane ruffling by a growth factor-receptor-independent mechanism. Proc. Nad. Acad. Sci. U.S.A. 90, 10390-10394. 111. Finlay, B. B., Ruschkowski, S., and Dedhar, S. (1991). Cytoskeletal rearrangements accompanying Salmonella entry into epithelial cells. / Cell Sci. 99, 283-296. 112. Finlay, B. B., and Falkow, S. (1988). Comparison of the invasion strategies used by Salmonella cholerae-suis. Shigella flexneri and Yersinia enterocolitica to enter cultured animal cells: Endosome acidification is not required for bacterial invasion or intracellular replication. Biochimie 70, 1089-1099. 113. Galan, J. E., Pace, J., and Hayman, M. J. (1992). Involvement of the epidermal growth factor receptor in the invasion of cultured mammalian cells by Salmonella typhimurium. Nature 357, 588-589.

7.


323

114. Garcia-del Portillo, F., and Finlay, B. (1994). Salmonella invasion of nonphagocytic cells induces formation of macropinosomes in the host cell. Infect. Immun. 62, 4641^645. 115. Leung, K. Y., and Finlay, B. B. (1991). Intracellular replication is essential for the virulence of Salmonella typhimurium. Proc. Natl Acad. Sci. U.S.A. 88, 11470-11474. 116. Garcia-del Portillo, F., Foster, J. W., Maguire, M. E., and Finlay, B. B. (1992). Characterization of the micro-environment of Salmonella typhimurium-contSLinmg vacuoles within MDCK epithelial cells. Mol. Microbiol. 6, 3289-3297. 117. Garcia-del Portillo, F, and Finlay, B. B. (1995). Targeting of Salmonella typhimurium to vesicles containing lysosomal membrane glycoproteins bypasses compartments with mannose 6-phosphate receptors. J. Cell Biol. 129, 81-97. 118. Garcia-del Portillo, F, Zwick, M. B., Leung, K. Y, and Finlay, B. B. (1993). Salmonella induces the formation of filamentous structures containing lysosomal membrane glycoproteins in epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 90, 10544-10548. 119. Ireton, K., Payrastre, B., Chap, H., Ogawa, W., Sakaue, H., Kasuga, M., and Cossart, P (1996). A role for phosphoinositide 3-kinase in bacterial invasion. Science 274, 780-782. 120. Rosenshine, L, Ruschkowski, S., Foubister, V., and Finlay, B. B. (1994). Salmonella typhimurium invasion of epithelial cells: Role of induced host cell protein tyrosine phosphorylation. Infect. Immun. 62, 4969-4974. 121. Zigmond, S. H. (1996). Signal transduction and actin filament organization. Curr Opin. Cell Biol. 8, 66-73. 122. Adam, T., Giry, M., Boquet, P., and Sansonetti, P. (1996). Rho-dependent membrane folding causes Shigella entry into epithelial cells. EMBO J. 15, 3315-3321. 123. Chen, L.-M., Hobbie, S., and Galan, J. E. (1996). Requirement of CDC42 for Salmonella-induced cytoskeletal and nuclear responses. Science 274, 2115-2118. 124. Hardt, W.-D., Chen, L.-M., Schuebel, K. E., Bustelo, X. R., and Galan, J. E. (1998). S. typhimurium encodes an activator of Rho GTPases that induces membrane ruffling and nuclear responses in host cells. Cell 93, 815-826. 125. Pace, J., Hayman, M. J., and Galan, J. E. (1993). Signal transduction and invasion of epithelial cells by S. typhimurium. Cell 72, 505-514. 126. Hobbie, S., Chen, L. M., Davis, R. J., and Galan, J. E. (1997). Involvement of mitogen-activated protein kinase pathways in the nuclear responses and cytokine production induced by Salmonella typhimurium in cultured intestinal epithelial cells. J. Immunol. 159, 5550-5559. 127. Weinstein, D. L., O'Neill, B. L., and Metcalf, E. S. (1997). Salmonella typhi stimulation of human intestinal epithelial cells induces secretion of epithelial cell-derived interleukin-6. Infect. Immun. 65, 395^04. 128. McCormick, B. A., Colgan, S. P, Delp-Archer, C , Miller, S. I., and Madara, J. L. (1993). Salmonella typhimurium attachment to human intestinal epithelial monolayers: Transcellular signalling to subepithelial neutrophils. J. Cell Biol. 123, 895-907. 129. Jepson, M. A., Collares-Buzato, C. B., Clark, M. A., Hirst, B. H., and Simmons, N. L. (1995). Rapid disruption of epithelial barrier function by Salmonella typhimurium is associated with structural modification of intercellular junctions. Infect. Immun. 63, 356-359. 130. McCormick, B. A., Parkos, C. A., Colgan, S. P, Games, D. K., and Madara, J. L. (1998). Apical secretion of a pathogen-elicited epithelial chemoattractant activity in response to surface colonization of intestinal epithelia by Salmonella typhimurium. J. Immunol. 160, 455^66. 131. McCormick, B. A., Miller, S. I., Delp-Archer, C , and Madara, J. T. (1995). Transepithelial signaling to neutrophils by Salmonellae: A novel virulence mechanism for gastroenteritis. Infect. Immun. 63, 2302-2309.

324


132. Bowe, F, Lipps, C. J., Tsolis, R. M., Groisman, E., Heffron, R, and Kusters, J. G. (1998). At least four percent of the Salmonella typhimuhum genome is required for fatal infection of mice. Infect. Immun. 66, 3372-3377. 133. Ochman, H., and Groisman, E. A. (1996). Distribution of pathogenicity islands in Salmonella spp. Infect. Immun. 64, 5410-5412. 134. Mills, D. M., Bajaj, V., and Lee, C. A, (1995). A 40 kilobase chromosomal fragment encoding Salmonella typhimuhum invasion genes is absent from the corresponding region of the Escherichia coli K-12 chromosome. Mol Microbiol. 15, 749-759. 135. Ochman, H., Soncini, F. C , Solomon, P., and Groisman, E. A. (1996). Identification of a pathogenicity island required for Salmonella survival in host cells. Proc. Natl. Acad. Sci. U.S.A. 93, 7800-7804. 136. Shea, J. E., Hensel, M., Gleeson, C., and Holden, D. W. (1996). Identification of a virulence locus encoding a second type III secretion system in Salmonella typhimurium. Proc. Natl. Acad. Sci. U.S.A. 93, 2593-2597. 137. Blanc-Potard, A. B., and Groisman, E. A. (1997). The Salmonella selC locus contains a pathogenicity island mediating intramacrophage survival. EMBO J. 16, 5376-5385. 138. Wood, M. W., Jones, M. A., Watson, P. R., Hedges, S., Wallis, T. S., and Galyov, E. E. (1998). Identification of a pathogenicity island required for Salmonella enteropathogenicity. Mol. Microbiol. 29,883-891. 139. Wong, K. K., McClelland, M., Stillwell, L. C , Sisk, E. C, Thurston, S. J., and Saffer, J. D. (1998). Identification and sequence analysis of a 27-kilobase chromosomal fragment containing a Salmonella pathogenicity island located at 92 minutes on the chromosome map of Salmonella enterica serovar typhimurium LT2. Infect. Immun. 66, 3365-3371. 140. Sukopolvi, S., Lorenz, R. G., Gordon, J. I., Bian, Z., Pfeifer, J. D., Normark, S., and Rhen, M. (1997). Expression of thin aggregative fimbriae promotes interaction of Salmonella typhimurium SR-11 with mouse small intestinal epithelial cells. Infect. Immun. 65, 5320-5325. 141. van der Velden, A. W. M., Baumler, A. J., Tsolis, R. M., and Heffron, R (1998). Multiple fimbrial adhesins are required for full virulence of Salmonella typhimurium in mice. Infect. Immun. 66, 2803-2808. 142. Hoiseth, S. K., and Stocker, B. A. D. (1981). Aromatic-dependent Salmonella typhimurium are non-virulent and effective as live vaccines. Nature 291, 238-239. 143. Tacket, C. O., Hone, D. M., Curtiss III, R., Kelly, S. M., Losonsky, G., Guers, L., Harris, A. M., Edelman, R., and Levine, M. M. (1992). Comparison of the safety and immunogenicity of AaroC AaroD and Acya Acrp Salmonella typhi strains in adult volunteers. Infect. Immun. 60, 536-541. 144. Curtiss III, R., and Kelly, S. M. (1987). Salmonella typhimurium deletion mutants lacking adenylate cyclase and cyclic AMP receptor protein are avirulent and immunogenic. Infect. Immun. 55, 3035-3043. 145. Baumler, A. J., Tsolis, R. M., and Heffron, R (1996). Contribution of fimbrial operons to attachment to and invasion of epithelial cell lines by Salmonella typhimurium. Infect. Immun. 64, 1862-1865. 146. Germanier, R., and Purer, E. (1971). Immunity in experimental salmonellosis, II: Basis for the avirulence and protective capacity of galE mutants of Salmonella typhimurium. Infect. Immun. 4, 663-673. 147. Hone, D. M., Attridge, S. R., Porrest, B., Morona, R., Daniels, D., LaBrooy, J. T, Bartholomeusz, R. C , Shearman, D. J., and Hackett, J. (1988). AgalE via (Vi antigen-negative) mutant of Salmonella typhi Ty2 retains virulence in humans. Infect. Immun. 56, 1326-1333. 148. Bajaj, V., Hwang, C , and Lee, C. A. (1995). hilA is a novel ompR/toxR family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715-727.

7.


325

149. Kaniga, K., Bossio, J. C , and Galan, J. E. (1994). The Salmonella typhimurium invasion genes invF and invG encode homologues of the AraC and PulD family of proteins. Mol Microbiol. 13, 555-568. 150. CoUazo, C. M., Zierler, M. K., and Galan, J. E. (1995). Functional analysis of the Salmonella typhimurium invasion genes invl and invJ and identification of a target of the protein secretion apparatus encoded in the inv locus. Mol. Microbiol. 15, 25-38. 151. Galan, J. E., Ginocchio, C., and Costeas, P. (1992). Molecular and functional characterization of the Salmonella invasion gene invA: Homology of Inv A to members of a new protein family. J. Bacteriol. 174, 4338^349. 152. Eichelberg, K., Ginocchio, C. C., and Galan, J. E. (1994). Molecular and functional characterization of the Salmonella typhimurium invasion genes invB and invC: Homology of InvC to the FyoFy, ATPase family of proteins. J. Bacteriol. 176, 4501-4510. 153. Baumler, A. J., Tsolis, R. M., and Heffron, F. (1996). The //?/fimbrial operon mediates adhesion oiSalmonella typhimurium to murine Peyer's patches. Proc. Natl. Acad. Sci. U.S.A. 93, 279-283. 154. Lockman, H. A., and Curtiss III, R. (1992). Virulence of non-type 1-fimbriated and nonfimbriated nonflagellated Salmonella typhimurium mutants in murine typhoid fever. Infect. Immun. 60, 491-^96. 155. De Groote, M. A., Testerman, T, Xu, Y., Stauffer, G., and Fang, F. C. (1996). Homocysteine antagonism of nitric oxide-related cytostasis in Salmonella typhimurium. Science 272, A\A-A\1. 156. Moncrief, M. B. C., and Maguire, M. E. (1998). Magnesium and the role of mgtC in growth of Salmonella typhimurium. Infect. Immun. 66, 3802-3809. 157. Dorman, C. J., Chatfield, S., Higgins, C. F., Hay ward, C., and Dougan, G. (1989). Characterization of porin and ompR mutants of a virulent strain of Salmonella typhimurium: ompR mutants are attenuated in vivo. Infect. Immun. 57, 2136-2140. 158. Baumler, A. J., Tsolis, R. M., Bowe, F A., Kusters, J. G., Hoffmann, S., and Heffron, F. (1996). The p^/fimbrial operon of Salmonella typhimurium mediates adhesion to murine small intestine and is necessary for fluid accumulation in the infant mouse model. Infect. Immun. 64, 61-68. 159. Miller, S. I., Kukral, A. M., and Mekalanos, J. J. (1989). A two-component regulatory system {phoP phoQ) controls Salmonella typhimurium virulence. Proc. Natl. Acad. Sci. U.S.A. 86, 5054-5058. 160. Guo, L., Lim, K., Gunn, J. S., Bainbridge, B., Darveau, R., Hackett, M., and Miller, S. I. (1997). Regulation of lipid A modifications by Salmonella typhimurium virulence genes phoP-phoQ. Science 216, 250-253. 161. Groisman, E, A., Chiao, E., Lipps, C. J., and Heffron, F. (1989). Salmonella typhimurium phoP virulence gene is a transcriptional regulator. Proc. Natl. Acad. Sci. U.S.A. 86, 7077-7081. 162. Wick, M. J., Harding, C. V., Twesten, N. J., Normark, S. J., and Pfeifer, J. D. (1995). The phoP locus influences processing and presentation of Salmonella typhimurium antigens by activated macrophages. Mol. Microbiol. 16, 465-476. 163. Hohmann, E. L., Oletta, C. A., and Miller, S. I. (1995). Evaluation of phoP/phoQ deleted awA-deleted live oral S. typhi vaccine in humans. Vaccine 14, 19-24. 164. Hohmann, E. L., Oletta, C. A., Killeen, K. P, and Miller, S. I. (1996). phoP/phoQ-ddetcd Salmonella typhi (TY800) is a safe and immunogenic single dose typhoid fever vaccine in volunteers. J. Infect. Dis. 173, 1408-1414. 165. Makela, P. H., Sarvas, M., Calcagno, S., and Lounatmaa, K. (1978). Isolation and characterization of polymyxin-resistant mutants of Salmonella. FEMS Microbiol. 3, 323-326. 166. McFarland, W. C., and Stocker, B. A. (1987). Effect of different purine auxotrophic mutations on mouse virulence of a Vi-positive strain of Salmonella dublin and of two strains of Salmonella typhimurium. Microb. Pathogen. 3, 129-141.

326


167. Heffeman, E. J., Harwood, J., Fierer, J., and Guiney, D. (1992). The Salmonella typhimurium virulence plasmid complement resistance gene rck is homologous to a family of virulence-related outer membrane protein genes, including pagC and ail. J. Bacterial. 174, 84-91. 168. Nickerson, C. A., and Curtiss III, R. (1997). Role of sigma factor RpoS in initial stages of Salmonella typhimurium infection. Infect. Immun. 65, 1814-1823. 169. Wilmes-Riesenberg, M. R., Foster, J. W., and Curtiss III, R. (1997). An altered rpoS allele contributes to the avirulence oi Salmonella typhimurium LT2. Infect. Immun. 65, 203-210. 170. Robbe-Saule, V., Coynault, C , and Norel, F. (1995). The live oral typhoid vaccine Ty21a is a rpoS mutant and is susceptible to various environmental stresses. FEMS Microbiol. Lett. 126, 171-176. 171. Hueck, C. J., Hantman, M. J., Bajaj, V., Johnston, C , Lee, C. A., and Miller, S. I. (1995). Salmonella typhimurium secreted invasion determinants are homologous to Shigella Ipa proteins. Mol Microbiol. 18, 479^90. 172. Kaniga, K., Tucker, S., Trollinger, D., and Galan, J. E. (1995). Homologs of the Shigella IpaB and IpaC invasins are required for Salmonella typhimurium entry into cultured epithelial cells. J. Bacteriol. Ill, 3965-3971. 173. Collazo, C. M., and Galan, J. E. (1997). The invasion-associated type III system oi Salmonella typhimurium directs the translocation of Sip proteins into the host cell. Mol. Microbiol. lA, 747-756. 174. Buchmeier, N., Bossie, S., Chen, C.-Y, Fang, R C , Guiney, D. G., and Libby, S. J. (1997). SlyA, a transcriptional regulator of Salmonella typhimurium, is required for resistance to oxidative stress and is expressed in the intracellular environment of macrophages. Infect. Immun. 65, 3725-3730. 175. Libby, S. J., Goebel, W., Ludwig, A., Buchmeier, N., Bowe, F., Fang, F. C , Guiney, D. G., Songer, J. G., and Heffron, F. (1994). A cytolysin encoded by Salmonella is required for survival within macrophages. Proc. Natl. Acad. Sci. U.S.A. 91, 489^93. 176. Norris, K A., Wilson, M. R, Wallis, T S., Galyov, E. E., and Majerus, R W. (1998). SopB, a protein required for virulence of Salmonella dublin, is an inositol phosphate phosphatase. Proc. Natl. Acad. Sci. U.S.A. 95, 14057-14059. 177. Galyov, E. E., Wood, M. W., Rosqvist, R., Mullan, R B., Watson, R R., Hedges, S., and Wallis, T. S. (1997). A secreted effector protein of Salmonella dublin is translocated into eukaryotic cells and mediates inflammation and fluid secretion in infected ileal mucosa. Mol. Microbiol. 25,903-912. 178. Jones, M. A., Wood, M. W, Mullan, R B., Watson, R R., Wallis, T S., and Galyov, E. E. (1998). Secreted effector proteins of Salmonella dublin act in concert to induce enteritis. Infect. Immun. 66, 5799-5804. 179. Wood, M. W, Rosqvist, R., Mullan, R B., Edwards, M. H., and Galyov, E. E. (1996). SopE, a secreted protein of Salmonella dublin, is translocated into the target eukaryotic cell via an Sip-dependent mechanism and promotes bacterial entry. Mol. Microbiol. 22, 327-338. 180. Groisman, E. A., and Ochman, H. (1993). Cognate gene clusters govern invasion of host epithelial cells by Salmonella typhimurium and Shigella flexneri. EMBO J. 12, 3779-3787. 181. Hardt, W.-D., Urlaub, H., and Galan, J. E. (1998). A substrate of the centisome 63 type III protein secretion system of Salmonella typhimurium is encoded by a cryptic bacteriophage. Proc. Natl. Acad Sci. U.S.A. 95, 2574-2579. 182. Kaniga, K., Uraill, J., Bliska, J. B., and Galan, J. E. (1996). A secreted protein tyrosine phosphatase with modular effector domains in the bacterial pathogen Salmonella typhimurium. Mol. Microbiol. 21, 633-641.

7.


327

183. Gulig, P. A., Caldwell, L., and Chiodo, V. A. (1992). Identification, genetic analysis, and DNA sequence of a 7.8-kilobase virulence region of the Salmonella typhimurium virulence plasmid. Mol Microbiol 6, 1395-1411. 184. Hensel, M., Shea, J. E., Gleeson, C , Jones, M. D., Dalton, E., and Holden, D. W. (1995). Simultaneous identification of bacterial virulence genes by negative selection. Science 269, 400-403. 185. Hensel, M., Shea, J. E., Waterman, S. R., Mundy, R., Nikolaus, T., Banks, G., Vazquez-Torres, A., Gleeson, C., Fang, F. C., and Holden, D. W. (1998). Genes encoding putative effector proteins of the type III secretion system of Salmonella pathogenicity island 2 are required for bacterial virulence and proliferation in macrophages. Mol. Microbiol. 30, 163-174. 186. Stone, B. J., and Miller, V. L. (1995). Salmonella enteritidis has a homologue of tolC that is required for virulence in BALB/c mice. Mol Microbiol 17, 701-712. 187. Ronson, C. W., Nixon, B. T., and Ausubel, F M. (1987). Conserved domains in bacterial regulatory proteins that respond to environmental stimuli. Cell 49, 579-581. 188. Gunn, J. S., Hohmann, E. L., and Miller, S. I. (1996). Transcriptional regulation oi Salmonella virulence: A PhoQ periplasmic domain mutation results in increased net phosphotransfer to PhoP. J. Bacteriol 178, 6369-6373. 189. Kier, L. K., Weppleman, R. M., and Ames, B. N. (1979). Regulation of nonspecific acid phosphatase in Salmonella: phoN and phoP genes. J. Bacteriol 138, 155-161. 190. Gunn, J. S., and Miller, S. I. (1996). PhoP/PhoQ activates transcription of pmrA/B, encoding a two-component system involved in Salmonella typhimurium antimicrobial peptide resistance. J. Bacteriol 178, 6857-6864. 191. Soncini, F C , and Groisman, E. A. (1996). Two-component regulatory systems can interact to process multiple environmental signals. J. Bacteriol 178, 6796-6801. 192. Pegues, D. A., Hantman, M. J., Behlau, I., and Miller, S. I. (1995). PhoP/PhoQ transcriptional repression of Salmonella typhimurium invasion genes: Evidence for a role in protein secretion. Mol Microbiol 17, 169-181. 193. Soncini, F C , Garcia-Vescovi, E., and Groisman, E. A. (1995). Transcriptional autoregulation of the Salmonella typhimurium phoPQ operon. J. Bacteriol 177, 4364-4371. 194. Valdivia, R. H., and Falkow, S. (1997). Fluorescence-based isolation of bacterial genes expressed within host cells. Science 277, 2007-2011. 195. Bearson, B. L., Wilson, L., and Foster, J. W. (1998). A low pH-inducible, PhoPQ-dependent acid tolerance response protects Salmonella typhimurium against inorganic acid stress. J. Bacteriol 180, 2409-2417. 196. Garcia-Vescovi, E., Soncini, F C , and Groisman, E. A. (1996). Mg^^ as an extracellular signal: Environmental regulation of Salmonella virulence. Cell 84, 165-174. 197. Waldburger, C. D., and Sauer, R. T. (1996). Signal detection by the PhoQ sensor-transmitter: Characterization of the sensor domain and a response-impaired mutant that identifies ligandbinding determinants. J. Biol Chem. 211, 26630-26636. 198. Garcia-Vescovi, E., Ayala, Y. M., Di Ceras, E., and Groisman, E. A. (1997). Characterization of the bacterial sensor protein PhoQ: Evidence for distinct binding sites for Mg^^ and Ca^^. J. Biol Chem. 272, 1440-1443. 199. Soncini, F C , Garcia Vescovi, E., Solomon, F, and Groisman, E. A. (1996). Molecular basis of the magnesium deprivation response in Salmonella typhimurium: Identification of PhoP-regulated genes. J. Bacteriol 178, 5092-5099. 200. Miller, S. I., and Mekalanos, J. J. (1990). Constitutive expression of the PhoP regulon attenuates Salmonella virulence and survival within macrophages. J. Bacteriol 172, 2485-2490.

328


201. Roland, K. L., Martin, L. E., Esther, C. R., and Spitznagel, J. K. (1993). Spontaneous pmrA mutants of Salmonella typhimurium LT2 define a new two-component regulatory system with a possible role in virulence. J. Bacteriol. 175, 4154-4164. 202. Groisman, E. A., Kayser, J., and Soncini, F. C. (1997). Regulation of polymyxin resistance and adaptation to low-Mg^^ environments. J. Bacteriol. 179, 7040-7045. 203. Gunn, J. S., Lim, K. B., Krueger, J., Kim, K., Guo, L., Hackett, M., and Miller, S. I. (1998). PmrA-PmrB-regulated genes necessary for 4-aminoarabinose lipid A modification and polymyxin resistance. Mol. Microbiol. 27, 1171-1182. 204. Valdivia, R. H., and Falkow, S. (1996). Bacterial genetics by flow cytometry: Rapid isolation of Salmonella typhimurium acid-inducible promoters by differential fluorescence induction. Mol. Microbiol. 22, 367-378. 205. Guo, L., Lim, K. B., Poduje, C. M., Daniel, M., Gunn, J. S., Hackett, M., and Miller, S. I. (1998). Lipid A acylation and bacterial resistance against vertebrate antimicrobial peptides. Cell 95, 189-198. 206. Parra-Lopez, C , Baer, M. T., and Groisman, E. A. (1993). Molecular genetic analysis of a locus required for resistance to antimicrobial peptides in Salmonella typhmurium. EMBO J. 12, 4053-^062. 207. Parra-Lopez, C., Lin, R., Aspedon, A., and Groisman, E. A. (1994). A Salmonella protein that is required for resistance to antimicrobial peptides and transport of potassium. EMBO J. 13, 3964-3972. 208. Makela, P H., Hovi, M., Saxen, H., Valtonen, M., and Valtonen, V. (1988). Salmonella, complement and mouse macrophages. Immunol. Lett. 19, 217-222. 209. Eisenstein, T. K., Meissler, J. J. J., Miller, S. I., and Stocker, B. A. D. (1998). Immunosuppression and nitric oxide production induced by parenteral live Salmonella vaccines do not correlate with protective capacity: AphoPwTnXO mutant does not suppress but does protect. Vaccine 16, 24-32. 210. Tao, T., Grulich, P F, Kucharski, L. M., Smith, R. L., and Maguire, M. E. (1998). Magnesium transport in Salmonella typhimurium: Biphasic magnesium and time dependence of the transcription of the mgtA and mgtCB loci. Microbiology 144, 655-664. 211. Smith, R. L., and Maguire, M. E. (1998). Microbial magnesium transport: Unusual transporters searching for identity. Mol. Microbiol. 28, 217-226. 212. Hohmann, E. L., and Miller, S. L (1994). The Salmonella PhoP virulence regulon. In "Phosphate in Microorganisms: Cellular and Molecular Biology" (A. Torriani-Gorini, E. Yagil, and S. Silver, eds.), pp. 120-125. American Society for Microbiology, Washington, DC. 213. Gunn, J. S., Belden, W. J., and Miller, S. I. (1998). Identification of PhoP-PhoQ activated genes within a duplicated region of the Salmonella typhimurium chromosome. Microb. Pathogen. 25, 77-90. 214. Galan, J. E., and Curtiss III, R. (1990). Expression of Salmonella typhimurium genes required for invasion is regulated by changes in DNA supercoiling. Infect. Immun. 58, 1879-1885. 215. Jones, B. D., and Falkow, S. (1994). Identification and characterization of a Salmonella typhimurium oxygen-regulated gene required for bacterial internalization. Infect. Immun. 62, 3745-3752. 216. Lee, C. A., Jones, B. D., and Falkow, S. (1992). Identification of a Salmonella typhimurium invasion locus by selection for hyperinvasive mutants. Proc. Natl. Acad. Sci. U.S.A. 89, 1847-1851. 217. Eichelberg, K., Kaniga, K., and Galan, J. (1996). Transcriptional regulation of SalmoneIla-sQcreted virulence determinants. Gen. Mtg. Am. Soc. Microbiol., 96th, American Society for Microbiology, New Orleans, Louisiana.

7.


329

218. Johnston, C , Pegues, D. A., Hueck, C. J., Lee, C. A., and Miller, S. I. (1996). Transcriptional activation of Salmonella typhimurium invasion genes by a member of the phosphorylated response-regulator superfamily. Mol. Microbiol. 11, 715-727. 219. Rakeman, J., Bonifield, H., and Miller, S. I. (1998). Unpublished obrservations. 220. Loewen, P. C , and Hengge-Aronis, R. (1994). The role of the sigma factor a' (KatF) in bacterial global regulation. Annu. Rev. Microbiol. 48, 53-80. 221. Guiney, D. G., Fang, F. C, Krause, M., Libby, S., Buchmeier, N. A., and Fierer, J. (1995). Biology and clinical significance of virulence plasmids in Salmonella serovars. Clin. Infect. Dis. 21, S146-S15I. 222. Bearson, S., Bearson, B., and Foster, J. W. (1997). Acid stress responses in enterobacteria. FEMS Microbiol. Lett. 147, 173-180. 223. Bearson, S. M., Benjamin Jr., W. H., Swords, W. E., and Foster, J. W. (1996). Acid shock induction of RpoS is mediated by the mouse virulence gene mviA of Salmonella typhimurium. J. Bacteriol. 178, 2572-2579. 224. Lee, I. S., Lin, J., Hall, H. K., Bearson, B., and Foster, J. W. (1995). The stationary-phase sigma factor a^ (RpoS) is required for a sustained acid tolerance response in virulent Salmonella typhimurium. Mol. Microbiol. 17, 155-167. 225. Foster, J. W., and Spector, M. P. (1995). How Salmonella survive against the odds. Annu. Rev. Microbiol. 49, 145-174. 226. Hall, H. K., and Foster, J. W. (1996). The role of Fur in the acid tolerance response of Salmonella typhimurium is physiologically and genetically separable from its role in iron acquisition. /. Bacteriol. 178,5683-5691. 227. Garcia-del Portillo, F., Foster, J. W., and Finlay, B. B. (1993). Role of acid tolerance response genes in Salmonella typhimurium virulence. Infect. Immun. 61, 4489-4492. 228. Wilmes-Riesenberg, M. R., Bearson, B., Foster, J. W., and Curtiss, R. I. (1996). Role of the acid tolerance response in virulence oi Salmonella typhimurium. Infect. Immun. 64, 1085-1092. 229. Chatfield, S. N., Dorman, C. J., Hay ward, C., and Dougan, G. (1991). Role of om/7/?-dependent genes in Salmonella typhimurium virulence: Mutants deficient in both OmpC and OmpF are attenuated in vivo. Infect. Immun. 59, 449-^52. 230. Mills, S. D., Ruschkowski, S. R., Stein, M. A., and Finlay, B. B. (1998). Trafficking of porin-deficient Salmonella typhimurium mutants inside HeLa cells: ompR and envZ mutants are defective for the formation oi Salmonella-\x\(^\xctdi filaments. Infect. Immun. 66, 1806-1811. 231. Collinson, S. K., Clouthier, S. C., Doran, J. L., Banser, R A., and Kay, W. W. (1996). Salmonella enteritidis agfBAC operon encoding thin, aggregative fimbriae. J. Bacteriol. 178, 662-667. 232. Romling, U., Bian, Z., Hammar, M., Sierralta, W. D., and Normark, S. (1998). Curii fibers are highly conserved between Salmonella typhimurium and Escherichia coli with respect to operon structure and regulation. J. Bacteriol. 180, 722-731. 233. Woodward, M. J., Allen-Vercoe, E., and Redstone, J. S. (1996). Distribution, gene sequence and expression in vivo of the plasmid encoded fimbrial antigen of Salmonella serotype Enteritidis. Epidemiol. Infect. Ill, 17-28. 234. Olsen, A., Jonsson, A., and Normark, S. (1989). Fibronectin binding mediated by a novel class of surface organelles on Escherichia coli. Nature 338, 652-655. 235. Giannasca, K. T., Giannasca, P. J., and Neutra, M. R. (1996). Adherence of Salmonella typhimurium to Caco-2 cells: Identification of a glycoconjugate receptor. Infect. Immun. 64, 135-145. 236. Hueck, C. J. (1998). Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Biol. Rev 62, 379^33.

330


237. Rosqvist, R., Magnusson, K. E., and Wolf-Watz, H. (1994). Target cell contact triggers expression and polarized transfer of Yersinia YopE cytotoxin into mammalian cells. EMBO J. 13, 964-972. 238. Li, J., Ochman, H., Groisman, E. A., Boyd, E. P., Solomon, P., Nelson, K., and Selander, R. K. (1995). Relationship between evolutionary rate and cellular location among the Inv/Spa invasion proteins of Salmonella enterica. Proc. Natl. Acad. Sci. U.S.A. 92, 7252-7256. 239. Pugsley, A. P. (1993). The complete general secretory pathway in gram-negative bacteria. Microbiol. Rev. 57, 50-108. 240. Crago, A. M., and Koronakis, V. (1998). Salmonella InvG forms a ring-like multimer that requires the InvH lipoprotein for outer membrane localization. Mol. Microbiol. 30, 47-56. 241. Daefler, S., and Russel, M. (1998). The Salmonella typhimurium InvH protein is an outer membrane lipoprotein required for the proper localization of InvG. Mol. Microbiol. 28, 1367-1380. 242. Ginocchio, C. C., Olmsted, S. B., Wells, C. L., and Galan, J. E. (1994). Contact with epithelial cells induces the formation of surface appendages on Salmonella typhimurium. Cell 76, 717-724. 243. Kubori, T., Matsushima, Y., Nakamura, D., Uralil, J., Lara-Tejero, M., Sukhan, A., Galan, J. E., and Aizawa, S.-I. (1998). Supramolecular structure of the Salmonella typhimurium Type III protein secretion system. Science 280, 602-605. 244. Hermant, D., Menard, R., Arricau, N., Parsot, C , and Popoff, M. Y. (1995). Functional conservation of the Salmonella and Shigella effectors of entry into epithelial cells. Mol. Microbiol 17, 781-789. 245. Fu, Y, and Galan, J. E. (1998). The Salmonella typhimurium tyrosine phosphatase SptP is translocated into host cells and disrupts the actin cytoskeleton. Mol. Microbiol. 27, 359-368. 246. Menard, R., Sansonetti, P., Parsot, C , and Vasselon, T. (1994). Extracellular association and cytoplasmic partitioning of the IpaB and IpaC invasins of S.flexneri. Cell 79, 515-525. 247. Menard, R., Prevost, M.-C, Gounon, P., Sansonetti, P, and Dehio, C. (1996). The secreted Ipa complex of Shigella flexneri promotes entry into mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 93, 1254-1258. 248. Hakansson, S., Schesser, S. K., Persson, C, Galyov, E. E., Rosqvist, R., Homble, P., and Wolf-Watz, H. (1996). The YopB protein of Yersinia pseudotuberculosis is essential for the translocation of Yop effector proteins across the target cell plasma membrane and displays a contact-dependent membrane disrupting activity. EMBO J. 15, 5812-5823. 249. Zychlinsky, A., Kenny, B., Menard, R., Prevost, M. C , Holland, I. B., and Sansonetti, P J. (1994). IpaB mediates macrophage apoptosis induced by Shigella flexneri. Mol. Microbiol. 11, 619-627. 250. Chen, Y, Smith, M. R., Thirumalai, K., and Zychlinsky, A. (1996). A bacterial invasin induces macrophage apoptosis by binding directly to ICE. EMBO J. 15, 3853-3860. 251. Hersh, D., Monack, D. M., Smith, M. R., Ghori, N., Falkow, S., and Zychlinsky, A. (1999). Proc. Natl. Acad Sci. U.S.A. 96, 2396-2401. 252. Hardt, W.-D., and Galan, J. E. (1997). A secreted Salmonella protein with homology to an avirulence determinant of plant pathogenic bacteria. Proc. Natl. Acad. Sci. U.S.A. 94, 98879892. 253. Zhou, D., and Galan, J. E. (1998). S. typhimurium encodes an actin-binding protein that modulates the bundling activity of plastin to promote bacterial entry into host cells. Ann. Mtg. Am. Soc. Cell Biol, 38th, American Society for Cell Biology, San Francisco. 254. Hong, K. H., and Miller, V. L. (1998). Identification of a novel Salmonella invasion locus homologous to Shigella ipgDE. J. Bacteriol 180, 1793-1802.

7.


331

255. Eckmann, L., Rudolf, M. T., Ptasznik, A., Schultz, C , Jiang, T, Wolfson, N., Tsien, R., Fierer, J., Shears, S. B., Kagnoff, M. R, and Traynor-Kaplan, A. E. (1997). D-myo-inositol 1,4,5,6tetrakisphosphate produced in human intestinal epithelial cells in response to Salmonella invasion inhibits phosphoinositide 3-kinase signaling pathways. Proc. Natl. Acad. Sci. U.S.A. 94, 14456-14460. 256. Hensel, M., Shea, J. E., Baumler, A. J., Gleeson, C , Blattner, P., and Holden, D. W. (1997). Analysis of the boundaries of Salmonella pathogenicity island 2 and the corresponding chromosomal region of Escherichia coli K-12. J. Bacteriol. 179, 1105-1111. 257. Cirillo, D. M., Valdivia, R. H., Monack, D., and Falkow, S. (1998). Macrophage-dependent induction of the Salmonella pathogenicity island 2 type III secretion system and its role in intracellular survival. Molec. Microbiol. 30, 175-188. 258. Donnenberg, M. S., Tzipori, S., McKee, M. L., O'Brien, A. D., Alroy, J., and Kaper, J. B. (1993). The role of the eae gene of enterohemorrhagic Escherichia coli in intimate attachment in vitro and in a porcine model. J. Clin. Invest. 92, 1418-1424. 259. Kenny, B., Lai, L. C , Finlay, B. B., and Donnenberg, M. S. (1996). EspA, a protein secreted by enteropathogenic Escherichia coli, is required to induce signals in epithelial cells. Mol. Microbiol. 20,313-323. 260. Kenny, B., DeVinney, R., Stein, M., Reinscheid, D. J., Frey, E. A., and Finlay, B. B. (1997). Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells. Ce//91, 511-520. 261. Knutton, S., Rosenshine, I., Pallen, M. J., Nisan, I., Neves, B. C , Bain, C , Wolff, C , Dougan, G., and Frankel, G. (1998). A novel EspA-associated surface organelle of enteropathogenic Escherichia coli involved in protein translocation into epithelial cells. EMBO J. 17, 2166-2176. 262. Reed, K. A., Clark, M. A., Booth, T. A., Hueck, C. J., Miller, S. I., Hirst, B. H., and Jepson, M. A. (1998). Cell-contact-stimulated formation of filamentous appendages by Salmonella typhimurium does not depend on the type III secretion system encoded by Salmonella pathogenicity island 1. Infect. Immun. 66, 2007-2017. 263. Stojiljkovic, I., Baumler, A. J., and Heffron, F. (1995). Ethanolamine utilization in Salmonella typhimurium: Nucleotide sequence, protein expression, and mutational analysis of the cchA cchB eutE eutJ eutG eutH gene cluster. J. Bacteriol. 177, 1357-1366. 264. Klose, K. E., and Mekalanos, J. J. (1997). Simultaneous prevention of glutamine synthesis and high-affinity transport attenuates Salmonella typhimurium virulence. Infect. Immun. 65, 587596. 265. Edwards, M. F, and Stocker, B. A. D. (1988). Construction of MroA his Apur strains of Salmonella typhi. J. Bacteriol. 170, 3991-3995. 266. Smith, B. R, Reina-Guerra, M., Hoiseth, S. K., Stocker, B. A. D., Habasha, R, Johnson, E., and Merritt, F. (1984). Aromatic-dependent Salmonella typhimurium as modified live vaccines for calves. Am. J. Vet. Res. 45, 59-66. 267. Tackett, C. O., Sztein, M. B., Losonsky, G. A., Wasserman, S. S., Nataro, J. P, Edelman, R., Pickard, D., Dougan, G., Chatfield, S. N., and Levine, M. M. (1997). Safety of live oral Salmonella typhi vaccine strains with deletions in htrA and aroC aroD and immune response in humans. Infect. Immun. 65, 452-456. 268. Buchmeier, N. A., Lipps, C. J., So, M. Y. H., and Heffron, F (1993). Recombination-deficient mutants of Salmonella typhimurium are avirulent and sensitive to the oxidative burst of macrophages. Mol. Microbiol. 7, 933-936. 269. Looney, R. J., and Steigbigel, R. T. (1986). Role of the Vi antigen of Salmonella typhi in resistance to host defense in vitro. J. Lab. Clin. Med. 108, 506-516. 270. Peterson, J. W. (1986). Salmonella toxins. In "Pharmacology of Bacterial Toxins" (F. Domer and J. Drews, eds.), pp. 227-234. Pergamon, New York.

332


271. Malik, P., Sharma, V. D., and Thapliyal, D. C. (1996). Partial purification and characterization of Salmonella cytotoxin. Vet. Microbiol. 49, 11-19. 272. Rumeu, M. T., Suarez, M. A., Morales, S., and Rotger, R. (1997). Enterotoxin and cytotoxin production by Salmonella enteritidis strains isolated from gastroenteritis outbreaks. J. Appl. Microbiol. 82, 19-31. 273. Guilloteau, L. A., Wallis, T. S., Gautier, A. V., Maclntyre, S., Piatt, D. J., and Lax, A. J. (1996). The Salmonella virulence plasmid enhances Salmonella-induced lysis of macrophages and influences inflammatory responses. Infect. Immun. 64, 3385-3393. 274. Fang, F. C , Libby, S. J., Buchmeier, N. A., Loewen, P. C , Switala, J., Harwood, J., and Guiney, D. G. (1992). The alternative sigma factor katF (rpoS) regulates Salmonella virulence. Proc. Natl. Acad. Sci. U.S.A. 89, 11978-11982. 275. Abe, A., Matsui, H., Danbara, H., Tanaka, K., Takahashi, H., and Kawahara, K. (1994). Regulation of spvR gene expression of Salmonella virulence plasmid pKDSC50 in Salmonella cholerae-suis serovar Choleraesuis. Mol. Microbiol. 12, 119-1^1. 276. Heithoff, D. M., Conner, C. R, Hanna, P C , Julio, S. M., Hentschel, U., and Mahan, M. J. (1997). Bacterial infection as assessed by in vivo gene expression. Proc. Natl. Acad. Sci. U.S.A. 94, 934-939. 277. Fang, F. C , Krause, M., Roudier, C , Fierer, J., and Guiney, D. G. (1991). Growth regulation of a Salmonella plasmid gene essential for virulence. J. Bacteriol. 173, 6783-6789. 278. Gulig, P A., Danbara, H., Guiney, D. G., Lax, A. J., Norel, F, and Rhen, M. (1993). Molecular analysis of spv virulence genes of the Salmonella virulence plasmids. Mol. Microbiol. 7, 825-830. 279. Cirillo, D. M., Heffeman, E. J., Wu, L., Harwood, J., Fierer, J., and Guiney, D. G. (1996). Identification of a domain in Rck, a product of the Salmonella typhimiirium virulence plasmid, required for both serum resistance and cell invasion. Infect. Immun. 64, 2019-2023. 280. Pulkkinen, W. S., and Miller, S. L (1991). A Salmonella typhimiirium virulence protein is similar to a Yersinia enterocolitica invasion protein and a bacteriophage lambda outer membrane protein. J. Bacteriol. 173, 86-93. 281. Ahmer, B. M. M., van Reeuwijk, J., Timmers, C. D., Valentine, P J., and Heffron, F. (1998). Salmonella typhimurium encodes an SdiA homolog, a putative quorum sensor of the LuxR family, that regulates genes on the virulence plasmid. J. Bacteriol. 180, 1185-1193. 282. Hurme, R., Bemdt, K. D., Normark, S. J., and Rhen, M. (1997). A proteinaceous gene regulatory thermometer in Salmonella. Cell 90, 55-64. 283. Colquhoun, J., and Weetch, R. S. (1950). Resistance to chloramphenicol developing during treatment of typhoid fever. Lancet 2, 621-623. 284. Gupta, A. (1994). Multidrug-resistant typhoid fever in children: Epidemiology and therapeutic approach. Pediatn Infect. Dis. J. 13, 134-140. 285. Threlfall, E. J., Frost, J. A., Ward, L. R., and Rowe, B. (1996). Increasing spectrum of resistance in multiresistant Salmonella typhimurium [letter]. Lancet 347, 1053-1054. 286. Wall, P G., Morgan, D., Lamden, K., Ryan, M., Griffin, M., Threlfall, E. J., Ward, L. R., and Rowe, B. (1994). A case control study of infection with an epidemic strain of multiresistant Salmonella typhimurium DTI04 in England and Wales. Commun. Dis. Rep. CDR Rev. 4, R130-R135. 287. CDC (1998). Preventing typhoid fever: A guide for travelers. Centers for Disease Control, http://www.cdc.gov/ncidod/diseases/bacter/typhoid.htm. 288. Levine, M. M., Black, R. E., Ferreccio, C , Clements, M, L., Lanata, C , Rooney, J., Germanier, R., and Chilean Typhoid Committee (1987). Field trials of the efficacy of attenuated Salmonella typhi oral vaccine strain Ty21a. In "Proceedings of the International Symposium on Bacterial Vaccines" (J. Robbins, ed.). Praeger, New York.

7.


333

289. Hassan, J. O., and Curtiss III, R. (1997). Efficacy of a live avirulent Salmonella typhimurium vaccine in preventing colonization and invasion of laying hens by Salmonella typhimurium and Salmonella enteritidis. Avian Dis. 41, 783-791. 290. Coloe, P. J., Alderton, M. R., Gerraty, N. L., Christopher, W., and Smith, S. C. (1995). Aromatic vitamin-dependent Salmonellae as vaccines in food animals: Efficacy and persistence. Dev. Biol Stand. 84, 263-267. 291. Chatfield, S., Roberts, M., Li, J., Stams, A., and Dougan, G. (1994). The use of live attenuated Salmonella for oral vaccination. Dev. Biol. Stand. 82, 35-42. 292. Chatfield, S. N., Charles, I. G., Makoff, A. J., Oxer, M. D., Dougan, G., Pickard, D., Slater, D., and Fairweather, N. F. (1992). Use of the nirB promoter to direct the stable expression of heterologous antigens in Salmonella oral vaccine strains: Development of a single-dose oral tetanus vaccine. Biotechnology (N.Y.) 10, 888-892. 293. Hohmann, E. L., Oletta, C. A., Loomis, W. P, and Miller, S. I. (1995). Macrophage-inducible expression of a model antigen in Salmonella typhimurium enhances immunogenicity. Proc. Natl. Acad. Sci. U.S.A. 92, 2904-2908. 294. Nakayama, K., Kelly, S. M., and Curtiss III, R. (1988). Construction of an Asd+ expressioncloning vector: Stable maintenance and high level expression of cloned genes in a Salmonella vaccine strain. Bio/Techology 6, 693-697. 295. Schodel, F, Kelly, S. M., Peterson, D. L., Milich, D. R., and Curtiss III, R. (1994). Hybrid hepatitis B virus core-pre-S proteins synthesized in avirulent Salmonella typhimurium and Salmonella typhi for oral vaccination. Infect. Immun. 62, 1669-1676. 296. VanCott, J. L., Staats, H. F, Pascual, D. W., Roberts, M., Chatfield, S. N., Yamamoto, M., Coste, M., Carter, P. B., Kiyono, H., and McGhee, J. R. (1996). Regulation of mucosal and systemic antibody responses by T helper cell subsets, macrophages, and derived cytokines following oral immunization with live recombinant Salmonella. J. Immunol. 156, 1505-1514. 297. Darji, A., Guzman, C , Gerstel, B., Wachholz, P, Timmis, D. M., Wehland, J., Chakraborty, T., and Weiss, S. (1997). Oral somatic transgene vaccination using attenuated S. typhimurium. Cell 91, 765-775. 298. Russmann, H., Shams, H., Poblete, F, Fu, Y., Galan, J. E., and Donis, R. O. (1998). Delivery of epitopes by the Salmonella type III secretion system for vaccine development. Science 281, 565-568. 299. Pawelek, J. M., Low, K. B., and Bermudes. D. (1997). Tumor-targeted Salmonella as a novel anticancer vector. Cancer Res. 51, 4537-4544.


CHAPTER 8

Shigellosis: From Disease Symptoms to Molecular and Cellular Pathogenesis PHILIPPE J. SANSONETTI COUMARAN EGILE CHRISTINE WENNERAS

I. II. III. IV. V.

VI. VII. VIII. IX.

X.

Introduction Bacteriology The Somatic Antigen Epidemiology and Transmission Disease Symptoms and Complications: Orientations for Future Research? A. A Spectrum of Disease Symptoms from Diarrhea to Dysentery B. Acute Complications of Shigellosis C. Long-Term Complications of Shigellosis Histopathology of Shigellosis: A Window on Pathogenesis Animal Models: Strengths and Weaknesses Cellular Models of Infection: The Contribution of Shigella to the Concept of Cellular Microbiology Pathogenic Mechanisms: In Vitro Expression of the Invasive Phenotype A. Molecular and Cellular Mechanisms of Shigella Invasion of Epithelial Cells: Basic Principles and Reviews B. Molecular and Cellular Biology of the Entry Process into Epithelial Cells C. Escape of Shigella into the Cell Cytoplasm D. Intracellular Motility and Cell-to-Cell Spreading of Shigella E. Actin-Based Intracellular Motility of Shigella E Apoptotic Killing of Macrophages and Induction of the Release of Mature IL-1 p by Shigella G. Activation of PMNs Also Reflects the Shigella Invasive Phenotype Pathogenic Mechanisms: In Vivo Expression of the Invasive Phenotype A. Does Shigella Express Organ-Specific Adhesins?


'^'^^

336 337 338 338 340 340 341 342 342 343 344 345 345 346 355 355 356 361 364 365 366

336

PHILLIPPE J. SANSONETTI, COUMARAN EGILE, AND PHILLIPPE JENNERAS

B. Crossing the Epithelial Lining by Shigella: Which Route is Best? C. Intestinal Inflammation during Human Infection XI. Role of Chromosomally Encoded Genes in the Virulence of Shigella A. Regulation of Plasmid Virulence Genes B. Lipopolysaccharide C. Toxins D. Other Virulence Factors XII. Conclusions References

366 368 369 369 370 371 372 373 373

/. Introduction Enteric bacterial pathogens may be divided into two pathovars: (1) The noninvasive pathovar, exemplified by Vibrio cholerae and enterotoxigenic Escherichia coli (ETEC), cause disease by adhering to the apical side of the small intestinal epithelium and by secreting enterotoxins, thereby causing massive water and electrolyte secretion (i.e., watery diarrhea). Likewise, enteropathogenic E. coli (EPEC) and enterohemorrhagic E. coli (EHEC) colonize and partially destroy the epithelium. EHEC may even cause bloody diarrhea due to the production of Shiga-like toxins. However, here again, in spite of significant epithelial alterations (i.e., attaching-effacing effect on epithelial villi), these bacteria are not significantly invasive and the syndrome they cause is still dominated by watery diarrhea [1]. (2) The invasive pathovar is exemplified by microorganisms such as Shigella and Salmonella, it is characterized by invasion of the intestinal mucosa. In the case of Salmonella typhi infection, the intestinal phase of the invasive process may lead to limited symptoms, but crossing of the epithelial barrier allows the microorganisms to enter a septicemic phase that causes the systemic symptoms and complications characteristic of typhoid fever [2]. In the case of Shigella, the invasive process remains localized to the colonic and rectal mucosa, thereby causing major inflammatory destruction that accounts for a dysenteric syndrome, thus the name bacillary dysentery [3]. In many cases, however, shigellosis causes only a watery diarrhea similar to that observed with noninvasive pathogens. Shigellosis is a disease of the poor, primarily affecting young children in the developing world. Epidemic outbreaks also occur in industrialized countries following accidental breaches in hygiene or sanitation. The term "dysentery" was introduced by Hippocrates, who noted the seasonal pattern of the disease. The disease entity and its probable bacterial cause were described by Widal and Chantemesse. One of the four etiological species, Shigella dysenteriae (Shiga bacillus), was first identified by Shiga a century ago. The global burden of Shigella infection has recendy been reevaluated [4].

8.

SHIGELLOSIS

337

//. Bacteriology The shigellae are Gram-negative, nonsporulating, facultative anaerobic bacilli belonging to the family Enterobacteriaceae. The genus Shigella comprises four different species: S.flexneri (6 serotypes), S. dysenteriae (16 serotypes), S. sonnet (1 serotype), and S. hoydii (8 serotypes). Shigella are identified based on their capacity to ferment various sugars and on the antigenic specificity conferred by LPS 0-sidechains (i.e., somatic antigen), which accounts for their specific serotype. Shigella is a close variant of Escherichia coli since Shigella species have more than 80% nucleotide sequence identity with E. coli [5]. Conjugation between Shigella and E. coli has been carried out successfully [6], thereby facilitating genetic studies, particularly those investigating pathogenesis. Shigella is a nonmotile bacterium, although the genes encoding the flagellar apparatus are present [7]. The significance of the crypticity of flagellar expression is presently unknown. However, it should be noticed that careful electron microscopic analysis has allowed identification of one to three polar flagella on certain fresh clinical isolates [8]. As described below, what really characterizes Shigella and maintains it as an independent genus is its invasive phenotype. This invasive phenotype reflects, in cell assay systems of infection, the ability of Shigella to enter into, invade, and destroy the colonic and rectal epithelial tissues. The invasive phenotype is encoded by a large 200-kb virulence plasmid found in all Shigella species. Coevolution of the chromosome with the virulence plasmid has led to coadaptation, and genetic modifications have progressively accumulated in the Shigella chromosome that enhance the efficiency of the plasmid-encoded invasive phenotype. Plasmid invasion genes falling under chromosomally encoded regulatory loops, acquisition of pathogenicity islands, and the absence of certain metabolic functions constituting "black holes" in the chromosome [9] support this concept, which is reinforced by the observation that enteroinvasive E. coli (EIEC), an E. coli pathovar considered an intermediate in evolution between E. coli and Shigella, is less virulent in humans since the oral infectious doses required in human volunteers to cause dysentery are 100 colony forming units (cfu) for Shigella and 10^ cfu for EIEC [10]. Shigella is present in the stools of patients at a concentration of 10^ to 10^ cfu per gram of feces during the first days of illness. Then, the number of colony-forming units decreases dramatically, and diagnosis may become difficult, especially because the microorganisms are fragile and no enrichment medium exists. Large numbers of polymorphonuclear leukocytes (PMNs) are present in stools at the early stage of the disease, reflecting the intense inflammation caused by the pathogen. Careful examination of the stools can rule out the presence of trophozoites of pathogenic Entamoeba histolytica, which, unlike Shigella, do not elicit massive luminal release of PMNs. Bacterial identification is confirmed by seroagglutination with sera elicited against the various Shigella somatic antigens corresponding to the 0-sidechains of the LPS.

338

PHILLIPPE J. SANSONETTI, COUMARAN EGILE, AND CHRISTINE WENNERAS

///. The Somatic Antigen LPS is the major constituent of the bacterial outer membrane. It is composed of three covalently Hnked moieties: lipid A, core, and 0-sidechains. The Shigella lipid A is identical to that of E. coli and mediates the endotoxicity of LPS. The core region consists of an oligosaccharide similar but not identical to that of E. coli. The 0-sidechains are composed of repeated sugar subunits which vary in their composition, thereby contributing to the serotypic diversity. For example, the tetrasaccharide 3P-A^-acetyl-glucosamine-a 1 -2-rhamnose-a 1 -2-rhamnoseal-3-rhamnose-l represents the basic tetrasaccharide repeating subunit of S. flexneri Y. Further changes, such as glycosylation of one of the rhamnoses, produce the serotype specific variations that yield serotypes 1-5 [11]. This explains the high level of crossreactivity among the S. flexneri serotypes, which extends in some cases to E. coli strains. On the other hand, S. sonnei does not crossreact with E. coli, but with one serotype of Pleisiomonas shigelloides [12]. In S. flexneri, the genes involved in the biosynthesis of LPS are mainly chromosomal, but some of them have been identified on the virulence plasmid (C. Parsot, unpublished data, 2000); however, whether these plasmid genes are functional has not yet been addressed. Some of the chromosomal genes that determine 0-sidechain specificity are carried by lysogenic bacteriophages [13]. In S. dysenteriae 1, the LPS biosynthesis genes are located both on the chromosome and on a 9-kb plasmid [14]. In S. sonnei, the genes encoding LPS 0-sidechains are all located on the virulence plasmid, which is easily lost on subculturing, thus generating stable rough Form II colonies from the unstable smooth Form I colonies carrying the plasmid [15, 16].

IV. Epidemiology and Transmission The most common Shigella species in the developing world are S. flexneri and S. dysenteriae 1. They accounted for 66 and 16% of hospitalized cases of shigellosis in Bangladesh, respectively, in the early 1980s [17]. S. flexneri is primarily responsible for the endemic form of the disease, whereas S. dysenteriae accounts for the epidemic form of the disease. S. boydii is rarely encountered and seems essentially associated with cases of shigellosis on the Indian subcontinent. In industrialized areas. Shigella epidemic outbreaks are dominated by S. sonnei. The transition from S. flexneri to S. sonnei is associated with economic development. The reasons underlying this interesting association between socioeconomic context and prevalence of a particular species of Shigella are currently unknown.

8.

SHIGELLOSIS

339

Also, the mechanisms that cause severe dysenteric forms of the disease in some patients and mild purely diarrheic forms in others are likely to respond to the nature and size of the infecting inoculum, as well as to the host immune status and the host genetic background. The exact incidence of shigellosis is difficult to assess. Recent extrapolations made from reliable epidemiological data selected on a worldwide basis indicate that there are about 145 million cases every year, 99% of which take place in the developing world, with children under the age of 5 years being the principal victims [4]. For instance, in a poor suburban area of Santiago (Chile), a child had a 67% chance of developing shigellosis in the first year of life when this extensive study was carried out in 1991 [18]. Mortality due to shigellosis reaches between 500,000 and 1.5 million every year [4] and is particularly associated with epidemics developing in a dramatic public health context. Rapid constitution of refugee camps such as those that recently formed in the Central Lake area of Africa is a typical example of such emergency situations. Lack of sanitation and of a safe water supply, absence of personal hygiene, stress, malnutrition, concurrent infections, and antibiotic resistance are likely to explain the high rate of mortality in those situations in which S. dysenteriae 1, the most virulent Shigella, is the etiological agent. Accurate mortality data have been obtained in Bangladesh, where Shigella infection may account for up to 20% of the total mortality among children between the ages of 1 and 4 years [19]. Studies carried out in the 1980s showed that the mortality rate in hospitalized Bangladeshi children with Shigella dysentery was 10%, with a surprisingly similar rate of death regardless of species or serotype [19]. During epidemics, which are often due to S. dysenteriae 1, attack rates have been calculated to range from 1 to 50%, and the mortality rate from 6 to 70 per thousand. The only natural hosts of Shigella are humans and monkeys. Most of the disease transmission occurs via person-to-person contact, the bacteria being able to survive on the skin. Shigella is also often transmitted by contaminated food and water. Flies can transmit Shigella from human feces to food [20]. Due to oral/anal and oral/genital sexual practices, shigellosis is also considered a potentially sexually transmitted pathogen that became very prevalent in some homosexual communities in the 1980s [21]. As already mentioned, Shigella is highly infectious. Striking examples are the occurrence of large outbreaks of shigellosis following accidental contact between a sewage pipe and an urban water supply system in Haifa (Israel), thereby causing 8,000 cases of shigellosis within a week [22], or a food contamination in the United States in which more than 50% of the 12,700 persons attending a mass gathering contracted S. sonnei infection. These striking epidemiological observations reflect the results of experimental infections carried out with volunteers in the United States, during the course of which it was established that as few as 10

340


microorganisms administered orally could cause infection [10], and a few hundred induced a high attack rate that was not increased in frequency and severity by further increasing the infectious dose [23]. Acid resistance may be part, but not all, the explanation for the high infectivity of Shigella, which is not seen with other enteric pathogens, such as V cholerae, which requires much higher doses (i.e., 106-10^ cfu).

V. Disease Symptoms and Complications: Orientations for Future Research!?

A. A Spectrum of Disease Symptoms from Diarrhea to Dysentery As already mentioned, there are essentially two patterns of illness caused by Shigella. The first is the classical bacillary dysentery, characterized by fever, major intestinal discomfort with intestinal cramps and tenesmus, as well as permanent emission of a fecal, bloody, mucopurulent stools. This clinical form is characteristic of infections caused by S. dysenteriae 1 and S. flexneri. Second is a more benign episode of watery diarrhea that may last several days, which is typically associated with S. sonnei. There are, however, severe cases of S. sonnei infection [19] and mild cases of S. flexneri infection; therefore, the intrinsic capacity to cause dysentery seems to be present in all strains, regardless of species and serotype. The watery diarrhea that often precedes the dysenteric form of the disease, as well as the purely diarrheic form of the illness, raise the yet-unsolved problem of the pathogenesis of the diarrheal component of this disease. It is well established that shigellosis leads to colonic dysfunction, which is characterized by net decreased absorption of water, increased secretion of potassium ions, and decreased absorption of chloride ions [24]. There are two, possibly complementary, explanations for these symptoms. First, diarrhea represents the minimal disease symptom if the invasive process remains limited in severity and extension. This situation is classically observed in inflammatory bowel diseases such as ulcerative colitis. Second, diarrhea reflects the production of one or several "classical" enterotoxins that have been recognized in S. flexneri and will be mentioned later in this chapter. It should also be emphasized that about 50% of Shigella infections are asymptomatic [25]. In experimental infections of volunteers, the classical "symptomatic triad" of shigellosis—fever, abdominal pain, and bloody mucopurulent stools—was seen in fewer than 50% of the volunteers who developed the illness [26].

8.

B.

SHIGELLOSIS

341

Acute Complications of Shigellosis

The most frequent causes of death among hospitaHzed children during shigellosis are septicemia, primarily seen in malnourished children [27], and hypoglycemia. Septicemic strains are not necessarily Shigella, because other Gram-negative microorganisms are observed in about half of cases [19]. This is due to rupture of the epithelial barrier facilitating passage of other bacteria from the intestinal flora. The hypoglycemia may result in blood glucose concentrations as low as 1 mM, regardless of the nutritional status of the child. As insulinemia is normal in this situation, a block in gluconeogenesis is likely to occur, as the concentrations of glucose-releasing hormones such as glucagon, epinephrine, and norepinephrine are high [28]. Relevant animal models mimicking these processes are required in order to study these situations at the experimental level. Toxic megacolon, a major widening of the colon that becomes atonic, possibly leading to intestinal perforation with peritonitis and severe sepsis, is of poor prognosis. The pathogenesis of this complication is not understood. It occurs regardless of the infecting species, thereby ruling out a major role for Shiga toxin. It is likely that severe underlying inflammation caused by mucosal invasion largely accounts for this complication. The relationship existing between inflammation and malfunction of the autonomous colonic nervous system needs to be further investigated. Among the classical complications of shigellosis are seizures. They may appear in the absence of hypoglycemia [28, 29], and their pathogenesis is not clear. They are seen in children regardless of the infecting strain, thus ruling out an exclusive role of Shiga toxin, which has been shown to be neurotoxic, in their pathogenesis. The hemolytic and uremic syndrome (HUS) occurs essentially as a complication of infections by S. dysenteriae 1, thereby suggesting a major role played by Shiga toxin in this process (see below), and in agreement with the phenotypes reported for its homologs in EHEC infections [30]. Ongoing experiments carried out in animal models of HUS should soon unravel the molecular and cellular mechanisms underlying this complication and possibly indicate prophylactic and therapeutic approaches [31]. Similarly, the pseudo-leukemoid syndrome in which white blood cells may reach 40 x 10^/L can be observed either in the presence or the absence of an HUS. It has no consequence in itself but reflects a serious condition and is associated with a poor prognosis [32]. It is likely to reflect a major inflammatory status, possibly with high circulating levels of cytokines (such as IL-6, G-CSF and GM-CSF) stimulating leucopoiesis. In the mouse model of pulmonary infection by Shigella, a strict correlation exists among the severity of the inflammatory process, the numbers of colony-forming units in lungs, and the levels of circulating IL-6 [33].

342


C. Long-Term Complications of Shigellosis Malnutrition is a severe consequence of serious cases of shigellosis. Shigellosis has been recognized as a major cause of chronic malnutrition in children [19]. The pathogenesis of this condition is probably multifactorial, associating lack of appetite and limited malabsorption, the latter having a weak impact as the small intestine is not affected by Shigella. However, it is likely that the persistence of intestinal inflammation with colonic protein loss and the systemic effect of elevated levels of TNF-a accounts for this condition. This is supported by the observation that the number of cells producing proinflammatory cytokines remains identical in rectal biopsy samples taken at the acute phase of shigellosis and a month later [34, 35]. Reactive arthritis following shigellosis is a condition more frequently observed in patients expressing the HLAB27 haplotype.

VL Histopathology of Shigellosis: A Window on Pathogenesis Colonoscopic examination of patients at the acute phase of shigellosis shows that the rectosigmoid area is constantly affected by the inflammatory process. The proximal segments of the colon, and the ileum, are less often affected. Segmentous localization is another characteristic that is shared between shigellosis and ulcerative colitis. However, the reason for this rather selective localization of the lesions is unknown. The intestinal lesions are diffuse and continuous; they comprise edema, erythema, focal hemorrhages, and often a white mucopurulent layer of adherent exudate resembling false membranes [36]. On histopathological examination of these lesions, it appears that the mucosa is primarily affected, whereas the submucosa is rather spared by the inflammatory infiltrate. When seen, the bacteria are usually localized to the epithelium of the surface and upper thirds of colonic crypts. The relationship between the degree of bacterial invasion of colonocytes and colonocyte damage is not obvious. In addition, when rectal biopsies are performed during the early stages of infection, the epithelium overlying lymphoid follicles (i.e., follicle-associated epithelium or FAE) is damaged by Shigella infection [37]. The corresponding macroscopic lesions may be aphthoid ulcers, as observed in Crohn's disease. These observations raise two major points that will be considered later: (1) the route of Shigella translocation through the epithelial lining at the initial stage of the disease that seems to correspond to the FAE, and (2) the mechanisms of the destructive process that extends far beyond the sites of

8.

SHIGELLOSIS

343

bacterial invasion, thus indicating that the development of shigellosis is characterized by an uncontrolled and diffuse inflammatory process responding to a cascade of amplification signals that still needs to be characterized. An exudate made of colonocytes, PMNs, bacteria, and red blood cells in a fibrin layer is observed in the intestinal lumen, the epithelium showing major signs of degeneration with damaged cells, sometimes in the process of shedding, and empty goblet cells. The epithelial lining is often infiltrated by white blood cells—particularly PMNs and lymphocytes [38]—and in some places it undergoes complete detachment. In the crypt epithelium, activated intraepithelial lymphocytes are seen as well as numerous PMNs. Monocytes and eosinophils can be seen, although at a later stage of the disease. The lamina propria is infiltrated by PMNs, monocytes/macrophages, and plasma cells. Widespread vascular lesions are seen, from swollen or pyknotic endothelial cells to total destruction of capillaries. Thrombi may be observed in larger local vessels [38]. At a later stage of the disease, dilated, elongated, and branched crypts may be observed, some of them heavily infiltrated by PMNs, thus forming crypt abscesses [36].

VIL Animal Models: Strengths and Weaknesses The oldest animal model for shigellosis is the keratoconjunctivitis assay or Sereny test [39]. This assay consists of instillating a suspension of bacteria in the keratoconjunctival sac of a guinea pig. Pathogenic shigellae invade the conjunctival layer, causing an acute destructive conjunctivitis that is characterized by redness of the eye followed by a keratitis with massive migration of inflammatory cells, particularly PMNs, which cause corneal turbidity and purulent discharge closing the eyelids [40]. Bacterial mutants that cannot invade epithelial cells, are unable to spread from cell to cell, or express a rough LPS, are negative in the Sereny test. This assay reflects the invasive phenotype of Shigella both in terms of invasion of cells and cell-to-cell spread, and in terms of triggering an inflammatory response. It does not reflect, however, the specificity and complexity that can be observed in histopathological studies in humans. During the 1950s, an impressive series of studies was carried out in order to identify animals (essentially mammals) that would develop a typical dysenteric syndrome in the presence of an oral or a gastric inoculum of Shigella [41]. Only monkeys (particularly macaques), young bears, and kittens developed the disease when orally infected with 10^^ microorganisms. From these studies, only macaque monkeys, who develop an invasive rectocolitis when orally or intragastrically infected with Shigella, have "survived" as one of the most faithful animal models.

344


This model has been extensively used for pathogenesis studies and for testing of live oral attenuated vaccines [42]. Yet, ethical and experimental drawbacks make it unsuitable for routine studies. These drawbacks include the high cost of animals and their housing. In addition, the infectious doses required for the animals to develop dysentery are so high (i.e., 10 million to 100 million times greater than the infectious dose in humans) that the relevance of the model can be questioned, particularly in the context of testing the tolerance of an attenuated vaccine candidate, or doing challenge experiments in vaccinated animals. Guinea pigs are infectable orally or intragastrically; however, they develop a dysentery-like syndrome only if starved 4 days before the inoculation, which must be preceded by administration of streptomycin in order to disrupt the barrier effect of the resident intestinal flora, and morphine in order to block intestinal peristaltism. Two models are of value for routine pathogenesis studies. The first is a mouse model in which the Shigella inoculum is administered intranasally, resulting in invasion of the tracheobronchial tract that causes a massive inflammatory bronchotracheal alveolitis [43]. Although this model is not relevant with regard to the organ specificity of Shigella infection, it has the advantage of making use of an animal in which the immune system has been explored in such detail that most of the tools are available to study the immunoinflammatory components of the disease, as well as some aspects of the systemic and local immune protection against Shigella infection [33,44,45]. Alternatively, the rabbit ligated-loop model of infection consists of using intestinal loops that are ligated after laparotomy under general anesthesia, the vasculature being carefully preserved. A large inoculum of bacteria is then injected into the loops, and the invasive and inflammatory processes can be followed by sacrificing animals at given times (usually between 2 and 16 hr). Histopathological studies can then be carried out, and proinflammatory cytokines dosed. This model is particularly useful for studying the role of specific cytokines, such as interleukin-1 (IL-1) and interleukin-8 (IL-8), in the development of intestinal invasion, mucosal inflammation, and tissue destruction [46, 47], and to better analyze the role of the follicular-associated epithelium in the initial steps of epithelial translocation [48, 49].

VIIL Cellular Models of Infection: The Contribution of Shigella to ttie Concept of Cellular Microbiology In v/Yro-cultured cells (e.g., HeLa cells, Henle cells, Hep-2 cells) have been used to study the capacity of Shigella to penetrate into cells that are not professional

8.

SHIGELLOSIS

345

phagocytes [42]. In the 1980s, these tests were refined and became quantitative. The "gentamicin protection assay," in which the antibiotic remains extracellular, thereby killing bacteria that have not entered cells but sparing intracellular bacteria, allowed to quantitate bacterial entry into host cells. The "plaque assay" [50] and the "infectious focus assay" [51 ] both allowed the study of cell-to-cell spread on confluent cell monolayers. Combined with molecular genetic analysis of bacterial pathogenicity, these tests have allowed identification of the bacterial factors involved in entry, escape into the cytoplasm, and intracellular motility/cellto-cell spread. They have also uncovered the major eukaryotic cell components supporting these processes, particularly the actin cytoskeleton [52]. These approaches helped establish the concept of cellular microbiology that analyzes microbial pathogenesis at its intersection between the prokaryotic and eukaryotic worlds [53]. According to this concept, virulent Shigella are characterized by expression of an "invasive phenotype" whose effect depends on the target cell: entry/escape into cytoplasm/intracellular motility/cell-to-cell spread in the presence of epithelial cells; the complete invasive phenotype being best observed when polarized epithelial cells such as Caco-2 or T84 cells are infected basolaterally [54]; apoptotic killing and maturation/release of IL-ip in macrophage [55, 56]; activation of PMNs adherence; and release of granule contents [57]. The cell-dependent aspect of this "invasive phenotype" is summarized in Figure 1. More recently, these cell assays have been made even more sophisticated by growing the epithelial cells on filters in order to induce their polarity, and by establishing complex systems combining epithelial cells and PMNs in order to mimic some aspects of the inflammatory response [58, 59].

IX. Pathogenic Mechanisms: In Vitro Expression of the Invasive Phenotype A. Molecular and Cellular Mechanisms of Shigella Invasion of Epithelial Cells: Basic Principles and Reviews The pathogenic factors involved in Shigella invasion of epithelial cells have been extensively studied and reviewed [52, 60-65]. These factors participate in entry into nonphagocytic cells by a macropinocytic event requiring massive cytoskeletal rearrangements [66], escape of the bacterium into the cytoplasm [67], intracellular multiplication, and cell-to-cell spread [51, 68, 69]. Shigella can enter several cell lines in vitro, regardless of the species and organ of origin.

346

PHILLIPPE J. SANSONETTI, COUMARAN EGILE, AND CHRISTINE WENNERAS BACTERIAL SURVIVAL

+++ (CYTOPLASMIC)

HOST CELL RESPONSE

^„.^„^,... ^.... EPITHELIAL CELL

INTERNALISATION PRODUCTION OF INFLAMMATORY CYTOKINES/CHEMOKINES (IL-8..) MONOCYTE/MACROPHAGE INTERNALISATION RELEASE OF ILlp APOPTOSIS

+++ (CTTOPLASMIC)

POLYMORPHONUCLEAR LEUCOCYTE INTERNALISATION RELEASE OF GRANULES (INTRAVACUOLAR)

ENHANCED EXPRESSION OF ADHERENCE MOLECULES

Fig. 1 Differential expression of the Shigella invasive phenotype depending on whether epithehal cells, macrophages, or polymorphonuclear cells are infected.

B.

Molecular and Cellular Biology of the Entry Process into Epithelial Cells 1.

THE PARADIGMS OF BACTERIAL ENTRY INTO EPITHELIAL CELLS

Two paradigms of the entry process have been described. First, the "zippering" process, which involves intimate contact between the bacterial surface and the host cell membrane, is mediated by a bacterial surface ligand binding with high affinity to a cell membrane receptor involved in cell adherence [70]. The Y. pseudotuberculosis invasin, which binds with high affinity to integrins of the (31 family [71], and Listeria monocytogenes intemalin A, which binds to E-cadherin [72], illustrate this "zippering" phagocytic process. Second, the "triggering" process, which is closely related to macropinocytosis, results in bacterial internalization in a vacuole that initially is loosely associated to the bacterial body. Shigella and Salmonella, which both secrete a set of homologous invasion proteins on contact with their cellular targets, induce important but localized rearrangements of the cell cytoskeleton at their site of interaction. Actin polymerization is essential for bacterial entry since it is abrogated by cytochalasins. Cellular extensions rise to 10 |Lim over the cell surface, forming a flower-like

8.

SHIGELLOSIS

347

Fig. 2 Entry of Shigella into epithelial (HeLa) cells, (a) Scanning electron microscopy showing an ongoing entry focus characterized by membrane projections getting organized in multiple ruffles that eventually form a macropinocytic vacuole. Reprinted through the courtesy of Ariel Blocker and Roger Webf, EMBL (Heidelberg, Germany), (b) Transmission electron microscopy showing the section of an entry focus with multiple projections characterized by massive rearrangements of the cell subcortical cytoskeleton. Reprinted through the courtesy of R Gounon and P. Sansonetti, Station Centrale de Microscopic Electronique, Institut Pasteur. Bars = 1 |im.

Structure (Fig. 2), both in the case of Salmonella [73] and Shigella. These projections merge and engulf the bacterial body within 5 to 10 minutes [74]. These membrane projections are supported by newly formed actin filaments. These filaments are oriented with their fast-growing "barbed" end facing the inner face of the cell cytoplasmic membrane, and bundled by plastin, which is necessary to

348


Stabilize the projections, and required for efficient bacterial entry [74]. The signals eliciting these massive cytoskeletal rearrangements are the result of a complex crosstalk between the bacterium and its target cell. In confluent, polarized epithelial layers (i.e., Caco2 or T84 colonic cell lines of human origin) reconstituted in vitro on a tissue culture dish, or on a filter. Shigella is unable to significantly enter cells. Only direct contact with the basolateral pole of these cells allows efficient entry [54]. This paradox has been solved in complex cellular models of infection, or in in vivo studies in which the invading bacteria and their respective mutants are traced, according to kinetics of infection of the intestinal mucosa.

2.

GENES AND GENE PRODUCTS REQUIRED FOR SHIGELLA ENTRY

Invasive strains of S.flexneri harbor a 220-kb plasmid that contains most of the identified virulence genes of this microorganism [15]. A 30-kb locus in this plasmid contains all the genes necessary for entry [75-77]. This region is divided into two divergently transcribed operons that encode two classes of proteins (Fig. 3). The mxi-spa locus comprises about 20 genes specifying a type III secretory apparatus [78-81], which is expressed and assembled at 37°C and activated on contact with the target cells [82, 83]. Homologs of this system are present in many Gram-negative pathogenic bacteria [84]. These systems seem essentially devoted to allow translocation of bacterial effectors (i.e., toxins) directly into the cytosol of the eukaryotic target cells; they are functionally conserved among these different species. In the presence of proper chaperones, heterologous secretion of virulence proteins has been shown in Shigella, Salmonella, and Yersinia [85, 86]. One of the best-characterized translocation processes is the Yersinia Ysc type III secreton, which, on contact with macrophages, allows secretion of Yop proteins through the Ysc complex. One of them, YopB, forms a pore across the host cell membrane through which other secreted proteins such as YopE and YopH are translocated into the cytoplasmic compartment of the host cell, resulting in

• t virB

^fe

ipgD

a M ipaA

ipaD

ipaC

ipgF mxiH mxU mx'iK mxiM mxiD ipgE mxiG mxU mxiL mxiE

ipaB ipgC ipgA icsB ipgB

mxiC

mxiA

T^ ipgD

spa47 spa32 spa33 spa9 spa40 spa 15 spalS spa24 spa29

Fig. 3 Genetic map of the Shigella flexneri 5 entry locus. This 30-kb region is located on the large virulence plasmid of this species, pWRlOO. It comprises two subloci transcribed in opposite directions. (Top) The ipa operon encodes the entry effectors genes (shaded in gray). (Bottom) The mxi and spa operons encoding the components of the type HI secretory apparatus (shaded in gray).

8.

SHIGELLOSIS

349

inhibition of phagocytosis [87]. In Shigella, a complex formed by the IpaB and IpaD proteins regulates secretion of the Ipa proteins [83]. Optimal secretion occurs during the exponential phase of growth [88]. The host compounds that activate the Mxi-Spa secretory system are now being identified. Fibronectin has been proposed as a possible candidate [89]. In addition, dyes such as Congo red and some of its structural equivalents are able to induce active secretion of the Ipa proteins, and may provide some insight into the structure of the components that are able to induce secretion [90]. The Mxi-Spa system of Shigella secretes about 15 proteins that share such common features as the lack of signal peptide and the capacity to aggregate in an extracellular milieu in large supramolecular structures [91]. The ipa locus, which essentially consists of an operon, encodes four proteins— IpaB (62 kDa), IpaC (42 kDa), IpaD (37 kDa), and IpaA (70 kDa)—which are secreted by the type III Mxi-Spa system on contact of bacteria with host cells. It also encodes IpgC, an 18-kDa cytoplasmic chaperone that binds IpaB and IpaC in the bacterial cytoplasm, preventing their aggregation and proteolytic degradation [92]. In-frame deletions of the ipaB, ipaC, and ipaD genes lead to complete inactivation of the bacterial entry phenotype, indicating their essential role in entry [93], whereas inactivation of the ipaA gene leads only to partial inactivation of entry [94]. Once secreted, IpaB and IpaC form a complex [92] interacting with the epithelial cell membrane [94a]. Latex beads coated with the IpaB-C complex are internalized by HeLa cells, indicating that this molecular complex acts as an entry effector [95]. However, the surface rearrangements observed during internalization of the beads are not as strong as those observed during bacterial entry, suggesting that the entry mechanism is only partially reproduced by the IpaB/C complex. There are two major possible explanations for these results: either some secreted proteins are missing from the complex, or immobilization of the complex on the bead surface allows only partial interaction of the secreted proteins with their cellular targets. The mode of action of these proteins is still unclear. The Ipa complex has been shown to bind a fibronectin receptor, a5pi integrin [96], but integrins do not appear to be exclusive receptors for the IpaB-C complex. Another cell surface receptor, CD44, the hyaluronate receptor, binds IpaB [96a]. These interactions may contribute to an adherence stage, which represents only a preliminary step in the entry process. It is possible that, after transient binding to one or several receptors, such as the a5pl integrin and/or CD44, the IpaB-IpaC complex inserts into the epithelial cell membrane, forming a pore or translocator structure that allows further injection of other effector proteins into the cell. This is based on analogies with the Ysc type III secretion system of the Yops and with the ability of IpaB and IpaC to insert into and destabilize lipid bilayers [97, 97a]. Insertion of this IpaB-IpaC complex may cause two different effects: (1) induction of nucleation and polymerization of actin filaments supporting membrane projections, and (2) translocation of other Ipa proteins such as IpaA and other Shigella proteins secreted on cell contact. Ipa proteins are likely to bypass the usual processes of outside signaling in which extracellular ligands such as growth

350


factors bind the extracellular domain of a transmembrane receptor that becomes activated and initiates a signal cascade causing cytoskeletal rearrangements. A current hypothetical model for the secretion and translocation of the Ipa proteins is shown in Figure 4. 3.

SIGNALING PATHWAYS INVOLVED DURING SHIGELLA ENTRY

Observation by transmission electron microscopy after SI-myosin decoration of actin filaments shows progressive accumulation of dense dots underneath the cytoplasmic membrane of the epithelial cell at the site of interaction with the bacteria. These dots correspond to actin nucleation zones from which filaments

0 Fig, 4 Hypothetical scheme showing activation of the Mxi-Spa apparatus in the presence of the target eukaryotic cell membrane, formation of a pore by the IpaB-IpaC complex, and injection of proteins such as IpaA. There is no current evidence that the IpaB-IpaD complex, which controls protein secretion, is located at the tip of the secretory apparatus.

8. SHIGELLOSIS

351

further extend [74]. These filaments support the formation of microspikes that progressively fuse to form membrane leaflets that engulf the microorganism in a large vacuole. As maturation of this entry focus proceeds, a dense meshwork of polymerized actin accumulates around the vacuole, forming a cup that may represent an essential step for internalization of the macropinocytic vacuole. These steps are summarized in Figure 5. The recruitment of most of the cytoskeletal-associated proteins seen in focal adherence complexes indicate that there is significant analogy between these structures and the actin cup. If one considers the subdomains of an entry focus, the cytoskeleton-associated proteins can be classified into two major categories: (1) proteins required for signaling such as the three small GTPases of the Rho family—Cdc42, Rac and Rho itself [98, 99, 99a]—and the protooncogene protein p60^"^'^^ [100], and (2) proteins likely to be required for structuring and stabilizing the entry focus architecture (i.e., scaffolding proteins), particularly those forming F-actin bundles such as plastin [74], a-actinin [94], ezrin [96a], and vinculin and talin, which are both involved in focal complexes formation [94]. The location of these structural, motility, scaffolding, and signaling molecules in the two major substructures of an entry focus is shown in Figure 6. 4.

INVOLVEMENT OF SMALL GTPASES

The Rho subfamily of small GTPases regulates specific rearrangements of the cytoskeleton [101]. In Swiss 3T3 fibroblasts, Cdc42 induces the formation of mi-

Filopodes Lamellipodes Adherence-plaque like structure

Fig. 5 The four steps of Shigella remodeling of the eukaryotic cell cytoskeleton leading to a fully functional entry focus. Initial filopodial projections supported by extension of actin filaments undergo a transition to the formation of curtains (ruffle-like) structures. An actin cup eventually forms around the entering vacuole.

352

fUopodesâmeiiipodes


n In Oil W ff r

actin myosin n a-actinin plastin ezrtn RhoB/C Cdc42

Rac focai-piaqiie like stmctiire

,^^

actin myosin 11 >v a-actlnin Ji::> vinculin • paxiiiin \ talin 1 src RhoA

structure motiliiy/scalfoldiiif scaffdidiiig

-

slgita!]zati0&

• stmctiire iiMrtility/scalfoidiiig scaffc^ding

-

signalizatlon

-

Fig. 6 Localization of structural, motility/scaffolding, and signalization-related cytoskeletal proteins to subdomains of the Shigella entry focus.

crospikes and filopodia; Rac is mostly responsible for lamellipodial formation, and Rho is implicated in the formation of focal adhesion complexes and stress fibers. Members of the Rho family act as molecular switches and cycle between their active GTP-bound form and their inactive GDP-bound form. These GTPases can be inhibited by overexpressing an inactive, dominant negative form of the GTPase bearing a mutation in the GTPase domain (asparagine 17 for Cdc42 and Rac, and asparagine 19 for Rho). Rho can also be inactivated by the C3 exoenzyme of Clostridium botulinum that ADP-ribosylates the 21 -kDa Rho molecule on asparagine 41 [102]. Using these approaches, Cdc42 and Rac were shown to be required for entry of Shigella into HeLa cells [99a]. Transient expression of a dominant negative form of Rac or Cdc42 inhibited Shigella entry approximately 70%, due to a dramatic decrease in actin polymerization at the level of entry foci. Rho is also required [98, 99] because both C3 treatment and transient expression of Rho N19 inhibited Shigella entry by 90 and 70%, respectively [99a]. Decoration of actin filaments with S1 -myosin in samples treated with C3 suggests that Rho is involved in the bundling, and possibly also in the elongation of actin filaments rather than in the nucleation of foci [98]. These observations are in contrast with the Salmonella entry process, in which Cdc42, but not Rho, appears to be required [103]. The pathway(s) by which the IpaB-IpaC complex activates the three GTPases is yet unknown. A current model is shown in Figure 7 along with other factors that are presented below. It is likely that the various small GTPases are used in a stepwise process providing initial actin polymerization (i.e., Cdc42 mediated) followed by dynamic remodeling of the actin-made entry focus (i.e., Rac mediated), which ends up in a mature structure (i.e., focal-plaque-like) able to efficiently internalize the bacterial body (i.e., Rho mediated).

8.

353

SHIGELLOSIS

IpaB/C

I

IpaB/C

f

LAMELLBPODIA

IpaA

\TNCUr.IN

FOCAL PLAQUE

Fig. 7 Signaling cascade causing the serial events of cytoskeletal remodeling that lead to a mature Shigella entry focus. Activation of the cascade of small GTPases initiates the actin polymerization process and remodeling of the entry focus. Both cellular and bacterial proteins participate in the remodeling, particularly ezrin, which increases filament extension, and IpaA, which participates in formation of the actin cup.

5.

INVOLVEMENT OF PP60^-^^^

Substrates of the src tyrosine kinase family and src itself are recruited at the site of Shigella entry. Current evidence indicates that c-src is activated since one of its major substrates, cortactin, becomes tyrosine phosphorylated as entry of Shigella proceeds [100]. It is not yet clear whether the cortactin-mediated kinase activity of c-src is required for entry, or if c-src simply recruits focal adhesion components such as FAK and/or paxillin through its SH2 and SH3 domains. Recent demonstration that expression of a dominant negative form of c-^rc inhibited cortactin phosphorylation as well as the efficiency of Shigella entry via a significant reduction of

354


the actin rearrangements in the entry focus would indicate that c-src activation is required for bacterial entry [104]. Src could therefore activate cytoskeletal components, or coordinate some key steps leading to the remodeling and maturation of the entry focus. Src and Rho are two major regulators of focal adhesions, consistent with the fact that the formation of SL Shigella entry focus shares similarity with focal adhesion plaque formation. Substrates of the tyrosine kinase Src, such as cortactin, which is the major tyrosine-phosphorylated substrate during Shigella entry, accumulate in the entry focus [100]. Other substrates found in focal adhesion complexes, such as paxillin and pl25FAK, are also present in the entry focus [96] (J. Mounier and P. J. Sansonetti, unpublished results, 1998). Finally, Rho is known to recruit c-src at the cell periphery via remodeling of the cytoskeleton [105]. Phosphorylation of pl25FAK has been shown to occur in conjunction with the interaction between the Ipa complex and a5pl integrins [96]. How important this signal is in the development of the entry focus remains to be determined. 6.

INVOLVEMENT OF "SCAFFOLDING" PROTEINS

Several cytoskeletal proteins are recruited at the entry site of Shigella. These include actin-bundling proteins, which may stabilize the actin extensions that support the entry focus, and has been demonstrated for plastin [74]. In addition, ezrin, which is a member of the ERM (ezrin, radixin, moesin) family is essential to the maturation and functional efficiency of the foci elicited by entering shigellae [96a]. A common characteristic of ERM proteins is to accumulate underneath the plasma membrane in subcellular structures such as microvilli, at cell-to-cell contact sites, and in structures that are dynamically regulated such as filopodia and lamellipodia [106]. In the dynamic contact of Shigella entry, ezrin seems to act not only as a membrane-cytoskeleton linker but also as a molecule mediating extension of cellular projections. 7.

THE "PSEUDO-FOCAL ADHESION" MODEL OF SHIGELLA ENTRY: A ROLE FOR VINCULIN

Several focal adhesion proteins are present in the entry focus [94, 107]. In addition, the Rho GTPase and src are both regulatory components of the architecture of focal contacts. Therefore, it is likely that formation of this "pseudo-focal plaque" is essential for Shigella entry. Vinculin shows a dense pattern of recruitment in close contact with the forming vacuole during the early stages of internalization. A similarly dense recruitment of vinculin can be observed at the level of contact points between the basal surface of the cell and the matrix in the focal adhesions, where the cytoskeleton is anchored to the membrane [108]. Vinculin is thought to play a major role in anchoring the cytoskeleton to the cell membrane. This molecule exists in two conformations: a folded conformation that represents the major pool of vinculin in the cytosol, and an active, open configuration, able to interact with F-actin and actin-binding proteins such as talin and a-actinin, which simultaneously bind the cytoplasmic domain of (31 -integrins [108]. IpaA is able to interact direcdy with

8.

SHIGELLOSIS

355

vinculin during Shigella entry into cells [94], presumably after translocation from extracellular bacteria into the cell. IpaA-vinculin interaction increases entry by a factor of 10. This interaction does not seem to play a role in the induction of the entry foci, but rather in their maturation. Entry foci induced by wild-type Shigella are characterized by a dense network of vinculin and F-actin, localized at the interface between the cell membrane and the bacterial body. This is in contrast to the foci promoted by a AipaA mutant that show massive and prolonged actin polymerization, and diffuse recruitment of vinculin, but no cup structure. It is possible that IpaA binding leads to vinculin activation, thus consolidating the formation of the pseudo-focal adhesion structure that could organize the actin network necessary to anchor the bacterial body in the entry focus. It may also allow efficient recruitment of a-actinin and other constituents that further participate in the assembly of the entry focus. It is likely that several bacterial components act in concert with cytoskeletal proteins or regulatory components of the host-cell cytoskeleton to build up the entry focus and achieve its maturation. C. Escape of Shigella into the Cell Cytoplasm Once intracellular. Shigella lyse the membrane-bound vacuole and gain access to the cytoplasm [67]. Ipa proteins account for the membrane-lysis process [109]. Both purified IpaB [109a] and IpaC [97] are independendy able to lyse lipid vesicles in a pH-dependent manner. Lysis of the vacuole is followed by bacterial multiplication at a rate similar to that observed in broth, thereby reaching a number of several hundred microorganisms per infected epithelial cells. This behavior differs from that displayed by Salmonella that remains intravacuolar, having to adapt to the hostile metabolic conditions that prevail inside this vacuole before they remodel this compartment and acquire the capacity to grow at a rate similar to that of Shigella [110]. Aside from rapid intracellular multiplication, another major consequence of establishing direct contact with cytosolic components is intracellular motility.

D. Intracellular Motility and Cell-to-Cell Spreading of Shigella 1.

The OLM PHENOTYPE

This phenotype is visible when fibroblasts are infected by Shigella [111]. After their escape from the vacuole, shigellae move along the stress cables that radiate from adhesion plaques toward the edge of the nucleus. In this area, they grow as microcolonies. This movement is similar to that of cell organelles and has thus been termed organelle-like movement (Olm). It is clearly visible in cells with a highly organized cytoskeleton such as fibroblasts. The molecular basis of this movement is still unknown. In epithelial cells, Olm movement is hard to detect, with bacteria moving almost immediately by a process in which they induce actin polymerization as shown below.

356


E. Actin-Based Intracellular Motility of Shigella 1.

MORPHOLOGICAL ASPECTS OF ACTIN-BASED MOTILITY

Although many bacterial pathogens are able to enter host cells by inducing cytoskeletal rearrangements, few of them invade the cytosol and subvert the host cytoskeleton in order to colonize the entire tissue. Shigella, Listeria, and Rickettsia are three bacterial species that escape from the phagocytic vacuole and use cytoplasmic cytoskeletal components to propel themselves inside the first infected cell before they reach the cell membrane and induce cellular protrusions. Engulfment of these cellular extensions that contain bacteria by neighboring cells leads to cell-to-cell spread of the pathogen [112, 113]. Among these three bacterial species, Shigella was the first pathogen identified as an intracellular spreading bacterium inside infected cells. Using time-lapse video microscopy, Ogawa and colleagues [114] showed that intracellular shigellae were able to move independently of cellular organelles and to induce formation of cellular extensions at the cell surface. Since this pioneering work, it has been shown that Shigella intracellular motility is associated with formation of actin tails at one pole of the bacterium (Fig. 8, see color plate) [69]. Most of the thermodynamic studies of bacterial intracellular motility have been originally done with Listeria and confirmed with Shigella. Using video microscopy and microinjection of labeled actin monomers, it has been shown that the actin tail remains stationary in the cytoplasm, trailing behind the moving bacterium, and that the rate of incorporation of actin monomers correlates directly with the speed of bacterial movement, suggesting that continuous actin polymerization at one pole of the bacterium is itself sufficient to generate the motile force [115]. Shigella movement inside the cytoplasm is random and rapid (6-60 |Lim/min) [116]. It occurs optimally at the stage of bacterial division [117]. Ultrastructural analysis of Shigella actin comet tails by transmission electron microscopy after SI myosin decoration reveals two types of comets. Intracellular comets are composed of short, crosslinked, randomly organized filaments with their fast growing (barbed) ends always oriented toward the bacterial surface. The density of actin filaments is high in the vicinity of the bacterial body and decreases at the distal part of the comet (Fig. 8, see color plate). Comets present in the protrusions are similar to the intracellular ones with a third supplemental distal portion mainly composed of long axial filaments [117a].

2. IcsA, A

SHIGELLA SURFACE PROTEIN INVOLVED IN

ACTIN-BASED MOTILITY

Genetic analysis of Shigella virulence factors has allowed the identification of icsA, a gene that is responsible for F-actin comet tail formation and intracellular movement [69]. The IcsA protein, also named VirG [68, 118], is required for inter-

8.

SHIGELLOSIS

357

cellular spread in in vitro assays and dissemination in the Sereny test. Thus, IcsA is critical for Shigella virulence, and AicsA mutants have been used for the development of live attenuated Shigella vaccine strains [119]. The icsA gene is present on the pWRlOO virulence plasmid of S. flexneri 5a strain outside the entry locus and 15 kb downstream from the spa operon [107]. The G+C content of the icsA gene, 41%, is different from the G+C content of the entry genes, 34%, suggesting independent acquisition of the icsA gene by Shigella. IcsA is an outer membrane protein. Translocation across both membranes and surface anchoring is independent of the Mxi/Spa type III secretion apparatus and is mediated by an autotransporter secretion pathway, similar to the one responsible for IgA-protease secretion [120]. The IcsA autotransporter domain, IcsAP (344 carboxy terminal residues), may form a p-barrel-like structure composed of antiparallel hydrophobic stretches allowing translocation and anchorage of the N-terminal IcsAa domain (710 residues) at the outer membrane [120] (Fig. 9). Apart from a series of 6 glycine-rich repeats (GRRs) of 32 to 34 residues at the N terminus, IcsAa does not exhibit any specific feature [121], nor does it exhibit sequence similarity to any known protein, and this includes cytoskeletal proteins, or the ActA protein that mediates actin-dependent bacterial motility in Listeria [122, 123]. The surface distribution of IcsA is unusual among bacterial proteins. In wild-type bacteria, IcsA is asymmetrically distributed, being present exclusively at the bacterial pole opposite to the septation furrow in dividing bacteria [121]. Inside infected cells, IcsA colocalizes with the base of the actin tail and determines the site of actin assembly and direction of movement. Two factors are required for polar localization of IcsA. A yet uncharacterized intrinsic property of IcsA may privilege addressing of the protein to the bacterial pole, although a significant proportion of it is still expressed over the entire cell surface. In addition, cleavage of 50% of the total amount of IcsA removes the fraction of this protein expressed laterally and leads to exclusive polar distribution. Cleavage occurs at the junction of the IcsAa and IcsAp domains [124] on a sequence previously shown by in vitro experiments to be the target of PKA mediated-phosphorylation: —S756SRRASS762— [125,126]. The bacterial surface protease SopA (IcsP), a member of the OmpT/OmpP family of serine proteases, is involved in cleavage of IcsA [127, 128]. The processing of IcsA by SopA(IcsP) at the bacterial surface leads to secretion of IcsAa into the culture medium. A AsopA mutant is unable to polarize IcsA and unable to induce comet tail formation, suggesting that polar distribution of IcsA is important for bacterial motility [127]. The IcsAa domain is exposed at the bacterial surface. Expression of IcsAa at the surface of a AompTE. coli K12 mutant induces actin comet tail formation and bacterial motility in cytoplasmic extracts of cell-free Xenopus eggs [129, 130]. Expression of IcsAa at the inner face of the eukaryotic plasma membrane induces subcortical actin polymerization and membrane ruffles [130a]. Thus, membrane presentation of IcsAa is necessary and sufficient to induce actin polymerization.

358


IcsAa

IcsAp

surface exposed domain (53-758)

membrane anchoring domain (759-1102)

SP ,

N-WASP binding site (103-433)

Autotransporter (623-1102)

Vinculin binding site (103-506)

B 1-Actln Nucleation

3-Capping and Cross-ilnking

2-Elongation

Actin niaments

Capping protein

Actin bundling protein

Proniin Actin

VASP

Vinculin

N-WASP

Fig. 9 (A) Map of the functional domains of the 120-kDa IcsA protein. SP = signal peptide. (B) Scheme summarizing the three major steps of IcsA-mediated actin polymerization leading to intracellular motility of Shigella.

Functional (domains of IcsAa involve(J in actin polymerization have been identified by expressing in-frame truncates on the surface of A/cM bacteria [131, 132]. Expression of the Rio3-Ala433 domain of IcsA, which encompasses the GRR domain, is sufficient to elicit F-actin accumulation [132]. The carboxy-terminal portion of IcsAa (residues I509-T720) seems to be involved in establishment of the asymmetric distribution of IcsA on the bacterial surface and in formation of the actin tail [131].

8. SHIGELLOSIS 3.

359

CELLULAR PARTNERS INVOLVED IN ACTIN-BASED MOTILITY

The mechanism of actin polymerization induced by IcsA is still unclear. As the IcsA protein does not interact directly with actin and has no actin-nucleating properties, recruitment of a cytosolic protein complex to the bacterial vicinity is likely to create the proper conditions for actin polymerization and formation of the comet tail [133]. The following steps are currendy considered: (1) actin nucleation, (2) elongation of actin filaments, and (3) capping of filament pointed ends and bundling of filaments, achieving comet stabilization. Several cytoskeletal proteins found in lamellipodial structures have been detected in Shigella actin tails by immunofluorescence analysis. These include T-plastin, a-actinin [134], VASP [135], vinculin [131, 136], Mena, Arp2/3 (C. Egile and P. J. Sansonetti, unpublished results, 1999), and neural Wiskott-Aldrich syndrome protein (N-WASP) [132]. However, the contribution of these proteins to actin-based motility has been demonstrated for only a few of them. Among them, vinculin and N-WASP are the only IcsA ligands so far identified [131, 132]. The role of cellular proteins involved in Listeria actin-based motility (Arp2/Arp3 complex, profilin, VASP, and ADF/Cofilin) [113] in IcsA-mediated motility is currently unknown. Vinculin is a structural protein involved in the crosslinking of actin cytoskeleton to the cell membrane via focal adhesion complexes [137]. Vinculin is present in the entire Shigella actin comet. The globular amino-terminal domain of vinculin interacts in vitro with the amino-terminal portion of IcsA (residues 53-506) [131]. However, the relevance of IcsA-vinculin interaction for Shigella actin-based motility is still debated. Microinjection of an FEFPPPPTDE peptide of ActA or of a (GPPPPP)3 peptide of VASP inhibits Shigella motility [116], and microinjection of the vinculin head portion in Shigella-infccitd cells leads to a threefold increase of bacterial rate of movement [136]. Based on these results, it was proposed that IcsA binding to vinculin may unmask a VASP interaction motif present in the vinculin head portion leading to recruitment of the VASP-profilinactin complex, which could then serve as a molecular scaffold for Shigella actin polymerization [136]. Listeria ActA is also able to recruit VASP at the bacterial surface for formation of the actin comet tail [135, 138]. These observations suggest that Listeria and Shigella have evolved a similar motility mechanism involving VASP recruitment directly in the case of Listeria and indirectly by an ActA analog, such as vinculin, in the case of Shigella. An alternative model was proposed in which vinculin head interaction with IcsA and tail interaction with actin filaments may serve as a link between the bacterial surface and the actin comet [131]. However, Shigella motility appears to be unaffected in a vinculindeficient murine cell line [139], raising the possibility that a functional homolog of vinculin is recruited by IcsA in this cell line. Vinculin does not seem to be involved in the nucleation step, but rather in elongation and stabilization of actin filaments (Fig. 9). Further studies are required in order to identify the precise role of vinculin in Shigella motility.

360


N-WASP belongs to the WASP-related proteins family [140-142], which includes WASP, the protein deficient in patients suffering of the Wiskott-Aldrich syndrome [143], the Dictyostelium SCAR protein [144], and its human homolog WAVE [145]. Both WASP and N-WASP interact with Cdc42Hs, but ectopic expression of these proteins generates different actin structures: WASP induces formation of cytoplasmic F-actin clusters [146], while N-WASP generates cortical actin polymerization [140]. In addition, N-WASP enhances Cdc42HsV12-induced formation of filopodia, while WASP antagonizes it [145]. WASP and N-WASP share several functional domains, including a verprolin-homology domain, a cofilin-homology domain, and an acidic domain. N-WASP may act as a severing or depolymerization factor of actin filaments. The acidic verprolin-cofilin (VCA) domain generates uncapped small actin filaments, and subsequent profilin-based elongation may induce polymerization of new actin filaments. The VCA domain is probably buried inside N-WASP, thus requiring the Cdc42Hs-N-WASP interaction to unmask it. Several lines of evidence suggest a role for N-WASP during the nucleation step of Shigella actin comet tail formation [132]. First, during Shigella infection, N-WASP is recruited to the bacterial pole. Recruitment of N-WASP requires interaction of the verprolin region of N-WASP with the glycin-rich repeats of IcsA (residues Rio3-Ala433). Second, Shigella motility is impaired in cells expressing an N-WASP form lacking four conserved residues of the cofilin-homology region, suggesting that this domain is required for comet tail formation. Third, immunodepletion of N-WASP in Xenopus extracts abolished bacterial motility, and add-back of recombinant N-WASP could only restore the nucleation step. Unexpectedly, N-WASP-mediated nucleation during Shigella motility does not seem to require Cdc42. Inactivation of Rho GTPases by TcdB toxin in infected cells or in Xenopus extracts has no effect on bacterial motility [99a]. Thus, it is likely that IcsA binding to the verprolin domain of N-WASP could itself unmask the VCA domain, thus bypassing the requirement for Cdc42Hs.

4.

CELL-TO-CELL SPREAD

When contact occurs between a moving organism and the inner face of the cytoplasmic membrane, a protrusion is formed that is phagocytosed by the adjacent cell [51, 147]. This process involves interaction with components of the cellular junction, allowing bacterial passage via an actin-driven protrusion into the adjacent cell. Expression of cadherins is a prerequisite to allowing phagocytosis of these protrusions by the adjacent cells [51]. Once absorbed by the adjacent cell, the bacteria are trapped inside a pocket surrounded by a double membrane, which is subsequently lysed, a process that needs the intervention of IcsB, a 57-kDa protein encoded by a gene located upstream from the ipa genes in the entry locus [148]. Because icsB mutants remain trapped inside large vacuoles, the

8.

361

SHIGELLOSIS

IcsB protein is likely to be required for membrane degradation, although this has not been direcdy established. In summary, in the context of an epithelial lining, the invasive phenotype of Shigella leads to an efficient process of intracellular colonization. It encompasses entry into nonphagocytic cells, escape to the cytoplasm, intracellular growth, intracellular motility, and cell-to-cell spread. This multifactorial process is a spectacular example of integradon of several defined steps leading to progression of infecdon in a sanctuary that is relatively protected from humoral and cellular effectors of the innate and adapdve immune response (Fig. 10). F.

Apoptotic Killing of Macrophages and Induction of the Release of Mature IL-ip by Shigella

Shigella are phagocytosed by macrophages, but, unlike their noninvasive isogenic mutants, the wild-type strains escape from the phagosome and induce apoptodc death of the host macrophage (Fig. 11), causing typical DNA fragmentadon after 2-3 hours of incubadon [55]. This observadon has been confirmed in vivo in infected Peyer's patches in the rabbit ligated-loop model. A background level of apoptosis is detected in this area when infection is carried out with a noninvasive mutant, whereas numerous apoptodc cells can be seen (Fig. 12) only if an invasive Shigella is used [149]. All invasive clinical isolates belonging to the different species of Shigella induce macrophage apoptosis [150], and the number of

Ipa^/Mxi-Spa

IcsB Ipa^/Mxi-Spa?

Fig. 10 Scheme of Shigella entry, escape into the cytoplasm, and intracellular and cell-to-cell spread, with reference to the effectors required for each step of invasion of a polarized epithelium.

362


Fig. 11 Shigella-induced apoptosis of J774 macrophages, (a) Invasive Shigella escaping into the cytoplasm and causing a typical process of apoptosis to its host macrophage after 2 hr of incubation. Note the shrinking nucleus with peripheral condensed chromatin and the formation of pits by the cytoplasmic membrane, preceding cellular fragmentation, (b) Noninvasive Shigella mutant phagocytosed by a J774 macrophage showing an intact host cell in spite of numerous intracellular bacteria and extended incubation period. Bacteria remain inside individual vacuoles. Reprinted through the courtesy of M. C. Prevost, A. Zychlinsky, and P. Sansonetti, Station Centrale de Microscopic Electronique, Institut Pasteur. Bar = 1 )im.

apoptotic cells appears significantly higher than the background level in the rectal mucosa of patients experiencing bacillary dysentery [151]. Apoptosis remains restricted to monocytes/macrophages and does not occur during epithelial cell infection [152].

8,

SHIGELLOSIS

363

Fig. 12 Evidence for apoptotic cells in the dome of lymphoid follicles infected by Shigella, (a) Infected macrophage at early stage of apoptosis. (b) Tingible bodies corresponding to phagocytosis of several apoptotic cells by a macrophage. Reprinted through the courtesy of M. C. Prevost, P. Gounon, and P. Sansonetti, Station Centrale de Microscopic Electronique, Institut Pasteur. Bars = I |Lim.

IpaB is required for apoptotic killing of macrophages, provided that the bacteria are phagocytosed and the vacuoles disrupted. The cytotoxicity of IpaB has been demonstrated both by a genetic approach [153] and, more direcdy, by microinjecting the purified protein into macrophages [154]. The mode of action of IpaB appears to reside essentially in its capacity to bind to Caspase-1 (ICE), the IL-ip-converting enzyme [155]. It belongs to the growing family of cysteine

364


proteases that act in a proteolytic cascade involving initiator and effector caspases, the latter achieving degradation of various substrates, eventually causing programmed cell death. All caspases are made as proenzymes and subsequently cleaved into a prodomain and two polypeptides that dimerize or tetramerize to form the active enzyme. Activation of these proteolytic cascades often responds to activation of cell surface receptors by their specific ligands (e.g., binding of Fas to FasL, TNF-a to TNF-Rl, or Apo3L to Apo3). This leads to activation of the initiating Caspase-8. Activation of caspases can also respond to the occurrence of internal damage, particularly in mitochondria, thus causing release of cytochrome c and activation of the initiating Caspase-9. However, Caspase-1 belongs to a group of proinflammatory caspases whose actual role appears to be cleavage of pro-IL-lp and IL-18. Its role in the induction of cell death is so far unknown, despite its homology with the Caenorhabditis elegans CED3 cell death protein [156]. Caspase-1 knockout mice develop normally and do not show any defect in the various processes that require apoptosis [157]. It is likely that the binding of IpaB on the Caspase-1 molecule [154, 158] activates its cryptic proapoptotic potential. The fact that Caspase-1 mediates IpaB-induced macrophage death is supported by the observation that the YVAD-CHO tetrapeptide, which inhibits Caspase-1 and closely related caspases, protects macrophages against Shigellainduced apoptosis [154]. Moreover, peritoneal macrophages from Caspase-1 knockout mice are not killed by invasive Shigella [158]. When activated by LPS, macrophages express cytokines such as TNF-a, IL-6, and IL-1. However, IL-ip remains intracellular as a promolecule until these macrophages are challenged by invasive Shigella. Shigella-infecied macrophages start producing and releasing large quantities of this strongly proinflammatory cytokine under its mature 17-kDa form [56, 158] before cell death occurs. Therefore, Shigella has evolved a remarkable strategy combining two potentials: (1) efficient killing of its front line predator, the macrophage, and (2) the use of this process to initiate inflammation, thereby destabilizing the intestinal tissue. This original strategy, compared to other pathways of programmed cell death, is summarized in Figure 13.

G. Activation of PMNs Also Reflects the Shigella Invasive Phenotype Unlike monocytes/macrophages, PMNs do not undergo apoptotic killing in the presence of Shigella. Moreover, in vitro, shigellae are unable to lyse the endocytic vacuole and are consequently killed [159]. A major effector of bacterial death is the bactericidal/permeability-increasing protein (BPI) [160]. This suggests that PMNs are the major phagocytic cells involved in the killing of Shigella. However, it has also been shown that infection of PMNs by invasive shigellae enhances their expression of adherence molecules and release of bactericidal granules [161]. Even though the outcome of the interaction is bacterial death, activation of

8.

365

SHIGELLOSIS

I

Fas/FasL

EC

ic

Alteration of Intracellular membranes and/or Mitochondria

C^^^

Apoptosis

^r

InilammatioiT Fig. 13 Molecular bases of IpaB induction of macrophage apoptosis and IL-1 [3-mediated initiation of inflammation. IpaB-mediated activation of the Caspase-1 pathway is compared to other established pathways of induction of apoptosis in cells.

PMNs is likely to enhance to the proinflammatory potential characteristic of the Shigella invasive phenotype, thus causing further tissue destruction.

X. Pathogenic Mechanisms: In Vivo Expression of the Invasive Phenotype After oral ingestion, Shigella progresses along the intestinal tract until it reaches the mucosal surface of the colon and rectum, which it eventually invades, causing inflammation and tissue destruction.

366


A. Does Shigella Express Organ-Specific Adhesins? S.flexneri carrying the virulence plasmid are more hydrophobic and adhere 10 times better to eukaryotic cells than plasmid-less mutants [162]. This observation cannot account for organ specificity, and the basis for colonic and rectal specificity of Shigella is currently unknown. Like almost all Enterobacteriaceae, Shigella harbors mannose-specific type I fimbriae [163], the expression of which can be induced in vitro by serial passages in aerobic conditions in liquid medium [164]. Type 1-piliated strains bind to isolated human colonic epithelial cells in a mannose-dependent manner [163]. Although no role has been ascribed to type 1 pili, one can speculate that they help survival in the environment, or may be involved in adherence to the colonic and rectal surfaces. Recent electron microscopic studies have shown that fresh clinical isolates oi Shigella can express at least two additional types of fimbriae [165]. Another study has shown that the binding of S. flexneri to colonic epithelial cells of guinea pigs could be inhibited by fucose, A^-acetyl neuraminic acid, and N-acetyl mannosamine [166]. Shigella may therefore express various carbohydrate-specific adhesins on its surface. Identification of an organ-specific adhesin has become a priority, and a strategy to identify such candidate molecule(s) needs to be defined. As described below, M cells, which are part of the follicular-associated epithelium, seem to represent the major entry route of Shigella. Lectins specific for M cells are known [167]. Previous work has also shown that lectins belonging to the guinea pig mucus could achieve bridging adherence between the 5. flexneri LPS and the glycosylated motives of mucus glycoproteins [168]. Finally, it is possible that this putative adhesin is expressed only in the intestinal lumen. In this case, methods such as signature-tagged transposon mutagenesis may be required in order to identify, in the in vivo context, mutants that have lost their colonizing capacities [169]. The tropism of Shigella for primates makes the experimental approach potentially extremely difficult.

B.

Crossing the Epithelial Lining by Shigella: Which Route is Best?

In vitro. Shigella is inefficient at invading an epithelial lining apically [54]. A similar situation is likely to prevail in vivo. There are potentially two ways for bacteria to reach the basolateral pole of the epithelium: (1) to cross the M cells of the follicular-associated epithelium (FAE) covering the mucosal lymph nodes associated with the mucosa, and (2) to take advantage of the inflammation, which disrupts epithelial integrity, to translocate and reach subepithelial tissues. Based on a combination of in vitro and in vivo models of Shigella infection, both mechanisms may occur, although the former may dominate during the early stages of invasion.

8.

SHIGELLOSIS

1.

367

TRANSLOCATION THROUGH THE FOLLICULAR-ASSOCIATED EPITHELIUM

Experiments carried out in macaque monkeys and in the rabbit ligated-intestinal-loop model have indicated that Shigella primarily translocated through the M cells of the FAE (Fig. 14) [48, 49, 119]. Once established in the dome region of the FAE, bacteria are phagocytosed by numerous macrophages present in this area. These macrophages quickly die (Fig. 12), many of them showing a typical pattern of apoptosis [49, 149]. A current hypothesis is that those macrophages that are permanently exposed to bacterial material from the intestinal flora may be in a chronic state of subactivation. In the presence of the cell death message they receive from invading shigellae, these macrophages release large amounts of IL-lp, which initiates inflammation [170]. In addition, during the first 4 hours of infection, the balance between the tissue concentration of IL-1 receptor antagonist (IL-lra) and IL-1 is tipped toward a very low ratio, reflecting poor expression of IL-lra [170a]. This enhances the proinflammatory capacity of Shigella. Infusion of IL-1 ra to rabbits prior to and during Shigella infection of ligated intestinal loops

Fig. 14 Entry of Shigella through M cells of the follicular-associated epithelium in a rabbit Peyer's patch. M and arrow point to the ruffle that has been elicited by the entering shigellae. E points to adjacent enterocytes. Reprinted through the courtesy of M. C. Prevost, J. Perdomo, and P. Sansonetti, Station Centrale de Microscopic Electronique, Institut Pasteur. Bar = 1 |Lim.

368


causes dramatic reduction of inflammation-mediated tissue destruction, thus confirming the major role played by IL-1 in pathogenesis of the disease [171]. 2.

INITIAL INFLAMMATION INCREASES PERMEABILITY OF THE EPITHELIAL BARRIER TO SHIGELLA

Transmigration of polymorphonuclear leukocytes through an epithelial lining loosens cellular junctions, thus facilitating Shigella entry. This has been shown both in vitro on human colonic epithelial cells cultured on filters and in vivo in the rabbit ligated-loop model of infection by Shigella [58, 172]. Inflammation may be initiated exclusively at the level of the follicular structures. It is also possible that the presence of Shigella on the apical side of the epithelium elicits transepithelial signaling sufficient to induce local transmigration of PMNs, thus causing the formation of multiple zones of bacterial invasion away from the follicular-associated epithelium. The vicious circle that involves crossing of the epithelial barrier and inflammation is summarized in Figure 15. More recent evidence indicates that expression of virulence plasmid genes is required to induce trafficking of neutrophils across polarized monolayers of the intestinal epithelium [173]. SepA, a 110-kDa surface protein, is encoded by the virulence plasmid. It is the protein that is most abundantly secreted by Shigella flexneri [ 174]. This protein has a signal peptide and is secreted in a way similar to IgA protease. Its secretion does not require the Mxi-Spa type III system. Even though one portion of this protein is very similar to the N-terminal portion of the IgA protease oi Haemophilus influenzae, SepA does not express an IgA-protease activity [175]. An sepPs. mutant produces weaker inflammation in the rabbit ligated-loop infection assay. While this is indicating a proinflammatory potential for this protease, no specific target has been identified so far for SepA. B. Intestinal Inflammation during Human Infection The first study on increased levels of cytokines elicited by shigellosis was performed on children from Sri Lanka infected with S. dysenteriae 1 [176]. These children showed elevated levels of IL-6 and TNF-a in their serum. Expression of all the pro- and antiinflammatory cytokines was observed after immunostaining of rectal biopsy sections of Bangladeshi patients developing shigellosis. However, severe cases were associated with higher levels of IL-1 p, IL-6, TNF-a, and IFN-y [34]. In another study, the plasma and stool levels of these cytokines, except for IFN-y, appeared elevated during the first 2 weeks of infection [35]. How these observations fit with the recent demonstration that IFN-y is essential for elimination of Shigella in a murine model of pneumonia remains to be established. Downregulation of IFN-a production at the early stage of infection may be an efficient strategy for Shigella to successfully resist and survive the initial response of the innate immune system [177]. These in vivo data are in general in agreement with in vitro data, although factors other than expression of the invasive phenotype may participate in the development of inflammation.

8.

369

SHIGELLOSIS

Shigella entry routes

J4

Fig. 15 General scheme summarizing the early steps of Shigella crossing of the follicular-associated epithelium and initiation of inflammation (i.e., mostly influx of PMNs), following induction of macrophage apoptosis, which destabilizes the architecture of the epithelial lining and enhances bacterial invasion, the process being subsequently amplified by the capacity of Shigella to spread from cell to cell.

XL Role of Chromosomally Encoded Genes in the Virulence of Shigella Aside from the armamentarium of genes encoded by the virulence plasmid, some chromosomal genes are also involved in virulence. They can be classified in regulatory genes such as virR, and structural gene-encoding factors such as LPS and toxins. A. Regulation of Plasmid Virulence Genes Shigella expresses an invasive phenotype when grown at 37°C, but not at 30°C [178]. Transcription of the ipa and mxi-spa genes is under the control of a dual

370


regulation system: a positive regulatory cascade involving two plasmid products—VirF and VirB—and a negative regulatory cascade involving a chromosomal regulator—VirR. The virF gene encodes a 30-kDa regulatory protein belonging to the AraC family of transcriptional activators and is required for expression of genes of the entry region, as well as for expression of icsA [179]. VirF operates essentially by controlling the transcription of virB, a gene encoded on the virulence plasmid, directly downstream of the ipa operon [180], which activates transcription of the ipa, mxi, and spa operons of the entry region. The screening of transposon-induced mutants for expression of entry genes at 30°C has led to identification of virR [181], a chromosomal gene that encodes the H-NS (HI) protein [182]. H-NS is a major component of a family of histone-like proteins, which control DNA supercoiling, thereby regulating expression of numerous genes [183]. Current evidence indicates that H-NS binds to the virB promoter at 30°C, thereby preventing its activation by VirF [184, 185]. Another chromosomal gene, vacB, regulates the invasive properties of Shigella [186]. A vacB mutant showed decreased expression of both icsA and ipaB gene products, despite unaltered transcription of these genes, indicating that vacB-mediated regulation may occur at the posttranscriptional level. In addition to temperature, osmolarity also modulates expression of invasion genes. Transcription of an mxi-lac gene fusion was enhanced in conditions of high osmolarity and reduced in tênvL and AenvZ-ompK mutants. These mutants are less invasive than their isogenic wild-type strain [187]. More recent evidence indicates that other regulatory systems exist. A set of plasmid-encoded proteins comprising VirA [188] and IpaH9.8 [189], which are secreted through the Mxi-Spa apparatus, are encoded by genes whose transcription depends on activation of this type III system, unlike transcription of the ipa genes, which does not depend on activation of protein secretion [190].

B.

Lipopolysaccharide

Smooth Shigella strains are considerably more virulent than rough mutants, thus indicating an essential role for LPS 0-sidechains in virulence. When the virulence plasmid and chromosomal sequences from S. flexneri were introduced into an E. coli K12 strain by conjugation, expression of a smooth LPS appeared essential for the exconjugants {E. coli K\2IShigella hybrids) to become virulent in in vivo conditions such as in the rabbit ligated-loop-infection assay and in the Sereny test [191]. This work confirmed previous data showing that rough Shigella mutants were consistently negative in the keratoconjunctivitis assay in guinea pigs and rabbits [192]. Although rough strains can invade epithelial cells in vitro, they are still impaired in their capacity to spread from cell to cell. This could be related to an indirect effect where the lack of 0-sidechains alters surface localization of

8. SHIGELLOSIS

371

IcsA, thereby preventing its polar localization, an important factor for optimal actin-driven motility [193]. Both mislocalization and insufficient proteolytic cleavage of IcsA may be taking place in the mutants lacking 0-sidechains. Aside from suffering alterations in their motility, rough mutants of Shigella may also be avirulent in in vivo assays due to their hypersensitivity to killing by complement [194]. The complement system protects the host against invading bacterial pathogens by generating opsonic, chemotactic, and lytic factors [195]. Its protein components are present in the serum, but also in inflammatory exudates, as it occurs in the mucosal tissues in the course of shigellosis. A recent study describes S. flexneri mutants with short 0-sidechains expressing a normal motiUty phenotype, but exhibiting susceptibility to serum-mediated kilUng [13].

C. Toxins Shigella produces various toxins, among which Shiga toxin is the best characterized [30]. Among the numerous species and serotypes of Shigella, only S. dysenteriae 1 produce this toxin. It is a holotoxin composed of five B subunits of 7 kDa each, and one A subunit of 32 kDa. The B subunit binds to the disaccharide Galal-4galp found in glycolipids such as the globotriaosylceramide Gbs. The A subunit is the toxic part of Shiga toxin that becomes enzymatically active on proteolytic cleavage, releasing a 27-kDa A1 fragment corresponding to the toxic moiety and a 4-kDa carboxy-terminal A2 portion [30]. Shiga toxin is an A^-glycosidase that cleaves adenine off one specific adenosine of the 28 S component of the 60 S ribosomal subunit, thus irreversibly destroying ribosomal functions [196]. Therefore, Shiga toxin is a potent inhibitor of protein biosynthesis, in both eukaryotic and prokaryotic systems [197, 198]. The function of Shiga toxin in eliciting diarrhea in the course of shigellosis due to S. dysenteriae 1 has not been convincingly demonstrated in humans, although its enterotoxicity in the rabbit ligated-loop model would indicate that it has this function [24]. As Gb3 is mainly detected in the nonepithelial fraction of colonic sections, it is not clear how the toxin could interact with the lumenal pole of epithelial cells to cause diarrhea. On the other hand, Shiga toxin is cytotoxic to intestinally derived epithelial cells in vitro [199] and could thereby participate in the cytotoxic effect to the epithelium. This could occur either by release of the toxin by extracellular or intracellular bacteria. In any event, considering the localization of Gb3, it is likely that the toxin has subepithelial targets. Macaque monkeys inoculated intragastrically either with a wild-type or with a stxA (knockout mutant lacking expression of the catalytic A subunit of Shiga toxin) mutant of 5. dysenteriae 1 developed dysentery. However, animals infected with the wild-type strain consistently showed the presence of blood in their dysenteric stools. In support of this difference in clinical symptoms, the histopathological analysis showed severe alterations of the capillaries in the lamina propria of

372


colonic tissues from animals infected with the toxin-producing strains [200]. These observations indicate that Shiga toxin may primarily behave as a toxin for the vascular endothelium, thereby adding a component of ischemic and hemorrhagic colitis to the dysentery caused by the invasive phenotype. Epidemiological observations confirm this finding, showing that the patients infected with S. dysenteriae 1 have more blood in their stools, compared to patients infected with other Shigella species [17]. The classical neurotoxic effect of Shiga toxin [201] is also likely to be due to brain capillary damage. Another severe condition, primarily observed in the context of S. dysenteriae 1 infection, is the hemolytic uremic syndrome (HUS). Although the pathogenesis of this acute (often fatal) renal failure is not fully understood, evidence points to Shiga toxin penetrating the blood vessels of the intestinal mucosa and reaching the kidneys via the bloodstream as a major etiological factor. The glycolipid Gb3 is found in high concentrations on human kidney endothelial cells, which are exquisitely sensitive to Shiga toxin [202]. Histopathological analysis shows thrombosis and destruction of the glomerular capillaries, and sometimes renal cortical necrosis [203]. It is possible that LPS acts in synergy with Shiga toxin to destroy renal blood vessels [204], Shigella produces toxins other than Shiga toxin. Two enterotoxins have recently been described that may account for the early diarrheal phase often observed during shigellosis. Shigella enterotoxin 1 is a chromosomally encoded, iron-dependent toxin of 55 kDa, mainly expressed by S.flexneri 2a [205]. Shigella enterotoxin 2 is a plasmid-encoded protein of 63 kDa [206].

D.

Other Virulence Factors

Other virulence factors encoded by the chromosome include the OmpC porin [207]. Also, all Shigella strains express siderophores and their corresponding outer membrane receptors. Siderophores are low-molecular-weight molecules with high affinity for ferric iron that they can extract from their physiological carriers, transferrin and lactoferrin. Most isolates of S. flexneri and S. sonnei express aerobactin siderophores, whereas 5". dysenteriae and 5". sonnei produce enterochelin siderophores. Some isolates can express both types [208]. Aerobactin-negative mutants are impaired in their growth in tissues but not inside cells [209]. Last, but not least, the interesting concept of "black holes" was presented [9] in 1998, according to which portions of the bacterial chromosome encoding enzyme that may affect pathogenicity have been deleted as the pathogenic species evolves toward maximum fit with the host. For example, lysine decarboxylase (LDC), which is present in more than 90% of E. coli isolates, is absent in Shigella and enteroinvasive E. coli. When cadA, the gene encoding lysine decarboxylase.

8. SHIGELLOSIS

373

was introduced into Shigella, invasiveness and enterotoxicity were both affected. Cadaverin, a byproduct of lysine decarboxylase, was identified as the inhibitor. Moreover, the chromosomal region encompassing cadA was shown to be deleted in Shigella and entero-invasive E. coli. Similarly, the E. coli ompTgQne attenuates virulence in Shigella, which normally lacks ompT [210].

XIL Conclusions Thanks to a combination of in vitro and in vivo approaches, the pathogenesis of shigellosis is progressively unraveling its secrets. The major contributions in this area over the last two decades have been: (1) the establishment of the genetic basis for cell invasion and its regulation by environmental cues such as temperature, (2) the description of the infectious cycle of the bacteria in epithelial cells and the recognition of the molecular crosstalks accounting for entry, intracellular movement, and cell-to-cell spread, (3) the description of apoptotic killing of macrophages with its dual implications for bacterial survival and elicitation of inflammation, and (4) the description of the role of inflammation in facilitating mucosal invasion. These concepts have also led to the development of a series of promising live attenuated vaccine candidates against shigellosis. Several questions remain open, such as: (1) understanding the colonic specificity of Shigella; (2) understanding the dramatically high infectiousness of the pathogen (i.e., 100 cfu absorbed orally); (3) deciphering the signaling pathways that, on expression of the invasive phenotype, lead either to entry into epithelial cells or to programmed cell death of macrophages; (4) confirming that translocation through the epithelium occurs essentially via M cells, since inflammation occurring at distant sites may also facilitate entry; (5) identifying the specificities of the signaling cascades that cause the particularly severe inflammation observed during shigellosis; and (6) understanding the bases of immune protection against the disease and developing live oral or subunit parenteral vaccines.

Acknowledgments We wish to thank Colette Jacquemin for outstanding editorial work on this manuscript. We also wish to express our gratitude to all past and present members of Unite de Pathogenie Microbienne Moleculaire whose work made a significant part of this manuscript possible.

374


References 1. Kaper, J. B. (1998). Enterohemorrhagic Escherichia coli. Cum Opin. Microbiol. 1, 103-108. 2. Finlay, B, B., and Falkow, S. (1997). Common themes in microbial pathogenicity revisited. Microbiol. Mol. Biol. Rev. 61, 136-169. 3. Takeuchi, A., Spring, H., LaBrec, E. H., and Formal, S. B. (1965). Experimental bacillary dysentery: An electron microscopic study of the response of intestinal mucosa to bacterial invasion. Am. J. Pathol. 4, 1011-1044. 4. Kotloff, K. L., Winickoff, J. P., Ivanoff, B., Clemens, J. D., Swerdlow, D. L., Sansonetti, P. J., Adak, G. K., and Levine, M. M. (1999). Global burden of Shigella infections: Implication for vaccine development and implementation of control strategies. WHO Bull. 77, 651-666. 5. Brenner, D. J., Fanning, G. R., Miklos, G. V., and Steigerwalt, A. G. (1973). Polynucleotide sequence relatedness among Shigella species. Int. J. Syst. Bacteriol. 23, 1-7. 6. Luria, S. E., and Burrows, J. W. (1957). Hybridization between Escherichia coli and Shigella. J. Bacteriol. 47, 461-476. 7. Tominaga, A., Mahmoud, M. A. H., Mukaihara, T., and Enomoto, M. (1994). Molecular characterization of intact, but cryptic, flagellin genes in the genus Shigella. Mol. Microbiol. 12, 277-285. 8. Giron, J. A. (1995). Expression of flagella and motility by Shigella. Mol. Microbiol. 18, 63-75. 9. Maurelli, A. T., Fernandez, R. E., Bloch, C. A., Rode, C. K., and Fasano, A. (1998). "Black holes" and bacterial pathogenicity: A large genomic deletion that enhances the virulence of Shigella spp. and enteroinvasive Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 95, 3943-3948. 10. DuPont, H. L., Levine, M. M., Hornick, R. B., and Formal, S. B. (1989). Inoculum size in shigellosis and implications for expected mode of transmission. J. Infect. Dis. 159, 1126-1128. 11. Simmons, D. A. R., and Romanowska, E. (1987). Structure and biology of Shigella flexneri O antigens. / Med. Microbiol. 23, 289-302. 12. Brahmbhatt, H. N., Lindberg, A. A., and Timmis, K. N. (1992). Shigella lipopolysaccharide: Structure, genetics, and vaccine development. Cum Topics Microbiol. Immunol. 180, 45-64. 13. Hong, M., and Payne, S. M. (1997). Effect of mutations in Shigella flexneri chromosomal and plasmid-encoded lipopolysaccharide genes on invasion and serum resistance. Mol. Microbiol. 24,779-791. 14. Watanabe, H., and Timmis, K. N. (1984). A small plasmid in Shigella dysenteriae 1 specifies one or more functions essential for O antigen production and bacterial virulence. Infect. Immun. 43, 391-396. 15. Sansonetti, P. J., Kopecko, D. J., and Formal, S. B. (1982). Involvement of a large plasmid in the invasive ability of Shigella flexneri. Infect. Immun. 35, 852-860. 16. Sansonetti, P. J., David, M,, and Toucas, M. (1980). Correlation entre la perte d'ADN plasmidique et le passage de la phase I virulente a la phase II avirulente chez Shigella sonnei. CR. Acad. Sci. 290, 879-882. 17. StoU, B. J., Glass, R. I., Huq, M. I., Khan, M. U., Banu, H., and Holt, J. (1982). Epidemiologic and clinical features of patients infected with Shigella who attended a diarrheal disease hospital in Bangladesh. J. Infect. Dis. 146, 177-183. 18. Ferreccio, C , Prado, V., Ojeda, A., Cayyazo, M., Abrego, P., Guers, L., and Levine, M. M. (1991). Epidemiologic patterns of acute diarrhea and endemic Shigella infections in children in a poor periurban setting in Santiago, Chile. Am. J. Epidemiol. 134, 614-627. 19. Bennish, M. L., and Wojtyniak, B. J. (1991). Mortality due to shigellosis: Community and hospital data. Rev. Infect. Dis. 13, S245-S251.

8.

SHIGELLOSIS

375

20. Levine, O. S., and Levine, M. M. (1991). Houseflies (Musca domestica) as mechanical vectors of shigellosis. Rev. Infect. Dis. 13, 688-696. 21. Tauxe, R. V., McDonald, R. C , Hargrett-Bean, N., and Blake, R A. (1988). The persistence of Shigella flexneri in the United States: Increasing role of adult males. Am. J. Public Health 78, 1432-1435. 22. Egoz, N., Shmilovitz, M., Kretzer, B., Lucian, M., Porat, V., and Raz, R. (1991). An outbreak of Shigella sonnei infection due to contamination of a municipal water supply in Northern Israel. /. Infect. 22, 87-93. 23. DuPont, H. L., Homick, R. B., Snyder, M. J., Libonati, J. P, Formal, S. B., and Gangarosa, E. J. (1972). Immunity in shigellosis, II: Protection induced by oral live vaccination or primary infection. / Infect. Dis. 125, 12-16. 24. Buder, T, Speelman, P., Kabir, I., and Banwell, J. (1986). Colonic dysfunction during shigellosis. J. Infect. Dis. 154, 817-824. 25. Guerrero, L., Calva, J. J., Morrow, A. L., Velazquez, F. R., Tuz-Dzib, F., Lopez-Vidal, Y., Ortega, H., Arroyo, H., Cleary, T. G., Pickering, L. K., and Ruiz-Palacios, G. M. (1994). Asymptomatic Shigella infections in a cohort of Mexican children younger than two years of age. Pediatr. Infect. Dis. J. 13, 597-602 26. DuPont, H. L., Homick, R. B., Dawkins, A. T., Snyder, M. J., and Formal, S. B. (1969). The response of man to virulent Shigella flexneri 2a. J. Infect. Dis. 119, 296-299. 27. Struelens, M. J., Patte, D., Kabir, I., Salam, A., Nath, S. K., and Bulier, T. (1985). Shigella septicemia: Prevalence, presentation, risk factors, and outcome. J. Infect. Dis. 152, 784-790. 28. Bennish, M. L., Azad, A. K., Rahman, O., and Phillips, R. E. (1990). Hypoglycemia during diarrhea in childhood. New Engl. J. Med. 322, 1357-1363. 29. Bennish, M. L. (1991). Potentially lethal complication of shigellosis. Rev. Infect. Dis. 13, S319-S324. 30. O'Brien, A. D., Tesk, V. L., Donohue-Rolfe, A., Jackson, M. P, Olsnes, S., Sandwig, K., Lindberg, A. A., and Keusch, G. T. (1992). Shiga toxin: Biochemistry, genetics, mode of action, and role in pathogenesis. Curr. Topics Microbiol. Immunol. 180, 65-94. 31. Karpman, D., Council, H., Svensson, M., Schentz, F., Aim, P, and Svanborg, C. (1997). The role of lipopolysaccharide and Shiga toxin in a mouse model of Escherichia coli 0157:H7 infection. /. Infect. Dis. 175, 611-620. 32. Butler, T, Islam, M. R., and Bardham, P. K. (1984). The leucemoid reaction in shigellosis. Am. J. Dis. Child. 138, 162-165. 33. Phalipon, A., Kaufmann, M., Michetti, P., Cavaillon, J. M., Huerre, M., Sansonetti, P. J., and Kraehenbuhl, J. P. (1995). Monoclonal immunoglobulin A antibody directed against serotypespecific epitope of Shigella flexneri lipopolysaccharide protects against murine experimental shigellosis. J. Exp. Med. 182, 769-778. 34. Raqib, R., Lindberg, A. A., Wretlind, B., Bardhan, P. K., Andersson, U., and Andersson, J. (1995). Persistance of local cytokine production in shigellosis in acute and convalescent stages. Infect. Immun. 63, 289-296. 35. Raqib, R., Wredind, B., Andersson, J., and Lindberg, A. A. (1995). Cytokine secretion in acute shigellosis is correlated to disease activity and directed more to stool than to plasma. / Infect. Dis. Ill, 376-384. 36. Speelman, P., Kabir, I., and Islam, M. (1984). Distribution and spread of colonic lesions in shigellosis: A colonoscopic study. J. Infect. Dis. 150, 899-903. 37. Mathan, M. M., and Mathan, V. I. (1991). Morphology of rectal mucosa of patients with shigellosis. Rev Infect. Dis. 13, S314-S318. 38. Mathan, M. M., and Mathan, V. I. (1986). Ultrastructural pathology of the rectal mucosa in Shigella dysentery. Am. J. Pathol. 123, 25-38.

376


39. Sereny, B. (1955). Experimental Shigella conjunctivitis. Acta Microbiol. Acad. Sci. Hung. 2, 293-295. 40. Piechaud, M., Sztuim-Rubinsten, S., and Piechaud, D. (1958). Evolution histologique de la kerato-conjonctivite a bacilles dysenteriques du cobaye. Ann. Inst. Pasteur 94, 298-309. 41. Watkins, H. M. S. (1960). Host-parasite relationship in shigellosis: Some attributes of virulence in Shigella. Ann. N.Y. Acad. Sci. 88, 1167-1187. 42. LaBrec, E. H., Schneider, H., Magnani, T. J., and Formal, S. B. (1964). Epithelial cell penetration as an essential step in the pathogenesis of bacillary dysentery. J. Bacteriol. 88, 1503-1518. 43. Voino-Yasenetsky, M. V., and Voino-Yasenetskaya, M. K. (1961). Experimental pneumonia caused by bacteria of the Shigella group. Acta Morphol. 11, 440-454. 44. Mallett, C , Van de Verg, L., Collins, H., and Hale, T. L. (1993). Evaluation of Shigella vaccine safety in an intranasal challenge mouse model. Vaccine 11, 190-196. 45. Way, S. S., Borcznk, A. C , Dominitz, R., and Goldberg, M. B. (1998). An essential role for gamma interferon in innate resistance to Shigella flexneri infection. Infect. Immun. 66, 1342-1348. 46. Sansonetti, R J., Arondel, J., Cavaillon, J. M., and Huerre, M. (1995). Role of IL-1 in the pathogenesis of experimental shigellosis. J. Clin. Invest. 96, 884-892. 47. Sansonetti, P. J., Arondel, J., Huerre, M., Harada, A., and Matsushima, K. (1999). Interleukin-8 controls bacteiial transepithelial translocation at cost of epithelial destruction in experimental shigellosis. Infect. Immun. 67, 1471-1480. 48. Wassef, J., Keren, D. P., and Mailloux, J. L. (1989). Role of M cells in initial bacterial uptake and in ulcer formation in the rabbit intestinal loop model in shigellosis. Infect. Immun. 57, 858-863. 49. Sansonetti, P. J., Arondel, J., Cantey, R. J., Prevost, M. C , and Huerre, M. (1996). Infection of rabbit Peyer's patches by Shigella flexneri: Effect of adhesive or invasive bacterial phenotypes on follicular-associated epithelium. Infect. Immun. 64, 2752-2764. 50. Oaks, E. V., Hale, T. L., and Formal, S. B. (1986). Serum immune response to Shigella protein antigens in rhesus monkeys and humans infected with Shigella spp. Infect. Immun. 53, 57-63. 51. Sansonetti, P. J., Mounier, J., Prevost, M. C , and Mege, R. M. (1994). Cadherin expression is required for the spread of Shigella flexneri between epithelial cells. Cell 76, 829-839. 52. Goldberg, M. B., and Sansonetti, P. J. (1993). Shigella subversion of the cellular cytoskeleton: A strategy for epithelial colonization. Infect. Immun. 61, 4941-4946. 53. Cossart, P., Boquet, P., Normark, S., and Rappuoli, R. (1996). Cellular microbiology emerging. Sc/ewc^ 271(5247), 315-316. 54. Mounier, J., Vasselon, T., Hellio, R., Lesourd, M., and Sansonetti, P. J. (1992). Shigella flexneri enters human colonic Caco-2 epithelial cells through their basolateral pole. Infect. Immun. 60, 237-248. 55. Zychlinsky, A., Prevost, M. C , and Sansonetti, P. J. (1992). Shigella flexneri induces apoptosis in infected macrophages. Nature 358, 167-169. 56. Zychlinsky, A., Fitting, C , Cavaillon, J. M., and Sansonetti, P. J. (1994). Interleukin-1 is released by macrophages during apoptosis induced by Shigella flexneri. J. Clin. Invest. 94, 1328-1332. 57. Renesto, P., Mounier, J., and Sansonetti, P. J. (1996). Induction of adherence and degranulation of polymorphonuclear leukocytes: A new expression of the invasive phenotype of Shigella flexneri. Infect. Immun. 64, 719-723. 58. Perdomo, J. J., Gounon, P., and Sansonetti, P. J. (1994). Polymorphonuclear leukocyte transmigration promotes invasion of colonic epithelial monolayer by Shigella flexneri. J. Clin. Invest. 93, 633-643.

8.

SHIGELLOSIS

377

59. Beatty, W. L., and Sansonetti, P. J. (1997). Role of lipopolysaccharide in signaling to subepithelial polymorphonuclear leukocytes. Infect. Immun. 65, 4395-4404. 60. Sansonetti, P. J. (1991). Genetic and molecular basis of epithelial cell invasion by Shigella species. Rev. Infect. Dis. 13, S285-S292. 61. Hale, T. L. (1991). Genetic bases of virulence in Shigella species. Microbiol. Rev. 55, 206-224. 62. Parsot, C., and Sansonetti, P. J. (1996). Invasion and the pathogenesis of Shigella infections. Curr. Topics Microbiol. Immunol. 209, 25-42. 63. Menard, R., Dehio, C., and Sansonetti, P. J. (1996). Bacterial entry into epithelial cells: The paradigm of Shigella. Trends Microbiol. 4, 220-226. 64. Tran Van Nhieu, G., and Sansonetti, P. J. (1998). In "Recent Advances in the Pathogenesis of Gastrointestinal Bacterial Infections" (P. Rampal and P. Boquet, eds.), pp. 83-95. John Libbey Eurotext, Paris. 65. Sasakawa, C., Buysse, J. M., and Watanabe, H. (1992). The large virulence plasmid oi Shigella. In "Pathogenesis of Shigellosis: Current Topics in Microbiology and Immunology" (P. J. Sansonetti, ed.), pp. 1\-AA. Springer-Verlag, Berlin-Heidelberg. 66. Clerc, P., and Sansonetti, P. J. (1987). Entry oi Shigella flexneri into HeLa cells: Evidence for directed phagocytosis involving actin polymerization and myosin accumulation. Infect. Immun. 55,2681-2688. 67. Sansonetti, P. J., Ryter, A., Clerc. P., Maurelli, A. T, and Mounier, J. (1986). Multiplication of Shigella flexneri within HeLa cells: Lysis of the phagocytic vacuole and plasmid-mediated contact hemolysis. Infect. Immun. 51, 461^69. 68. Makino, S., Sasakawa, C, Kamata, K., Kurata, T, and Yoshikawa, M. (1986). A genetic determinant required for continuous reinfection of adjacent cells on large plasmid in S. flexneri 2a. C^//46, 551-555. 69. Bemardini, M. L., Mounier, J., d'Hauteville, H., Coquis-Rondon, M., and Sansonetti, P. J. (1989). Identification of/cÂ, a plasmid locus of Shigella flexneri which governs bacterial intraand intercellular spread through interaction with F-actin. Proc. Natl. Acad. Sci. U.S.A. 86, 3867-3871. 70. Isberg, R. R. (1991). Discrimination between intracellular uptake and surface adhesion of bacterial pathogens. Science 252, 934-938. 71. Tran Van Nhieu, G., and Isberg, R. R. (1993). Bacterial internalization mediated by beta 1 chain integrins is determined by ligand affinity and receptor density. EMBO J. 12, 1887-1895. 72. Mengaud, J., Ohayon, H., Gounon, P., Mege, R. M., and Cossart, P. (1995). E-cadherin is the receptor for internalin, a surface protein required for entry of L. monocytogenes into epithelial cells. Cell 84, 923-932. 73. Francis, C. L., Ryan, T. A., Jones, B. D., Smith, S. J., and Falkow, S. (1993). Ruffles induced by Salmonella and other stimuli: Direct micropinocytosis of bacteria. Nature 364, 639-642. 74. Adam T, Arpin, M., Prevost, M. C , Gounon, P, and Sansonetti, P J. (1995). Cytoskeletal rearrangements and the functional role of T-plastin during entry of Shigella flexneri into HeLa cells. J. Cell Biol. 129, 367-381. 75. Maurelli, A. T, Baudry, B., d'Hauteville, H., Hale, T. L., and Sansonetti, R J. (1985). Cloning of plasmid DNA sequences involved in invasion of HeLa cells by Shigella flexneri. Infect. Immun. 49, 164-171. 76. Sasakawa, C , Kamata, K., Sakai, T, Makino, S. I., Yamada, M., Okada, N., and Yoshikawa, M. (1988). Virulence associated genetic regions comprising 30 kilobases of the 230-kilobase plasmid in Shigella flexneri 2a. J. Bacteriol. 170, 2480-2484. 77. Venkatesan, M., Buysse, J. M., and Kopecko, D. J. (1988). Characterization of invasion plasmid antigen (ipaBCD) genes from Shigella flexneri: DNA sequence analysis and control of gene expression. Proc. Natl. Acad. Sci. U.S.A. 85, 9317-9321.

378


78. Andrews, G. P., Horomockyj, A. E., Coker, C , and Maurelli, A. T. (1991). Two novel virulence loci, mxiA and mxiB, in Shigella flexneri 2a faciliate excretion of invasion plasmid antigens. Infect. Immun. 59, 1997-2005. 79. Venkatesan, M. M., Buysse, J. M., and Oaks, E. V. (1992). Surface presentation of Shigella flexneri invasion plasmid antigen requires the products of the spa locus. J. Bacteriol. 174, 1990-2001. 80. Sasakawa, C , Komatsu, K., Tobe, T, Suzuki, T., and Yoshikawa, M. (1993). Eight genes in region 5 that form an operon are essential for invasion of epithelial cells by Shigella flexneri 2a. J. Bacteriol. 175, 2334-2346. 81. Allaoui, A., Sansonetti, P. J., and Parsot, C. (1993). MxiD: An outer membrane protein necessary for the secretion of the Shigella flexneri Ipa invasion. Mol. Microbiol. 7, 59-68. 82. Parsot, C , and Sansonetti, P. J. (1996). Invasion and the pathogenesis of Shigella infections. Curr Topics Microbiol. Immunol. 209, 2 5 ^ 2 . 83. Menard, R., Sansonetti, P. J., and Parsot, C. (1994). The secretion of the Shigella flexneri Ipa invasins is induced by the epithelial cell and controlled by IpaB and IpaD. EMBO J. 13, 5293-5302. 84. Van Gisjegem, F., Genin, S., and Boucher, C. (1993). Conservation of secretion pathways for pathogenicity determinants of plant and animal bacteria. Trends Microbiol. 1, 175-180. 85. Groisman, E. A., and Ochman, H. (1993). Cognate gene clusters govern invasion of host epithelial cells by Salmonella typhimurium and Shigella flexneri. EMBO J. 12, 3779-3787. 86. Rosqvist, R., Hakansson, S., Forsbery, A., and Wolf-Watz, H. (1995). Functional conservation of the secretion and translocation machinery for virulence proteins of yersiniae, salmonellae and shigellae. EMBO J. 14, 4187-4195. 87. Comelis, G., and Wolf-Watz, H. (1997). The Yersinia Yop virulon: A bacterial system for subverting eukaryotic cells. Mol. Microbiol. 23, 861-867. 88. Mounier, J., Bahrani, F. K., and Sansonetti, P. J. (1997). Secretion of Shigella flexneri Ipa invasins upon contact with epithelial cells is growth-stage dependent and starts in vicinity of the bacterial septation furrow. Infect. Immun. 65, 774-782. 89. Watarai, M., Tobe, T, Yoshikawa, M., and Sasakawa, C. (1995). Contact of Shigella with host cells triggers release of Ipa invasins and is an essential function of invasiveness. EMBO J. 14, 2461-2470. 90. Bahrani, F. K., Sansonetti, P. J., and Parsot, C. (1997). Secretion of Ipa proteins by Shigella flexneri: Inducing molecules and kinetics of activation. Infect. Immun. 65, 4(X)5^010. 91. Parsot, C , Menard, R., Gounon, P., and Sansonetti, P. J. (1995). Enhanced secretion through the Shigella flexneri Mxi-Spa translocon leads to assembly of extracellular proteins into macromolecular structures. Mol. Microbiol. 16, 291-300. 92. Menard, R., Sansonetti, P. J., Parsot, C , and Vasselon, T. (1994). Extracellular association and cytoplasmic partitioning of the IpaB and IpaC invasins oi Shigella flexneri. Cell 79, 515-525. 93. Menard, R., Sansonetti, P. J., and Parsot, C. (1993). Nonpolar mutagenesis of the ipa genes defines IpaB, IpaC and IpaD as effectors of Shigella flexneri entry into epithelial cells. J. Bacteriol. 175, 5899-5906. 94. Tran Van Nhieu, G., Ben Ze'ev, A., and Sansonetti, P. J. (1997). Modulation of bacterial entry in epithelial cells by association between vinculin and the Shigella IpaA invasin. EMBO J. 16, 2717-2729. 94a. Blocker, A., Gounon, P., Larquet, E., Niebuhr, K., Cabiaux, V., Parsot, C , and Sansonetti, P. J. (1999). Role of Shigella's type III secretion system in insertion of IpaB and IpaC into the host membrane. / Cell Biol. 147, 683-693.

8.

SHIGELLOSIS

379

95. Menard, R., Prevost, M. C , Gounon, P., Sansonetti, P J., and Dehio, C. (1996). The secreted Ipa complex of Shigella flexneri promotes entry into mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 93, 1254-1258. 96. Watarai, M., Funato, S., and Sasakawa, C. (1996). Interaction of Ipa proteins of Shigella flexneri with alpha5betal integrin promotes entry of the bacteria into mammalian cells. J. Exp. Med. 183, 991-999. 96a. Skoudy, A., Tran Van Nhieu, G., Mantis, N., Arpin, M., Mounier, J., Gounon, P., and Sansonetti, P. J. (1999). A functional role for Ezrin during Shigella entry into epithelial cells. J. Cell Sciences 112, 2059-2069. 97. de Geyter, C., Vogt, B., Benjelloun-Touimi, Z., Sansonetti, P. J., Ruysschaert, J.-M., Parsot, C , and Cabiaux, V. (1997). Interaction of IpaC, a protein involved in entry of S. flexneri into epithelial cells, with lipid membranes. FEBS Lett. 400, 149-154. 97a. Blocker, A., et al. (2000). Manuscript in preparation. 98. Adam, T., Giry, M., Boquet, P., and Sansonetti, P. J. (1996). Rho-dependent membrane folding causes Shigella entry into epithelial cells. EMBO J. 15, 3315-3321. 99. Watarai, M., Kamata, Y., Kozaki, S., and Sasakawa, C. (1997). rho, a small GTP-binding protein, is essential for Shigella invasion of epithelial cells. J. Exp. Med. 185, 281-292. 99a. Mounier, J., Laurent, V., Hall, A., Fort, P, Carlier, M.-F., Sansonetti, P J., and Egile, C. (2000). Rho family GTPases control entry of Shigella flexneri into epithelial cells but not intracellular motility. J. Cell Sciences 112, 2069-2080. 100. Dehio, C , Prevost, M. C , and Sansonetti, P. J. (1995). Invasion of epithelial cells by Shigella flexneri induces tyrosine phosphorylation of cortactin by a pp60^'''' mediated signalling pathway. EMBO J. 14,2471-2482. 101. Nobes, C. D., and Hall, A. (1995). GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia and filopodia. Cell 81, 53-62. 102. Ridley, A. J., and Hall, A. (1992). The small GTP-binding protein Rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70, 389-399. 103. Chen, L. M., Hobbie, S., and Galan, J. E. (1996). Requirement of Cdc42 for Salmonella-inductd cytoskeletal and nuclear responses. Science 274, 2115-2118. 104. Dumenil, G., Olivo, J. C., Pellegrini, S., Fellous, M., Sansonetti, P J., and Tran Van Nhieu, G. (1998). Interferon a inhibits a Src-mediated pathway necessary for Shigella-'mduccd cytoskeletal rearrangements in epithelial cells. J. Cell Biol. 143, 1-10. 105. Finchan, V. J., Unlu, M., Bruntou, V. G., Pitts, J. D., Wyke, J. A., and Frame, M. C. (1996). Translocation of src kinase to the cell periphery is mediated by the actin cytoskeleton under the control of the rho family of small G proteins. J. Cell Biol. 135, 1551-1564. 106. Franck, Z., Gary, R., and Bretscher, A. (1993). Moesin, like ezrin, colocalizes with actin in the cortical cytoskeleton in cultured cells, but its expression is more variable. J. Cell Sciences 105, 219-231. 107. Menard, R., Dehio, C., and Sansonetti, P. J. (1996). Bacterial entry into epithelial cells: The paradigm of Shigella. Trends Microbiol. 4, 220-226. 108. Craig, S. W., and Johnson, R. P. (1996). Assembly of focal adhesions: Progress, paradigms, and portents. Curr Opin. Cell Biol. 8, 74-85. 109. High, N., Mounier, J., Prevost, M. C , and Sansonetti, P J. (1992). IpaB of Shigella flexneri causes entry into epithelial cells and escape from the phagocytic vacuole. EMBO J. 11, 1991-1999. 109a. de Geyter, C , Vogt, B., Benjelloun-Touimi, Z., Sansonetti, P J., Ruysschaert, J.-M., Parsot, C , and Cabiaux, V. (1997). Interaction of IpaC, a protein involved in entry of S. flexneri into epithelial cells, with lipid membranes. FEBS Lett. 400, 149-154.

380


110. Garcia del Portillo, F, Zwick, M. B., Leung, K. Y., and Finlay, B. B. (1993). Salmonella induces the formation of filamentous structures containing lysosomal membrane glycoproteins in epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 90, 10544-10548. 111. Vasselon, T., Mounier, J., Hellio, R., and Sansonetti, P. J. (1992). Movement along actin filaments of the perijunctional area and de novo polymerization of cellular actin are required for Shigella flexneri colonization of epithelial Caco-2 cell monolayers. Infect. Immun. 60, 1031-1040. 112. Theriot, J. A. (1995). The cell biology of infection by intracellular bacterial pathogens. Annu. Rev. Cell. Dev. Biol. 11, 213-239. 113. Dramsi, S., and Cossart, P. (1998). Intercellular pathogens and actin cytoskeleton. Annu. Rev. Cell. Dev. Biol. 14, 137-166. 114. Ogawa, H., Nakamura, A., and Nakaya, R. (1968). Cinemicrographic study of tissue cell cultures infected with Shigella flexneri. Jpn. J. Med. Sci. Biol. 21, 259-273. 115. Theriot, J. A., Mitchison, T. J., Tilney, L. G., and Portnoy, D. A. (1992). The rate of actin-based motility of intracellular Listeria monocytogenes equals the rate of actin polymerization. Nature 357, 257-260. 116. Zeile, W. L., Purich, D. L., and Southwick, F. S. (1996). Recognition of two classes of oligoproline sequences in profilin-mediated acceleration of actin based Shigella motility. J. Cell Biol. 133, 49-59. 117. Goldberg, M. B., Theriot, J. A., and Sansonetti, P. J. (1994). Regulation of surface presentation of IcsA, a Shigella protein essential to intracellular movement and spread, is growth phase dependent. Infect. Immun. 62, 5664-5668. 117a. Gouin, E., Gantelet, H., Egile, C., Lasa, I., Ohayon, H., Villiers, V., Gounon, P, Sansonetti, P J., and Cossart, P. (1999). A comparative study of the actin-based motilities of the pathogenic bacteria Listeria monocytogenes. Shigella flexneri and Rickettsia conorii. J. Cell Sciences 112, 1697-1708. 118. Lett, M. C , Sasakawa, C , Okada, N., Sakai, T, Makino, S., Yamada, M., Komatsu, K., and Yoshikawa, M. (1988). virG, a plasmid-coded virulence gene of Shigella flexneri: Identification of the VirG protein and determination of the complete coding sequence. J. Bacteriol. Ill, 353-359. 119. Sansonetti, P. J., Arondel, J., Fontaine, A., d'Hauteville, H., and Bemardini, M. L. (1991). ompB (osmo-regulation) and icsA (cell-to-cell spread) mutants of Shigella flexneri: Vaccine candidates and probes to study the pathogenesis of shigellosis. Vaccine 9, 416-422. 120. Suzuki, T, Lett, M. C , and Sasakawa, C. (1995). Extracellular transport of VirG protein in Shigella. J. Biol. Chem. 270, 30874-30880. 121. Goldberg, M. B., Barzu, O., Parsot, C , and Sansonetti, P. J. (1993). Unipolar localization and ATPase activity of IcsA, a Shigella flexneri protein involved in intracellular movement. Infect. Immun. 175,2189-2196. 122. Kocks, C , Gouin, E., Tabouret, M., Berche, P, Ohayon, H., and Cossart, P (1992). Listeria monocytogenes-'mducQd actin assembly requires the actA gene product, a surface protein. Cell 68,521-531. 123. Vazquez-Boland, J.-A., Kocks, C , Dramsi, S., Ohayon, H., Geoffroy, C , Mengaud, J., and Cossart, P. (1992). Nucleotide sequence of the lecithinase operon of Listeria monocytogenes and possible role of lecithinase in cell-to-cell spread. Infect. Immun. 60, 219-230. 124. Fukuda, I., Suzuki, T, Munakata, H., Hayashi, N., Katayama, E., Yoshikawa, M., and Sasakawa, C. (1995). Cleavage of Shigella surface protein VirG occurs at a specific site, but the secretion is not essential for intracellular spreading. J. Bacteriol. Ill, 1719-1726.

8.

SHIGELLOSIS

381

125. d'Hauteville, H., and Sansonetti, P. J. (1992). Phosphorylation of IcsA by cAMP-dependent protein kinase and its effect on intercellular spread of Shigella flexneri. Mol. Microbiol 6, 833-841. 126. d'Hauteville, H., Dufourcq-Lagelouse, R., Nato, F., and Sansonetti, P. J. (1996). Lack of cleavage of IcsA in Shigella flexneri causes aberrant movement and allows demonstration of a cross reactive eukaryotic protein. Infect. Immiin. 64, 511-517. 127. Egile, C , d'Hauteville, H., Parsot, C , and Sansonetti, P. J. (1997). SopA, the outer membrane protease responsible for polar localization of IcsA in Shigella flexneri. Mol. Microbiol. 23, 1063-1074. 128. Shere, K. D., Sallustio, S., Manessis, A., d'Aversa, T. G., and Golberg, M. B. (1997). Disruption of IcsP, the major Shigella protease that cleaves IcsA, accelerates actin-based motility. Mol. Microbiol. 25,451-462. 129. Goldberg, M. B., and Theriot, A. J. (1995). Shigella flexneri surface protein IcsA is sufficient to direct actin-based motility. Proc. Natl. Acad. Sci. U.S.A. 92, 6572-6576. 130. Kocks, C., Marchand, J. B., Gouin, E., d'Hauteville, H., Sansonetti, P J., Carlier, M. E, and Cossart, P. (1995). The unrelated surface proteins ActA of Listeria monocytogenes and IcsA of Shigella flexneri are sufficient to confer actin-based motility to L. innocua and E. coli, respectively. Mol. Microbiol. 18, 413^23. 130a. Egile, C., Loisel, T. P., Laurent, V., Rong Li, Pantaloni, D., Sansonetti, P. J., and Carlier, M. F. (1999). Activation of CDC42 effector N-WASP by the Shigella icsA protein promotes actin nucleation by Arp2/3 complex and bacterial actin-based motility. J. Cell Biol. 146, 1319-1332. 131. Suzuki, T., Shinsuke, S., and Sasakawa, C. (1996). Functional analysis of Shigella VirG domains essential for interaction with vinculin and actin-based motility. J. Chem. Biol. 271,21878-21885. 132. Suzuki, T., Miki, H., Takenawa, T., and Sasakawa, C. (1998). Neural Wiskott-Aldrich syndrome protein is implicated in the actin-based motility of Shigella flexneri. EMBO J. 17, 2767-2776. 133. Lasa, I., Egile, C., Cossart, P., and Sansonetti, P. J. (1997). Motilite intracellulaire et polymerisation de I'actine par Listeria monocytogenes et Shigella flexneri. Ann. Inst. Pasteur 8, 163-172. 134. Prevost, M. C , Lesourd, M., Arpin, M., Vernel, E, Mounier, J., Hellio, R., and Sansonetti, P. J. (1992). Unipolar reorganisation of F-actin layer at bacterial division and bundling of actin filaments by plastin correlate with movement of Shigella flexneri within Hela cells. Infect. Immun. 60, 4088^099. 135. Chakraborty, T., Ebel, E, Domman, E., Niebuhr, K., Gerstel, B., Pistor, S., Temm-Grove, C. J., Jockush, B. M., Reinhard, M., Walter, U., and Wehland, J. (1995). A focal adhesion factor directly linking intracellularly motile Listeria monocytogenes and Listeria ivanovii to the actin based cytoskeleton of mammalian cells. EMBO J. 14, 1314-1321. 136. Laine, R. O., Zeile, W., Kang, E, Purich, D. L., and Southwick, E S. (1997). Vinculin proteolysis unmasks an ActA homolog for actin-based Shigella motility. / Cell Biol. 138, 1255-1264. 137. Jockush, B. (1995). The molecular architecture of focal adhesions. Annu. Rev. Cell. Dev. Biol. 11, 379-416. 138. Niebuhr, K., Ebel, E, Ronald, E, Reinhard, M., Domann, E., Carl, U. D., Ulrich, W., Gerder, E B., Wehland, J., and Chakraborty, T. (1997). A novel proline-rich motif present in ActA of Listeria monocytogenes and cytoskeletal proteins is the ligand for the EVHl domain, a protein module present in the Ena/VASP family. EMBO J. 16, 2793-2802. 139. Goldberg, M. B. (1997). Shigella actin-based motility in the absence of vinculin. Cell. Motil. Cytoskeleton 37, 44-53. 140. Miki, H., Miura, K., and Takenawa, T. (1996). N-WASP, a novel actin-depolymerizing protein regulates the cortical cytoskeletal rearrangements in a PIP2-dependent manner downstream of tyrosine kinases. EMBO J. 15, 5326-5335.

382


141. Miki, H., Sasaki, T., Takai, Y., and Takenawa, T. (1998). Induction of filopodium formation by a WASP-related actin-depolymerizing protein N-WASP. Nature 391, 93-96. 142. Miki, H., and Takenawa, T. (1998). Direct binding of the verprolin-homology domain in N-WASP to actin is essential for cytoskeletal reorganization. Biochem. Biophys. Res. Commun. 243, 73-78. 143. Derry, J. M. J., Ochs, H. J., and Francke, U. (1994). Isolation of a novel gene mutated in Wiskott-Aldrich syndrome. Cell 78, 635-644. 144. Bear, J. E., Rawls, J. R, and Saxe III, C. (1998). SCAR, a WASP-related protein, isolated as a suppressor of receptor defects in late Dictyostelium development. J. Cell Biol 142, 1325-35. 145. Miki, H., Suetsugu, S., and Takenawa, T. (1998). WAVE, a novel WASP-family protein involved in actin reorganization induced by Rac. EMBO J. 17, 6932-6941. 146. Symons, M., Derry, J. M. J., Karlak, B., Jiang, S., Lemahieu, V., McCormick, P., Francke, U., and Abo, A. (1994). Wiskott-Aldrich Syndrome Protein, a novel effector for the GTPases Cdc42Hs, is implicated in actin polymerization. Cell 84, 723-734. 147. Prevost, M. C , Lesourd, M., Arpin, M., Vemel, F., Mounier, J., Hellio, R., and Sansonetti, P. J. (1992). Unipolar reorganization of F-actin layer at bacterial division and bundling of actin filaments by plastin correlate with movement of Shigella flexneri within HeLa cells. Infect. Immun. 60, 4088-4099. 148. Allaoui, A., Mounier, J., Prevost, M. C , Sansonetti, P J., and Parsot, C. (1992). icsB: A Shigella flexneri virulence gene necessary for the lysis of protrusions during intercellular spread. Mol. Microbiol. 6, 1605-1616. 149. Zychlinsky, A., Thirumalai, K., Arondel, J., Cantey, J. R., Aliprantis, A. O., and Sansonetti, P. J. (1996). In vivo apoptosis in Shigella flexneri infections. Infect. Immun. 64, 5357-5365. 150. Guichon, A., and Zychlinsky, A. (1997). Clinical isolates of Shigella species induce apoptosis in macrophages. J. Infect. Dis. 175, 470-473. 151. Islam, D., Veress, B., Bardhan, P. K., Lindberg, A. A., and Christensson, B. (1997). In situ characterization of inflammatory responses in the rectal mucosae of patients with shigellosis. Infect. Immun. 65, 739-749. 152. Mantis, N., Prevost, M. C , and Sansonetti, P. J. (1996). Analysis of epithelial cell stress responses during infection by Shigella flexneri. Infect. Immun. 64, 2474-2482. 153. Zychlinsky, A., Kenny, B., Menard, R., Prevost, M. C , Holland, I. B., and Sansonetti, P. J. (1994). IpaB mediates macrophage apoptosis induced by Shigella flexneri. Mol. Microbiol. 11, 619-627. 154. Chen, Y., Smith, M. R., Thirumalai, K., and Zychlinsky, A. (1996). A bacterial invasin induces macrophage apoptosis by binding directly to ICE. EMBO J. 15, 3853-3860. 155. Thomberry, N. A., and Lazebnik, Y. (1998). Caspases: Enemies within. Science 281, 1312-1316. 156. Yuan, J., Shalam, S., Ledoux, S., Ellis, H. M., and Horvitz, H. R. (1993). The C elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta-converting enzyme. Cell 75, 641-652. 157. Ghayur, T, Banerjee, S., Hugunin, M., Butler, D., Herzog, L., Carter, A., Quintal, L., Sekut, L., Talanian, R., Paskind, M., Wong, W., Kamen, R., Tracey, D., and Allen, H. (1997). Caspase-1 processes IFN-gamma-inducing factor and regulates LPS-induced IFN-gamma production. Nature 386, 619-623. 158. Hilbi, H. J., Moss, E., Hersh, D., Chen, Y, Arondel, J., Banerjee, R. A., Flavell, J., Yuan, J., Sansonetti, P. J., and Zychlinsky, A. (1998). Shigella-'induccd apoptosis is dependent on Caspase-1 which binds to IpaB. J. Biol. Chem. 273, 32864. 159. Mandic-Mulec, I., Weiss, J., and Zychlinsky, A. (1997). Shigella flexneri remains in polymorphonuclear leucocyte vacuoles where it is efficiently killed. Infect. Immun. 65, 110-115. 160. Weiss, J. (1994). Leukocyte-derived anti-microbial proteins. Curr Opin. Hematol. 1, 78-84.

8.

SHIGELLOSIS

383

161. Renesto, P., Sansonetti, P. J., and Guillen, N. (1997). Interaction between Entamoeba histolytica and intestinal epithelial cells involves a CD44 cross-reactive protein expressed on the parasitic surface. Infect. Immun. 65, 4330-4333. 162. Pal, T., and Hale, T. L. (1989). Plasmid-associated adherence of Shigella flexneri in a HeLa cell model. Infect. Immun. 57, 2580-2582. 163. Duguid, J. P., and Gillies, R. R. (1957). Fimbriae and adhesive properties in dysentery bacilli. J. Pathol. Bacteriol. 74, 397-411. 164. Snellings, N. J., Tall, B. D., and Venkatesan, M. M. (1997). Characterization oi Shigella type 1 fimbriae: Expression, FimA sequence, and phase variation. Infect. Immun. 65, 2462-2467. 165. Utsunomiya, A., Naito, T., Ehara, M., Ichinose, Y, and Hamamoto, A. (1992). Studies on novel pili from Shigella flexneri, I: Detection of pili and hemagglutination activity. Microbiol. Immunol. 36, 803-813. 166. Guhathakurta, B., Sasmal, D., and Datta, A. (1992). Adhesion oi Shigella dysenteriae type 1 and Shigella flexneri to guinea pig colonic epithelial cells in vitro. J. Med. Microbiol. 36, 403^05. 167. Giannesca, P J., Giannesca, K. T., Falk, P, Gordon, J. I., and Neutra, M. R. (1994). Regional differences in glycoconjugates of intestinal M cells in mice: Potential targets for mucosal vaccines. Am. J. Physiol. 161, G1108-G1121. 168. Izhar, M., Nuchamowitz, Y, and Mirelman, D. (1982). Adherence oi Shigella flexneri to guinea pig intestinal cells is mediated by mucosal adhesion. Infect. Immun. 35, 1110-1118. 169. Shea, J., Hensel, M., Gleeson, C , and Holden, D. (1996). Identification of a virulence locus encoding a second type III secretion system in Salmonella typhimurium. Proc. Natl. Acad. Sci. U.S.A. 93, 2593-2597. 170. Zychlinsky, A., and Sansonetti, P. J. (1997). Apoptosis as a proinflammatory event: What can we learn from bacteria-induced cell death? Trends Microbiol. 5, 201-204. 170a. Arondel, J., et al. (2000). Manuscript submitted. 171. Sansonetti, P J., Arondel, J., Cavaillon, J. M., and Huerre, M. (1995). Role of IL-1 in the pathogenesis of experimental shigellosis. J. Clin. Invest. 96, 884-892. 172. Perdomo, J. J., Cavaillon, J. M., Huerre, M., Ohayon, H., Gounon, P., and Sansonetti, P. J. (1994). Acute inflammation causes epithelial invasion and mucosal destruction in experimental shigellosis. / Exp. Med. 180, 1307-1319. 173. McCormick, B. A., Siber, A. M., and Maurelli, A. T. (1998). Requirement of the Shigella flexneri virulence plasmid in the ability to induce trafficking of neutrophils across polarized monolayers of the intestinal epithelium. Infect. Immun. 66, 4237^243. 174. Benjelloun-Touimi, Z., Sansonetti, P. J., and Parsot, C. (1995). SepA, the major extracellular protein of Shigella flexneri: Autonomous secretion and involvement in tissue invasion. Mol. Microbiol. 17, 123-135. 175. Benjelloun-Touimi, Z., Si Tahar, M., Montecucco, C , Sansonetti, P J., and Parsot, C. (1998). SepA, the 110 kDa protein secreted by Shigella flexneri: Two-domain structure and proteolytic activity. Microbiology 144, 1815-1822. 176. Harendra de Silva, D. G., Mendis, L. N., Sheron, N., Alexander, G. J. M., Candy, D. C. A., Chart, H., and Rowe, B. (1993). Concentrations of interleukin-6 and tumor necrosis factor in serum and stools of children with Shigella dysenteriae 1 infection. Gut 34, 194-198. 177. Way, S. S., Borczuk, A. C , Dominitz, R., and Goldberg, M. B. (1998). An essential role for gamma interferon in innate resistance to Shigella flexneri infection. Infect. Immun. 66, 1342-1348. 178. Maurelli, A. T., Blackmon, B., Curtiss III, R. (1984). Temperature-dependent expression of virulence genes in Shigella species. Infect. Immun. 43, 195-201.

384


179. Sakai, T., Sasakawa, C , and Yoshikawa, M. (1986). Expression of the four virulence antigens oi Shigella flexneri is positively regulated at the transcriptional level by the 30 kD VirF protein. Mol Microbiol 2, 589-597. 180. Adler, B., Sasakawa, C , Tobe, T., Makino, S., Komatsu, K., and Yoshikawa, M. (1989). A dual transcriptional activation system for the 230-kb plasmid genes coding for virulence associated antigens oi Shigella flexneri. Mol. Microbiol. 3, 627-635. 181. Maurelli, A. T., and Sansonetti, P. J. (1988). Identification of a chromosomal gene controlling temperature regulated expression of Shigella virulence. Proc. Natl. Acad. Sci. U.S.A. 85, 2820-2824. 182. Dorman, C. J., Ni Bhriain, N., and Higgins, C. F. (1990). DNA supercoiling and environmental regulation of virulence gene expression in Shigella flexneri. Nature 344, 789-792. 183. Hulton, C. S. J., Seiraf, A., Hinton, J. C. D., Sidebotham, J. M., Waddel, L. Pavitt, G. D., Owen-Hughes, T., Spassky, A., Buc, H., and Higgins, C. F. (1990). Histone-like protein JHl (H-NS), DNA supercoiling and genes expression in bacteria. Cell 63, 631-642. 184. Tobe, N., Nagai, S., Okada, N., Adler, B., Yoshikawa, M., Sasakawa, C. (1991). Temperatureregulated expression of invasion genes in Shigella flexneri is controlled through the transcriptional activation of the virB gene on the large plasmid. Mol. Microbiol. 5, 887-893. 185. Tobe, T, Yoshikawa, M., Mizumo, T, and Sasakawa, C. (1993). Transcriptional control of the invasion regulatory gene virB of Shigella flexneri: Activation by virF and repression by H-NS. J. Bacteriol. 175, 6142-6149. 186. Sasakawa, C. (1995). Molecular basis of pathogenicity of Shigella. Rev. Med. Microbiol. 6, 257-266. 187. Bemardini, M. L., Fontaine, A., and Sansonetti, P.J. (1990). The two-component, osmo-dependent, regulatory system (ompR-ompZ) controls the virulence of Shigella flexneri. J. Bacteriol. 172,6274-6281. 188. Uchiya, K., Tobe, T, Komatsu, K., Suzuki, T, Watarai, M., Fukuda, I., Yoshikama, M., and Sasakawa, C. (1995). Identification of a novel virulence gene, virA, on the large plasmid of Shigella, involved in invasion and intercellular spreading. Mol. Microbiol. 17, 241-250. 189. Venkatesan, M. M., Buysse, J. M., and Hartman, A. B. (1991). Sequence variation in two ipaH genes of Shigella flexneri 5 and homology to the LRG-like family of proteins. Mol. Microbiol. 5, 2435-2445. 190. Demers, B., Sansonetti, P. J., and Parsot, C. (1998). Induction of type III secretion in Shigella flexneri is associated with differential control of transcription of genes encoding secreted proteins. EMBO J. 17(10), 2894-2903. 191. Sansonetti, P J., Hale, T L., Dammin, G. J., Kapfer, C., Colins, H., and Formal, S. B. (1983). Alterations in the pathogenicity of Escherichia coli K-12 after transfer of plasmid and chromosomal genes from Shigella flexneri. Infect. Immiin. 39, 1392-1402. 192. Okamura, N., and Nakaya, R. (1977). Rough mutant of Shigella flexneri 2a that penetrates tissue culture cells but does not evoke keratoconjunctivitis in guinea pigs. Infect. Immun. 17, 4-8. 193. Sandlin, R. C , Lampel, K. A., Keasler, S. P, Goldberg, M. B., Stolzer, A. L., and Maurelli, A. T (1995). Avirulence of rough mutants of Shigella flexneri: Requirement of O antigen for correct unipolar localization of IcsA in the bacterial outer membrane. Infect. Immun. 63, 229-237. 194. Rowley, D. (1968). Sensitivity of rough Gram-negative bacteria to the bactericidal action of serum. J. Bacteriol. 95, 1647-1650. 195. Joiner, K. A. (1988). Complement evasion by bacteria and parasites. Annu. Rev. Microbiol. 42, 201-230. 196. Endo, Y, Tsurugi, K., Yutsudo, K., Takeda, T, Ogasawara, T, and Igararshi, K. (1988). Site of action of verotoxin (VT2) from Escherichia coli 0157:H7 and of Shiga toxin on eukaryotic ribosomes: RNA N-glycosidase activity of the toxins. Eur J. Biochem. 171, 45-50.

8.

SHIGELLOSIS

385

197. Thompson, M. R., Steinberg, M. S., Gemski, P., Formal, S. B., and Doctor, B. P. (1976). Inhibition of in vitro protein synthesis by Shigella dysenteriae 1 toxin. Biochem. Biophys. Res. Commun. 71, 783-788. 198. Olenick, J. G., and Wolfe, A. D. (1980). Shigella toxin inhibition of binding and translation of polyuridylic acid by Escherichia coli ribosomes. J. Bacteriol. 141, 1246-1250. 199. Moyer, M. P, Dixon, P S., Rothman, S. W., and Brown, J. E. (1987). Cytotoxicity of Shiga toxin for primary cultures of human colonic and ileal epithelial cells. Infect. Immun. 55, 1533-1535. 200. Fontaine, A., Arondel, J., and Sansonetti, P. J. (1988). Role of Shiga toxin in the pathogenesis of bacillary dysentery as studied with a Tox-mutant of Shigella dysenteriae 1. Infect. Immun. 56,3099-3109. 201. Bridgwater, F. A. J., Morgan, R. S., Rowson, E. K. E. K., and Payhing Wright, G. (1955). The neurotoxin of Shigella shigae: Morphological and functional lesions produced in the central nervous system of rabbits. Br J. Exp. Pathol. 36, 447^53. 202. Obrig, T. G., Louise, C. B., Lingwood, C. A., Boyd, B., Barley-Maloney, L., and Daniel, T. O. (1993). Endothelial heterogeneity in Shiga toxin receptors and responses. J. Biol. Chem. 268, 15484-15488. 203. Koster, F., Levin, J., Walker, L., Tung, K. S., Gilman, R. H., Rahaman, M. M., Majid, M. A., Islam, S., and WilUiams Jr., R. C. (1978). Hemolytic-uremic syndrome after shigellosis: Relation to endotoxemia and circulating immune complexes. New Engl. J. Med. 298, 927-933. 204. Heyderman, R. S., Fitzpatrick, M. M., and Robin Barclay, G. (1994). Haemolytic-uraemic syndrome. L««c^r 343, 1042. 205. Noriega, F. R., Liao, F. M., Formal, S. B., Fasano, A., and Levine, M. M. (1995). Prevalence of Shigella enterotoxin 1 among Shigella clinical isolates of diverse serotypes. J. Infect. Dis. Ill, 1408-1410. 206. Nataro, J. P., Seriwatana, J., Fasano, A., Maneval, D. R., Guers, L. D., Noriega, F., Dubovsky, F., Levine, M. M., and Morris Jr., J. G. (1995). Identification and cloning of a novel plasmid-encoded enterotoxin of enteroinvasive Escherichia coli and Shigella. Infect. Immun. 63, 4721-4728. 207. Bernardini, M. L., Sanna, M. G., Fontaine, A., and Sansonetti, P. J. (1993). OmpC is involved in invasion of epithelial cells by Shigella flexneri. Infect. Immun. 61, 3625-3635. 208. Lawlor, K. M., Daskaleros, P. A., Robinson, R. E., and Payne, S. M. (1987). Virulence of iron transport mutants oi Shigella flexneri and utilization of host iron compounds. Infect. Immun. 55, 594-599. 209. Nassif, X., Mazert, M. C , Mounier, J., and Sansonetti, P. J. (1987). Evaluation with an iuc::TnlO mutant of the role of aerobactin production in the virulence of Shigella flexneri. Infect. Immun. 55, 1963-1967. 210. Pal, T, Newland, J. W, Tall, B. D., Formal, S. B., and Hale, T. L. (1989). Intracellular spread of Shigella flexneri associated with the kcpA locus and a 140 kDa protein. Infect. Immun. 18, 94-98.


CHAPTER 9

Pathogenic Escherichia

coli

JOSE L. PUENTE B. BRETT FINLAY

I. Introduction 11. Enterotoxigenic E. coli (ETEC) A. Disease B. Virulence Factors C. Virulence Gene Regulation III. Enteroinvasive E. coli (EIEC) A. Disease B. Virulence Factors C. Virulence Gene Regulation IV. Enteropathogenic E. coli (EPEC) A. Disease B. Virulence Factors C. Virulence Gene Regulation V. Enterohemorrhagic E. coli (EHEC) A. Disease B. Virulence Factors C. Virulence Gene Regulation VI. Enteroaggregative E. coli (EAEC) A. Disease B. Virulence Factors VII. Diffusely Adhering E. coli (DAEC) A. Disease B. Virulence Factors VIII. Uropathogenic £. co// A. Disease B. Virulence Factors IX. E. coli That Cause Sepsis and Meningitis A. Disease B. Virulence Factors X. Conclusions References


'^87

388 390 390 390 395 396 396 396 398 398 398 400 407 408 408 409 414 414 414 415 417 417 417 418 418 418 426 426 426 428 428

388

JOSE L. PUENTE AND B . BRETT FINLAY

/- Introduction Escherichia coli is the most extensively studied microorganism. It has been a model system for the study of bacterial metabolism, the cell division process, cell wall biosynthesis, chemotaxis, bacterial genetics, and the physiological role of enteric bacteria as part of the normal fecal flora [1]. Despite the vast knowledge that has been accumulated over the years, the recent release of its full genomic composition has made it obvious that there are still many things to learn about this microorganism [2]. Analysis of the E. coli K-12 genome sequence also shows that about 2% of its DNA consists of mobile genetic elements, including phages, plasmids, and transposons [2]. These elements are responsible for the continuous evolution of the bacterial genomic repertoire, providing significant diversity in E. coli strains. In this regard, pathogenic E. coli appears to have evolved from nonpathogenic strains by acquiring new virulence factors by the horizontal transfer of accessory DNA, which is often organized in clusters (pathogenicity islands) in the chromosome or on plasmids [3]. The high genetic diversity of the E. coli genome is also reflected by the large variation in DNA content between different strains [4-6] and by the distribution or genomic location (insertion site) of different virulence determinants [7, 8]. In this context, it seems that most pathogenic E. coli strains do not have a single evolutionary origin, but instead have emerged as a result of different events of DNA transfer, and that even strains capable of causing the same disease do not constitute a monophyletic group [9]. E. coli pathogenic variants are represented by strains of specific serogroups possessing a particular set of virulence factors, which are responsible for the different clinical manifestations that characterize E. coli infections. Pathogenic E. coli cause various diseases in humans, including several types of diarrhea, urinary tract infections, sepsis, and meningitis (Table I). E. coli strains that cause human diarrhea of varying severity have been divided into six major categories: enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), enteroaggregative E. coli (EAEC), and diffuse adhering E. coli (DAEC). E. coli strains causing urinary tract infections are known as uropathogenic E. coli (UPEC), while E. coli Kl are often responsible for cases of meningitis or sepsis (Table I). Different bacterial virulence attributes appear to dictate the types of interactions that occur between the pathogenic organism and its host cells, and where in the body these interactions occur. Tissue tropism plays an important role in disease— for example, UPEC infects the urinary tract and kidneys, EPEC the small bowel, and EHEC the large bowel. For nearly all pathogenic E. coli, colonization of a particular host surface is mediated by fimbriae or pili, which are often called colonization factors [10-12]. Common adhesins are frequently found within several E. coli types. For example, type I pili are found in most of the different pathogenic E. coli, making it difficult to assign a specific role for this adhesin in disease, although it has been suggested to be important for spreading and

9.

389

PATHOGENIC ESCHERICHIA COLI

Table I

E. coli That Are Pathogenic for Humans

Type of E. coli

Disease

Virulence factors

Enterotoxigenic (ETEC)

Watery to cholera-like diarrhea

Heat-labile toxin (LT), heat-stable toxin (ST), colonization factors (CFs)

Enteroinvasive (EIEC)

Watery diarrhea to dysentery

Ipas, type III secretion (Mxi and Spa), VirG/IcsA

Enteropathogenic (EPEC)

Watery diarrhea

Esps, type III secretion (Sep and Esc), intimin, Tir, and BFP

Enterohemorrhagic (EHEC)

Hemorrhagic colitis, hemolytic uremic syndrome (HUS)

Above EPEC factors and Shiga toxin, hemolysin

Enteroaggregative (EAEC)

Watery to mucoid diarrhea

AAF adhesins, EAST-1, Pet, Pic, hemolysin

Diffusely adhering (DAEC)

Watery diarrhea

F1845 and AIDA-I fimbriae

Uropathogenic (UPEC)

Urinary tract infections

Type I pili, P pili, Afimbrial adhesins (Afa), hemolysin, CNF-1

Septic (SEC)

Neonatal sepsis, meningitis

Capsule, type I pili, S-fimbrial adhesin, IbeA and IbeB (invasion proteins)

colonization by commensal E. coli [13, 14] or colonization of the urinary tract [15, 16]. Once localized to a particular tissue, the molecular interactions that occur between pathogenic E. coli and their host cells follow specific steps, and are quite different between different pathogenic types. Some strains adhere to mucosal surfaces and secrete specific toxins that either intoxicate localized epithelial cells or spread systemically to affect distant host cells. Other strains interact more intimately with host cell surfaces, and this intimate interaction results in disease. Finally, other strains actually enter host cells and live as intracellular pathogens, or penetrate host barriers and live systemically within the human host, resulting in septic disease [17, 18]. The wide diversity of virulence factors identified and characterized in different pathogenic E. coli resemble many of the virulence mechanisms found in other pathogens [17, 19]. ETEC utilizes a cholera-like toxin to cause cholera-like disease [20]. EIEC behaves as Shigella, in that it contains the same virulence factors (e.g., type III secretion system, invasins, and intracellular spread mechanism) that are responsible for producing a dysentery-like disease [21]. EHEC produces a Shiga-like toxin (similar to that found in Shigella dysenteriae) that seems to be involved in causing the hemolytic uremic syndrome in a proportion of cases [22]. EHEC and EPEC utilize a type III secretion system, similar to those seen in Salmonella, Shigella, Yersinia, and other Gram-negative pathogens, to inject E. C(9//-specific factors into the host cell. These factors induce actin

390


rearrangements and activation of particular signal transduction pathways that result in disease [23]. As with other pathogens that cause systemic disease and meningitis, E. coli Kl also produces a polysaccharide capsule that prevents clearance by phagocytic cells [24]. Expression of the genes encoding this variety of virulence factors is often modulated in response to a series of environmental cues such as temperature, ion concentrations, osmolarity, iron levels, pH, carbon source availability, growth phase, and oxygen levels [25]. Virulence gene expression is determined by a consensus response to a mixture of these different biochemical and physical parameters that allows the bacterial cell to identify and exploit a particular extracellular or intracellular niche. Mechanistically, it often involves the interplay of regulatory proteins acting independently or as a cascade; these proteins share similarity with members of different families of regulatory proteins [17, 26]. The aim of this chapter is to examine our present understanding of the molecular basis of E. coli pathogenesis and the function and regulation of the various virulence determinants that distinguish each category in the context of their contribution to disease. An excellent comprehensive review was published in 1998 that examines the epidemiology, clinical symptoms, detection, diagnosis, and virulence of the diarrheagenic E. coli [18].

//. Enterotoxigenic E. coli (ETEC) A.

Disease

ETEC causes a watery diarrhea, ranging in severity from mild and self-limiting to a severe cholera-like profuse diarrhea. Diarrhea is usually without mucus, blood, pus, fever, or vomiting, consistent with it being an intoxication (i.e., toxin mediated), rather than a systemic infection. ETEC can affect adults and is often seen in travelers in developing countries (thus its name "traveler's diarrhea"), being contracted through contaminated food or water [27]. Serious (life-threatening) disease is seen in infants in developing countries. ETEC is responsible for more than 650 million cases of diarrhea and between 700,000 and 800,000 deaths in children under the age of 5 years. Antibiotics decrease the severity and duration of diarrhea, but antibiotic-resistant strains of ETEC are increasingly common. Like cholera, therapy is mainly rehydration (usually oral). With proper hydration, the disease is usually self-limiting [18]. B. Virulence Factors ETEC strains cause diarrhea through the action of thermolabile and thermostable enterotoxins (reviewed in [20]). Approximately 30% of the ETEC strains express a heat-labile toxin (LT), 35% produce a heat-stable toxin (ST), and the rest express both. In addition, ETEC strains produce one or more colonization factors

9.

391


(adhesins) that mediate attachment to intestinal mucosal surfaces [28-30]. Figure 1 summarizes the current proposed action mechanisms of ETEC toxins.

-CF

ETEC

GSa

Fig. I ETEC interactions with intestinal epithelial cells. ETEC adherence to intestinal cells is mediated by different colonization factors (CFs). Once established in the proximal small intestine, ETEC strains produce the heat-labile toxin (LT) and/or the heat-stable toxin (ST). The LT holotoxin, consisting of an A subunit and a pentamer of B subunits, is internalized by endocytosis on binding to its ganglioside GM] receptor. The A subunit is proteolitically cleaved into the A] and A2 subunits, which remain linked by a disulfide bond. The Ai subunit ADP-ribosylates the alpha subunit of the GTP-binding protein, Gs, inhibiting its intrinsic GTPase activity, resulting in constitutive activation of adenylate cyclase at the basolateral membrane. This activation leads to increased levels of intracellular cyclic AMP, activation of a cAMP-dependent A kinase, and supranormal phosphorylation of intestinal epithelial cell chloride channels, such as CFTR. These events result in inhibition of NaCl absorption and stimulation of chloride secretion. ST acts by binding to guanylate cyclase type C (GC-C), localized in the brush-border membrane of intestinal epithelial cells. Activation of GC-C results in increased levels of intracellular cyclic GMP that stimulates chloride secretion and/or inhibits NaCl absorption, resulting in net intestinal fluid secretion. In vivo chloride secretion may occur through activation of a cGMP-dependent protein kinase (G-kinase), which ultimately activates the chloride channel CFTR.

392 1.

JOSE L. PUENTE AND B . BRETT FINLAY HEAT-LABILE TOXIN (LT)

There are two forms of LT: LT-I and LT-II [31]. LT-I, the predominant form, is quite similar to CT at the sequence level (above 80% identity) [32], and is thought to act mechanistically in an identical fashion (see [33] and Chapter 10 for details on Vibrio cholerae, its toxin, and mechanism of action). LT-I is oligomeric in structure with one enzymatic A subunit and five identical B subunits [34]. The five B subunits are arranged symmetrically in a ring-like structure that binds the ganglioside GMi and weakly to ganglioside GDlb [35]. The A subunit is proteolytically cleaved into two domains—A] and A2—that remain linked by a disulfide bond and span the center of the ring [36, 37]. The toxin is endocytosed, and the A subunit reaches the basolateral (bottom) surface of the epithelial cell after escaping the endocytic vesicle (reviewed in [20, 33]). The Al peptide transfers an ADP-ribosyl group from NAD to the a-subunit of the GTP binding protein G^. This modification of Gâ inhibits its intrinsic GTPase activity, which results in permanent activation of adenylate cyclase, leading to accumulation of intracellular levels of cyclic AMP (cAMP). As cAMP accumulates inside intestinal cells, it is thought that a cAMP-dependent kinase (A kinase) is activated, which then results in phosphorylation of apically located chloride channel proteins such as CFTR (the channel affected in cystic fibrosis patients). This causes channel opening, and chloride ion efflux out of cells along with a block in ion and fluid absorption into cells, resulting in a net osmotic imbalance (Fig. 1). The result is watery diarrhea. It has been proposed that the reason for the high prevalence of cystic fibrosis in the Caucasian population is that a defective CFTR provides a protective mechanism against CT- (and LT-) mediated disease [38, 39], though other data suggest it confers protection to S. ry/?/2/-mediated typhoid fever. Although the above explanation is logical, several other more complex mechanisms have also been implicated in LT- and CT-mediated disease (reviewed in [18]). These include promotion of the production of prostaglandins, stimulation of a mild inflammatory response, and activation of the enteric nervous system. It is likely that the watery diarrhea resulting from ETEC infection is a combination of the classic mechanism described above along with one or more of these other events. The LT-II form shows less identity to LT-I and CT (ca. 57%) and basically no identity to the B subunit [40, 41]. These differences are also reflected in their specificity for binding to gangliosides, since LT-II binds best to gangliosides GDlb or GDI a [35] and is found associated mainly with animal, but not human, disease [42]. Like ST and the colonization factors, LTs are usually encoded on a plasmid. 2.

HEAT-STABLE TOXIN (ST)

As mentioned above, about one-third of ETEC strains expresses ST, and another third expresses ST and LT Thus, ST alone is capable of causing watery diarrhea, without LT being present. In contrast to LT, ST is heat stable (thus its

9. PATHOGENIC ESCHERICHIA COLI

393

name), a property provided by its intramolecular disulfide bonds. There are two major STs: STa and STb (reviewed in [20, 43]). STa has been studied in more detail and provides an excellent bacterial example of a hormone-like peptide that affects normal host cell function [44]. STa enterotoxins are small, cysteine-rich molecules (18-19 amino acids) that can form three intramolecular disulfide bonds [45]. These toxins are made from larger precursors (72 amino acids) that are cleaved as the molecule transits out of the bacterium [46]. The eukaryotic membrane receptor for STa is guanylate cyclase C (GCC), which is located in the apical membrane of intestinal cells [47]. Because of the apical location of guanylate cyclase, STa-mediated cell activation is quite rapid. It has been suggested that STa receptors are more abundant on enterocytes of infants or young animals, but that their numbers decrease in older individuals. This variation in receptor abundance may determine the severity of the secretory response, and give a plausible explanation for the high susceptibility of human infants and newborn animals to STa-mediated diarrhea [48-50]. GCC activation triggers a cascade of events (reviewed in [18, 20]) (Fig. 1), including the accumulation of intracellular cGMP levels, the cGMP-dependent activation of protein kinase A (PKA), and the PKA-dependent phosphorylation and activation of the cystic fibrosis transmembrane conductance regulator (CFTR), which finally leads to increased chloride secretion and blockage of sodium chloride uptake, resulting in diarrhea [38, 51] (Fig. 1). GCC-null mice were protected against an infection with ETEC, further evidencing the importance of this receptor in Sta-mediated diarrhea [52]. Several other intracellular signals are activated in response to STa, but the contribution of these signals to diarrhea has not been fully defined. STa behaves like guanylin, the endogenous intestinal peptide hormone that binds to guanylate cyclase [53, 54]. Guanylin regulates ion and fluid levels by modulating cGMP levels, thereby mediating intestinal homeostasis. It is a 15-aa peptide that contains four cysteines (two disulfide bonds) and, ironically, is less efficient than STa in activating guanylate cyclase. The discovery of guanylin was one of the first demonstrations that bacteria have evolved the ability to mimic eukaryotic endogenous functions to take advantage of host cells and, also, an indication of how the study of microbial pathogens is teaching us important aspects of the cell biology of eukaryotes. Although STb is primarily found in animal pathogens, it has also been isolated from human ETEC isolates. STb bears no sequence similarity to STa, although it does have four cysteine residues that form disulfide bridges [55]. It is also initially synthesized as a larger precursor of 71 aa, which is then proteolitically processed to a 48-aa mature protein [56]. Sulfatide has been suggested as a functional receptor for STb [57, 58]. STb affects neither cAMP or cGMP levels nor stimulates chloride secretion [59]; instead, it seems to elicit an intestinal response characterized by the secretion of bicarbonate, to which human cell lines seem to be insensitive [60].

394

3.


ADHESINS

ETEC attachment to an intestinal surface is mediated by colonization factor antigens (CFAs), coli surface antigens (CSs), and putative colonization factors (PCFs), which are generally referred to as colonization factors (CFs) (reviewed in [11, 30]. These structures are essential for ETEC to colonize the small intestine, a central step in ETEC's virulence. At least 20 different and antigenically distinct CFs have been described in human ETEC, and these are found in varying combinations along with LT, ST, or both [30]. The main CFs associated with human ETEC strains include CFA/I, constituted by a single fimbrial structure, and CFA/II and CFA/IV, which can be a combination of a particular set of CSs. CFA/II strains can express CS3 alone or in combination with CSl or CS2 [61, 62], while CFA/IV strains can express CS6 alone or mixed with CS4 or CS5 [63]. It is thought that CFs dictate host and tissue specificity, since animal ETEC CFs are not found in human ETEC isolates. Morphologically, CFs can be subdivided into four major groups: rigid rods (e.g., CFA/I), bundle-forming (longus), fibrillar, and nonfimbrial adhesins [11, 30]. The genetics of ETEC CFs have been studied extensively, and most are encoded within standard fimbrial operons, much like that seen with type I and Pap pili (see below). These operons usually encode 4-8 proteins whose functions include regulation, the major subunit that forms the adhesin, and accessory factors that include a periplasmic chaperone and an outer membrane molecular usher (Fig. 2) [61, 62, 64]. One of the best-studied systems is CSl, which has served as a prototype to study the mechanisms of assembly of CF [65]. The coo operon is composed of the cooB, cooA, cooC, and cooD genes, which are required for expression of functional CS1. CooB has been shown to act as a chaperone-like protein, which is required for pilus assembly but is not present in the final pili structure [66]. CooA is the major structural subunit [67]. CooC is an outer membrane protein likely to be involved in transmembrane transport of CSl [68]. CooD, which is a located at the tip of the pilus and potentially involved in adherence, determines the initiation of pilus assembly and modulates the number of CSl pili on bacterial cells [69, 70]. Unlike Pap and type I pili, which produce minor tip adhesins, the major structural subunit (the stalk protein) also functions as the major adhesin. Operons encoding CFs have a low G+C content, are usually flanked by transposons, and are mainly contained on plasmids that also contain LT and/or ST, suggesting that horizontal transfer and transposition events have been responsible for generating the combination of virulence factors found in ETEC strains [30]. The intestinal receptors that CFs bind include sialoglycolipids such as GM2, sialic acid containing glycoconjugates, asialogangliosides, and several other glycoconjugates (glycolipids and glycoproteins) found on the cell surface (reviewed in [11]). The oligosaccharides expressed on mammalian cell surfaces vary widely, providing an extensive range of options that might contribute to host and tissue specificity for the ETEC CFs.

9.

395


Rns

/ /

/ .

oo-.o«

Fig. 2 Genetic organization and regulation of the ETEC CSl fimbrial operon. The plasmid-encoded coo operon contains four genes that are co-expressed from a promoter located upstream from the cooB gene. Transcriptional activation of this operon requires the product of the plasmid-encoded rns gene, which exhibits sequence similarity to the AraC family of regulatory proteins. Rns seems to positively activate its own transcription. Transcription of the coo operon and rns is negatively regulated by H-NS, a small histone-like protein that potentially binds transcriptional silencer sequences present in both the cooB and rns upstream regulatory regions and open reading frames. Mutations in H-NS abolish negative regulation at low temperatures, suggesting that H-NS antagonizes Rns-mediated activation.

4.

OTHER VIRULENCE FACTORS

Some ETEC strains are capable of invading epithelial cell lines, although the role of invasion in ETEC pathogenesis remains undefined [71]. Two separate chromosomally encoded invasion loci, designated da and tib (toxigenic invasion loci A and B), direct noninvasive E. coli to adhere to and invade intestinal epitheUal cell lines [71,72]. The tib locus directs the synthesis of Tib A, a 104-kDa outer membrane protein that has been correlated with the adherence and invasion phenotypes [73]. Tib A is synthesized as a 100-kDa precursor (preTibA) that is subsequently glycosylated to render the active form [74]. Tib A shares similarity with members of the autotransporter family of outer membrane proteins (afimbrial adhesins) that play an important role in the virulence of different Gram-negative bacteria [74]. C. Virulence Gene Regulation Expression of the CSl and CS2 adhesins requires the product of the plasmid-encoded regulatory gene rns (Fig. 2) [75]. Rns belongs to a family of transcriptional regulators known as the AraC family and is considered the prototype of a number of AraC-like proteins involved in regulation of virulence determinants [76]. The

396


binding sites identified by Rns by footprinting analysis are neither palindromic nor repeated sequences, as has been observed for other AraC-like proteins [77]. Likewise, CFA/I expression requires CfaR/CfaD, a close homolog of Rns [78, 79]. The expression of these adhesins is negatively regulated at low temperatures. Interestingly, mutations in the gene coding for the histone-like protein H-NS abolish this negative regulation, suggesting that Rns and CfaR are required to overcome the repression mediated by H-NS on the expression of these adhesins [80, 81]. It is likely that, in the same way, rns positively regulates its own transcription [82]. Transcriptional silencer regions have been described for the coo operon and for rns [80, 82]. Some of these silencer regions, which are potential DNA-binding sites for H-NS, overlap the cooB and rns open reading frames, but act in conjunction with upstream sequences [80, 82].

///. Enteroinvasive E. coli (EIEC) A.

Disease

EIEC causes a watery diarrhea that often resembles that caused by ETEC. However, some patients do experience a dysentery-like disease, with mucus, blood, and pus in the stool, along with fever. Since the virulence factors in EIEC are virtually identical to those in Shigella species (see below and Chapter 8 herein), disease symptoms are often similar to those caused by S.flexneri (significant inflammation, ulcer formation, and clinical dysentery) [21, 83]. The infectious dose of EIEC is much larger than that of Shigella. Unlike S. dysenteriae, EIEC does not contain a Shiga toxin and so does not cause hemolytic uremic syndrome. The incidence of EIEC is low in developed countries, but foodbome outbreaks have been reported [84]. B. Virulence Factors Both EIEC and Shigella species invade the colonic epithelium. To achieve this, they follow a series of steps as they interact with the intestinal mucosa: invasion of colonic epithelial cells, lysis of the endocytic vacuole, bacterial multiplication, spread to adjacent cells, and host cell killing by apoptosis if the host cell is a macrophage [21, 83]. The virulence factors required for these multiple steps are encoded on a 140-MDa plasmid (pinv), although some other factors, such as regulators, are encoded on the chromosome [21, 85]. Nearly all we know about EIEC virulence mechanisms is based on extrapolation from the Shigella systems, so only an overview of ElEC/Shigella invasion and the necessary bacterial factors is provided here (Fig. 3). The homologous invasion mechanism of Shigella species is described in detail in Chapter 8 in this volume.

9.


397

Fig. 3 EIEC interactions with intestinal epithelial cells. EIEC strains secrete Ipa proteins via a type III secretion system onto and into host cells, causing actin rearrangements and membrane ruffling, resulting in bacterial internalization. Once inside a vacuole in the host cell, the IpaB protein degrades the vacuole, releasing the bacterium into the cytosol, where IcsA polymerizes actin. This action propels the organism through the cell and into neighboring cells.

1.

INVASINS (IPAS)

Like Shigella, EIEC secretes Ipas (invasion plasmid antigens) A-D into the host cell. Ipas mediate invasion by triggering several events in host cells resulting in membrane ruffling, macropinocytosis, actin rearrangements, and bacterial engulfment (reviewed in [83]). Once accomplished, internalized bacteria are surrounded by a membrane-bound inclusion in the cytoplasm of host cells. Unlike phagocytosis, this process can occur in nonphagocytic (e.g., epithelial) cells. Like most Gram-negative pathogens, EIEC and Shigella utilize a specialized secretion system, designated the type III system, to export the invasion plasmid antigens (Ipa proteins) out of the bacteria and into their host cells [23]. This system is encoded by more than 20 mxi and spa genes located on the large plasmid [21]. 2.

VIRG(ICSA)

Once free in the cytoplasm, EIEC and Shigella trigger an event that is responsible for their cell-to-cell spreading. Analogous to the process described for Listeria monocytogenes (see Chapter 16), these bacteria trigger nucleation of actin

398


at one pole of the bacterial cell, which propels the bacterium through the cytoplasm at the head of this "comet tail" [86]. The VirG protein (also called IcsA) is critical and sufficient for triggering of this actin-mediated motion [87], probably by binding cytoskeletal components such as N-WASP [88]. When the moving bacteria encounter the cytoplasmic face of the host cell membrane, they propel outward, pushing into adjacent cells, thereby avoiding extracellular exposure. C. Virulence Gene Regulation Shigella, and presumably EIEC, coordinately regulate their virulence factors in response to certain environmental conditions, including temperature, osmolarity, and DNA supercoiling. In addition, secretion of Ipas is increased in the presence of mammalian cells or serum [89]. VirR is a chromosomally encoded protein that regulates the various virulence factors in concert with the plasmid-encoded AraC-like transcriptional activator VirR VirR is a histone-like protein, functioning like H-NS (see Chapter 3 for details), that modulates the expression of virulence genes by repressing transcription at 30°C or in low osmolarity [90-93]. Interestingly, mutations in the hns gene derepress transcription of invasion genes at 30°C only in the presence of VirF [90, 91, 93], indicating that the role of VirF is not simply to overcome the negative influence of H-NS, as has been suggested for CfaD in ETEC (see above). In addition, H-NS also seems to control expression of virF in a temperature-dependent manner [94, 95].

IV. Enteropathogenic Escherichia coli (EPEC) A.

Disease

Enteropathogenic E. coli (EPEC) is the predominant cause of infant diarrhea worldwide and represents a major endemic health threat to children under 6 months of age living in developing countries [96]. In addition, isolated outbreaks in day care centers, nurseries, and pediatric wards, as well as among adults in developed countries, have also been reported. It has been estimated that EPEC kills several hundred thousand children each year worldwide. EPEC disease is characterized by prolonged watery diarrhea of varying severity, with vomiting and low fever often accompanying fluid loss (reviewed in [18, 96, 97]). Despite our increasing knowledge of the bacterial factors and host molecules that mediate EPEC interactions with epithelial cells, the actual molecular mechanisms that cause diarrhea remain undefined [98]. Unlike other diarrheas caused by pathogenic E. coli strains such as ETEC (see above), no toxin has been implicated in EPEC-mediated diarrhea. Instead, EPEC binds to intestinal surfaces of the small bowel, causing a characteristic histological lesion called the attaching and effacing (A/E) lesion [99-101]. A/E lesions are marked by dissolution of the intestinal brush-border surface and loss of epithelial microvilli (effacement) at the sites of bacterial attachment. Once bound, EPEC reside on cup-like projections


399

or pedestals formed by cytoskeletal rearrangements in host actin. EPEC-associated diarrhea has been attributed to the loss of absorptive surface area due to effacement of epithelial cell microvilli [102]. Although this mechanism remains to be proven, a series of recent in vitro studies have indicated that a more complex, yet not well-defined, multifactorial mechanism might be involved. Active ion secretion has been involved in the development of infectious diarrheal diseases [20]. It has been reported that EPEC causes significant depolarization of cultured HeLa and Caco-2 epithelial cells, altering the distribution of ions across the cell membrane [103]. In addition, EPEC infection of Caco-2 cell monolayers induced rapid but transient increases in short-circuit current (Isc) and CI" secretion [104]. Occurrence of such a process in the gut would reduce the electrochemical gradient available for sodium ion absorption from the gut lumen, thereby contributing to ionic imbalance, fluid loss, and diarrhea. However, other reports suggest that this may not be the case by demonstrating that EPEC-infected epithelial cell monolayers show a diminution in net ion transport, reflected as a decrease in short-circuit current (Isc), without revealing any difference in Cl~ channel activity [105, 106]. Early studies have also indicated that EPEC infection of HEp-2 cells increased the intracellular concentration of free calcium ([Ca^în) [107-109]. It has been hypothesized that this rise in [Ca^"^]in leads to disruption of the microvillus actin cytoskeleton by activating a calcium-dependent actin-severing protein, providing a plausible explanation for microvilli effacement [107]. However, in contrast to these data, a significant increase in [Ca^'^]in was not observed on EPEC infection in HEp-2 cells, using calcium-imaging fluorescence microscopy, which allows spatial and temporal measurements of [Ca^"^]in in live cells [110]. Furthermore, chelation of intracellular calcium with BAPTA did not inhibit pedestal formation, as previously reported [107, 109], suggesting that calcium fluxes are probably due to EPEC-mediated cytotoxicity [110]. In addition to morphological rearrangements that occur on the apical surface of cells, EPEC causes a large decrease in transepithelial electrical resistance in polarized epithelial cell monolayers [111]. EPEC infection induces phosphorylation of myosin light chain [112], an event that also leads to a decrease in transepithehal electrical resistance by disrupting tight junction integrity [113, 114]. EPEC infection also causes transmigration of polymorphonuclear leukocytes across the epithelial cell monolayers, which may contribute to disruption of epithelial barrier function [115]. Such transepithelial disruptions may occur in vivo, and this could lead to ionic imbalances, an increase of intestinal permeability, and possibly diarrhea. By using EPEC mutant strains, it has been shown that some of the virulence factors required for the formation of A/E lesions in vitro are also required to induce some of the events described above and are also needed to cause disease in human volunteers or animal models [103, 116-118]. What is not yet clear is how these virulence factors actually cause disease. Although EPEC can enter (invade) tissue culture cells [119, 120], it does not normally cause invasive disease and rarely penetrates the intestinal barrier.

400


EPEC belongs to a family of related pathogenic organisms that form A/E lesions, including enterohemorrhagic E. coli (EHEC), several EPEC-like animal pathogens that cause disease (e.g., REPEC in rabbits and PEPEC in pigs), Citrobacter rodentium, and Hafnia alvei [98, 121]. These organisms all cause cytoskeletal rearrangement and pedestal formation on host epithelial cells, and encode conserved proteins that mediate these effects.

B. Virulence Factors

The development of an A/E lesion has been divided in three stages (Fig. 4) (reviewed in [18, 98, 122, 123]). The first stage is characterized by the initial nonintimate attachment of EPEC to the epithelial cell surface in a pattern termed localized adherence (LA) [124-127]. LA is associated with the production of a type IV fimbriae known as bundle-forming pili (BFP) [128]. During the second stage, a set of EPEC-secreted proteins (Esps) triggers the activation of signal transduction pathways leading to a complex response by the epithelial cell and cytoskeletal rearrangements [129-132]. Finally, during stage three, an outer membrane protein called intimin allows EPEC to attach intimately to the host cell membrane on interaction with its translocated intimin receptor call Tir [133, 134]. Like many other pathogens, EPEC's virulence factors are found in clusters or pathogenicity islands. EPEC possess a large 69-kb plasmid that encodes two major virulence attributes—a bundle-forming pilus (BFP) needed for initial localized adherence to host cells and bacterium-bacterium interactions, and the regulatory locus Per (BfpTVW), which regulates expression of EPEC virulence factors (see below) [135]. All the factors necessary to form A/E lesions are clustered in a 35.5-kb region in the EPEC chromosome, inserted at the selenocysteine {selQ tRNA gene [136, 137]. This region of DNA (called the locus for enterocyte effacement, or LEE region) encodes >40 open reading frames, and has a low G-i-C content (38% versus 51% for E. coli), indicating it was acquired by horizontal transfer [138]. The LEE can be divided into three functional regions: one encoding a type III secretion system, a region needed for intimate binding to host cells (encodes Tir, CesT and intimin), and the one harboring the genes for E. coli effector proteins (Esps) that are secreted by the type III system and their chaperones (Fig. 4). The Esps trigger signals and actin rearrangements in host cells and are needed for A/E lesions (see below). In addition to the Esps, EPEC also insert another molecule (Tir) into host cell membranes [134, 139]. Upon insertion, Tir is tyrosine phosphorylated and serves as the receptor for intimin, an EPEC surface protein [133, 134]. Tir-intimin binding leads to underlying cytoskeletal organization and resultant pedestal formation (Fig. 4). Thus, this enteric pathogen uses a very specialized sequence of events to establish an intimate interaction with host cells, which then results in disease.

9.

401


tight junction^ disruption

PMN

Fig. 4 EPEC interactions with epithelial cells. The initial nonintimate attachment of EPEC microcolonies to the epithelial cell surface (the localized adherence phenotype) is mediated by type IV fimbriae known as BFP. After this initial interaction, secretion and translocation of the EspA, EspB, EspD, and Tir proteins onto and into the host cell cytosol is mediated by the type III secretion apparatus. EspA is assembled into a large filamentous organelle that is essential for translocation of EspB. Upon translocation, Tir is phosphorylated and inserted into the plasma membrane, where it binds to the bacterial OM protein intimin. At this stage, Tir nucleates cytoskeletal components including actin, alpha-actinin, talin, and ezrin. These cytoskeletal rearrangements form actin-rich pedestals beneath the adhering EPEC. Translocated proteins trigger the activation of signal transduction pathways that lead to a complex cellular response. Enhancement of membrane-associated protein kinase C (PKC) activity might stimulate CI" secretion. Tyrosine phosphorylated phospholipase Cy converts phosphatidyinositol-4,5-biphosphate (PIP2) into inositol triphosphate (IP3) and diacylglycerol (DAG), which might stimulate the release of intracellular calcium ([Ca^"^]i from IP3-sensitive stores. Activation of the transcription factor NF-KB initiates interleukin 8 (IL-8) transcription, which in turn stimulates transmigration of polymorphonuclear cells (PMNs). Myosin light-chain kinase (MLK) activation leads to tight junction phosphorylation and increased intestinal permeability.

402 1.


BUNDLE-FORMING PILUS ( B F P )

EPEC initial adherence to host cells involves the generalized nonintimate interaction of bacterial microcolonies in a pattern known as localized adherence (LA) [124]. This pattern of attachment requires the 69-kb EPEC adherence factor (EAF) plasmid [125-127], which is considered a common property of the classic EPEC serotypes [140]. The EAF plasmid contains a cluster of 14 genes that is sufficient to direct synthesis of BFP, a 7-nm-in-diameter type IV fimbriae associated with microcolony formation and interbacterial interactions, that tends to aggregate into bundles (Fig. 5) [128, 141-143]. Several of the proteins encoded by the bfp genes exhibit similarity to other proteins involved in the biogenesis of

Niicl«atide

F t l t i i SrtibiiRit

Type IV {Hliii*like

LI|io|>rotrl8

EAF plasmid

pmA B 0

PerC t, (BfpW)

1/

^

H>NS

^'^

~' LEEI

fCWfl llW

c««0 •%cC4 (roff9,B

arff2 0SCVN orfi5,1M$

tirc€MTmt0

LEE region Type III secretion

Intimate attachment

Off2$

Secreted proteini

Fig. 5 EPEC virulence genes: organization and transcriptional regulation. Two major groups of genes are involved in EPEC virulence. The EAF plasmid contains the bfp operon, consisting of 14 genes that encode the functions required for BFP biosynthesis. Transcription of the bfp operon requires the product of the perA (bfpT) gene. PerA (BfpT) belongs to the AraC family of transcriptional activators and is required for autoactivation of the per (bfpTVW) operon. The genes encoding the secreted proteins, the type III secretion system, and the proteins involved in intimate attachment are located in the LEE region organized in at least five operons, denominated LEEI-LEE5. Expression of these operons, except LEEI, requires the product of ler, the first gene of the LEEI operon, to overcome the negative effect of H-NS. Interestingly, Ler shares significant similarity with H-NS. In addition, PerC (BfpW) seems to positively influence the expression of ler.

9.

PATHOGENIC ESCHERICHIA COU

403

Other type IV pilus such as TCP in Vibrio cholerae and the Pseudomonas aeruginosa piH [142, 143]. In addition to the proteins encoded by the bfp cluster, BFP biosynthesis requires DsbA, an enzyme that mediates disulfide bond formation [144]. EPEC can encode at least two other types of fimbriae, although the role of these in pedestal formation and disease has not been characterized [145]. The role of BFP in the development of EPEC-mediated diarrhea has not been clearly established. Experiments with pediatric small intestinal tissue in in vitro organ culture have suggested that BFP are not involved in initial adherence but mediate interbacterial interactions that allow formation of three-dimensional bacterial aggregates at a later stage of infection [146]. However, it has been demonstrated that a functional bfpA gene, the first gene of the bfp operon that encodes the major structural subunit of BFP [ 147, 148], is required for production of BFP and full virulence in human volunteers [149]. A mutation in bfpF that encodes one of the putative nucleotide binding proteins that provide energy for pilus biosynthesis significandy affected the capacity of EPEC to cause diarrhea in humans [149]. Mutations in BfpF increased piliation, enhanced the localized adherence phenotype, and abolished twitching motility, affecting the dispersal phase of microcolony formation, which is associated with dramatic alterations in the structure of BFP bundles [ 149-151]. Other A/E pathogens do not contain BFP, but instead encode other adhesins that mediate adherence to host tissues [152]. For example, AF/Rl and AF/R2 fimbriae mediate the initial adherence of rabbit EPEC strains RDEC-1 and REPEC O103, respectively [153-156]. These differences in initial adherence might relate to their host and tissue specificity, as EPEC is a human-specific pathogen and does not infect animals [12]. 2.

EPEC-SECRETED PROTEINS (ESPS)

When EPEC interact with cultured epithelial cells, several signal transduction pathways are activated in the epithelial cells, including release of the eukaryotic secondary messenger IP3 [109, 129, 157], activation of phospholipase Cy [158], protein kinase C [159], and NF-KB [160], and phosphorylation of host proteins [161] (Fig. 4). In addition, tyrosine dephosphorylation of several host proteins, which is observed following EPEC infection, correlates with inhibition of bacterial uptake by macrophage-like cell lines [162]. EPEC binding to cultured epithelial cells also causes tyrosine phosphorylation of Tir (formerly Hp90) [133, 134]. The key to EPEC activation of eukaryotic signal transduction pathways is effector proteins secreted by the type III secretion system (described below). Strains containing mutations in genes encoding the EPEC-secreted proteins EspA {E. coli secreted protein A), EspB (formerly EaeB), or EspD do not stimulate signal transduction, cytoskeletal rearrangements, or the antiphagocytic phenotype [129, 131, 132, 162]. As for other type III effectors in bacterial pathogens, the Esps lack an amino-terminal signal sequence characteristic of sec-dependent secreted proteins [23] and require a chaperone for proper secretion [163].

404


EspA is a secreted protein with an apparent molecular weight of 25 kDa [131] that constitutes a major component of a bacterial surface organelle [164]. It was proposed that this structure serves as a delivery system for other Esps and Tir, as mutants lacking EspA cannot deliver these molecules into host cells [134, 164]. EspA does not appear to be injected into host cells. These filaments provide an essential first step in the molecular crosstalk between the bacterium and the host cell. Furthermore, the role of EspA is critical for virulence, as mutations in the espA gene render rabbit EPEC avirulent [118]. EspB (38 kDa, formerly EaeB) is translocated directly into host cells, where it is distributed between the mammalian cell membranes and cytosol, where it presumably mediates its effects [164-166]. Strains lacking EspB are unable to activate host signals, rearrange actin, or form A/E lesions [165, 167]. This molecule is central for EPEC pathogenesis, as mutants in EspB are not virulent, and are unable to form A/E lesions in an animal model [118]. How EspB mediates its effects has not been clearly defined, but experiments with HeLa cell clones transfected with the espB gene have suggested that EspB may act as a cytoskeletal toxin disrupting filamentous actin distribution and function [168]. Alternatively, it may form a pore in the host membrane to allow other EPEC effectors such as Tir to pass into the host cell [130]. EspB does not seem to be a component of the EspA filament, since this filament is produced in an espB mutant strain and anti-EspB antiserum does not stain the filaments [164]. EspD (40 kDa) is the third secreted protein involved in A/E lesion formation [132]. Upon secretion, EspD is inserted into the host cell membrane, but apparently not translocated into the cytoplasm [169]. By analogy with the Yersinia YopB protein, it has been suggested that EspD is part of the putative EPEC translocation apparatus [169]. Its role in virulence has not been established, but, given the results with EspA and EspB, it is likely to be required for disease. Another EPEC-secreted protein (EspF) was identified in 1998 [170]. This protein requires the type III secretion system for its secretion; however, EPEC mutants lacking EspF still formed pedestals on cultured cells that were indistinguishable from those seen with the parental strain [170]. Although no role for EspF could be identified using tissue culture cells, it is possible that it may play a role in vivo. Another major EPEC protein, EspC (110 kDa), is still secreted in strains lacking the type III secretion system, and instead mediates its own secretion using an autotransporter mechanism [171]. It is quite homologous to several other members of the autotransporter family, including EspP in EHEC [172], Pet in EAEC [173], and Sep A in S.flexneri [174], and shares some similarity to Neisseria gonorrhoeae'^ IgA protease and Haemophilus influenzae^ high-molecular-weight adhesins [175]. Autotransporters utilize a sec-dependent secretion system to cleave a signal peptide off as it passes to the periplasm. These proteins then insert their C terminus into the outer membrane in a (J-barrel, through which the N terminus passes out of the bacterium [176] (see Chapter 2 in this volume). Several

9.


405

members of this family, including EspC, cleave themselves, releasing a smaller secreted protein [177]. The role of EspC in EPEC pathogenesis is unclear, as mutants in EspC still form pedestals in tissue culture [171]. An EspC mutation has not been tested for virulence in a relevant animal model, and the rabbit EPEC strains do not produce an EspC homolog. EspC is not encoded within the LEE region. 3.

THE TYPE III SECRETION SYSTEM

As with many other Gram-negative pathogens, EPEC use a type III system to secrete effector proteins out of the bacteria and translocate these proteins into host cells [23] (see chapter by Silhavy and Harper for details). This type III system is encoded by >20 esc {E. coli secretion) and sep (secretion of E. coli proteins) genes, conforming at least three potential operons (Fig. 5) [138]. Work with mutant strains has shown that some of the genes are important for secretion [178, 179]; however, the specific role and function of all the predicted components of EPEC's type III secretion system remains undetermined. As with other type III secretion systems, there are also predicted chaperones encoded within the LEE to assist in secretion of the Esp proteins. For example, a chaperone for EspD, CesD (chaperone for E. coli secretion D), has been described [163], and CesT has been shown to be a chaperone for Tir [180, 181]. The main function of the type III system is to transport Tir and the Esps (other than EspC) out of the bacteria and into the host cell, although some of the Esps (such as EspA) also play a role in establishing the process of translocation into the host cell (see above). 4.

INTIMIN

Intimin, the product of the eae locus within the LEE, is a 94-kDa outer membrane protein needed for intimate adherence [182] and full virulence in both human and animal models [117, 183-186]. Intimin mutants form immature A/E lesions, characterized by diffuse actin accumulated near adherent bacteria, that is not focused and reorganized into defined pedestals beneath adherent organisms; however, epithelial signals are still activated, since the Esp effectors are still delivered to host cells [103, 133, 161]. Intimin molecules direct the final condensation and reorganization of the underlying host cytoskeleton by binding to Tir [133, 134]. The amino-terminal region of intimin is needed for export to the outer membrane and shares homology with Yersinia invasin (see Chapter 6 by Boyd and Comelis), suggesting that invasin and intimin share secretion and membrane insertion mechanisms. In contrast, its cell binding domain is located at the highly divergent C-terminal 280 amino acids (Int280) [187], which are sufficient for adherence to the EPEC protein Tir [134]. Sequence comparison of the different intimin types, which have been designated alpha, beta, delta, and gamma, has

406


revealed a nonrandom clustering of polymorphic sites mainly in the C-terminal domain, suggesting that protein divergence has been accelerated by recombination and diversifying selection [188, 189]. It has also been reported that, like invasin, intimin binds P-1 integrins [190, 191]; however, the significance of these findings remains unclear, since a more recent report indicates that P-1 integrins are not essential for intimin-mediated cell attachment and A/E lesion formation [192]. To date, a Tir-independent cell-binding activity for intimin cannot be ruled out [102]. Immunological and amino-acid sequence analysis of the cell-binding domain of different intimins has revealed the existence of at least five different intimin subtypes that may be responsible for tissue tropism [188, 193]. The global fold of Int280 has been determined by nuclear magnetic resonance, revealing that it is structured in three domains: two immunoglobulin-like domains and a C-type lectin-like module [194].

5.

TRANSLOCATED INTIMIN RECEPTOR (TIR)

The A/E lesion (or pedestal) formed by EPEC on interaction with epithelial cells is associated with the assembly of highly organized cytoskeletal structures in the epithelial cells immediately beneath adherent bacteria [98]. These structures contain actin, a-actinin, talin, ezrin, myosin light chain, and other molecules associated with polymerized actin structures (Fig. 4) [112, 195, 196]. Although this pedestal usually raises the bacterium slightly above the epithelial cell surface, EPEC can trigger extended pseudopod formation, with projections extending up to 10 microns above the epithelial cell surface, with the bacteria located extracellularly at the tip of these extensions [196]. Extended pedestals are not seen when strains containing mutations in eae, espB, tir, or type III loci are used, reinforcing the linkage between signal transduction events and cytoskeletal rearrangement. Interestingly, in contrast with other enteric pathogens that trigger cytoskeletal rearrangements, small GTP-binding proteins such as Rac, Rho, and Cdc42 do not seem to be involved in pedestal formation [197, 198]. At the tip of these structures in the epithelial membrane is Tir, where it interacts with intimin on the bacterial surface, linking the bacterium intimately to the host cell. Tir (formerly Hp90, also called EspE) is a bacterial protein that is inserted into host membranes, requiring the type III secretion system and Esps for its delivery to host cells [133, 134, 199]. Tir is predicted to be an integral membrane protein with two transmembrane domains, both N- and C-terminal domains probably facing toward the cell cytosol and a central extracellular loop [122, 134, 200]. Tir is secreted out of EPEC as a 78-kDa unphosphorylated protein. Once inserted into the host membrane, it is tyrosine phosphorylated and its apparent molecular weight shifts to 90 kDa. Alkaline phosphatase treatment reduces the 90-kDa form to 78 kDa. Unlike the Esp proteins and other type III secreted proteins, its amino-terminal methionine is missing from the bacterial secreted protein [134].

9.


407

The role of tyrosine phosphorylation of Tir in the host cell is unclear. The bacterial secreted form of Tir (unphosphorylated, 78 kDa) binds intimin in a dose-dependent, saturatable, and competitive manner. Additionally, EHEC Tir is not tyrosine phosphorylated (see later), yet it binds intimin and is interchangeable with EPEC Tir [200]. The phosphorylation of tyrosine 474 of Tir seems to be essential for actin nucleation activity, but not for the increase in apparent molecular mass observed in target cells, suggesting additional modifications [201]. Deletion analysis of Tir has led to identification of the intimin-binding domain, an extracellular loop that resides between the two previously predicted membrane-spanning regions [191, 201, 202]. Tir binding to intimin triggers additional signals in host cells, including activation of phospholipase C-y. Thus, it seems that Tir has several functions: to bind intimin, to focus the cytoskeletal rearrangements induced by EPEC, and to potentially send additional signals to the host cell [122, 134].

C. Virulence Gene Regulation

Despite the current knowledge about EPEC pathogenesis, little is known about the regulation of virulence gene expression in this microorganism. Transcriptional regulation of BFP expression occurs selectively during the exponential phase of growth in tissue culture media at 37°C, where it is modulated by ammonium concentration and temperature [203]. The coordinate regulation of the genes contained in the bfp operon is controlled by a promoter located upstream of bfpA [141, 203, 204]. Activation of the bfpA promoter requires the product of the bfpT gene (BfpT), which is the first gene of the bfpTVW operon (previously identified as the perABC locus), localized 18 kb downstream of bfpA on the EAF plasmid [205, 206]. BfpT (PerA) belongs to the AraC/XylS family of transcriptional factors [76, 205, 206], and its expression is autoregulated and modulated by the same environmental signals that regulate bfpA expression [207] (Fig. 5). The genes encoding the type III secretion apparatus {esc and sep), the EPEC-secreted effectors {esp) and the proteins involved in intimate attachment {tir, cesT, and eae) are organized in at least five different polycistronic units or operons {LEEl to LEE5) (Fig. 5) [208, 209]. Except for LEEl, expression of these operons requires the product of the ler gene, which is located at the beginning of the LEEl operon [208-210]. The Ler protein exhibits amino acid similarity with H-NS, a histone-like DNA binding protein that has been involved in negative regulation of virulence factors [138]. Ler is required to overcome the repression carried out by H-NS on expression of the promoters located upstream of the LEE2, LEE3, and LEE4 operons. These promoters are Ler independent in the absence of H-NS or on removal of negative regulatory sequences located upstream of the putative -35 boxes [210].

408


The bJpTVW (per) operon is also involved in the regulation of genes contained within the LEE region [205, 208, 211], and in the production and/or secretion of EspB and other secreted proteins, which also respond to conditions similar to those found in the gastrointestinal tract [130, 205, 212]. However, in contrast to its direct role in bfpA and Z?^7 activation [204, 206, 207], it is still not clear how this locus participates in expression of LEE-encoded genes, since transcriptional activation of LEE2 to LEE5 promoters is similar in both EPEC wild-type and EPEC strains lacking the EAF plasmid [209, 210]. In addition, it has also been suggested that initial contact with HeLa cells induces de novo protein synthesis by EPEC and activation of its type III secretion system [165, 213].

V. Enterohemorrhagic E. coli (EHEC) A.

Disease

Over the last 20 years, EHEC strains have emerged as the cause of a major health problem, particularly in developed countries [22, 214, 215]. Following ingestion of EHEC-contaminated food or water and transit to the large bowel, there is an incubation period of 3 to 4 days. The onset of disease is a nonbloody watery diarrhea, abdominal pain, and fever. Vomiting may also occur at this stage. As the disease progresses, abdominal pain increases and bloody diarrhea commences. In the majority of cases, the bloody diarrhea subsides and symptoms resolve. However, in 10-20% of cases (especially in the pediatric and geriatric populations), EHEC infections can lead to the development of serious life-threatening complications such as hemolytic uremic syndrome (HUS) (hemolytic anemia), thrombotic thrombocytopenic purpura (TTP) (decreased platelets in the blood), and renal failure [216]. It has a fatality rate of 5%, while about 25% of patients will have permanent kidney damage. The typing of EHEC strains is often based on their O (LPS) and H (flagella) antigens (e.g., 0157:H7). However, there is no evidence that either of these antigens is involved in disease. Despite the broad variety of EHEC serotypes found in the gastrointestinal tract of domestic animals, only a limited number of them (particularly 0157, 0111, and 026) are associated with the serious clinical manifestations seen during human EHEC infections [22, 216]. EHEC is acid resistant, and it is thought that, like Shigella species, acid resistance accounts for its low infectious dose (10-100 organisms) [217-219]. EHEC outbreaks have been caused by contamination of various foodstuffs, including beef, radishes, lettuce, sprouts, apple juice, salami, yogurt, and even chlorinated water [220-224]. Cattle can be asymptomatic EHEC carriers, and contamination with beef products or manure can often be traced as the source of EHEC [225-228].

9.


B.

409

Virulence Factors

The disease is currently associated with three major virulence attributes: the capacity to cause formation of A/E lesions, mediated by the genes encoded within the LEE; the expression of Shiga toxin (Stx); and the presence of a 60-MDa plasmid that encodes a hemolysin (see below) [18, 22, 229]. However, it has been observed that other pathogenic serotypes (such as 026, 0103, and 0111) have a distinctive pattern of virulence factors, with respect to that of E. coli 0157:H7 [230]. In addition, there are "atypical EHEC" strains that express Stx but do not produce A/E lesions nor possess the plasmid. It is thought that the capacity to attach tighdy to enterocytes and elicit the formation of A/E lesions contributes to the nonwatery diarrhea; this process appears necessary for intestinal colonization. Once established, EHEC secrete the Stx, which has systemic effects, causing the bloody diarrhea and HUS. 1.

SHIGA TOXIN

Also known as Shiga-like toxins (SLTs) or verotoxins (VTs), the Shiga toxins produced by EHEC strains possess high similarity to the cytotoxins produced by 5. dysenteriae [231]. The production of Stx constitutes a key element to EHEC pathogenesis and a distinctive characteristic that distinguishes EHEC strains from EPEC. Because of this, EHEC is also called STEC (for Stx-producing E. coli) or VTEC (for verotoxigenic E. coli). Two subgroups of serologically distinguishable toxins, Stxl and Stx2, have been recognized in EHEC. While Stxl is identical to S. dysenteriae Stx, Stx2 is 56% identical to the other toxins and presents a number of variant forms, such as Stx2c, Stx2d, and Stx2e [232-235]. The sequence variability, which is mainly observed in the B subunit, is reflected not only antigenically, but also in receptor binding and toxicity for tissue culture cells or in animal models [236-239]. Stxl and Stx2 toxins (except Stx2e) are encoded by bacteriophages (i.e., toxin-converting bacteriophages), which are able to spread stx genes among enteric E. coli strains [244]. Interestingly, some antimicrobial agents have been shown to cause prophage induction, which could increase the copy number and transcription of the stx genes and cell lysis-mediated toxin release [245, 246]. This process is mediated by the Rec A protein, because recA mutants showed a significant reduction in toxin synthesis and were deficient of specific phage production [246,247]. Sequence and genetic analysis has shown that the stx genes are part of an apparent Q-dependent late transcript, suggesting that toxin production and phage release would be regulated by the Q gene product, rendering maximal expression during the lytic growth of the phage [246,248,249]. Despite the common elements found for Stxl and Stx2 prophages, they show sequence and morphological differences (Stx 1 phages are more closely related to lambda), as well as different phage immunity and receptor affinity [248-250]. As mentioned above, the Stx produced by EHEC is identical to that produced by S. dysenteriae and uses the same mechanism (reviewed in [20]). Like LT, CT, and several other toxins, Stx (Stx) is an AB5 toxin, with one 32-kDa catalytic

410


subunit (A) and five 7.7-kDa binding subunits (B) [251]. The A subunit is nicked into two products (Ai and A2) that are Hnked by a disulfide bond. Toxin binding to its specific glycolipid receptor, globotriaosylceramide, or GB3, on cell surfaces, occurs by the B subunits [252]. Following binding, toxin uptake occurs by endocytosis, followed by retrograde transport to the Golgi apparatus and endoplasmic reticulum (litde is known about this unconventional endocytic routing in cells) [253]. The A subunit enters the cytoplasm, where the Al peptide acts as a specific ^-glycosidase that cleaves an adenine from the 28S ribosomal RNA [254, 255], inhibiting elongation factor 1 (EF-1 )-dependent aminoacyl tRNA binding [256]. This action blocks protein synthesis, resulting in death of intoxicated cells (Fig. 6). Intestinal cell death may result in hemorrhagic colitis (bloody diarrhea) due to a breach in the intestinal barrier, while the pathogenesis of HUS and TTP is characterized by Stx-mediated destruction of endothelial cells in venules and Stx hoioloxln

R»e#ptor blncHiia

Eitdocytosls

B sylHtfitt ^

Rdlrogracte transport C«ll mmf^mmm

Ai sylHifilt

NiekliiQ by pmtmm Fufifi

InNlbHI^m 1:^ prc^ln synthesis

28SmHA €i«purlfuitk>fi

RIlKHiom^

0oif i appififeiis

Emfopiasinlis r^tlisuliim

activity At subunit r»l««s«

Fig. 6 Structure and mode of action of Stx. The Stx holoenzyme is composed of a single catalytic subunit (A) that is associated with a pentameric ring formed by B subunits. The B subunit binds the toxin to a specific glycolipid receptor, globotriaosylceramide (Gb3). Once bound to the cell membrane, the toxin molecules seem to be internalized by receptor-mediated endocytosis, through clathrin-coated pits. Vesicles containing toxin-receptor complexes are transported to the Golgi apparatus and then to the endoplasmic reticulum, before being translocated to the cytosol (retrograde transport). The A subunit is proteolitically nicked during this process, generating the catalytically active A] amino-terminal portion, which remains linked to the C terminus by a disulfide bond. The A| subunit has an A^-glycosidase activity that depurinates the 28S rRNA of 60S, resulting in inhibition of protein synthesis and cell death.

9.


411

arterioles, which results in a thrombotic microangiopathy [257]. This damage is also believed to be due to circulating host-derived cytokines such as tumor necrosis factor alpha (TNF-a), TNF-P, interleukin-ip (IL-IP) and IL-6, which sensitize endothelial cells (ECs) to the cytotoxic action of the toxins [258-260]. This enhanced sensitivity to the toxins is probably achieved by inducing or increasing the expression of the glycolipid Gb3 receptor, thus increasing Stx binding [257, 261-266]. Animals differ in their susceptibility to Stx probably due to varying expression of Gb3. For example, rabbits infected with Stx-expressing RDEC-1 (a rabbit EPEC that encodes an LEE) showed more serious histological changes, including edema and inflammation (much like hemorrhagic colitis) than those infected with RDEC-1 alone [267]. This study suggested that Stx is responsible for the bloody diarrhea and HUS seen in EHEC outbreaks, presumably due to intoxication of intestinal and renal cells. In contrast, Stx is not an essential factor to produce disease in piglets [268]. In addition, experiments with T- or B-cell lines from different origins have led to the conclusion that lymphocytes are also susceptible to Stx and that these toxins contribute to the pathogenesis of EHEC-associated diarrhea by suppressing the mucosa-associated immune response [269, 270]. It has been suggested that Stx2 expressing strains are more likely to be associated with the development of HUS [271], although this is not always the case. Using human intestinal microvascular endothelial cells (HIMECs), it was observed that the binding affinity of Stxl was 50-fold greater than that of Stx2. Nonetheless, Stx2 was more toxic to HIMECs than an equivalent amount of Stxl, a feature that may explain the higher association of Stx2-producing STEC with cases of hemorrhagic colitis and its systemic complications [272]. 2. LEE EHEC possesses an LEE that is functionally and structurally similar to the LEE found in EPEC [136]. Interestingly, a recent characterization of the EHEC 0157:H7 LEE revealed that the molecules that are thought to be on the bacterial surface or interact with host cells (i.e., those that would be exposed to the host immune system)—such as the Esps, Tir, and intimin—are much more diverse than the others when compared to the EPEC products [273]. In addition, the EHEC LEE also encodes a cryptic prophage of the P4 family [273]. The LEE region in 0157:H7 is inserted at the same position as EPEC within the selenocysteine tRNA gene. However, other EHEC serotypes contain an LEE inserted at different positions such as the pheU locus and other undefined insertion sites [274]. Intriguingly, in contrast to what was observed with the EPEC LEE [137], the LEE from EHEC 0157:H7 was unable to confer on an E. coli K-12 strain the capacity to form A/E lesions or to secrete Esp proteins [275]. Like EPEC, EHEC 026:H~ produces a Tir (also called EspE) protein that becomes tyrosine phosphorylated on translocation into host cells, serves as the

412

JOSE L. PUENTE AND B. BRETT FINLAY

intimin receptor, and focuses cytoskeletal components beneath the adherence site [199]. In contrast, EHEC 0157:H7 Tir is not tyrosine phosphorylated and acts as the primary determinant of bacterial adherence to epithelial cells [200, 276]. As for EPEC and EHEC 026:H-, EHEC 0157:H7 Tir binds intimin and focus cytoskeletal rearrangements, suggesting that tyrosine phosphorylation is not needed for pedestal formation. Despite the remarkable heterogeneity found in the amino-acid sequence of Tir proteins from different EHEC serotypes [277], as well as between the C-terminal domain of different intimins [278], it has been shown that EHEC and EPEC intimins are functionally interchangeable; although EHEC Tir shows a much greater affinity for EHEC intimin than for EPEC intimin [200]. In vivo studies have shown that E. coli 0157:H7 requires intimin to efficiendy colonize the intestinal tract, to cause diarrhea and A/E lesions in neonatal calves, and to cause colonic edema and A/E lesions in piglets [184, 186, 279], as well as for determining the site of intestinal colonization [185]. As mentioned above, the LEE region of EHEC also encodes proteins homologous to EspA, EspB, and EspD [273], which are secreted through the type III apparatus [280-282]. As for EPEC, EHEC EspA is essential for bacterial attachment and is a part of filamentous appendages that appear during the early stages of the attachment process and are necessary for protein translocation of other effector proteins [283]. Similarly, EHEC EspD is required to obtain efficient bacterial attachment to target cells and to establish a direct link between bacteria and eukaryotic cells via EspA-containing surface appendages [284]. EspD is transferred to the cytoplasm and is also found as an integral protein of the cytoplasmic membrane of infected cells. By interacting with EspB to form a pore in the cytoplasmic membranes of the target cells, it may also facilitate translocation of the effector proteins required for A/E lesion and intimate attachment, resembling the interaction between Yersinia YopB and YopD proteins [284].

3.

THE 6 0 - M D A PLASMID

In addition to the Stx phages and the LEE region (see above), EHEC also possess a 60-MDa plasmid termed p0157 that is not found in EPEC. The contribution of this plasmid to EHEC adherence or colonization has not been clearly established. While some authors have described a plasmid-dependent adherence to epithelial cells in vitro or intestinal colonization in vivo [285-287], others have been unable to confirm these observations [268, 288, 289]. Despite the lack of consistent experimental evidence about the role of p0157 in disease, the determination of its full nucleotide sequence confirmed that it encodes several potential virulence factors, including a hemolysin (HlyA), a catalase-peroxidase (KatP), a serine protease (EspP) and a type II secretion system, as well as a protein containing a putative active site shared with the large clostridial toxin (LCT) [290, 291].

9.

PATHOGENIC E^c/zE/J/c/fM COL/

413

a. The hemolysin. The EHEC hemolysin encoded within the p0157 plasmid is highly conserved between many EHEC strains of different serotypes isolated from outbreaks [292-296]. Although its role in pathogenesis has not been clearly established, it has been suggested that it may stimulate bacterial growth in the gut by releasing hemoglobin from red blood cells, thus providing a source of iron [297]. Alternatively, one can speculate that erythrocyte lysis would lead to cellular debris that could affect renal function. The EHEC-hlyCABD (also called ehxCABD) operon codes for the EHEC hemolysin (HlyA), a protein required for HlyA activation (HlyC) [298], and the proteins that constitute the transport mechanism (HlyB and HlyD); this operon shares around 60% of identity with the alpha-hemolysin operon of uropathogenic E. coli [299-301]. EHEC hemolysin and alpha-hemolysin (discussed below) belong to the RTX family of pore-forming cytolysin toxins [299], which are widely distributed among Gram-negative bacteria [302]. b. The serine protease (EspP). EspP is a 104-kDa extracellular protein that shares significant similarity with a group of surface-associated or secreted bacterial proteins that are also known as autotransporters, which includes the IgAl protease of Neisseria gonorrhoeae, the Pet protein of EAEC (see below), and the EspC secreted protein of EPEC [172]. In contrast to the hlyA gene (see above), the gene coding for EspP is less widely conserved between EHEC isolates. Although present in EHEC 0157:H7 and 026:H~ strains, it was not detected in a significant number of EHEC isolates belonging to different serotypes (e.g., 0157:H-) [296]. This serine protease cleaves the coagulation factor V and is cytotoxic for Vero cells [172, 303]. These features have led to the suggestion that it might have a synergistic effect during the development of the hemorrhagic disease [172], a role that is supported by the presence of EspP antibodies in patients with EHEC infections [172].

4.

SIGNAL TRANSDUCTION

EHEC 0157:H7 induces rearrangements of cytoskeletal proteins such as F-actin and alpha-actinin independently of detectable tyrosine phosphorylation of Tir or host cell proteins. This suggests that EHEC (in contrast to EPEC) uses different signal transduction mechanisms to produce A/E lesions or that tyrosine phosphorylation is not important for this phenomena [200, 304]. EHEC infection of T84 cells decreases transmonolayer resistance and increases intercellular permeability, events that seem to be a consequence of the EHEC-mediated disruption of ZO-1 distribution in the tight junctions [305]. The increased epithelial permeability could lead to an alteration of the electrochemical gradients in the intestinal epithelium, resulting in diarrhea or initiation of inflammatory

414


responses [305]. As for EPEC (see above), EHEC infection also results in the release of host second messenger molecules, including Ca^"^ and inositol triphosphate [276]. Signaling responses also include PKC and MLCK activation, which seem to be involved in altering tight junction permeability in T84 cells [305]. A possible additional mechanism involved in diarrheal disease has been suggested through the study of C57BL/6J mice infected with EHEC, which showed a progressive increase in both NF-KB activation and galanin-1 receptor (Gall-R) expression by epithelial cells lining the colon. On activation, Gall-R causes Cl~ secretion, promoting fluid secretion [306]. This observation was reproduced in vitro, where galanin also increases short-circuit current (Isc) in EHEC-infected T84 cells, in contrast to uninfected cells [306].

C. Virulence Gene Regulation As in EPEC, production and secretion of secreted proteins, such as EspA and EspB, is enhanced in tissue culture media at 37°C [280]. Likewise, Tir synthesis is stimulated in bacteria grown in defined media M9 or tissue culture media, but not in complex media LB [200]. Activation of esp operon transcription is favored at high osmolarities, modulated by temperature, and influenced by the degree of DNA supercoiling [307]. As for other virulence factors, the H-NS protein and sigma S factor participate in its regulated expression, which also seems to be switched off in tighdy attached bacteria [307]. EspP secretion is reduced in tissue culture media [280] but is optimally produced in nutrient broth at 37°C and pH 7, showing a dramatic reduction at 20°C and pH 5 [172].

Vl. Enteroaggregative E. coli (EAEC) A.

Disease

EAEC is associated with persistent pediatric diarrhea in developing countries and is characterized by its aggregative adherence pattern (for recent reviews see [308, 309]). EAEC causes a watery secretory diarrhea, often mucoid in nature. There is often low-grade fever, but no vomiting [310, 311]. Grossly bloody stools can occur, although the majority of cases do not involve bloody diarrhea [312]. Formation of a thick mucus gel on the intestinal mucosa and mucosal damage probably mediated by a mucosa damaging toxin are pathogenic features of EAEC histopathology [313, 314]. EAEC often affects children [310, 312], although adults cases have also been reported [315, 316].

9.

415


B. Virulence Factors EAEC strains are a heterogeneous collection of pathogenic E. coli that share certain chromosomal and plasmid-bome genes [18, 308, 317]. Their defining feature is that they adhere to cultured HEp-2 cells in small clumps or aggregates ("aggregative adherence," or AA), resembling a stacked-brick configuration [318] (Fig. 7). As expected, adherence to host cells and neighboring bacteria is mediated by fimbrial adhesins encoded on a large plasmid. EAEC can increase mucus secretion, leading to a blanket of adherent bacteria trapped in a layer of mucus. Some EAEC can cause tissue damage, resulting in villus atrophy and other cytotoxic effects that are probably mediated by toxins, although the contribution of each toxin to disease has not been defined (see below). Invasiveness of tissue culture cells by some EAEC strains has been suggested [319]. However, although EAEC seem to be able to colonize many regions of the gastrointestinal tract, as studied in human intestinal explants, these explants do not show bacterial intemaHzation [313].

Kkf m»!brbr

Mucîst icHvtty

cNanttft

SeriMii r««l«t*nc«^^î Pic H«fii»9gltitlfi«tlofi

V, Fig. 7 EAEC interactions with intestinal cells. AAF fimbriae mediate the initial adherence of EAEC strains to the intestinal mucosa in a stacked-brick configuration (aggregative adherence). Colonization enhances mucus production, leading to accumulation of a thick mucus layer where bacterial cells remain embedded. During this process, EAEC delivers different toxins that damage the mucosa and promote intestinal secretion. The heat-stable enterotoxin (EASTl) is related to ETEC ST and may act similarly (see Fig. 1). Pet and Pic belong to the SPATE (serine protease autotransporters of the Enterobacteriaceae) subfamily.

416


1.

ADHERENCE FACTORS

EAEC produce at least two fimbrial adhesins: aggregative adherence fimbriae I and 11 (AAF/I and AAF/II, respectively). AAF/I mediates adherence to tissue culture cells and human erythrocytes and is encoded by two regions on the 60-MDa plasmid [320, 321]. One region encodes the fimbrial structural gene and assembly genes (such as chaperones) [322], while the other is a regulatory region encoding an AraC-like regulator named AggR [323]. AAF/I is a member of the Dr family of adhesins, which adhere to the Dr blood group antigen. AAF/I are flexible fimbriae 2-3 nm in diameter and are capable of forming bundles, although they do not share homology with the type IV bundle-forming pilus of EPEC [320]. AAF/I fimbriae are produced by a small number of EAEC strains. AAF/II are a second fimbriae produced by some EAEC strains [324]. Like AAF/I, they are encoded by two regions (separated by 12 kb), although, unlike AAF/I, the structural subunit and the assembly genes are encoded in separate regions (a characteristic of the Dr family of adhesins) [325]. The AAF/I and AAF/II subunits are 25% identical. AAF/II fimbriae are filaments 5 nm in diameter arranged in semirigid bundles that may play an important role in adherence, as strains lacking AAF/II are no longer capable of adhering to human intestinal tissue [324]. 2. EAST-1 Many EAEC strains produce a plasmid-encoded heat-stable toxin designated enteroaggregative E. coli heat-stable toxin 1 or FASTI [326]. This toxin shares identity with STa (see above) and probably works in a similar manner by activating guanylate cyclase, leading to secretory diarrhea. It is composed of 38 amino acids (4.1 kDa) and contains four cysteine residues (like guanylin, but differing from STa, which has 6) that form disulfide bonds to stabilize the toxin [327, 328]. However, the contribution to disease by FASTI has not been established. Many E. coli strains produce FASTI, including 0157, several ETEC and even EPEC [326] strains; however, clinical isolates of EAEC that do not produce FASTI are common, and even nonpathogenic E. coli produce FASTI, indicating that it alone is not sufficient to cause disease [326]. Much like ETEC, it is likely that EAEC possess multiple adhesins and toxins that, in various combinations, contribute to disease.

3.

PET

AND P I C

EAEC also produce a 108-kDa plasmid-encoded enterotoxin (Pet) that may contribute to disease [173]. Purified Pet is capable of causing rises in short-circuit current (Isc) and falls in tissue resistance, characteristic of toxins that causes secretory diarrhea [329]. In addition, EAEC causes increased mucus release and exfoliation of cells, processes seen in EAEC disease [329], and is also able to

9.

PATHOGENIC E5cw£/?/cwM COL/

417

elicit cytoskeletal changes in epithelial cells [330]. The cytopathic and enterotoxic effects induced by Pet are ascribed to its protease activity [330]. However, as with EASTl, not all EAEC strains produce this toxin [317], although it may contribute to the pathogenicity of some strains during infection [315]. Some EAEC strains also produce and secrete a chromosomally encoded 110-kDa protein denominated Pic (protein involved in intestinal colonization), which is synthesized as a 146.5-kDa precursor molecule that is processed at the N and C termini during secretion. Pic is presumably involved in mucinase activity, serum resistance, and hemagglutination [331]. Pet and Pic belong to the autotransporter class of bacterial proteins that mediate their own secretion, exhibiting most sequence similarity to a subgroup termed the SPATE subfamily (serine protease autotransporters of the Enterobacteriaceae) that includes EspC from EPEC and EspP from EHEC (see above) [175, 331].

VII. Diffusely Adhering E. coli (DAEC) A.

Disease

E. coli strains showing a diffuse adherence pattern in tissue culture cell assays are known as DAEC [125, 318, 332]. With the characterization of EAEC, DAEC is now recognized as a separate class of E. coli, although little is known about their virulence mechanism and their association with diarrheal disease remains controversial. While some studies clearly establish a link between DAEC strains and diarrheal disease [333-335], others suggest that there is no such association [312, 336]. DAEC strains are also considered a heterogeneous group that comprises strains with different pathogenic potential due to the presence of variable virulence factors (see below). DAEC strains have been more frequently associated with persistent watery diarrhea in children between the ages of 2 and 5 years [337, 338]. B. Virulence Factors Diffuse adherence has been associated with four different adhesins, while toxins have not been described in any detail. DAEC strain CI845 expresses the F1845 fimbria [339], a member of the Dr family of adhesins [340]. This fimbrial adhesin uses the membrane-associated decay-accelerating factor (DAF) as receptor [341]. Interaction of F1845-expressing strains with DAF induces elongation and nucleation of microvilli in differentiated Caco-2 cells [341] and formation of long thin membrane projections in Hep2 cells [342, 343], and promotes F-acting rearrangements in INT407 cells, a process that involves recruitment of signal transduction molecules [344]. It has also been shown that some daaC positive strains (daaC codes for a molecular usher of F1845 fimbriae [345]) secrete Esp homologs, which are necessary to induce signal transduction events and the A/E phenotype in EPEC and EHEC [346]. The expression mechanism of the F1845 fimbrial genes, which may be

418


encoded on a plasmid or the chromosome, has been a model system for the study of fimbrial-regulated expression by mRNA processing [339, 345, 347]. According to some studies, F1845 positive strains seem to be widely distributed around the world [18]; however, others suggest that this fimbria is rare among DAEC strains [348]. DAEC strain 2787 of serotype 0126:H27 expresses a 100-kDa outer membrane afimbrial adhesin denominated AIDA-I [349, 350], which belongs to the family of outer membrane autotransporters [351]. This plasmid-encoded adhesin is synthesized as a 132-kDa precursor molecule that is processed at the C terminus by an autocatalytic cleavage mechanism, and seems to require a 45-kDa cytoplasmic protein for its correct maturation [352, 353]. In addition, a 57-kDa mannose-resistant hemagglutinin [354] and two major surface proteins of 16 and 29 kDa [333] have been associated with diffuse adherence.

VIIL Uropathogenic E. coli A.

Disease

Uropathogenic E. coli (UPEC) strains are responsible for approximately 80% of community-acquired and 30% of nosocomial-acquired urinary tract infections (UTIs). Females under 10 years of age, or between 18 and 40, are at the highest risk for community-acquired infections. Infections in children are often due to blockages in the urinary tract, resulting in pools of stagnant urine. Similarly, nosocomial infections are usually associated with indwelling urinary catheters, resulting in loss of flushing action of urine, which, in some cases, can proceed to a systemic infection. Aside from the distal tip of the urethra, the urinary tract is usually sterile. UPEC can reside in the colon and then be introduced into the urethra. UTIs result from ascending colonization of the urinary tract by these strains. Infections can occur in the urethra (urethritis), bladder (cystitis), and kidneys (pyelonephritis), and, under some conditions, the microorganisms enter the blood stream. Disruptions of the normal flora (such as the use of vaginal spermicides) or direct inoculation (e.g., during intercourse) can cause infections. The clinical symptoms of UTIs vary depending on the region of the urinary tract that is colonized. Painful or difficult urination (dysuria) and sometimes suprapubic pain characterize cystitis. Pyelonephritis includes the above symptoms and fever, flank pain, shivers, and sometimes vomiting. Treatment includes antibiotic therapy, although there is increased incidence of antibiotic-resistant organisms, especially in nosocomial infections. Catheter removal facilitates treatment in most cases. B. Virulence Factors In order to successfully colonize and establish a UTI, UPEC strains take advantage of an assortment of virulence properties [355, 356]. Urine flow is the

9.

419


largest obstacle facing E. coli attempting to colonize the urinary tract. Not surprisingly, UPEC encode several adhesins, both fimbrial and afimbrial in nature, which facilitate adherence to uroepithelial cells. However, like most adhesins, it has been difficult to precisely define the role of any particular adhesin due to overlapping function (redundancy). Type I and P fimbriae, the most common fimbriae found in UPEC strains, enhance virulence and are involved in initial urethral colonization (see below) (Fig. 8). In addition, many UPEC produce hemolysin, which may be involved in kidney disease, and the CNF-1 toxin. Certain UPEC strains possess iron sequestration systems to assist in growth, whereas others produce a capsule (usually Kl or K5) that may help avoid clearance from the urinary tract. Lipopolysaccharide also

Inm acqittiittciii s?«l«iii

f

& meinbmne ruffliiig

Uroi^lakiii

Mr^M^!*?iMi5^H R0l«ase of cytokines

Fig. 8 UPEC interaction with uroepithelial cells. Most UPEC strains express P or type I pili. P pili bind to a glycolipid Gala(l,4)Gal, while type I pili bind to mannose-containing glycoprotein receptors, known as uroplakins. The components of both pili share functional similarities, although the tip fibrillum of P pili is more complex. Coupling of the P pili PapG adhesin with its surface receptor activates expression of a bacterial iron-acquisition system by inducing the expression of AirS, a sensor-regulator protein. In addition, this interaction also stimulates the intracellular release of ceramides, which leads to cytokine production after activation of protein kinases. In addition, the toxic activity of hemolysin and CNF-1 may contribute to the kidney damage seen in pyelonephritis. When hemolysin binds calcium, it can be inserted into host membranes to form a pore, which eventually lyses the host cell. CFN-1 induces profound changes on the cytoskeleton of epithelial cells, such as actin reorganization and membrane ruffling, by modifying the GTPase activity of Rho.

420


appears to be involved in cytokine induction [357]. In general, it appears that the more virulence factors a uropathogenic E. coli strain possesses, the more severe the disease symptoms. 1.

TYPEIPILI

Type I pili, originally identified by their ability to mediate mannose-sensitive agglutination, are expressed by the majority of E. coli strains derived from patients with cystitis and pyelonephritis. Despite the extensive knowledge on type I pili, their role in disease is controversial. It is unlikely that type I pili play a role in gastrointestinal disease, although many enteric E. coli pathogens and nonpathogenic fecal isolates produce this type of pili. However, there is good evidence that it is involved in colonization of the oropharynx that contributes to invasive diseases (sepsis and meningitis) in neonates. There is also good evidence that type I pili contribute to infections of the lower urinary tract, such as cystitis, by mediating colonization of the bladder mucosa, promoting bacterial persistence, and enhancing the inflammatory response to infection, thus increasing UPEC virulence [15]. Type I pili have a composite structure of 7 nm in width and 1-2 microns in length and consist of a long rigid rod and a distal thin fibrillar structure [358]. Expression of type I pili requires at least nine genes in ihtfim cluster, encoding products that are assembled in a manner very similar to P pili (see below). FimH is the adhesin protein responsible for binding to mannosylated glycoproteins [359, 360] and is located at the distal tip of the heteropolymeric type-1 pilus rod, which is predominandy constituted by FimA subunits [358]. FimG and FimF are minor components that are probably needed as adaptors, initiators, or terminators, FimC is the chaperone and FimD the outer membrane usher [358, 361, 362]. FimH mediates E. coli binding to mannose-containing glycoprotein receptors, known as uroplakins, that are located on the luminal surface of the bladder epithelial cells [16]. Antibodies to FimH reduce colonization of the bladder mucosa and disease in a murine cystitis model, suggesting that a FimH-based vaccine may provide a means of preventing these infections [363-365]. FimH may have another role in promoting bacterial survival in the host by mediating binding to macrophages. This interaction, probably through CD48 receptors located on the macrophage surface, triggers a different endocytic pathway that appears to increase the ability of E. coli to survive within them, possibly by affecting phagolysosomal fusion with vacuoles containing UPEC [366]. 2.

PAP

One of the best-studied examples of pili and its associated assembly machinery is P pili (pyelonephritis-associated pili), which are encoded by pap genes found within pathogenicity islands (see below). As its name suggests, this adhesin is critical for upper UTI infections (pyelonephritis). The pap operon is a useful

9.

PATHOGENIC £'5C//£/?/c//M COL/

421

example of pilus assembly since it contains many conserved features that are found among various pilus operons, including type I pili (reviewed in [367-369]). Two molecules guide newly synthesized pilus components to the bacterial surface (reviewed in [370, 371]). PapD, a conserved chaperone molecule with an immunoglobulin-like domain, is necessary to transport several pilus subunits from the cytoplasmic membrane to the outer membrane [372, 373]. PapD-subunit complexes are targeted to the PapC outer membrane (OM) usher, which forms a pore through which the pili are translocated across the OM [374]. The major subunit of the pilus is PapA, which is assembled into a 6.8-nm thick helical rod that is anchored in the outer membrane by PapH [375-377]. At the distal end of the pilus rod is a 2-nm linear tip fibrillum composed of PapE [378], which is adapted to the PapA rod by PapK [379]. The actual molecule that mediates adherence (i.e., tip adhesin), PapG, is joined to the PapE tip fibrillum by the adapter protein PapE [379]. PapG mediates binding to the a-D-galactopyranosyl-(l-4)-P-D-galactopyranoside (Gala(l,4)Gal) moiety present in a globoseries of glycolipids found on host cells lining the upper urinary tract and erythrocytes [380-382]. There are three adhesin variants of PapG—G-I, G-II, and G-III—which recognize three different but related Gala(l,4)Gal receptors. It is thought that the distribution of these receptors differs among hosts and tissues, and differential expression of the PapG adhesins at the pilus tip could determine tissue and host specificity. Although the host receptor varies for different bacterial pili, the general features of the P pilus operon are conserved in many other pilus systems, and components are often interchangeable. For example, the PapD chaperone can modulate the assembly of type I pili, which mediate binding to mannose-containing molecules on the host cell surface. The gene organization of related pilus operons is also usually conserved. Thus, while type I and P pili are encoded by similar operons and functionally analogous sequences that can be aligned, they bind to quite different carbohydrates on the cell surface (reviewed in [367, 370]). One of the most interesting features of PapG-mediated interactions with its Gala(l,4)Gal-containing glycolipid receptor is the ability to activate specific responses in the bacteria and in the epithelial cell that promote virulence [383]. For example, AirS is a sensor-regulator protein that is expressed only when PapG binds to its receptor, thereby allowing activation of the UPEC iron-acquisition system. Efficient iron acquisition allows UPEC to grow in urine, an otherwise growth-limiting environment [384]. In addition, this interaction also triggers the intracellular release from receptor glycolipids of ceramide, an important second messenger that can activate cytokine production, through the activation of serine/threonine protein kinases and phosphatases [385-387]. 3.

AFIMBRIAL ADHESINS

E. coli produce a family of adhesins that bind to the same mammalian receptor. Members of this family include F1845 and Dr, which are fimbrial adhesins found

422


in gastrointestinal pathogens, and two nonfimbrial adhesins (afimbrial adhesins Afa-I and Afa-III) expressed by many uropathogenic E. coli [339, 388, 389]. All four of these adhesins use as receptor the Dr^ blood group antigen present on decay accelerating factor (DAF) on erythrocytes and other cell types, although they appear to recognize different epitopes of the Dr antigen [340]. Much like other fimbrial adhesins, expression and production of these adhesins requires 5-6 gene products. These include a periplasmic chaperone, an outer membrane anchor protein, 1 to 2 transcriptional regulators, and the adhesin [388, 390, 391]. At least for Afa-III, there appear to be two adhesins (AfaD and AfaE) encoded within this operon [392]. The AfaE and AfaD confer upon an E. coli laboratory strain the ability to bind to HeLa cells and to mediate internalization of the adherent bacteria, respectively [393]. Some adhesins of this family (F1845 and DR) form fimbriae, whereas others form nonfimbrial adhesins on the bacterial surface (Afa-I and Afa-III). This appears to be dictated by the sequence of the adhesin molecule because switching the genes encoding these adhesins switches the adhesin type [392]. It is possible that the E. coli afimbrial adhesins have evolved from the related fimbrial adhesins, but have been altered such that the properties needed to polymerize a pilus are missing, yet the adhesin domain remains anchored on the bacterial surface. 4.

HEMOLYSIN

Approximately half of UPEC strains that cause upper UTIs, about a third of those that cause lower UTIs, and only about 10% of fecal isolates produce a hemolysin (HlyA) that belongs to the RTX (repeats in toxin) family and shares similarity with the hemolysin described in EHEC (see above) [299, 302]. Four genes (hlyABCD) are required for the production and export of the hemolysin: hlyA encodes the hemolysin structural gene; HlyC acylates HlyA posttranslationally, adding fatty acids to two internal lysine residues via amide linkages [394]. It is thought these acyl chains anchor HlyA to host cell lipid bilayers or assist in HlyA oligomerization in host bilayers. The HlyB and HlyD proteins are needed for HlyA secretion out of E. coli, as is the TolC protein. When calcium binds to HlyA, this toxin is capable of inserting into host membranes to form a pore, which eventually lyses the host cell. Such toxic activity may contribute to the kidney damage seen in pyelonephritis. 5.

CYTOTOXIC NECROTIZING FACTOR

1 (CNF-1)

About one-third of UPEC strains produce cytotoxic necrotizing factor 1 (CNF-1), a 1014-aa cytotoxin that is also produced by some gastrointestinal E. coli (see above) [395, 396]. CFN-1 induces profound changes in the cytoskeleton of epithelial cells, including actin reorganization and membrane ruffling [397399], and impairs migration and proliferation of bladder cells that could interfere with repair of the bladder epithelium [400]. It has been demonstrated that CFN-1


423

interacts with Rho, a GTP-binding protein whose GTPase activity becomes constitutive on CFN-1-mediated deamidation of glutamine 63 [401^04]. However, a role for this toxin in urinary tract infection has not been defined. 6.

PATHOGENICITY ISLANDS

Like the LEE region seen in EPEC and EHEC, uropathogenic E. coli contain pathogenicity islands, which provides an excellent example of the role these regions play in acquisition of specific virulence factors [3, 405, 406]. The uropathogenic strain 536 contains two large unstable pathogenicity islands: PAI-I (70 kb) and PAI-II (190 kb) (Table II) [407]. Both pathogenicity islands are flanked by short (16-18 bp) direct repeats, which are likely responsible for their deletion due to recombination at a frequency of 10~^ [408]. PAI-I is inserted at the same site as LEE in EPEC and 0157:H7 EHEC (immediately downstream of selC at 82 min on the chromosome) and encodes, among other genes, the hly hemolysin operon. PAI-II is inserted at the leuX tRNA locus at 97 min on the chromosome, and encodes another hly operon and the prf (P-related fimbriae) pilus operon [3]. Like the LEE region, the G-HC content of these regions is lower (41%) than that of E. coli K-12 (51%). Similarly, another uropathogenic strain of E. coli (J96) contains two pathogenicity islands inserted in two different tRNA genes (pheV at 64 min for PAI-IV and pheR at 94 min for PAI-V) [409]. Each of these islands encodes an hly hemolysin operon; this is in addition to single copies of a pap operon (PAI-IV) and prs (P-related sequence) and cnfl for PAI-V. In addition, E. coli CFT073 possess the smallest of the five PAIs described (50 kb), which is inserted in the vicinity of the metV gene and carries an hly operon [410]. 7.

VIRULENCE GENE REGULATION

Like most virulence factors. Pap pili expression is tightly regulated in response to several environmental and nutritional factors, including temperature, amino Table II

Pathogenicity Islands of Uropathogenic E. coli

Strain

Pathogenicity island

536 536 J96 J96

Pail Paill PailV Pai V

CFT073

Pai VI

Genes hly hly, prf hly, pap hly, prs, cnfl hly

Size (kb)

Map location (min)

tRNA

70 190 170 110

82 97 64 94

selC leuX pheV pheR

50

26/63

metV

424


acids, and glucose, and modulated by different regulatory factors such as the leucine-responsive regulatory protein (Lrp), CRP (cAMP-binding protein), H-NS, Papl, and PapB. Pap pili expression is also controlled by a methylation-dependent phase variation mechanism (on/off switching) in response to these conditions [412, 413]. The promoter region of the pap operon contains two GATC sites that can be methylated by deoxyadenosine methylase (Dam) (Fig. 9B). GATC boxes I and II overlap with the DNA-binding sites for the leucine-responsive regulatory

Major i sutHintt I

Pap operon

I Milortip I comf>ofient

OM chaperone Periplasmic ch«{>«fOfi«

Rugulatton

RN«9e E

B

D»fti

Pha$# ON

Dam

papl Phase OFR

Vhkm

I RNA polymera^ê # l>apB HCIRP # Pa{>l r/-colonized persons never develop any clinical symptoms, persistent colonization can also be viewed as a commensal process, perhaps analogous to colonization by Bacteroides spp., alpha streptococci, and lactobacilli. In addition, there has been speculation that //. pylori colonization could be beneficial for humans in some circumstances. These different perspectives illustrate the complexity of interactions between H. pylori and humans, and the lack of current consensus about how these bacteria should be viewed. Regardless of whether the term "infection" or "colonization" is used, the salient feature is that //. pylori are extremely well adapted to life in the human stomach. //. pylori occupy an interesting position in the history of medicine and biomedical research. Bacteria were detected by microscopy in human gastric tissue more than a century ago, but these observations were essentially forgotten. Only since 1982 has there been interest in the idea that these bacteria might cause disease [1, 2]. Reports that gastric inflammation and peptic ulcers might be the consequences of a bacterial infection [3] were initially met with considerable skepticism in the medical community. However, human volunteer studies have shown that //. pylori ingestion indeed results in gastric inflammation [4-6], and most gastroenterologists now agree that //. pylori play a major role in the pathogenesis of peptic ulcers [7] and gastric cancer [8]. In this chapter, we review our current understanding of how //. pylori persistently colonize the human gastric mucosa and discuss factors that determine whether infections progress to clinically evident disease or remain asymptomatic.

//. Epidemiology H. pylori are present in nearly all human populations throughout the world. In the United States, about 30-50% of adults are persistendy colonized, whereas in

11.

511

H. PYLORI PATHOGENESIS

Table I

Histological Features and Clinical Diseases Associated with H. pylori

Condition

Definition

Estimated lifetime incidence in H. pylohpositive persons in the U.S.

Predisposing factor

Gastritis

Inflammation of the gastric mucosa

Duodenal ulcer

Defect in the duodenal mucosa

Gastric ulcer

Defect in the gastric mucosa

Atrophic gastritis

Loss of gastric glands

5-10%

Prolonged H. pylori carriage

Intestinal metaplasia

Intestinal epithelium in gastric mucosa

5-10%

Prolonged H. pylori carriage

Gastric epithelial dysplasia

Abnormal epithelial morphology and organization

n-infected humans is not known.

B. Role of the cag Pathogenicity Island A number of cytokines—including IL-lp, IL-2, IL-6, IL-8, IL-10, and TNF-a— are present in higher concentrations in the gastric mucosa of //. /7j/or/-positive persons than of//. /?y/on-negative persons [188, 189]. Several of these cytokines are proinflammatory, and IL-8 in particular has potent chemotactic and stimulatory properties for neutrophils. Levels of IL-8 are significandy higher in the

11.


531

gastric mucosa of patients infected with cag"" H. pylori strains than in patients infected with cag' strains [190], a difference that accounts, at least in part, for a more intense neutrophilic mucosal inflammatory response in patients with cag'^ strains compared to those with cag~ strains [101, 190]. The capacity of H. pylori to induce cytokine production has been explored by studying interactions of H. pylori with epithelial cells in vitro [104-108]. Direct contact of viable H. pylori with gastric epithelial cells in vitro results in the synthesis and release of several proinflammatory cytokines, including IL-8, and H. pylori strains containing the cag pathogenicity island are much more efficient in stimulating epithelial IL-8 expression in vitro than are cag~ strains. Adherence of cag'^ H. pylori strains to epithelial cells stimulates a cascade of signalling events that include activation of N F - K B , which leads to induction of IL-8 gene transcription [191-195]. Insertion mutations in many (but not all) genes of the cag island markedly diminish this induction of cytokine production [104-108]. Five proteins encoded by the cag pathogenicity island have substantial homology to components of type IV bacterial secretion systems, including Ptl proteins of Bordetella pertussis (which mediate secretion of pertussis toxin), and Vir proteins of Agrobacterium tumefaciens (which mediate T-DNA transport into plant cells). It is currently thought that the cag pathogenicity island encodes products that secrete an IL-8-inducing factor, but this putative factor is yet to be identified. Genes in the cag pathogenicity island do not have the same conserved contiguous arrangement found in the vir and/?^/ operons, and homologs of many Vir/Ptl genes are not seen in the H. pylori genome. Whether H. pylori homologs of Vir and Ptl have functional roles similar to those of the corresponding proteins in Bordetella and Agrobacterium, and what other H. pylori proteins are involved, is not known.

/X. Interactions of H. pylori witti thie Gastric Epittielium A.

Cytoskeletal Changes and Tyrosine Phosphorylation

Adherence of H. pylori to gastric epithelial cells in vitro can result in several changes in cell architecture, including the formation of adherence pedestals and effacement of microviUi at the site of bacterial attachment [107, 108, 196]. Actin, alpha-actinin, and talin cytoskeletal elements are rearranged direcdy beneath adherent bacteria [107, 108]. These changes seem similar to those caused by adherence of enteropathogenic E. coli (EPEC) to intestinal epithelial cells. However, H. pylori does not possess genes with significant homology to the EPEC genes that mediate these effects.

532

TIMOTHY L. COWER ETAL

Adherence of cag'^ H. pylori to cultured cells is also associated with tyrosine phosphorylation of a 145-kDa protein [107, 108]. This 145-kDa protein has been identified as H. pylori CagA, which is delivered into epithelial cells by the cag type IV secretion system [196a]. The kinase inhibitor staurosporine inhibits IL-8 induction but not tyrosine phosphorylation, which suggests that these represent two distinct pathways [107, 108]. B.

Apoptosis

Integrity of the mucosa in the gastrointestinal tract depends on a balance between production of new cells (proliferation) and cell loss. Apoptosis (programmed cell death) is an important mechanism for limiting unrestricted proliferation of epithelial cells, and many bacterial and viral pathogens can affect this process. The rate of apoptosis is increased in the gastric epithelium of H. pylori-infccied persons, compared with noninfected persons [197]. Apoptosis also is detected when gastric epithelial cells bind H. pylori in vitro [172, 198, 199]. In KatoIII cells (a gastric epithelial line), apoptosis results from binding of H. pylori to class II MHC components on the cell surface [172]. The expression of class II MHC components, in turn, is upregulated by adherence of H. pylori [200]. That there may be multiple pathways by which H. pylori induces apoptosis is indicated by the finding that H. pylori induces apoptosis in AGS cells (a gastric epithelial cell line that does not express class II MHC components) [201]. The rate of gastric epithelial cell proliferation is found to be higher in H. pylori-inftcied persons than in noninfected persons in some [202] but not all studies [203]. These seemingly conflicting reports might reflect important differences, such as presence or absence of the cag pathogenicity island, among H. pylori strains. In one study, cagA'^ strains were associated with higher rates of epithelial proliferation and lower rates of apoptosis than cagA' strains [204]. Both the rates of cell division and apoptosis may be modified by the complex cytokine network. For example, increased inflammation might induce enhanced cell proliferation. These effects of H. pylori on epithelial proliferation and apoptosis may be relevant to the development of gastric neoplasms that arise during the course of long-standing H. pylori infection.

X. Vacuolating Cytotoxin A.

Structure

Among H. pylori isolates from patients in the United States and Western Europe, about half produce detectable vacuolating toxic activity in a HeLa cell assay [205] (Fig. 4). This effect is mediated by a ~90-kDa secreted protein (VacA) [206-208]. Sequence analyses of the vacA gene revealed an ORF that encodes a predicted

11.


533

Fig. 4 Vacuolation of HeLa cells induced by the H. pylori vacuolating toxin (VacA). HeLa cells were incubated with H. pylori culture supernatant containing VacA, and then stained with crystal violet. Prominent intracellular vacuoles form in response to VacA.

product 140 kDa in size [209-211]. Proteolytic cleavage of the VacA protoxin at its N and C termini yields a 33-aa signal sequence, a mature 90-kDa secreted protein, and a C-terminal fragment that remains localized in the bacterial cell [209-211]. VacA lacks extensive homology with other known bacterial toxins, but its processing and secretion resemble those of the IgA protease family of secreted proteins [209,211]. Purified VacA migrates as a ~90-kDa protein under denaturing conditions, but in its nondenatured state it exists as a large six- or seven-sided complex, comprised of 12 or 14 identical 90-kDa subunits [212, 213]. During prolonged storage, the 90-kDa monomers undergo specific cleavage into 34- and 58-kDa fragments, which remain physically associated in an oligomeric complex [210213]. These two fragments may represent distinct domains or subunits of VacA.

B.

Allelic Variation in vacA

Most H. pylori strains that lack vacuolating activity for HeLa cells in vitro nevertheless produce and secrete an immunoreactive vacA product, albeit in lesser

534

TIMOTHY L. COVER ETAL

amounts than tox+ strains [92, 214]. These apparently tox~ strains also commonly elicit anti-VacA antibody responses in vivo [215]. Decreased production and secretion of VacA by tox~ strains seems to be attributable in part to either decreased vacA transcription or decreased stability of vacA transcripts in tox~ strains [92, 214]. In addition, there are potentially important differences between the vacA sequences of tox^ strains and tox~ strains [92]. Two major families of alleles (designated si and s2) are distinguished by differences in vacA signal sequences [92]. Essentially all isolates with type s2 vacA signal sequences fail to produce detectable vacuolation of HeLa cells. In addition, two families of alleles (designated ml and m2) are distinguished by striking sequence differences in the mid-portion of vacA [92]. Strains with type ml alleles typically produce more prominent vacuolation of HeLa cells than strains with type m2 alleles. Some strains with type m2 alleles produce vacuolation of RK-13 cells but not HeLa cells, which suggests that variation in VacA sequences may be related to cell-type specificity [216].

C. Mechanism of Action Purified VacA induces cell vacuolation when microinjected into HeLa cells, and, similarly, HeLa cells transfected with plasmids containing the vacA gene develop intracellular vacuoles [217]. These results provide strong evidence that vacuole formation is the consequence of an interaction between VacA and an intracellular target. In general, bacterial toxins that interact with intracellular targets are thought to act via a series of events: binding to the plasma membrane, internalization and translocation into the cytoplasm, and enzymatic modification of the intracellular target [218]. This framework provides a useful approach for understanding VacA action. Binding of VacA to cells probably involves interaction with a specific cell-surface receptor, as well as with an assortment of negatively charged lipids [219-221]. Probably both types of binding interactions are mediated primarily by amino-acid sequences within the domain of VacA corresponding to its 58-kDa proteolytic fragment [222]. How VacA gains access to the cytoplasm of cells is not yet known. VacA can insert into membrane vesicles, which suggests that toxin translocation might occur direcdy across the plasma membrane [221]. However, VacA is internalized slowly into vesicular compartments at 37°C [222], which suggests entry by receptor-mediated endocytosis. Nearly all bacterial toxins that act intracellularly do so by enzymatically modifying intracellular target molecules. No enzymatic activity has yet been identified for VacA. Interestingly, transfection of HeLa cells with plasmids encoding either of the two putative VacA subunits (34- and 58-kDa fragments) fails to induce cell vacuolation [217, 223], which suggests that a large portion of the 90 kDa VacA molecule is required for any putative enzymatic function.

11.

/ / . PKLO/?/PATHOGENESIS

535

The activity of VacA is increased markedly by exposure of the protein to acidic pH before addition to cells [224]. Acid activation is associated with conformational changes, including disassembly of the dodecameric VacA structure into monomeric subunits [213] and increased exposure of hydrophobic domains [225]. Acidification of VacA enhances its insertion into lipid membranes [225], and is associated with the formation of ion-conductive channels in lipid bilayers [220, 221, 226]. These phenomena may help to explain why low pH increases the toxin's vacuolating activity. The effects of low pH on VacA activity may be particularly relevant in the acidic gastric environment, where secreted VacA could be activated before interacting with epithelial cells. The cell vacuoles induced by VacA become visible within 2 hours after addition of high concentrations of the toxin to cells in vitro. The vacuole membranes contain both the small GTP-binding protein Rab7 (a late endosomal marker) and the membrane glycoprotein LgpllO (a lysosomal marker), which suggests that the vacuoles represent postendosomal hybrid compartments [227, 228]. HeLa cells overexpressing dominant negative Rab7 mutants do not develop vacuoles on exposure to VacA, indicating that functional Rab7 is required for VacA-induced vacuolation [229]. Accordingly, VacA may disrupt normal membrane trafficking at or near the level of late endosomes [230], perhaps by modifying a cellular constituent that normally regulates membrane trafficking within the endocytic pathway [231].

D. Role of VacA in Vivo As a general rule, protein toxins produced by bacterial pathogens may facilitate colonization of the host, enhance transmission to new hosts, or cause damage to host tissue. In one study, the rate of colonization of gerbils by H. pylori vacA null mutants was slighdy lower than by wild-type strains [232], but both types of strains persisted for several weeks in this model. In contrast, no differences between mutant and wild-type strains were detectable in short-term colonization of gnotobiotic piglets [233]. These data suggest that VacA plays a relatively minor role in early colonization events. Nevertheless, immunization of mice with VacA provides effective protective immunity against subsequent experimental challenges with H. pylori [174]. It seems more likely that VacA may play a role in promoting persistence of H. pylori in the gastric mucosa. VacA interferes with the processing involved in antigen presentation in vitro [234], which suggests that one action of this toxin may be to enable H. pylori to resist clearance by the host immune system. VacA also selectively increases the permeability of polarized epithelial cell monolayers for small molecules with molecular mass n-infected persons, polymorphonuclear neutrophils and mononuclear cells are found predominantly in the lamina propria rather than in the gastric mucus layer. Limited contact between these cells and //. pylori may be another factor that helps to explain why //. pylori are not eradicated. //. pylori also produce factors that impair phagocytic activity [238] and possess enzymes such

11.


537

as catalase and superoxide dismutase [239, 240], which may confer resistance to killing by phagocytes. The differentiation of CD4"^ T cells into Thl- or Th2-type cells is an important determinant of whether cell-mediated or humoral immunity will predominate. Optimal clearance of H. pylori should require an efficient secretory IgA response, which is dependent on recruitment of helper T cells belonging to the Th2 subset. In contrast, H. pylori infection is associated with mucosal recruitment of predominantly Thl-type T cells [241, 242], which are best suited for boosting cell-mediated immunity rather than secretory immune responses to extracellular pathogens. How H. pylori infection preferentially elicits a Thl-type response is not yet understood, but this phenomenon may help to explain persistence of the organism. The expression of a broad repertoire of surface antigens by H. pylori may represent another important mechanism for evading host defenses. Analysis of the H. pylori 26695 genome reveals genes encoding at least 32 putative outer membrane proteins, many of which are closely related [84]. Recombination between these genes could potentially generate proteins with new amino-acid sequences and novel antigenic properties, perhaps analogous to the antigenic diversity seen in Neisseria gonorrhoeae, Borrelia spp., Campylobacter fetus, and Mycoplasma genitalium. In addition to antigenic diversity arising via recombination, on/off (phase variation) mechanisms may be operative. Several genes encoding putative surface structures contain homopolymeric tracts or dinucleotide repeats, which tend to change in number as a result of mutations arising via slipped-strand mispairing [84, 243, 244]. This translational regulation of outer membrane protein expression would contribute to phase variation in bacterial surface components, and thus permit evasion of immune responses directed specifically against any of these antigens.

B.

Characteristics of H, pylori Lipopolysaccharide

The important role of LPS from organisms such as E. coli and Salmonella in inducing inflammation (and ultimately bacterial clearance from host tissue) has led to interest in analyzing the LPS structure of H. pylori. Gram-negative LPS typically consists of an innermost portion (lipid A) embedded in the outer membrane, an adjacent core oligosaccharide, and an external surface-exposed polysaccharide composed of repeating oligosaccharide units (0-antigen). H. pylori LPS has a general structural organization similar to that of LPS from Enterobacteriaceae, but has several unique characteristics [245, 246]. In comparison with E. coli lipid A, H. pylori lipid A is mono- rather than diphosphorylated, has six rather than four fatty-acid sidechains, and has longer fatty-acid sidechains (16 to 18 carbons instead of 14) [247-250]. In E. coli. Salmonella, and related genera of the family Enterobacteriaceae, lipid A has potent endotoxic and immunomodulatory properties, including the capacity to induce fever (pyrogenicity) and activate macrophages. H. pylori LPS has from

538


500- to 30,000-fold less endotoxic activity than either E. coli or Salmonella LPS [251-254], depending on the strain and the method of assay. The differences in lipid A structure may be responsible for the low toxicity of //. pylori LPS, and may help to explain the relatively mild mucosal inflammatory responses that accompany H. pylori infection (compared to infections with organisms such as Salmonella), which in turn could contribute to persistence. Consistent with this view, the LPS of Bacteroides species, which also persistently colonize the human gastrointestinal tract, has very low endotoxic activity. The number of 0-antigen structures is much more limited in //. pylori than in most genera of Enterobacteriaceae. Most //. pylori strains express either the fucosylated trisaccharide Lewis^, or its related fucosylated tetrasaccharide Lewis^ [245-256]. Three genes encoding fucosyl transferases, which catalyze production of these antigens, have been identified [84]. Polymeric cytosine tracts are present near the 5' ends of these genes, which suggests the potential for DNA slippage and on/off phase variation [244]. Other 0-antigen components include the mono-fucosylated H antigen, and the nonfucosylated backbone (i). The full repertoire of 0-antigens has not yet been established, but within a single clonal population of H. pylori cells there are often subclones that express more than one of these structures. In addition to being present as components of//, pylori LPS, Lewis x and Lewis y antigens are present on host tissue, including gastric epithelial cells [257]. The presence of these antigens on the surface of //. pylori may be a form of camouflage or molecular mimicry that serves to diminish host immune responses and thereby facilitates bacterial persistence, because of host tolerance to these "self antigens. There is evidence in some (but not all) studies that the predominant Lewis expression of the //. pylori population and of the particular colonized host are related [258, 259].

C. Bacteria-Host Equilibrium In acute bacterial infections, the intensity of inflammation typically increases progressively until pathogenic bacteria are eradicated, or until bacterial proliferation leads to destruction of tissue and/or death of the host. In the case of //. pylori, persistent infection must involve an equilibrium in which //. pylori persist in the stomach without any overall change in bacterial population size [236], and gastric inflammation neither progresses nor resolves. It seems possible that over hundreds of thousands of years there may have been selection for human hosts that can control //. pylori proliferation without extensive accompanying gastric inflammation and damage, as well as for //. pylori strains that elicit relatively mild inflammation [260]. Thus, comparisons can be drawn between persistent colonization of the human stomach by //. pylori and colonization of the intestine by commensal organisms.

11.


539

XIL Factors Influencing Development of Clinically Evident Disease In most H. pylori-mfQciod persons, colonization of the gastric mucosa and associated inflammation are tolerated for decades without causing any symptoms. Why serious gastroduodenal illnesses occur in a subset of infected persons, in an apparendy unpredictable or sporadic fashion, is not yet understood. However, it is currendy thought that characteristics of individual H. pylori strains, characteristics of individual human hosts, and environmental factors are each important determinants of clinical outcome [261].

A. Bacterial Factors

Due to extensive allelic polymorphism and frequent genetic recombination, each H. pylori strain has unique characterisdcs. Therefore, a spectrum of clinical outcomes can be reasonably attributed to the diversity that exists among H. pylori strains. To identify markers for strains associated with adverse clinical outcomes, investigators have compared H. pylori strains isolated from padents with ulcer disease with strains isolated from asymptomatic padents. Four bacterial markers that predict clinical outcomes have been idendfied thus far: the cag pathogenicity island [100, 101, 104, 262], type si vacA signal sequences [92, 262], the Le^ binding phenotype [117], and type iceAl alleles [99, 262]. The former three markers are frequently found together, probably as coadapted traits, in H. pylori strains associated with an increased risk for pepdc ulcer disease. Strains containing the cag island also have been associated with an increased risk for distal gastric adenocarcinoma [263-265]. iceAl alleles seem to constitute an independent marker for ulcerogenic strains [99, 262]. These markers have proven somewhat useful for predicting clinical outcome in the United States and Western Europe, but have limited udlity in many parts of Asia, where nearly all H. pylori isolates are cag'^ and possess type si vacA alleles [266]. Moreover, it is clear that only a minority of persons infected with putatively "ulcerogenic strains" (i.e., cag'^, vacA type si, type iceAl, with Le^-binding properties) ever develop ulcer disease or gastric cancer. Thus, these genedc markers provide important insights into the pathogenesis of H. pylori-sissociated ulcer disease and gastric cancer, but do not endrely account for differences in clinical outcome.

540

B.


Host Factors

Considerable diversity exists in the human population, and host differences are likely to influence clinical outcome. Consistent with this hypothesis is the epidemiologic observation that peptic ulcer disease and gastric cancer occur significandy more frequently in males than in females [43]. The increased incidence of these diseases in males is not attributable to a higher prevalence of H. pylori infection and is unlikely to result from gender-related differences in H. pylori strains. Clustering of pepdc ulcer disease in families occurs [43, 267], and could be due to either host-related genetic predispositions or to transmission of ulcerogenic strains within family groups. Higher rates of concordance for pepdc ulcer disease between monozygodc twins than between dizygodc twins suggest that host factors are indeed relevant [43]. Geographic variations in the incidence of ulcer disease and gastric cancer also may reflect host differences [43], but it is difficult to control for differences in H. pylori strains or environmental exposures in epidemiologic surveys. Of particular relevance to the clinical outcomes of H. pylori infection are host variables that affect bacterial growth or localization in the gastric mucosa. Levels of gastric acid production clearly vary among humans. High levels of gastric acidity are associated with growth of metaplastic gastric tissue in the duodenum, which provides a site for H. pylori colonization [33-35]. The risk of duodenal ulceradon is increased in padents whose duodenal mucosa is colonized by H. pylori compared to H. /?_y/on-infected patients without duodenal colonizadon [33-35]. Thus, by promoting growth of//, pylori in the duodenum, high levels of gastric acid producdon may predispose to duodenal ulcer disease. Low levels of gastric acid producdon are associated with increased //. pylori growth in the gastric corpus, which could be an important initiating step in the pathogenesis of gastric ulcers. The risk of gastric cancer is significandy increased in patients who have a history of gastric ulcers compared to patients without any history of pepdc ulcer disease, whereas the risk of gastric cancer is significantly decreased in padents with a history of duodenal ulcers [268]. This suggests that duodenal ulcer disease and gastric ulcer disease/gastric atrophy/gastric cancer represent two disdnct pathways in //. pylori-'mftcitd persons. Potentially these two pathways are determined by underlying host characteristics, such as high and low gastric acid producdon, respecdvely. An alternate hypothesis is that the time of life when //. pylori is acquired (early in childhood or later) is an important determinant of clinical outcome [269]. The latter model holds that early childhood infection predisposes to development of atrophic gastritis, gastric ulcers, and distal gastric cancer in adult years, and that acquisition of //. pylori later in life is associated with development of duodenal ulcers. Humans differ in expression of histo-blood group andgens, one of which (Le^) is a host cell receptor for H. pylori adherence. Presence or absence of Le^ may influence the extent to which //. pylori adhere to gastric epithelium. Studies in

11.


541

mice indicate that bacterial adherence to Le^ receptors is associated with the development of parietal cell autoantibodies and parietal cell loss [163], which would lead to reduced gastric acid production. Thus, an increased proportion of adherent H. pylori may also result in enhanced autoimmune-related phenomena or enhanced inflammatory responses in humans. Immune responses to bacteria can vary considerably among individuals. Immune responses to //. pylori seem to be generally ineffective in eradicating the organisms, but the intensity or nature of the immune response may influence growth or localization of H. pylori. For example, vigorous immune responses might be very effective in controlling bacterial growth, but may have undesired consequences related to tissue injury. Conversely, ineffective immune responses might lead to gastric mucosal damage due to excessive bacterial growth. C. Environmental Factors Environmental exposures probably play an important role in modulating clinical outcomes of H. pylori infection. In particular, the striking temporal changes in rates of ulcer disease and gastric cancer in many Western countries over the past two centuries are most easily explained by changing environmental conditions, rather than changes in bacterial or host characteristics. Determining which environmental factors are most relevant is a difficult task, but several factors are currently known to influence clinical outcome. Cigarette smoking and nonsteroidal antiinflammatory drugs are two well-documented risk factors for development of peptic ulcer disease. The invention of automated methods for manufacturing cigarettes in the late 1800s and the associated increases in smoking throughout the United States and Europe may have contributed to the high rates of ulcer disease earlier in this century. Dietary factors also may be relevant. In particular, gastric cancer has been associated with high salt intake and with low consumption of fruits and vegetables [54]. D. Perspectives on H, pj/ori-Related Diseases At least 20 different Helicobacter species are currently known, and various species colonize the gastrointestinal tracts of many different mammals, including several different primate species. Thus, it seems likely that individual Helicobacter species have evolved within the alimentary tracts of different mammals over a very long time period, and, in particular, //. pylori have probably colonized human stomachs for hundreds of thousands of years or longer [260, 261]. Currently, about 60% of the world's human population is infected with H. pylori, and this proportion is probably considerably lower than in previous centuries. Thus, even though the presence of H. pylori is a risk factor for potentially fatal diseases such as peptic ulcers and distal gastric cancer, the human species has proliferated quite successfully. One relevant factor may be that mortality due to peptic ulcer disease and gastric cancer occurs primarily in old age, and, therefore.

542


these illnesses do not substantially alter survival of children or reproductive-age adults [261]. In addition, it is quite possible that mortality rates from H. pylori-rt\2iitd diseases have been higher during the twentieth century than during most previous periods. During the last decades of the twentieth century, there has been a gradual decrease in the prevalence of H. pylori in many developed countries, and this trend will probably continue worldwide as sanitation and hygiene continue to improve and as family sizes shrink. Thus, we are in the midst of a long-term natural experiment that will help to clarify the role of H. pylori in human health and disease. The continuing loss of//, pylori from much of the human population should lead to commensurate declines in peptic ulcer disease and distal gastric cancer. The possibility of other less fortunate consequences, such as an increased rate of proximal gastric cancer and esophageal adenocarcinoma [63-66], must also be followed closely. Therefore, at present it seems prudent to recognize that //. pylori and humans have shared a long evolutionary history, and that our current understanding of "disease" in the context of this ancient relationship is quite limited.

Acknowledgments Research on H. pylori in our laboratories is supported by grants from the National Institutes of Health (ROl AI39657, DK 53623, DK 53727, AI 38166, DK 53707, and AI 25567) and the Medical Research Service of the Department of Veterans Affairs.

References 1. Warren, J. R. (1983). Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet i, 1273. 2. Marshall, B. J. (1983). Unidentified curved bacillus on gastric epithelium in active chronic gastritis. L««c^r i, 1273-1275. 3. Marshall, B. J., and Warren, J. R. (1984). Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet i, 1311-1313. 4. Marshall, B. J., Armstrong, J. A., McGechie, D. B., and Glancy, R. J. (1985). Attempt to fulfil Koch's postulates for pyloric Campylobacter. Med. J. Aust. 142, 436-439. 5. Morris, A., and Nicholson, G. (1987). Ingestion of Campylobacter pyloridis causes gastritis and raised fasting gastric pH. Am. J. Gastroenterol. 82, 192-199. 6. Morris, A. J., Ali, M. R., Nicholson, G. I., Perez-Perez, G. I., and Blaser, M. J. (1991). Long-term follow-up of voluntary ingestion of Helicobacter pylori. Ann. Intern. Med. 114, 662-663. 7. NIH Consensus Development Panel on Helicobacter pylori in Peptic Ulcer Disease (1994). Helicobacter pylori in peptic ulcer disease. JAMA 272, 65-69.

11. H. PYLORI PATHOGENESIS

543

8. Infection with Helicobacter pylori (1994). In "lARC Working Group on the Evaluation of Carcinogenic Risks to Humans: Schistosomes, Liver Flukes and Helicobacter pylori'' lARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 61, pp. 177-240. International Agency for Research on Cancer, Lyon, France. 9. Taylor, D. N., and Blaser, M. J. (1991). The epidemiology of Helicobacter pylori infections. Epidemiol Rev. 13, 42-59. 10. Pounder, R. E., and Ng, D. (1995). The prevalence oi Helicobacter pylori in different countries. Aliment. Pharmacol. Ther 9 (Suppl.), S33-S39. 11. Parsonnet, J. (1995). The incidence oi Helicobacter pylori infection. Aliment. Pharmacol. Ther 9 (Suppl.), 45-51. 12. Parsonnet, J., Blaser, M. J., Perez-Perez, G. I., Hargrett-Bean, N., and Tauxe, R. V. (1992). Symptoms and risk factors of Helicobacter pylori infection in a cohort of epidemiologists. Gastroenterology 102, 41-46. 13. Perez, G. I., Dworkin, B. M., Chodos, J. E., and Blaser, M. J. (1988). Campylobacter pylori antibodies in humans. Ann. Intern. Med. 109, 11-17. 14. Banatvala, N., Mayo, K., Megraud, F., Jennings, R., Decks, J. J., and Feldman, R. A. (1993). The cohort effect and Helicobacter pylori. J. Infect. Dis. 168, 219-221. 15. Mitchell, H. M., Li, Y. Y., Hu, R J., Liu, Q., Chen, M., Du, G. G., Wang, Z. J., Lee, A., and Hazell, S. L. (1992). Epidemiology of Helicobacter pylori in Southern China: Identification of early childhood as the critical period for acquisition. J. Infect. Dis. 166, 149-153. 16. Torres, J., Leal-Herrera, Y, Perez-Perez, G., Gomez, A., Camorlinga-Ponce, M., Cedillo-Rivera, R., Tapia, Conyer, R., and Munoz, O. (1998). A community-based seroepidemiologic study of Helicobacter pylori infection in Mexico. J. Infect. Dis. 178, 1089-1094. 17. Staat, M. A., Kruszon-Moran, D., McQuillan, G. M., and Kaslow, R. A. (1996). A populationbased serologic survey of Helicobacter pylori infection in children and adolescents in the United States. J. Infect. Dis. 174, 11210-11223. 18. Drumm, B., Perez-Perez, G. I., Blaser, M. J., and Sherman, P. M. (1990). Intrafamilial clustering of Helicobacter pylori infection. New Engl. J. Med. 322, 359-363. 19. Suerbaum, S., Maynard Smith, J., Bapumia, K., Morelli, G., Smith, N. H., Kunstmann, E., Dyrek, I., and Achtman, M. (1998). Free recombination within Helicobacter pylori. Proc. Natl. Acad Sci. U.S.A. 95, 12619-12624. 20. Kersulyte, D., Chalkauskas, H., and Berg, D. E. (1999). Emergence of recombinant strains of Helicobacter pylori during human infection. Mol. Microbiol. 31, 31-43. 21. Ferguson Jr., D. A., Li, C , Patel, N. R., Mayberry, W. R., Chi, D. S., and Thomas, E. (1993). Isolation of Helicobacter pylori from saliva. J. Clin. Microbiol. 31, 2802-2804. 22. Thomas, J. E., Gibson, G. R., Darboe, M. K., Dale, A., and Weaver, L. T. (1992). Isolation of Helicobacter pylori from human faeces. Lancet 340, 1194-1195. 23. Drazek, E. S., Dubois, A., and Holmes, R. K. (1994). Characterization and presumptive identification of Helicobacter pylori isolates from rhesus monkeys. J. Clin. Microbiol. 32, 1799-1804. 24. Akopyanz, N., Bukanov, N. O., Westblom, T. U., Kresovich, S., and Berg, D. E. (1992). DNA diversity among clinical isolates of Helicobacter pylori detected by PCR-based RAPD fingerprinting. Nucleic Acids Res. 20, 5137-5142. 25. Fujimoto, S., Marshall, B., and Blaser, M. J. (1994). PCR-based restriction fragment length polymorphism typing of Helicobacter pylori. J. Clin. Microbiol. 34, 331-334. 26. Solnick, J. V., O'Rourke, J., Lee, A., Paster, B. J., Dewhirst, F E., and Tompkins, L. S. (1993). An uncultured gastric spiral organism is a newly identified Helicobacter in humans. J. Infect. Dis. 168, 379-385.

544


27. Anderson, L. P., Norgaard, A., Hoick, S., Blom, J., and Elsborg, L. (1996). Isolation of a ''Helicobacter heilmcmir-Wkt organism from the human stomach. Eur. J. Clin. Microbiol. Infect. Dis. 15, 95-96. 28. Eisenberg, M. M., Fondacaro, P. F., and Dunn, D. H. (1995). Applied anatomy and anomalies of the stomach. In "Bockus Gastroenterology' 5th ed. (W. S. Haubrich, F. Schaffer, and J. E. Berk, eds.), pp. 561-581. Saunders, Philadelphia. 29. Quigley, E. M. M., and Turnberg, L. A. (1987). pH of the microclimate lining human gastric and duodenal mucosa in vivo. Gastroenterology. 92, 1876-1883. 30. Bhaskar, K. R., Garik, P, Turner, B. S., Bradley, J. D., Bansil, R., Stanley, H. E., and LaMont, J. T. (1992). Viscous fingering of HCl through gastric mucin. Nature 360, 458^61. 31. Hazell, S. L., Lee, A., Brady, L., and Hennessy, W. (1986). Campylobacter pyloridis and gastritis: Association with intercellular spaces and adaptation to an environment of mucus as important factors in colonization of the gastric epithelium. J. Infect. Dis. 153, 658-663. 32. Hessey, S. J., Spencer, J., Wyatt, J. I., Sobala, G., Rathbone, B. J., Axon, A. T, and Dixon, M. F. (1990). Bacterial adhesion and disease activity in Helicobacter pylori associated chronic gastritis. Gw/31, 134-138. 33. Wyatt, J. I., Rathbone, B. J., Dixon, M. F, and Heatley, R. V. (1987). Campylobacter pyloridis and acid induced gastric metaplasia in the pathogenesis of duodenitis. J. Clin. Pathol. 40, 841-848. 34. Carrick, J., Lee, A., Hazell, S., Ralston, M., and Daskalopoulos, G. (1989). Campylobacter pylori, duodenal ulcer, and gastric metaplasia: Possible role of functional heterotopic tissue in ulcerogenesis. Gut 30, 790-797. 35. Wyatt, J. I., Rathbone, B. J., Sobala, G. M., Shallcross, T, Heatley, R. V., Axon, A. T, and Dixon, M. F. (1990). Gastric epithelium in the duodenum: Its association with Helicobacter pylori and inflammation. J. Clin. Pathol. 43. 981-986. 36. Dixon, M, F., Genta, R. M., Yardley, J. H., and Correa, P. (1996). Classification and grading of gastritis—the updated Sydney system. Am. J. Surg. Pathol. 20, 1161-1181. 37. Rauws, E. A. J., Langenberg, W., Houthoff, H. J., Zanen, H. C , and Tytgat, G. N. (1988). Campylobacterpyloridis-a^ôciated chronic active antral gastritis. Gastroenterology. 94, 33-40. 38. Drumm, B., Sherman, P., Cutz, E., and Karmali, M. (1987). Association of Campylobacter pylori on the gastric mucosa with antral gastritis in children. New Engl. J. Med. 316, 1557-1561. 39. Kuipers, E. J., Uyterlinde, A. M., Pena, A. S., Roosendaal, R., Pals, G., Nelis, G. F., Festen, H. P. M., and Meuwissen, S. G. M. (1995). Long-term sequelae of Helicobacter pylori gastritis. Lancet. 345, 1525-1528. 40. Correa, P. (1992). Human gastric carcinogenesis: A multistep and multifactorial process. First American Cancer Society Award Lecture on cancer epidemiology and prevention. Cancer Res. 52, 6735-6740. 41. Graham, D. Y, Alpert, L. C , Smith, J. L., and Yoshimura, H. H. (1988). Iatrogenic Campylobacter pylori infection is a cause of epidemic achlorhydria. Am. J. Gastroenterol. 83, 974-980. 42. Ramsey, E. J., Carey, K. V., Peterson, W. L., Jackson, J. J., Murphy, F K., Read, N. W, Taylor, K. B., Trier, J. S., and Fordtran, J. S. (1979). Epidemic gastritis with hypochlorhydria. Gastroenterology. 76, 1449-1457. 43. Lam, S. K., Hui, W. M., and Ching, C. K. (1995). Peptic ulcer disease: Epidemiology, pathogenesis, and etiology. In "Bockus Gastroenterology," 5th ed. (W. S. Haubrich, F. Schaffer, and J. E. Berk, eds.), pp. 561-581. Saunders, Philadelphia. 44. Monson, R. R., and MacMahon, B. (1969). Peptic ulcer in Massachusetts physicians. New Engl. J. Med. 281, 11-15. 45. Jennings, D. (1940). Perforated peptic ulcer. Changes in age-incidence and sex-distribution in the last 150 years. Lcmcet. i, 395-398, 444-447.

11.


545

46. Blaser, M. J., Perez-Perez, G. I., Lindenbaum, J., Schneidman, D., Van Deventer, G., MarinSorensen, M., and Weinstein, W. M. (1991). Association of infection due to Helicobacter pylori with specific upper gastrointestinal pathology. Rev. Infect. Dis. 13, S704-S708. 47. Blaser, M. J. (1990). Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. J. Infect. Dis. 161, 626-633. 48. Nomura, A., Stemmermann, G. N., Chyour, P.-H., Perez-Perez, G. I., and Blaser, M. J. (1994). Helicobacter pylori infection and the risk for duodenal and gastric ulceration. Ann. Intern. Med. 120,977-981. 49. Forbes, G. M., Glaser, M. E., Cullen, D. J., Warren, J. R., Christiansen, K. J., Marshall, B. J., and Collins, B. J. (1994). Duodenal ulcer treated with Helicobacter pylori eradication: Seven-year follow-up. Lancet 343, 258-260. 50. Hentschel, I., Brandstatter, G., Dragosics, B., Hirschl, A. M., Nemec, H., Schutze, K., Taufer, M., and Wurzer, H. (1993). Effect of ranitidine and amoxicillin plus metronidazole on the eradication of Helicobacter pylori and the recurrence of duodenal ulcer. New Engl. J. Med. 328, 308-312. 51. Graham, D. Y., Lew, G. M., Klein, P D., Evans, D. G., Evans Jr., D. J., Saeed, Z. A., and Malaty, H. M. (1992). Effect of treatment of Helicobacter pylori infection on the long-term recurrence of gastric or duodenal ulcer: A randomized, controlled study. Ann. Intern. Med. 116, 705-708. 52. Fuchs, C. S., and Mayer, R. J. (1995). Gastric carcinoma. New Engl. J. Med. 333, 32-40. 53. Neugut, A. I., Hayek, M., and Howe, G. (1996). Epidemiology of gastric cancer. Semin. Oncol. 23,281-291. 54. Howson, C. P., Hiyama, T., and Wynder, E. L. (1986). The decline in gastric cancer: Epidemiology of an unplanned triumph. Epidemiol. Rev. 8, 1-27. 55. Ihamaki, T., Saukkonen, M., and Siurala, M. (1978). Long-term observation of subjects with normal mucosa and with superficial gastritis: Results of 23-27 years' follow-up examinations. Scand. J. Gastroenterol. 13, 11 \-115. 56. Ihamaki, T., Kekki, M., Sipponen, P., and Siurala, M. (1985). The sequelae and course of chronic gastritis during a 30- to 34-year bioptic follow-up study. Scand. J. Gastroenterol. 20, 485^91. 57. Siurala, M., Lehtola, J., and Ihamaki, T. (1974). Atrophic gastritis and its sequelae: Results of 19-23 years' follow-up examinations. Scand. J. Gastroenterol. 9, 441^46. 58. Nomura, A., Stemmermann, G. N., Chyou, P.-H., Kato, I., Perez-Perez, G. I., and Blaser, M. J. (1991). Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. New Engl. J. Med 325, 1132-1136. 59. Parsonnet, J., Friedman, G. D., Vandersteen, D. P., Chang, Y., Vogelman, J. H., Orentreich, N., and Sibley, R. K. (1991). Helicobacter pylori infection and the risk of gastric carcinoma. New Engl. J. Med. 325, 1127-1131. 60. The Eurogast Study Group (1993). An international association between Helicobacter pylori infection and gastric cancer. Lancet 341, 1359-1362. 61. Forman, D., Sitas, F, Newell, D. G., Stacey, A. R., Boreham, J., Peto, R., Campbell, T. C , Li, J., and Chen, J. (1990). Geographic association of Helicobacter pylori antibody prevalence and gastric cancer mortality in rural China. Int. J. Cancer 46, 608-611. 62. Watanabe, T., Tada, M., Nagai, H., Sasaki, S., and Nakao, M. (1998). Helicobacter pylori infection induces gastric cancer in Mongolian gerbils. Gastroenterology 115, 642-648. 63. Blot, W. J., Devesa, S. S., Kneller, R. W., and Frauemi Jr., J. F (1991). Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA 265, 1287-1289. 64. Labenz, J., Blum, A. L., Bayerdorffer, E., Meining, A., Stolte, M., and Borsch, G. (1997). Curing Helicobacter pylori infection in patients with duodenal ulcer may provoke reflux esophagitis. Gastroenterology 112, 1442-1447.

546


65. Chow, W.-H., Blaser, M. J., Blot, W. J., Gammon, M. D., Vaughan, T. L., Risch, H. A., Perez-Perez, G. I., Schoenberg, J. B., Stanford, J. L., Rotterdam, H., West, A. B., and Fraumeni Jr., J. F. (1998). An inverse relation between cagA^ strains of Helicobacter pylori infection and risk of esophageal and gastric cardia adenocarcinoma. Cancer Res. 58, 588-590. 66. Vicari, J. J., Peek, R. M., Falk, G. W., Goldblum, J. R., Easley, K. A., Schnell, J., Perez-Perez, G. I., Halter, S. A., Rice, T. W., Blaser, M. J., and Richter, J. E. (1998). The seroprevalence of cflÂ-positive Helicobacter pylori strains in the spectrum of gastroesophageal reflux disease. Gastroenterology 115, 50-57. 67. Wotherspoon, A, C. (1998). Gastric lymphoma of mucosa-associated lymphoid tissue and Helicobacter pylori. Anna. Rev. Med. 49, 289-299. 68. Wotherspoon, A. C , Ortiz-Hidalgo, C , Falzon, M. R., and Isaacson, P G. (1991). Helicobacter pyon-associated gastritis and primary B-cell gastric lymphoma. Lancet 338, 1175-1176. 69. Parsonnet, J., Hansen, S., Rodriguez, L., Gelb, A. B., Wamke, R. A., Jellum, E., Orentreich, N., Vogelman, J. H., and Friedman, G. D. (1994). Helicobacter pylori infection and gastric lymphoma. New Engl. J. Med. 330, 1267-1271. 70. Zucca, E., Bertoni, F, Roggero, E., Bosshard, G., Cazzaniga, G., Pedrinis, E., Biondi, A., and Cavalli, F (1998). Molecular analysis of the progression from //. py/on-associated chronic gastritis to mucosa-associated lymphoid-tissue lymphoma of the stomach. New Engl. J. Med. 338,804-810. 71. Wotherspoon, A. C , Doglioni, C , Diss, T. C , Pan, L., Moschini, A., de Boni, M., and Isaacson, P. G. (1993). Regression of primary low-grade B-cell gastric lymphoma of mucosa-associated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 342, 575-577. 72. Bayerdorffer, E., Neubauer, A., Rudolph, B., Thiede, C , Lehn, N., Sidt, S., Stolte, M. (1995). Regression of primary gastric lymphoma of mucosa-associated lymphoid tissue type after cure oi Helicobacter pylori infection. Lancet 345, 1591-1594. 73. Roggero, E., Zucca, E., Pinotti, G., Pascarella, A., Capella, C , Savio, A., Pedrinis, E., Paterlini, A., Venco, A., and Cavalli, F (1995). Eradication o\ Helicobacter pylori infection in primary low-grade gastric lymphoma of mucosa-associated lymphoid tissue. Ann. Intern. Med. 122, 767-769. 74. Matsumoto, T., lida, M., and Shimizu, M. (1997). Regression of mucosa-associated lymphoidtissue lymphoma of rectum after eradication of Helicobacter pylori. Lancet 335, 115-116. 75. Hussell, T., Isaacson, P. G., Crabtree, J. E., and Spencer, J. (1993). The response of cells from low-grade B-cell gastric lymphomas of mucosa-associated lymphoid tissue to Helicobacter pylori. Lancet 342, 571-574. 76. Fox, J. (1997a). Helicobacters: The next generation. Baillieres Clin. Infect. Dis. 4, 449-471. 77. Fox, J. G. (1997b). The expanding genus oi Helicobacter: Pathogenic and zoonotic potential. Sem. Gastrointest. Dis. 8, 124-141. 78. Quinn, T. C , Goodell, S. E., Fennel, C , Wang, S. P, Schuffler, M. D., Holmes, K. K., and Stamm, W. E. (1984). Infections with Campylobacter jejuni and Campylobacter-like organisms in homosexual men. Ann. Intern. Med.. 101, 187-192. 79. Fox, J. G., Dewhirst, F E., Shen, Z., Feng, Y., Taylor, N. S., Paster, B. J., Ericson, R. L., Lau, C. N., Correa, P., Araya, J. C , and Roa, I. (1998). Hepatic Helicobacter species identified in bile and gallbladder tissue from Chileans with chronic cholecystitis. Gastroenterology 114, 755-763. 80. Ward, J. M., Fox, J. G., Anver, M. R., Haines, D. C , George, C. V., Collins Jr., M. J., Gorelick, P. L., Nagashima, K., Gonda, M. A., and Gilden, R. V. (1994). Chronic active hepatitis and associated liver tumors in mice caused by a persistent bacterial infection with a novel Helicobacter species. J. Natl. Cancer Inst. 86, 1222-1227.

11.


547

81. Reynolds, D. J., and Penn, C. W. (1994). Characteristics of Helicobacter pylori growth in a defined medium and determination of its amino acid requirements. Microbiology 140, 26492656. 82. Benaissa, M., Babin, P., Quellard, N., Pezennec, L., Cenatiempo, Y., and Fauchere, J. L. (1996). Changes in Helicobacter pylori ultrastructure and antigens during conversion from the bacillary to the coccoid form. Infect. Immun. 64, 2331-2335. 83. Kusters, J. G., Gerrits, M. M., van Strijp, J. A. G., and Vandenbroucke-Grauls, C. M. J. E. (1997). Coccoid forms of Helicobacter pylori are the morphologic manifestation of cell death. Infect. Immun. 65, 3672-3679. 84. Tomb, J.-R, White, O., Kerlavage, A. R., Clayton, R. A., Sutton, G. G., Fleischmann, R. D., Ketchem, K. A., Klenk, H. P, Gill, S., Dougherty, B. A., Nelson, K., Quackenbush, J., Zhou, L., Kirkness, E. F., Peterson, S., Loftus, B., Richardson, D., Dodson, R., Khalak, H. G., Glodek, A., McKenney, K., Fitzgerald, L. M., Lee, N., Adams, M. D., Venter, J. C , et al. (1997). The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388, 539-547. 84a. Aim, R. A., Ling, L.-S. L., Moir, D. T, King, B. L., Brown, E. D., Doig, P C , Smith, D. R., Noonan, B., Guild, B. C , de Jonge, B. L., Carmel, G., Tummino, P. J., Caruso, A., Uria-Nickelsen, M., Mills, D. M., Ives, C , Gibson, R., Merberg, D., Mills, S. D., Jiang, Q., Taylor, D. E., Vovis, G. F., and Trust, T. J. (1999). Genomic sequence comparison of two unrelated isolates of the human gastric pathogen Helicobacter pylori. Nature 397, 176-180. 85. Huynen, M., Thomas, D., and Bork, P. (1998). Differential genome analysis applied to the species-specific features of Helicobacter pylori. FEBS Lett. 426, 1-5. 86. Berg, D. E., Hoffman, P S., Appelmelk, B. J., and Kusters, J. G. (1997). The Helicobacter pylori genome sequence: Genetic factors for long life in the gastric mucosa. Trends Microbiol. 5, 468-474. 87. Heuermann, D., and Haas, R. (1995). Genetic organization of a small cryptic plasmid of Helicobacter pylori. Gene 165, 17-24. 88. Kleanthous, H., Clayton, C. L., and Tabaqchali, S. (1991). Characterization of a plasmid from Helicobacter pylori encoding a replication protein common to plasmids in Gram-positive bacteria. Mol. Microbiol. 5, 2377-2389. 89. Heintschel von Heinegg, E., Nalik, H. P., and Schmid, E. N. (1993). Characterization of a Helicobacter pylori phage (HPl). J. Med. Microbiol. 38, 245-249. 90. Kansau, I., Raymond, J., Bingen, E., Courcoux, P., Kalach, N., Bergeret, M., Braimi, N., Dupont, C , and Labigne, A. (1996). Genotyping of Helicobacter pylori isolates by sequencing of PCR products and comparison with the RAPD technique. Res. Microbiol. 147, 661-669. 91. Gamer, J. A., and Cover, T. L. (1995). Analysis of genetic diversity in cytotoxin-producing and non-cytotoxin-producing Helicobacter pylori strains. J. Infect. Dis. 172, 290-293. 92. Atherton, J. C , Cao, P, Peek Jr., R. M., Tummuru, M. K. R., Blaser, M. J., and Cover, T. L. (1995). Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori. Association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270, 17771-17777. 93. Pan, Z. J., Berg, D. E., van der Hulst, R. W. M., Su, W.-W., Raudonikiene, A., Xiao, S.-D., Dankert, J., Tytgat, G. N. J., and van der Ende, A. (1998). Prevalence of vacuolating cytotoxin production and distribution of distinct vacA alleles in Helicobacter pylori from China. J. Infect. Dis. 178, 220-226. 94. van Doom, L.-J., Figueiredo, C , Sanna, R., Pena, S., Midolo, P, Ng, E. K., Atherton, J. C , Blaser, M. J., and Quinn, W. G. (1998). Expanding allelic diversity of Helicobacter pylori vacA. J. Clin. Microbiol. 36, 2697-2603.

548


95. Kansau, I., Guillain, F., Thiberge, J. M., and Labigne, A. (1996). Nickel binding and immunological properties of the C-terminal domain of the Helicobacter pylori GroES homologue (HspA). Mol. Microbiol. 22, 1013-1023. 96. Cao, P., and Cover, T. L. (1997). High-level genetic diversity in the vapD chromosomal region of Helicobacter pylori. J. Bacteriol. 179, 2852-2856. 97. van der Ende, A., Pan, Z.-J., Bart, A., van der Hulst, R. W. M., Feller, M., Xiao, S.-D., Tytgat, G. N. J., and Dankert, J. (1998). cagA-po^'it'we Helicobacter pylori populations in China and the Netherlands are distinct. Infect. Immiin. 66, 1822-1826. 98. Yamoaka, Y, Kodama, T., Kashima, K., Graham, D. Y, and Speulveda, A. R. (1998). Variants of the y region of the cagA gene in Helicobacter pylori isolates from patients with different H. /7>'/on-associated diseases. J. Clin. Microbiol. 36, 2258-2263. 99. Peek Jr., R. M., Thompson, S. A., Donahue, J. R, Tham, K. T, Atherton, J. C , Blaser, M. J., and Miller, G. G. (1998). Adherence to gastric epithelial cells induces expression of a Helicobacter pylori gene, iceA that is associated with clinical outcome. Proc. Assoc. Am. Phys. 110,531-544. 100. Cover, T. L., Dooley, C. P., and Blaser, M. J. (1990). Characterization of and human serologic response to proteins in Helicobacter pylori broth culture supernatants with vacuolizing cytotoxin activity. Infect. Immiin. 58, 603-610. 101. Crabtree, J. E., Taylor, J. D., Wyatt, J. I., Heatley, R. V., Shallcross, T. M., Tompkins, L. S., and Rathbone, B. J. (1991). Mucosal IgA recognition oi Helicobacter pylori 120 kDa protein, peptic ulceration, and gastric pathology. Lancet 338, 332-335. 102. Tummuru, M. K. R., Cover, T. L., and Blaser, M. J. (1993). Cloning and expression of a high-molecular-mass major antigen of Helicobacter pylori: Evidence of linkage to cytotoxin production. Infect. Immun. 61, 1799-1809. 103. Covacci, A., Censini, S., Bugnoli, M., Petracca, R., Burroni, D., Macchia, G., Massone, A., Papini, E., Xiang, Z., Figura, N., et al. (1993). Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. U.S.A. 90, 5791-5795. 104. Censini, S., Lange, C , Xiang, Z., Crabtree, J. E., Ghiara, P., Borodovsky, M., Rappuoli, R., and Covacci, A. (1996). cag, a pathogenicity island of Helicobacter pylori, encodes type I-specific and disease-associated virulence factors. Proc. Natl. Acad. Sci. U.S.A. 93, 14648-14653. 105. Akopyants, N. S., Clifton, S. W., Kersulyte, D., Crabtree, J. E., Youree, B. E., Reece, C. A., Bukanov, N. O., Drazek, E. S., Roe, B. A., and Berg, D. E. (1998). Analyses of the cag pathogenicity island of Helicobacter pylori. Mol. Microbiol. 28, 37-53. 106. Tummuru, M. K. R., Sharma, S. A., and Blaser, M. J. (1995). Helicobacter pylori picB, a homologue of the Bordetella pertussis toxin secretion protein, is required for induction of IL-8 in gastric epithelial cells. Mol. Microbiol. 18, 867-876. 107. Segal, E. D., Falkow, S., and Tompkins, L. S. (1996). Helicobacter pylori attachment to gastric cells induces cytoskeletal rerrangements and tyrosine phosphorylation of host cell proteins. Proc. Natl. Acad. Sci. U.S.A. 93, 1259-1264. 108. Segal, E. D., Lange, C , Covacci, A., Tompkins, L. S., and Falkow, S. (1997). Induction of host signal transduction pathways by Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 94, 7595-7599. 109. Akopyants, N. S., Fradkov, A., Diatchenko, L., Hill, J. E., Siebert, P D., Lukyanov, S. A., Sverdlov, E. D., and Berg, D. E. (1998). PCR-based subtractive hybridization and differences in gene content among strains of Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 95, 13108-13113.

II.


549

110. Go, M. F., Kapur, V., Graham, D. Y., and Mussser, J. M. (1996). Population genetic analysis of Helicobacter pylori by multilocus enzyme electrophoresis: Extensive allelic diversity and recombinational population structure. J. Bacteriol. 178, 3934-3938. III. Jorgensen, M., Daskalopoulos, G., Warburton, V., Mitchell, H. M., and Hazell, S. L. (1996). Multiple strain colonization and metronidazole resistance in Helicobacter /n'/or/-infected patients: Identification from sequential and multiple biopsy specimens. J. Infect Dis. 174, 631-635. 112. Nedenskov-Soresne, P., Bukholm, G., and Bovre, K. (1990). Natural competence for genetic transformation in Campylobacter pylori. J. Infect. Dis. 161, 365-366. 113. Odenbreit, S., Till, M., and Haas, R. (1996). Optimized BlaM-transposon shuttle mutagenesis of Helicobacter pylori allows the identification of novel genetic loci involved in bacterial virulence. Mol. Microbiol. 20, 361-373. 114. Hofreuter, D., Odenbreit, S., Henke, G., and Haas, R. (1998). Natural competence for DNA transformation in Helicobacter pylori: Identification and genetic characterization of the comB locus. Mol. Microbiol. 28, 1027-1038. 115. Ando, T., Israel, D. A., Kusugami, K., and Blaser, M. J. (1998). HP333, a dprA homolog, is required for DNA transformation of Helicobacter pylori. Gastroenterology 114 (Suppl. 4), A920. 116. Kuipers, E. J., Israel, D. A., Kusters, J. G., and Blaser, M. J. (1998). Evidence for a conjugation-like mechanism of DNA transfer in Helicobacter pylori. J. Bacteriol. 180, 2901-2905. 117. liver, D., Arnqvist, A., Ogren, J., Prick, I.-M., Kersulyte, D., Incecik, E. T., Berg, D. E., Covacci, A., Engstrand, L., and Boren, T. (1998). Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science 279, 373-377. 118. Campbell, S., Eraser, A., Holliss, B., Schmid, J., and O'Toole, P W. (1997). Evidence for ethnic tropism of Helicobacter pylori. Infect. Imnnin. 65, 3708-3712. 119. Ito, Y, Azuma, T., Ito, S., Miyaji, H., Hirai, M., Yamazaki, Y, Sato, R, Kato, T., Kohli, Y, and Kuriyama, M. (1997). Analysis and typing of the vacA gene from cY/Â-positive strains of Helicobacter pylori isolated in Japan. J. Clin. Microbiol. 35, 1710-1714. 120. Wang, H.-J., Kuo, C.-H., Yeh, A. A. M., Chang, P C. L., and Wang, W.-C. (1998). Vacuolating toxin production in clinical isolates of Helicobacter pylori with different vacA genotypes. J. Infect. Dis. 178,207-212. 121. Bukanov, N. O., and Berg, D. E. (1994). Ordered cosmid library and high-resolution physicalgenetic map of Helicobacter pylori strain NCTC 11638. Mol. Microbiol. 11, 509-523. 122. Jiang, Q., Hiratsuka, K., and Taylor, D. E. (1996). Variability of gene order in different Helicobacter pylori strains contributes to genome diversity. Mol. Microbiol. 20, 833-842. 123. Jones, D. M. and Curry, A. (1989). The ultrastructure of Campylobacter pylori. In "Campylobacter Pylori and Gastroduodenal Disease" (J. Rathbone and R. V. Headey, eds), pp. 48-62. Blackwell Scientific, Boston. 124. Geis, G., Suerbaum, S. Forsthoff, B., Leying, H., and Opferkuch, W. (1993). Ultrastructure and biochemical studies of Helicobacter pylori flagellar sheath. J. Med. Microbiol. 38, 371-377. 125. Kostrzynska, M., Betts, J. D., Austin, J. W., and Trust, T. J. (1991). Identification, characterization, and spatial localization of two flagellin species in Helicobacter pylori flagella. / Bacteriol. 173, 937-946. 126. Suerbaum, S., Josenhans, C , and Labigne, A. (1993). Cloning and genetic characterization of the Helicobacter pylori and Helicobacter mustelae flaB flagellin genes and construction of H. pylori flaA- and flaB-negative mutants by electroporation-mediated allelic exchange. J. Bacteriol. 175, 3278-3288. 127. Leying, H., Suerbaum, S., Geis, G., and Haas, R. (1992). Cloning and genetic characterization of a Helicobacter pylori flagellin gene. Mol. Microbiol. 6, 2863-2874.

550


128. Josenhans, C , Labigne, A., and Suerbaum, S. (1995). Comparative ultrastructural and functional studies of Helicobacter pylori and Helicobacter mustelae flagellin mutants: Both flagellin subunits, FlaA and FlaB, are necessary for full motility in Helicobacter species. J. Bacteriol. Ill, 3010-3020. 129. O'Toole, P. W., Kostrzynska, M., and Trust, T. J. (1994). Non-motile mutants oi Helicobacter pylori and Helicobacter mustelae defective in flagellar hook production. Mol. Microbiol. 14, 691-703. 130. Eaton, K. A., Morgan, D. R., and Krakowka, S. (1992). Motility as a factor in the colonisation of gnotobiotic piglets by Helicobacter pylori. J. Med. Microbiol. 37, 123-127. 131. Eaton, K. A., Suerbaum, S., Josenhans, C , and Krakowka, S. (1996). Colonization of gnotobiotic piglets by Helicobacter pylori deficient in two flagellin genes. Infect. Immiin. 64, 2445-2448. 132. Schmitz, A., Josenhans, C, and Suerbaum, S. (1997). Cloning and characterization of the Helicobacter pylori flbA gene, which codes for a membrane protein involved in coordinated expression of flagellar genes. J. Bacteriol. 179, 987-997. 133. Mizote, T., Yoshiyama, H., and Nakazawa, T. (1997). Urease-independent chemotactic responses of Helicobacter pylori to urea, urease inhibitors, and sodium bicarbonate. Infect. Immun. 65, 1519-1521. 134. Mobley, H. L. T., Island, M. D., and Hausinger, R. R (1995). Molecular biology of microbial ureases. Microbiol. Rev. 59, 451-480. 135. Hu L.-T., and Mobley, H. L. T. (1990). Purification and N-terminal analysis of urease from Helicobacter pylori. Infect. Immun. 58, 992-998. 136. Dunn, B. E., Campbell, G. P, Perez-Perez, G. I., and Blaser, M. J. (1990). Purification and characterization of urease from Helicobacter pylori. J. Biol. Chem. 265, 9464-9469. 137. Labigne, A., Cussac, V., and Courcoux, P. (1991). Shuttle cloning and nucleotide sequences of Helicobacter pylori genes responsible for urease activity. J. Bacteriol. 173, 1920-1931. 138. Cussac, V., Ferrero, R. L., and Labigne, A. (1992). Expression of Helicobacter pylori urease genes in Escherichia coli grown under nitrogen-limiting conditions. J. Bacteriol. 174, 24662473. 139. Phadnis, S. H., Parlow, M. H., Levy, M., liver, D., Caulkins, C. M., Connors, J. B., and Dunn, B. E. (1996). Surface localization of Helicobacter pylori urease and a heat shock protein homolog requires bacterial autolysis. Infect. Immun. 64, 905-912. 140. Dunn, B. E., Vakil, N. B., Schneider, B. G., Zitzer, J. B., Puetz, T., and Phadnis, S. H. (1997). Localization of Helicobacter pylori urease and heat shock protein homolog in human gastric biopsies. Infect. Immun. 65, 1181-1188. 141. Mobley, H. L. T., Cortesia, M. J., Rosenthal, L. E., and Jones, B. D. (1988). Characterization of urease from Campylobacter pylori. J. Clin. Microbiol. 26, 831-836. 142. Ferrero, R. L., Cussac, V., Courcoux, P., and Labigne, A. (1992). Construction of isogenic urease-negative mutants oi Helicobacter pylori by allelic exchange. J. Bacteriol. 174, AIMAIM. 143. Eaton, K. A., and Krakowka, S. (1994). Effect of gastric pH on urease-dependent colonization of gnotobiotic piglets by Helicobacter pylori. Infect. Immun. 62, 3604-3607. 144. Meyer-Rosberg, K., Scott, D. R., Rex, D., Melchers, K., and Sachs, G. (1996). The effect of environmental pH on the proton motive force of Helicobacter pylori. Gastroenterology 111, 886-900. 145. Matin, A., Zychlinsky, E., Keyhan, M., and Sachs, G. (1996). Capacity of Helicobacter pylori to generate ionic gradients at low pH is similar to that of bacteria which grow under strongly acidic conditions. Infect. Immun. 64, 1434-1436.

11. H. PYLORI PATHOGENESIS

551

146. McGowan, C. C , Cover, T. L., and Blaser, M. J. (1996). Helicobacter pylori and gastric acid: Biological and therapeutic implications. Gastroenterology 110, 926-938. 147. Perez-Perez, G. I., Olivares, A. Z., Cover, T. L., and Blaser, M. J. (1992). Characteristics of Helicobacter pylori variants selected for urease deficiency. Infect. Immun. 60, 3658-3663. 148. McGowan, C. C , Cover, T. L., and Blaser, M. J. (1994). The proton pump inhibitor omeprazole inhibits acid survival of Helicobacter pylori by a urease-independent mechanism. Gastroenterology 107, 1573-1578. 149. Foster, J. W., and Hall, H. K. (1991). Inducible pH homeostasis and the acid tolerance response of Salmonella typhimurium. J. Bacteriol. 173, 5129-5135. 150. Karita, M., and Blaser, M. J. (1998). Acid-tolerance response in Helicobacter pylori and differences between cagA'' and cagA~ strains. J. Infect. Dis. 178, 213-219. 151. McGowan, C. C , Necheva, A., Thompson, S. A., Cover, T. L., and Blaser, M. J. (1998). Acid-induced expression of an LPS-associated gene in Helicobacter pylori. Mol. Microbiol. 30, 19-31. 152. Fox, J. G., Otto, G., Taylor, N. S., Rosenblad, W., and Murphy, J. C. (1991). Helicobacter py/on-induced gastritis and elevated gastric pH in the ferret (Mustela putorius furo). Infect. Immun. 59, 1875-1880. 153. McColl, K. L. E., El-Omar, E., and Gillen, D. (1997). Helicobacterpylori-'mduced alteration in gastric acid secretion. In "Pathogenesis and Host Response in Helicobacter pylori Infections" (A. P. Moran and C. A. O'Morain, eds.), pp. 119-127. Normed Verlag, Englewood, NJ. 154. Cave, D. R., and Vargas, M. (1989). Effect of a Campylobacter pylori protein on acid secretion by parietal cells. Lancet ii, 187-189. 155. Baume, P. E., Nicholls, A., and Baxter, C. H. (1967). Inhibition of gastric acid secretion by a purified bacterial lipopolysaccharide. Nature 215, 59-60. 156. Beales, K. L. P., and Calam, J. (1997). Effect of A^-alpha-methyl-histamine on acid secretion in isolated culture rabbit parietal cells: Implications for Helicobacter pylori associated gastritis and gastric physiology. Gut 40, 14-19. 157. Courillon-Mallet, A., Launay, J.-M., Roucayrol, A.-M., Callebert, J., Emond, J.-P., Tabuteau, F, and Cattan, D. (1995). Helicobacter pylori infection: Physiopathologic implications of A^-alphamethyl-histamine. Gastroenterology 108, 959-966. 158. Wallace, J. L., Cucala, M., Mugridge, K., and Parente, L. (1991). Secretagogue-specific effects of interleukin-1 on gastric acid secretion. Am J. Physiol. 261, G559-G564. 159. Moss, S. F, Legon, S., Bioshop, A. E., Polak, J. M., and Calam, J. (1992). Effect of Helicobacter pylori on gastric somatostatin in duodenal ulcer disease. Lancet 340, 930-932. 160. Mossi, S., Meyer-Wyss, B., Renner, E. L., Merki, H. S., Gamboni, G., and Beglinger, C. (1993). Influence of Helicobacter pylori, sex, and age on serum gastrin and pepsinogen concentrations in subjects without symptoms and patients with duodenal ulcers. Gut 34, 752-756. 161. Mulholland, G., Ardill, J. E. S., Fillmore, D., Chittajallu, R. S., Fullarton, G. M., and McColl K. E. (1993). Helicobacter pylori-re\a.ttd hypergastrinaemia is the result of a selective increase in gastrin 17. Gut 34, 757-761. 162. Calam, J., Biggons, A., Healey, Z. V., Bliss, P, and Arebi, N. (1997). How does Helicobacter pylori cause mucosal damage? Its effect on acid and gastrin physiology. Gastroenterology 113, S43-S49. 163. Guruge, J. L., Falk, P G., Lorenz, R. G., Dans, M., Wirth, H.-P, Blaser, M. J., Berg, D. E., and Gordon, J. I. (1998). Epithelial attachment alters the outcome of Helicobacter pylori infection. Proc. Natl. Acad. Sci. U.S.A. 95. 3925-3930.

552


164. Claeys, D., Faller, G., Appelmelk, M. J., Negrini, R., and Kirchner, T. (1998). The gastric H+,K+-ATPase is a major autoantigen in chronic Helicobacter pylori gastritis with body mucosa atrophy. Gastroenterology 115, 340-347. 165. Falk, P., Roth, K. A., Boren, T, Westblom, T U., Gordon, J. I., and Normark, S. (1993). An in vitro adherence assay reveals that Helicobacter pylori exhibits cell lineage-specific tropism in the human gastric epithelium. Proc. Natl. Acad. Sci. U.S.A. 90, 2035-2039. 166. Boren, T, Falk, R, Roth, K. A., Larson, G., and Normark, S. (1993). Attachment of//6'//a;/7flc7^r pylori to human gastric epithelium mediated by blood group antigens. Science 262, 1892-1895. 167. Falk, P. G., Holgersson, J., and Gordon, J. I. (1995). Expression of a human alpha-1,3/4-fucosyltransferase in the pit cell lineage of FVB/N mouse stomach results in production of Le'^-containing glycoconjugates: A potential transgenic mouse model for studying Helicobacter pylori infection. Proc. Natl. Acad. Sci. U.S.A. 92, 1515-1519. 168. Evans, D. G., Karjalainen, T K., Evans Jr., D. J., Graham, D. Y, and Lee, C. H. (1993). Cloning, nucleotide sequence, and expression of a gene encoding an adhesin subunit protein of Helicobacter pylori. J. Bacteriol. 175, 674-683. 169. Jones, A. C, Logan, R. P, Foynes, S., Cockayne, A., Wren, B. W., and Penn, C. W. (1997). A flagellar sheath protein of Helicobacter pylori is identical to HpaA, a putative A^-acetylneuraminyllactose-binding hemagglutinin, but is not an adhesin for AGS cells. J. Bacteriol. 179, 5643-5647. 170. O'Toole, P W., Janzon, L., Doig, P, Huang, J., Kostrzynska. M., and Trust, T. J. (1995). The putative neuraminyllactose-bindng hemagglutinin HpaA of Helicobacter pylori CCUG 17874 is a lipoprotein. J. Bacteriol. Ill, 6049-6057. 171. Odenbreit, S., Till, M., and Haas, R. (1996). Optimized BlaM-transposon shuttle mutagenesis of Helicobacter pylori allows the identification of novel genetic loci involved in bacterial virulence. Mol. Microbiol. 20, 361-373. 172. Fan, X., Crowe, S. E., Behar, S., Gunasena, H., Ye, G., Haeberle, H., van Houten, N., Gourley, W. K., Ernst, P B., and Reyes, V. E. (1998). The effect of class II major histocompatibility complex expression on adherence of Helicobacter pylori and induction of apoptosis in gastric epithleial cells: A mechanism for T helper cell type 1-mediated damage. J. Exp. Med. 187, 1659-1669. 173. Eaton, K. A., Morgan, D. R., and Krakowka, S. (1989). Campylobacter pylori virulence factors in gnotobiotic piglets. Infect. Imnnm. 57, 1119-1125. 174. Marchetti, M., Arico, B., Burroni, D., Figura, N., Rappuoli, R., and Ghiara, P (1995). Development of a mouse model of Helicobacter pylori infection that mimics human disease. Science 261, 1655-1658. 175. Dubois, A., Berg, D. E., Incecik, E. T, Fiala, N., Heman-Ackah, L. M., Perez-Perez, G. I., and Blaser, M. J. (1996). Transient and persistent experimental infection of nonhuman primates with Helicobacter pylori: Implications for human disease. Infect. Immun. 64, 2885-2891. 176. Ernst, P. B., Crowe, S. E., and Reyes, V. E. (1997). How does Helicobacter pylori cause mucosal damage? The inflammatory response. Gastroenterology 113, S35-S42. 177. Mai, U. E. H., Perez-Perez, G. I., Allen, J. B., Wahl, S. M., Blaser, M. J., and Smith, P D. (1992). Surface proteins from Helicobacter pylori exhibit chemotactic activity for human leukocytes and are present in gastric mucosa. J. Exp. Med. 175, 517-525. 178. Mai, U. E. H., Perez-Perez, G. I., Wahl, L. M., Wahl, S. M., Blaser, M. J., and Smith, P D. (1991). Soluble surface proteins from Helicobacter pylori activate monocytes/macrophages by a lipopolysaccharide-independent mechanism. J. Clin. Invest. 87, 894-900. 179. Harris, P R., Mobley, H. L., Perez-Perez, G. I., Blaser, M. J., and Smith, P D. (1996). Helicobacter pylori urease is a potent stimulus of mononuclear phagocyte activation and inflammatory cytokine production. Gastroenterology 111, 419^25.

11.

//. PK/.0/?/PATHOGENESIS

553

180. Makristathis, A., Rokita, E., Labigne, A., Willinger, B., Rotter, M. L., and Hirschl, A. M. (1998). Highly significant role of Helicobacter pylori urease in phagocytosis and production of oxygen metabolites by human granulocytes. J. Infect. Dis. Ill, 803-806. 181. Tufano, M. A., Rossano, F., Catalanotti, R, Liguori, G., Capasso, C , Ceccarelli, M. T., and Marinelli, R (1994). Immunobiological activities of Helicobacter pylori porins. Infect. Immun. 62, 1392-1399. 182. Evans Jr., D. J., Evans, D. G., Takemura, T., Nakano, H., Lampert, H. C , Graham, D. Y., Granger, D. N., and Kvietys, P. R. (1995). Characterization of d. Helicobacter pylori neutrophilactivating protein. Infect. Ininnin. 63, 2213-2220. 183. Enders, G., Brooks, W., von Jan, N., Lehn, N., Bayerdorffer, E., and Hatz, R. (1995). Expression of adhesion molecules on human granulocytes after stimulation with Helicobacter pylori membrane proteins: Comparison with membrane proteins from other bacteria. Infect. Immun. 63, 273-2477. 184. Yoshida, N., Granger, D. N., Evans, D. G., Graham, D. Y, Anderson, D. C , Wolf, R. E., and Kvietys, P. R. (1993). Mechanisms involved in Helicobacter pylori-induced inflammation. Gastroenterology 105, 1431-1440. 185. Appelmelk, B. J., Simoons-Smit, I., Negrini, R., Moran, A. P, Aspinall, G. O., Forte, J. G., De Vries, T., Quan, H., Verboom, T., Maaskant, J. J., Ghiara, P., Kuipers, E. J., Bloemena, E., Tadema, T. M., Townsend, R. R., Tyagarajan, K., Crothers Jr., J. M., Monteiro, M. A., Savio, A., and De Graaff, J. (1996). Potential role of molecular mimcry between Helicobacter pylori lipopolysaccharide and host Lewis blood group antigens in autoimmunity. Infect. Immun. 64, 2031-2040. 186. Negrini, R., Lisato, L., Zanella, I., Cavazzini, L., Gullini, S., Villancci, V., Poiese, C, Albertini, A., and Ghielmi, S. (1991). Helicobacter pylori infection induces antibodies cross-reacting with human gastric mucosa. Gastroenterology 101, 437-445. 187. Blanchard, T. G., Czinn, S. J., Nedrud, J. G., and Redline, R. W. (1995). Helicobacter-?i?>ôc\ciiQd gastritis in SCID mice. Infect. Immun. 63, 1113-1115. 188. Crabtree, J. E., Shallcross, T. M., Heatley, R. V., and Wyatt, J. I. (1991). Mucosal tumour necrosis factor alpha and interleukin-6 in patients with Helicobacter pylori-associated gastritis. Gut 32, 1473-1477. 189. Crabtree, J. E., Wyatt, J. I., Trejdosiewicz, L. K., Peichl, P, Nichols, P H., Ramsay, N., Primrose, J. N., and Lindley, I. J. (1994). Interleukin-8 expression in Helicobacter pylori infected, normal, and neoplastic gastroduodenal mucosa. J. Clin. Pathol. 47, 61-66. 190. Peek Jr., R. M., Miller, G. G., Tham, K. T., Perez-Perez, G. I., Zhao, X., Atherton, J. C , and Blaser, M. J. (1995). Heightened inflammatory response and cytokine expression in vivo to cagA'^ Helicobacter pylori strains. Lab. Invest. 73, 760-770. 191. Crowe, S. E., Alvarez, L., Dytoc, M., Hunt, R. H., Muller, M., Sherman, P Patel, J., Jin, Y, and Ernst, P. B. (1995). Expression of interleukin 8 and CD54 by human gastric epithelium after Helicobacter pylori infection in vitro. Gastroenterology 108, 65-74. 192. Munzenmaier, A., Lange, C, Glocker, E., Covacci, A., Moran, A., Bereswill, S., Baeuerle, P. A., Kist, M., and Pahl, H. L. (1997). A secreted/shed product of Helicobacter pylori activates transcription factor nuclear factor-kappa B. J. Immunol. 159, 6140-6147. 193. Sharma, S. A., Tummuru, M. K. R., Blaser, M. J., and Kerr, L. D. (1998). Activation of IL-8 gene expression by Helicobacter pylori is regulated by transcription factor nuclear factor-kB in gastric epithelial cells. J. Immunol. 160, 2401-2407. 194. Aihara, M., Tsuchimoto, D., Takizawa, H., Azuma, A., Wakebe, H., Ohmoto, Y, Imagawa, K., Kikuchi, M., Mukaida, N., and Matsushima, K. (1997). Mechanisms involved in Helicobacter /7v76>r/-induced interleukin-8 production by a gastric cancer cell line, MKN45. Infect. Immun. 65,3218-3224.

554


195. Beales, I. L., and Calam, J. (1997). Stimulation of IL-8 production in human gastric epithelial cells by Helicobacter pylori, IL-lbeta and TNF-alpha requires tyrosine kinase activity, but not protein kinase C. Cytokine 9, 514-520. 196. Smoot, D. T., Resau, J. H., Naab, T., Desbordes, B. C , Gilliam, T., Bull-Henry, K., Curry, S. B., Nidiry, J., Sewchand, J., Mills-Robertson, K., et al (1993). Adherence oi Helicobacter pylori to cultured human gastric epithelial cells. Infect. Immiin. 61, 350-355. 196a. Segal, E. D., Cha, J., Lo, J., Falkow, S., and Tompkins, L. S. (1999). Altered states: Involvement of phosphorylated CagA in the induction of host cellular growth changes by Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 96, 14559-14564. 197. Jones, N. L., Shannon, R T, Cutz, E., Yeger, H., and Sherman, R M. (1997). Increase in proliferation and apoptosis of gastric epithelial cells early in the natural history of Helicobacter pylori infection. Am. J. Pathol. 151, 1695-1703. 198. Wagner, S., Beil, W., Westermann, J., Logan, R. R, Bock, C. T, Trautwein, C , Bleck, J. S., and Manns, M. R (1997). Regulation of gastric epithelial cell growth by Helicobacter pylori: Evidence for a major role of apoptosis. Gastroenterology 113, 1836-1847. 199. Rudi, J., Kuck, D., Strand, S., Vonherbay, A., Mariani, S. M., Krammer, R H., Galle, R R., and Stremmel, W. (1998). Involvement of the CD95 (apo-1/fas): Receptor and ligand system in Helicobacter pylori-induced gastric epithelial apoptosis. J. Clin. Invest. 102, 1506-1514. 200. Engstrand, L., Scheynius, A., Pahlson, C , Grimelius, L., Schwan, C. A., and S. Gustavsson. (1989). Association of Campylobacter pylori with induced expression of class II transplantation antigens on gastric epithelial cells. Infect. Immun. 57, 827-832. 201. Chen, G., Sordillo, E. M., Ramey, W. G., Reidy, J., Holt, R R., Krajewski, S., Reed, J. C , Blaser, M. J,, and Moss, S. F. (1997). Apoptosis in gastric epithelial cells is induced by Helicobacter pylori and accompanied by increased expression of BAK. Biochem. Biophys. Res. Commun. 239, 626-632. 202. Brenes, F., Ruiz, B., Correa, R, Hunter, F, Rhamakrishnan, T, Fontham, E., and Shi, T Y. (1993). Helicobacter pylori causes hyperproliferation of the gastric epithelium: Pre- and post-eradication indices of proliferating cell nuclear antigen. Am. J. Gastroenterol. 88, 1870-1875. 203. Chow, K. W., Bank, S., Ahn, J., Roberts, J., Blumstein, M., and Kranz, V. (1995). Helicobacter pylori infection does not increase gastric antrum mucosal cell proliferation. Am. J. Gastroenterol. 90, 64-66. 204. Peek Jr., R. M., Moss, S. F., Tham, K. T, Perez-Perez, G. I., Wang, S., Miller, G. G., Atherton, J. C , Holt, P. R., and Blaser, M. J. (1997). Helicobacter pylori cagA"" strains and dissociation of gastric epithelial cell proliferation from apoptosis. J. Natl. Cancer Inst. 89, 863-868. 205. Leunk, R. D., Johnson, R T, David, B. C , Kraft, W. G., and Morgan, D. R. (1988). Cytotoxic activity in broth culture filtrates of Campylobacter pylori. J. Med. Microbiol 26, 93-99. 206. Cover, T. L., and Blaser, M. J. (1992). Purification and characterization of the vacuolating toxin from Helicobacter pylori. J. Biol. Chem. 267, 10570-10575. 207. Cover, T. L. (1996). The vacuolating cytotoxin of Helicobacter pylori. Mol. Microbiol. 20, 241-246. 208. Cover, T. L., Cao, R, Murthy, U. K., Sipple, M. S., and Blaser, M. J. (1992). Serum neutralizing antibody response to the vacuolating cytotoxin of Helicobacter pylori. J. Clin. Invest. 90, 913-918. 209. Cover, T. L., Tummuru, M. K. R., Cao, R, Thompson, S. A., and Blaser, M. J. (1994). Divergence of genetic sequences for the vacuolating cytotoxin among Helicobacter pylori strains. J. Biol. Chem. 269, 10566-10573.

11.

//. pyLO/?/PATHOGENESIS

555

210. Telford, J. L., Ghiara, P., Dell'Orco, M., Comanducci, M., Burroni, D., Bugnoli, M., Tecce, M. R, Censini, S., Covacci, A., Xiang, Z., Papini, E., Montecucco, C , Parente, L., and Rappuoli, R. (1994). Gene structure of the Helicobacter pylori cytotoxin and evidence of its key role in gastric disease. J. Exp. Med. 179, 1653-1658. 211. Schmitt W., and Haas R. (1994). Genetic analysis of the Helicobacter pylori vacuolating cytotoxin: Structural similarities with the IgA protease type of exported protein. Mol. Microbiol. 12,307-319. 212. Lupetti, P., Heuser, J. E., Manetti, R., Massari, P., Lanzavecchia, S., Bellon, P. L., Dallai, R., Rappuoli, R., and Telford, J. L. (1996). Oligomeric and subunit structure of the Helicobacter pylori vacuolating cytotoxin. / Cell Biol. 133, 801-807. 213. Cover, T. L., Hanson, P. I., and Heuser, J. E. (1997). Acid-induced dissociation of VacA, the Helicobacter pylori vacuolating cytotoxin, reveals its pattern of assembly. J. Cell Biol. 138, 759-769. 214. Forsyth, M. H., Atherton, J. C., Blaser, M. J., and Cover, T. L. (1998). Heterogeneity in levels of vacuolating cytotoxin gene (vacA) transcription among Helicobacter pylori strains. Infect. Immun. 66, 3088-3094. 215. Perez-Perez, G. I., Peek Jr., R. M., Atherton, J. C , Blaser, M. J., and Cover, T L. (1999). Detection of anti-VacA antibody responses in serum and gastric juice samples using type si/ml and s2/m2 Helicobacter pylori VacA antigens. Clin. Diag. Lab. Immunol. 64, 489-493. 216. Pagliaccia, C , de Bernard, M., Lupetti, P., Ji, X., Burroni, D., Cover, T. L., Papini, E., Rappuoli, R., Telford, J. L., and Reyrat, J-M. (1998). The m2 form of the Helicobacter pylori cytotoxin has cell type-specific vacuolating activity. Proc. Natl. Acad. Sci. U.S.A. 95, 10212-10217. 217. de Bernard, M., Arico, B., Papini, E., Rizzuto, R., Grandi, G., Rappuoli, R., and Montecucco, C. (1997). Helicobacter pylori toxin VacA induces vacuole formation by acting in the cell cytosol, Mol. Microbiol. 26, 665-674. 218. Montecucco, C , Papini, E., and Schiavo, G. (1994). Bacterial protein toxins penetrate cells via a four-step mechanism. FEBS Lett. 346, 92-98. 219. Massari, P, Manetti, R., Burroni, D., Nuti, S., Norais, N., Rappuoli, R., and Telford, J. L. (1998). Binding of the Helicobacter pylori vacuolating cytotoxin to target cells. Infect. Immun. 66, 3981-3984. 220. Czajkowsky, D. M., Iwamoto, H., Cover, T. L., and Shao, Z. (1999). The vacuolating toxin from Helicobacter pylori forms hexameric pores in lipid bilayers at low pH. Proc. Natl. Acad. Sci. U.S.A. 96, 2001-2006. 221. Moll, G., Papini, E., Colonna, R., Burroni, D., Telford, J., Rappuoli, R., and Montecucco, C. (1995). Lipid interaction of the 37-kDa and 58-kDa fragments of the Helicobacter pylori cytotoxin. Eun J. Biochem. 234, 947-952. 222. Gamer, J. A., and Cover, T. L. (1996). Binding and internalization of Helicobacter pylori cytotoxin by epithelial cells. Infect. Immun. 64, 4197-4203. 223. de Bernard, M., Burroni, D., Papini, E., Rappuoli, R., Telford, J., and Montecucco, C. (1998). Identification of the Helicobacter pylori VacA toxin domain active in the cell cytosol. Infect. Immun. 66, 6014-6016. 224. de Bernard, M., Papini, E., de Filippis, V, Gottardi, E., Telford, J., Manetti, R., Fontana, A., Rappuoli, R., and Montecucco, C, (1995). Low pH activates the vacuolating toxin of Helicobacter pylori, which becomes acid and pepsin resistant. J. Biol. Chem. 270, 23937-23940. 225. Molinari, M., Galli, C , de Bernard, M., Norais, N., Ruysschaert, J.-M., Rappuoli, R., and Montecucco, C. (1998). The acid activation oi Helicobacter pylori toxin VacA: Structural and membrane binding studies. Biochem. Biophys. Res. Commun. 248, 334-340.

556


226. Tombola, F, Carlesso, C , Szabo, I., de Bernard, M., Reyrat, J. M., Telford, J. L., Rappuoli, R., Montecucco, C , Papini, E., and Zoratti, M. (1999). Helicobacter pylori vacuolating toxin forms anion-selective channels in planar lipid bilayers: Possible implications for the mechanism of cellular vacuolation. Biophys. J. 76, 1401-1409. 227. Papini, E., deBernard, M., Milia, E., Bugnoli, M., Zerial, M., Rappuoli, R., and Montecucco, C. (1994). Cellular vacuoles induced by Helicobacter pylori originate from late endosomal compartments. Proc. Natl. Acad. Sci. U.S.A. 91, 9720-9724. 228. Molinari, M., Galli, C , Norais, N., Telford, J. L., Rappuoli, R., Luzio, J. P., and Montecucco, C. (1997). Vacuoles induced by Helicobacter pylori toxin contain both late endosomal and lysosomal markers. J. Biol. Chem. 212, 25339-25344. 229. Papini, E., Satin, B., Bucci, C , de Bernard, M., Telford, J. L.. Manetti, R., Rappuoli, R., Zerial, M., and Montecucco, C. (1997). The small GTP binding protein rab7 is essential for cellular vacuolation induced by Helicobacter pylori cytotoxin. EMBO J. 16, 15-24. 230. Satin, B., Norais, N., Telford, J., Rappuoli, R., Murgia, M., Montecucco, C , and Papini, E. (1997). Effect oi Helicobacter pylori vacuolating toxin on maturation and extracellular release of procathepsin D and on epidermal growth factor degradation. J. Biol. Chem. 272, 2502225028. 231. Montecucco, C , Papini, E., and Schiavo, G. (1996). Bacterial protein toxins and cell vesicle trafficking. Experientia 52, 1026-1032. 232. Wirth, H.-P, Beins, M. H., Yang, M., Tham, K. T, and Blaser, M. J. (1998). Experimental infection of Mongolian gerbils with wild-type and mutant Helicobacter pylori strains. Infect. Immun. 66, 4856-4866. 233. Eaton, K. A., Cover, T L., Tummuru, M. K. R., Blaser, M. J., and Krakowka, S. (1997). Role of vacuolating cytotoxin in gastritis due to Helicobacter pylori in gnotobiotic piglets. Infect. Immun. 65, 3462-3464. 234. Molinari, M., Salio, M., Galli, C, Norais, N., Rappuoli, R., Lanzavecchia, A., and Montecucco, C. (1998). Selective inhibition of li-dependent antigen presentation by Helicobacter pylori toxin Vac A. J. Exp. Med. 187, 135-140. 235. Papini, E., Satin, B., Norais, N., de Bernard, M., Telford, J. L., Rappuoli, R., and Montecucco, C. (1998). Selective increase of the permeability of polarized epithelial cell monolayers by Helicobacter pylori vacuolating toxin. J. Clin. Invest. 102, 813-820. 236. Kirschner, D. E., and Blaser, M. J. (1995). The dynamics of Helicobacter pylori infection of the human stomach. / Theor Biol. 176, 281-290. 237. Blaser, M. J. (1992). Hypotheses on the pathogenesis and natural history of Helicobacter pylori-induced inflammation. Gastroenterology 102, 720-727. 238. Knipp, U., Birkholz, S., Kaup, W., and Opferkuch, W. (1993). Immune suppressive effects of Helicobacter pylori on human peripheral blood mononuclear cells. Med. Microbiol. Immunol. 182, 63-76. 239. Hazell, S. L., Evans, D. J., and Graham, D. Y. (1991). Helicobacter pylori catalase. J. Gen. Microbiol. 137,57-61. 240. Spiegelhalder, C , Gerstenecker, B., Kersten, A., Schiltz, E., and Kist, M. (1993). Purification of Helicobacter pylori superoxide dismutase and cloning and sequencing of the gene. Infect. Immun. 61,5315-5325. 241. Bamford, K. B., Fan, X., Crowe, S. E., Leary, J. F, Gourley, W. K., Luthra, G. K., Brooks, E. G., Graham, D. Y, Reyes, V. E., and Ernst, P. B. (1998). Lymphocytes in the human gastric mucosa during Helicobacter pylori have a T helper cell 1 phenotype. Gastroenterology 114, 482-492.

11.

//. PVLO/?/PATHOGENESIS

557

242. Sommer, F., Faller, G., Konturek, P., Kirchner, T, Hahn, E. G., Zeus, J., Rollinghoff, M., and Lohoff, M. (1998). Antrum- and corpus mucosa-infiltrating CD4^ lymphocytes in Helicobacter pylori gastritis display a Thl phenotype. Infect. Immiin. 66, 5543-5546. 243. Saunders, N. J., Peden, J. F, Hood, D. W., and Moxon, E. R. (1998). Simple sequence repeats in the Helicobacter pylori genome. Mol. Microbiol. 27, 1091-1098. 244. Appelmik, B. J., Shiberu, B., Trinks, C , Tapsi, N., Zheng, P Y., Verboom, T., Maaskant, J., Hokke, Ch., Schiphorst, W. E., Blanchard, D., Simoons-Smit, I. M., van den Eijnden, D. H., and Vandenbrouche-Grauls, C. M. (1998). Phase variation in Helicobacter pylori lipopolysaccharide. Infect. Immun. 66, 70-16. 245. Moran, A. P., and Aspinall, G. O. (1998). Unique structural and biological features of Helicobacter pylori lipopolysaccharides. Prog. Clin. Biol. Res. 397, 37-49. 246. Moran, A. P. The role of lipopolysaccharide in Helicobacter pylori pathogenesis. Aliment. Pharmacol. Then 10 (Suppl. 1), 39-50. 247. Moran, A. P, Lindner, B., and Walsh, E. J. (1997). Structural characterization of the lipid A component of Helicobacter pylori rough- and smooth-form lipopolysaccharides. J. Bacteriol. 179, 6453-6463. 248. Suda, Y., Ogawa, T., Kashihara, W., Oikawa, M., Shimoyama, T., Hayashi, T., Tamura, T., and Kusumoto, S. (1997). Chemical structure of lipid A from Helicobacter pylori strain 206-1 lipopolysaccharide./ Biochem. 121, 1129-1133. 249. Aspinall, G. O., and Monteiro, M. A. (1996). Lipopolysaccharides of Helicobacter pylori strains P466 and MO 10: Structures of the O antigen and core oligosaccharide regions. Biochemistry 35, 2498-2504. 250. Aspinall, G. O., Monteiro, M. A., Pang, H., Walsh, E. J., and Moran, A. P (1996). Lipopolysaccharide of the Helicobacter pylori type strain NCTC 11637 (ATCC 3504): Structure of the O antigen chain and core oligosaccharide regions. Biochemistry 35, 2489-2497. 251. Moran, A. P., Helander, I. M., and Kosunen, T. U. (1992). Compositional anlysis oiHelicobacter pylori rough-form lipopolysaccharides. J. Bacteriol. 174, 1370-1377. 252. Mattsby-Baltzer, L, Mielniczuk, Z., Larsson, L., Lindgren, K., and Goodwin, S. (1992). Lipid A in Helicobacter pylori. Infect. Immun. 60, 4383-4387. 253. Birkholz, S., Knipp, U., Netzki, C , Adamek, R. J., and Opferkuch, W. (1993). Immunological activity of lipopolysaccharide of Helicobacter pylori on human peripheral mononuclear blood cells in comparison to lipopolysaccharides of other intestinal bacteria. FEMS Immunol. Med. Microbiol. 6, 317-324. 254. Perez-Perez, G. I., Shepherd, V. L., Morrow, J. D., and Blaser, M. J. (1995). Activation of human THP-1 and rat bone marrow-derived macrophages by Helicobacter pylori lipopolysaccharide. Infect. Immun. 63, 1183-1187. 255. Sherburne, R., and Taylor, D. E. (1995). Helicobacter pylori expresses a complex surface carbohydrate, Lewis X. Infect. Immun. 63, 4564-4568. 256. Wirth, H.-R, Yang, M., Karita, M., and Blaser, M. J. (1996). Expression of the human cell surface glycoconjugates Lewis X and Lewis Y by Helicobacter pylori isolates is related to cagA status. Infect. Immun. 64, 4598^605. 257. Monteiro, M. A., Chan, K. H., Rasko, D. A., Taylor, D. E., Zheng, P Y, Appelmik, B. J., Wirth, H.-P, Yang, M., Blaser, M. J., Hynes, S. O., Moran, A. P, and Perry, M. B. (1998). Simultaneous expression of type 1 and 2 Lewis blood-group antigens by Helicobacter pylori lipopolysaccharides: Molecular mimicry between H. pylori lipopolysaccharides and human gastric epithelial cell surface glycoforms. / Biol. Chem. IIX 11533-11543. 258. Wirth, H. P, Yang, M., Peek Jr., R. M., Tham, K. T, and Blaser, M. J. (1997). Helicobacter pylori Lewis expression is related to the host Lewis phenotype. Gastroenterology 113, 1091-1098.

558


259. Taylor, D. E., Rasko, D. A., Sherburne, R., Ho, C , and Jewell, L. D. (1998). Lack of correlation between Lewis antigen expression by Helicobacter pylori and gastric epithelial cells in infected patients. Gastroenterology 115, 1113-1122. 260. Blaser, M. J. (1997). Ecology oi Helicobacter pylori in the human stomach. J. Clin. Invest. 100, 759-762. 261. Blaser, M. J. (1998). Helicobacters are indigenous to the human stomach: Duodenal ulceration is due to changes in gastric microecology in the modern era. Gut 43, 721-727. 262. van Doom, L. J., Figueiredo, C , Sanna, R., Plaisier, A., Schneeberger, R, de Boer, W., and Quint, W. (1998). Clinical relevance of the cagA, vacA, and iceA status oi Helicobacter pylori. Gastroenterology 115, 58-66. 263. Blaser, M. J., Perez-Perez, G. L, Kleanthous, H., Cover, T. L., Peek, R. M., Chyou, P H., Stemmermann, G. N., and Nomura, A. (1995). Infection with Helicobacter pylori strains possessing cagA is associated with an increased risk of developing adenocarcinoma of the stomach. Cancer Res. 55, 2111-2115. 264. Kuipers, E. J., Perez-Perez, G. I., Meuwissen, S. G., and Blaser, M. J. (1995). Helicobacter pylori and atrophic gastritis: Importance of the cagA status. J. Natl. Cancer Inst. 87, 1777-1780. 265. Parsonnet, J., Friedman, G. D., Orentreich, N., and Vogelman, H. (1997). Risk for gastric cancer in people with CagA positive or CagA negative Helicobacter pylori infection. Gut 40, 297-301. 266. Pan, Z.-J., van der Hulst, R. W. M., Feller, M., Xiao, S.-D., Tytgat, G. N. J., Dankert, J., and van der Ende, A. (1997). Equally high prevalences of infection with c«Â-positive Helicobacter pylori in Chinese patients with peptic ulcer disease and those with chronic gastritis-associated dyspepsia. J. Clin. Microbiol. 35, 1344-1347. 267. Brenner, H., Rothenbacher, D., Bode, G., and Adler, G. (1998). The individual and joint contributions of Helicobacter pylori infection and family history to the risk for peptic ulcer disease. J. Infect. Dis. Ill, 1124-1127. 268. Hansson, L.-E., Nyren, O., Hsing, A. W., Bergstrom, R., Josefsson, S., Chow, W.-H., Fraumeni, J. F, and Adami, H.-O. (1996). The risk of stomach cancer in patients with gastric or duodenal ulcer disease. New Engl. J. Med. 335, 242-249. 269. Blaser, M. J., Chyou, P. H., and Nomura, A. (1995). Age at establishment oi Helicobacter pylori infection and gastric carcinoma, gastric ulcer, and duodenal ulcer risk. Cancer Res. 55, 562-565.

CHAPTER 12

Neisseria SCOTT D . GRAY-OWEN CHRISTOPH DEHIO THOMAS RUDEL MICHAEL NAUMANN THOMAS F. MEYER 1. Introduction A. Neisserial Morphology and Physiology B. Neisserial Infections II. Natural Competence for Transformation III. Surface Structures A. Lipooligosaccharide B. The Meningococcal Capsule IV. Tissue Colonization A. Neisserial Type 4 Pilus B. Opa-Mediated Interactions C. Opc-Mediated Interactions D. Interactions Mediated by a Novel Multiple Adhesin Family V. PorB A. Influence of PorB on Cellular Interactions B. PorB Induction of Apoptosis VI. IgAl Protease VII. Iron Acquisition in Vivo VIII. Immune Response A. Cellular Response to Neisserial Infections B. Humoral Response to Neisserial Infection IX. Summary References

559 560 560 566 567 567 569 570 570 576 585 585 586 587 589 590 592 594 594 597 599 600

/. Introduction The genus Neisseria contains two human pathogenic species, as well as a number of other species that are either pathogenic to animals or are normal flora in either humans or animals. These fascinating organisms are exquisitely adapted to life within their host, a fact that is manifested by their limited biosynthetic capabilities as compared to enteric or free-living bacteria. Because of this, each species is Principles of Bacterial Pathogenesis Copyright © 2001 by Academic Press All rights of reproduction in any form reserved. ISBN 0-12-304220-8

SSQ

560

SCOTT D. GRAY-OWEN ETAL

usually able to colonize only a single host species. The pathogenic Neisseria spp. are also capable of evading the massive immune response that is typical of neisserial disease. In part, this is due to their ability to undergo both the frequent change of immunodominant epitopes (i.e., antigenic variation) and the high-frequency, reversible switching "on" and "off of the expression of various surface-exposed virulence factors (i.e., phase variation). In addition to presenting the immune system with an alternative antigenic make-up, each of these mechanisms can also result in a change in a bacterium's functional characteristics. For example, the phase-variable expression of adhesin molecules may result in generation of bacteria capable of invading tissues that the parental phenotype could not. This could potentially allow a subpopulation of the bacteria to become sequestered in a site that is relatively more protected from the immune system. The highly efficient exchange of chromosomal sequences between neisserial bacteria strains and species during mixed infections also provides a mechanism for the strain to adapt over a longer time-frame. Together, these features provide the Neisseria with a powerful arsenal of weapons to persist within their target population. A. Neisserial Morphology and Physiology The members of genus Neisseria, which belongs to the family Neisseriaceae, are Gram-negative diplococci with adjacent sides flattened. Since the bacteria divide in two planes, tetrads can also be observed. Single bacteria range from 0.6 to 1.5 |Lim in size, are nonmotile, and do not produce endospores. They are typically considered to be aerobic, with optimal growth occurring between 35 and 37°C in the presence of high humidity and 5% CO2. A^. gonorrhoeae can, however, also survive in an anaerobic environment. Some species may express a carbohydrate capsule, while others cannot. All species are oxidase positive, and most are catalase positive. Some strains are also highly sensitive to fatty acids, often necessitating incorporation of soluble starch into the growth medium. Consistent with their limited host specificity and the fact that they are obligate parasites, Neisseria have limited metabolic capacities. The species can therefore be differentiated based on their varying abilities to produce polysaccharide from sucrose, their catalase and DNase activities, and their ability to oxidatively (i.e., not fermentatively) produce acid from various carbohydrates, reduce nitrate and nitrite, oxidize fatty acids, and produce certain enzymes [1]. B. Neisserial Infections Despite their close evolutionary relationship [2], N. gonorrhoeae primarily infects the urogenital or anorectal mucosa following intimate sexual contact, while A^. meningitidis instead colonizes the nasopharynx after the inhalation of infected respiratory droplets. These associations are at least partially the result of their respective modes of transmission rather than to a tropism for these loci, since

12. NEISSERIA

561

gonococcal pharyngitis and meningococcal anogenital infections have also been described [3, 4]. The meningococcal polysaccharide capsule (Fig. 1 A) is a major virulence factor in this respect, since it does contribute to this organism's abihty to be spread via the aerosol route, while the absence of such a capsule makes the gonococci highly susceptible to drying when outside of the host. Although they are infrequently described as being the etiologic agents of various opportunistic infections, other neisserial species are generally considered as normal flora of the oro- and nasopharynges. The virulence mechanisms and primary interactions that occur between the pathogenic Neisseria and various mucosal surfaces are generally quite similar, and are depicted schematically in Figure IB. In this chapter, we will thus discuss neisserial virulence mechanisms in general terms, with specific differences between N. gonorrhoeae and N. meningitidis being highlighted as appropriate. 1.

NEISSERIA GONORRHOEAE

N gonorrhoeae is the second leading cause of sexually transmitted disease in the United States, one of the few countries in which gonorrhea is reportable. Over 325,000 cases were reported to the American Centers for Disease Control in 1996; however, this is considered a conservative estimate due to significant underreporting. Gonorrhea is also a major concern in non-Western countries, as evidenced by the fact that it is the most common sexually transmitted disease in China [5]. The incidence of gonococcal infection generally correlates with low socioeconomic status, being highest in poor regions of both developed and developing nations. In Africa, gonococci remain endemic in poorer regions, with a recent report showing 3.4% of men in a study group of transport workers in Kenya to have a urethral gonococcal infection [6]. For epidemiological studies, gonococcal strains are typically characterized by auxotyping and/or serotyping. Auxotyping is based on the different growth requirements of various gonococcal strains for specific nutrients or cofactors. Using different chemically defined growth media, over 30 different auxotypes have been identified [7]. A more common typing scheme is based on the antigenic characterization of protein I (PorB; Fig. lA), since variant alleles of this constitutively expressed protein are generally stably maintained in different strains. Currendy, the major serogroups of protein I, termed lA and IB, have been further subdivided into 26 and 31 serovars, respectively (e.g., serovar IA-21) [8, 9]. In men, gonorrhea typically presents as an acute urethritis after an incubation period of 2-5 days, with symptoms including purulent discharge and dysuria. Acute epididymitis is the most common complication of untreated gonococcal infection; however, disseminated gonococcal disease can also occur. In women, the majority of disease results from uncomplicated infections of the lower genital tract, with the primary site of infection being the endocervix. Estimates suggest that up to 80% of women may be either asymptomatic or possess minor symptoms that do not induce them to seek medical attention [10-121. Symptomatic disease is thought to present within 10 days of infection, and is characterized by cervicitis

562

SCOTT D . GRAY-OWEN ETAL

niys

Porin

Hiivosiofi ^Fhogosoitte mcityriitidii LOS

Opo J3H -Aiili#r»ii€iil '-Imm^lm

\

^

Y*' L^^

^ ^ ^ ^ ^ . ^ ^ \ \

^S«iiiiti r0slsten:c« {if si«ilyicite S O

Wj C^

(U ^ .:

w I .3f ^

is

^

r=

i ^ ^ I "i

•- ^ ^ o so • •S ^ L4 2

i2 *-* o : 1 (u o U ^ 1- C ^

ingomy«iiii

Cytoskeletai rearrangements I'hogocytosis

DAG

Chong

A| A

Helo/CHO HEp-2

Fig. 4 Intracellular signaling events that follow neisserial contact. (A) Heparan sulfate proteoglycan-dependent invasion into cultured epithelial cell lines. Depending on the cultured epithelial cell type being used, Opaso-expressing gonococci can be engulfed by at least three different processes, (a) HSPG-mediated invasion into Chang conjunctiva epithelial cells involves activation of phosphatidylcholine-dependent phospholipase C (PC-PLC), which generates the second messenger diacylglycerol (DAG) from phosphatidylcholine (PC). DAG activates the acidic sphingomyelinase (ASM), which then generates ceramide from sphingomyelin. By an uncharacterized process, ceramide seems to mediate cytoskeletal rearrangements that result in bacterial uptake via a mechanism that resembles conventional phagocytosis, (b) Efficient bacterial uptake into HeLa cervical carcinoma and Chinese hamster ovary (CHO) cells relies on the ability of Opaso to bind both the cellular HSPG receptors and the extracellular matrix protein vitronectin (VN). The subsequent binding of VN to its a^ integrincontaining receptors coligates the HSPG and VN receptor, resulting in bacterial engulfment by a process that requires the activity of protein kinase C (PKC). (c) In Hep-2 larynx carcinoma cells, the efficient uptake of gonococci instead requires Opaso binding to the extracellular matrix protein fibronectin (FN). Presumably, this would result in bacterial uptake being mediated by the coligation of the HSPG and integrin-containing FN receptors in a manner reminiscent to that outlined in (b).

ceramide, respectively (Fig. 4A) [191]. In many other epithelial cell lines (i.e., HeLa human cervical carcinoma and Chinese hamster ovary (CHO) cells), this signaling pathway appears to be less prominent and bacterial entry is poor. In these cell lines, however, an alternative pathway of HSPG-dependent invasion is triggered in the presence of serum. One serum-derived factor that stimulates this phenomenon is the extracellular matrix protein vitronectin (VN) [192, 193]. VN binds specifically to Opaso-expressing gonococci and stimulates bacterial uptake into HeLa cells in an ttvPs and avp5 integrin-dependent manner.

12.

579

NEISSERIA

flÊ CEACAM1 CEACAM6 CEACAM3

Hcr/par--^

« « '

P "lyr

-frrô'ittU

PAK j JNK i TranscriptioR

1 Phagoi;ytosls

JOSK-M/PMN F/g. 4 (B) CEACAM-dependent invasion into phagocytic cells. The opsonin-independent phagocytosis of Neisseria by polymorphonuclear neutrophils and by the myelomonocytic cell line JOSK-M is mediated by Opa binding to the cellular CEACAMl (biliary glycoprotein, BGP), CEACAM6 (nonspecific crossreacting antigen, NCA), and/or CEACAM3 (CEAgene family member 1, CGMla) receptors. Interactions mediated by the CEACAM-specific Opa proteins (e.g., Opa52) result in activation of the Src-family nonreceptor protein tyrosine kinases Hck and Fgr. This results in an increased level of cellular protein tyrosine phosphorylation and activation of the small GTP-binding protein Racl, which has been implicated in the cytoskeletal rearrangements that ultimately lead to the phagocytic uptake of bound bacteria. Rac 1 also activates p21 -activated protein kinase (PAK) and Jun N-terminal kinase (JNK), probably leading to subsequent activation of nuclear transcription of stress response mediators.

'"*. ?

w . GTPoses

t

PAK

Cytoskeletal rearrangemefits

MKKKs

Epitiieliaiceli

t MKK4 JNL

-'

^

Cytokines/ chem«kliies

Nucleus

1 AP-IJ|NF-KB| X

/

Fig. 4 (C). Opa-independent induction of a cellular cytokine response. Gonococcal binding to epithelial cells also triggers a signaling cascade that is independent of Opa expression or cellular invasion, as evidenced by the fact that it is also seen following infection by nonopaque, noninvasive gonococci. The response involves activation of Rho-family GTPases Rac 1 and/or Cdc42, followed by the activation of a PAK kinase. Subsequent activation of MAP kinase kinase kinase (MKKK), MAP kinase kinase 4 (MKK4), and c-Jun N-terminal kinase (JNK) ultimately results in induction of the immediate early response: basic leucine zipper transcription factor activator protein 1 (AP-1). Whether the resulting synthesis and release of immune response mediators, including some inflammatory cytokines, benefits the host or the infecting microbe is still uncertain.

580


This signaling process also appears to depend on the activity of protein kinase C (PKC) (Fig. 4A) [194]. A specific role for HSPG ligation in these two distinct Opa5o-dependent bacterial uptake mechanisms has been confirmed using latex beads coated with antibodies directed against the HS-GAG sidechains of HSPGs, thereby mimicking the binding activity of Opa5(). Consistent with bacterial uptake, these beads were efficiently internalized by Chang cells in the absence of any additional factor, while serum or purified VN (which nonspecifically associates with the beads) was necessary to stimulate efficient uptake into HeLa cells [195]. Aside from VN, the functionally related extracellular matrix protein fibronectin (FN) is also bound by OpasQ-expressing gonococci. This interaction triggers bacterial uptake into HEp-2 cells by coligating HSPGs and the FN-integrin receptor a5 (31 (Fig.4A) [196]. The core protein of different HSPGs may either possess a transmembrane and intracellular domain, composing the syndecan receptor family, or may be glycosylphosphatidylinositol (GPI)-anchored to the membrane, composing the glypican family of HSPGs [197]. Syndecan-4 is widely expressed by many epithelial cell lines. Interestingly, the overexpression of this receptor by HeLa cells increases both the VN-triggered uptake of Opaô-expressing gonococci and the basal level of uptake in the absence of VN. Overexpression of a mutant form of syndecan-4 that carries a deletion of the cytoplasmic domain instead reduces bacterial uptake [197a]. Syndecan-4 thus seems to play a major role in mediating bacterial uptake, and the cytoplasmic domain appears to be critical for this process. The finding that PKC activity is necessary for bacterial uptake into HeLa cells [194] is consistent with this premise, since PKC binds direcUy to the Syndecan-4 cytoplasmic domain resulting in an increase in its kinase activity [197b]. The roles of other syndecans (e.g., Syndecan-1) and the glypicans in mediating gonococcal interaction with epithelial cells still remain to be investigated. b. HSPG-Mediated Uptake of Opa^Q-expressing Neisseria into other Cell Types. Owing to the ubiquitous expression of HSPGs on eukaryotic cells, Opamediated interactions with HSPGs are not limited to epithelial cells. Opa5o-expressing bacteria bind strongly to endothelial cells, but uptake is inefficient in the absence of additional factors. Recruitment of either VN or FN to the surface of bacteria triggers efficient bacterial internalization in an integrin-dependent manner [197c]. Such a process could contribute to neisserial entry into the bloodstream and/or extravascularization of the bacteria, leading to colonization of nonmucosal tissues during disseminated disease. Fibroblasts, being potential targets in the submucosal tissue, are also bound and invaded by Opa5o-expressing gonococci, albeit with lower efficiency [191]. Although Opa5o expression has little effect on gonococcal interactions with polymorphonuclear leukocytes, it does mediate the efficient phagocytosis of these bacteria by monocytes. Moreover, measurements of luminol-enhanced chemiluminescence also demonstrated that

12. NEISSERIA

581

the phagocytosis of Opa5()-expressing gonococci by monocytes was accompanied by a release of oxygen-reactive metabohtes [155].

3.

OPA-MEDIATED INTERACTIONS WITH THE CELLULAR

CEACAM

RECEPTORS

Recently, several groups have clearly shown that most neisserial Opa variants specifically bind to the CEACAM receptors that are differentially expressed on multiple tissues within the human host. CEACAM 1 (biliary glycoprotein, BGP), CEACAM6 (nonspecific crossreacting antigen, NCA), CEACAM3 (CEA gene family member 1, CGMl) and CEACAM5 (carcinoembryonic antigen, CEA) can all function as receptors for N. gonorrhoeae and N. meningitidis, while CEACAM8 (CGM6) is not recognized by any Opa protein tested to date. As shown in Table I using the well-characterized Opa variants of N. gonorrhoeae strain MS 11 as an example, individual gonococcal and meningococcal Opa proteins may display various patterns of reactivity with one or more of the CEACAM receptors [198-201]. Although most Opa proteins bind to either CEACAM or HSPG glycoproteins, a few appear to interact with both types of these cellular receptors [106]. The gonococcal Opa variants that do bind to both receptors, however, are only able to mediate cellular invasion via CEACAM [176, 189]. CEACAM proteins represent a subset of the CEA receptor family, which itself belongs to the immunoglobulin superfamily. Each receptor consists of a single, highly conserved amino-terminal immunoglobulin variable region- (Igy-) like domain, followed by a variable number of immunoglobulin Igc2-like constant domains exposed at the cell surface. The carboxy-terminal domains of CEACAM 1 and CEACAM3 contain transmembrane and cytoplasmic domains, while anchorage of CEACAM8, CEACAM6, and CEACAM5 occurs via a glycosylphosphatidylinositol (GPI) anchor attached to the protein's carboxyl terminus [202]. Although the primary role of CEACAM receptors in vivo is still unclear, they appear to mediate intercellular adhesion via both homotypic (CEACAM 1, CEACAM6, and CEACAM5) and heterotypic (CEACAM8-CEACAM6 and CEACAM6-CEACAM5) interactions [203, 204]. This function may or may not be related to their apparent influence on cell cycle control and cellular differentiation [205, 206]. CEACAM 1 and CEACAM6 have also been shown to present the sialyl Lewis^ (sLe^) blood group antigen to E-selectin, and CEACAM6 can stimulate the upregulation of CDl 8-integrins on Huvecs. Both of these processes should facilitate the adherence of PMNs to inflammatory cytokine-stimulated endothelial cells in vivo [207], perhaps targeting the phagocyte to loci of infection. Interestingly, carbohydrate structures expressed by CEACAM 1, CEACAM6, and CEACAM5 have also been shown to function as a cellular receptor for the type 1 fimbriae of E. coli and Salmonella strains. This interaction has been proposed to faciUtate colonic colonization by these species [208, 209].

582

SCOTT D. GRAY-OWEN ETAL.

More than 95% of Opa-expressing mucosal and disease isolates of N. gonorrhoeae and A^. meningitidis have been shown to bind CEACAMl [198], thus implying the importance of these interactions for neisserial infection. It is also significant that meningococcal strains expressing both capsule and sialylated lipopolysaccharide are still able to bind CEACAMl [198], since these structures had previously been thought likely to sterically hinder any potential Opa-mediated interactions with a cellular receptor. Although the individual CEACAM proteins are highly glycosylated, these sugar structures do not influence binding by the Opa proteins [210]. Due to their close relationship to immunoglobulin, a three-dimensional structure for the CEACAM protein backbone has been proposed [211]. Neisserial binding studies using transfected epithelial cells that express chimeras of CEACAM8 and CEACAM6 indicate that Opa proteins bind to protein sequences located on a surface composed of four (3-strands (CC'FG) in the amino-terminal domain of CEACAM [201a, 201b, 211a]. The differential specificity of Opa variants to some CEACAM receptors is determined by a divergent tripeptide sequence within this region [211a]. The CC'FG p-sheet has also been proposed to function as the ligand-binding site of other members of the immunoglobulin superfamily, including the closely related membrane-bound CD2 [212] and CD4 [213] receptor proteins. It is thus interesting to speculate that the natural ligand for CEACAM receptors may also bind here. a. CEACAM'Dependent Interactions with Epithelial and Endothelial Cells. Individual CEACAM receptors are differentially distributed on human tissues [202, 214, 215], suggesting that the distinct specificities of individual Opa variants may influence both tissue tropism and the cellular response to bacterial binding. Characterization of Opa binding specificities has been achieved using stably transfected cell lines that express each CEACAM receptor in isolation. In each case, adherence leads to the subsequent engulfment of bound bacteria, thus demonstrating that each family member is itself capable of facilitating cellular invasion [199-201]. The singular importance of Opa proteins in this process is indicated by the fact that E. coli strains expressing recombinant Opa proteins are also efficiently internalized [200, 201]. Several cell lines that naturally express CEACAM receptor proteins have also been identified, and some of these have been employed in infection assays to determine whether the Opa-expressing bacteria are also capable of invading into cells in a somewhat less artificial system. Consistent with what was seen with the transfected cell lines, CEACAM receptors expressed by human colonic (HT29) and lung (A549) epithelial cell lines can mediate meningococcal binding and engulfment [107], and Opa-expressing gonococci have been shown to be taken up by the CEACAM5-expressing LS174T colonic adenocarcinoma cells [200]. We have shown that bacteria expressing CEACAM-specific Opa proteins are capable of passing from the apical to the basolateral surface of polarized T84

12. NEISSERIA

583

epithelial cell line monolayers without disrupting its transepithelial barrier function [216]. Electron microscopy of infected monolayers clearly shows that this transmigration occurs via the transcytotic route, since gonococci were evident within a tightly adherent phagocytic vacuole in the cell cytoplasm. This process depends on one or more of the CEACAMl, CEACAM6, and CEACAM5 receptors that were seen to be expressed exclusively on the apical surface of the monolayers, since CEACAM-specific antisera can block both bacterial binding and cellular invasion [201, 216, 217]. This remarkable process suggests that the presence of CEACAM on epithelia of the cervix and uterus [215] may mediate an analogous penetration of the epithelia to allow gonococcal dissemination into submucosal tissues. Although primary endothelial cells (Huvecs) grown in culture typically express only very low amounts of CEACAMl, there is a substantial upregulation of its expression following stimulation of these cells with the proinflammatory cytokine tumor necrosis factor alpha (TNF-a) [201, 217]. The upregulated expression of CEACAMl results in a marked increase in binding by recombinant A^. gonorrhoeae strains that express CEACAMl-specific Opa proteins [201], and our preliminary results indicate that this adherence may lead to bacterial uptake by the Huvecs [217a]. b. CEACAM'Dependent Interactions with Phagocytic Cells. Clinical specimens from patients with gonorrhea typically display human polymorphonuclear neutrophils (PMNs) containing intracellular gonococci. In vitro, neisserial binding to PMNs results in opsonin-independent phagocytosis of these bacteria [218220]. CEACAMl, CEACAM8, CEACAM6, and CEACAM3 receptors are thought to be constitutively expressed at low levels on the surface of PMNs, and activation of these cells results in further release of large amounts of CEACAM proteins from primary and secondary granules [221, 222]. Consistent with this, neisserial binding to PMNs increases significandy following their stimulation either by adherence to glass or treatment with phorbol myristate acetate [223, 224]. Opa-mediated binding to CEACAM receptors on PMNs correlates with generation of an enhanced respiratory burst in comparison to that which is stimulated by nonopaque or piliated gonococci, however, one that is reduced when compared to that seen with commensal Neisseria spp. [219, 225, 226]. The presence of Opa-specific F(ab')2 antibody fragments prevents bacterial binding and the subsequent oxidative response, confirming a role for these adhesins in both of these events [219]. The fact that purified Opa protein can also abrogate this response suggests that the induction of an oxidative response requires multiple cellular CEACAM receptors to be ligated by the high density of Opa proteins expressed on the bacterial surface [227]. Whether, and in which way, Opa-directed phagocytosis via CEACAM receptors could provide a survival

584


advantage to the bacterial population is currently still a matter for intriguing speculation. c. CEACAM-Dependent Intracellular Signaling. The cytoplasmic domains of CEACAMl and CEACAM3 contain sequences that have homology to the immunoreceptor tyrosine-based inhibitory (ITIM) and activation (ITAM) motifs, respectively [228, 229]. The presence of these structures on other receptors has been shown to modulate the response of various immune cells. For example, the clustering of ITAM-containing T-cell receptor complexes results in Src kinasemediated activation of T cells, whereas the coligation of an ITIM-containing receptor abrogates this response [230, 231]. Similarly, the stimulation of B cells can be blocked by the inhibitory action of FcyRIIb receptors [232]. Consistent with this [233], CEACAMl has been shown to associate with the Src-family kinases Lyn and Hck [234], and the Opa-mediated binding of A^. gonorrhoeae to neutrophils does result in activation of Hck and Fgr in a process that is essential for bacterial phagocytosis [235]. The Src homology 2-containing tyrosine phosphatase 1 (SHP-1) also interacts with CEACAMl [229] and appears to be inactivated by receptor crosslinking [235a]. Subsequent to these events, the GTPase Racl, the p21-activated protein kinase (PAK), and Jun N-terminal kinase (JNK) have been shown to be stimulated. Bacterial uptake can be reduced either by the presence of Src kinase inhibitors or by the preincubation of cells with anti-sense oligonucleotides, which downregulate the expression of Rac 1, indicating that both activities are essential for this process [235]. This is consistent with the role of the small G-protein Racl in regulating cytoskeletal rearrangements [236], such as those that might be required for neisserial phagocytosis. The stress-activated protein kinase JNK has previously been implicated in the induction of activator protein 1 (AP-l)-regulated transcription [237], suggesting that neisserial binding to CEACAM receptors may trigger a long-term adaptive response. Figure 4B summarizes the signaling cascade mediated by the interaction of Opa and cellular CEACAM receptors in phagocytic cells. It is important to note that the JNK/AP-1 pathway may also be activated in epithelial cells by a mechanism that is independent of Opa expression by the bacteria. This alternate cascade has been shown to involve the Rho family of GTPases and the cellular kinases PAK, MKKK and MKK4 [238] (Fig. 4C; see §VIII.A.2). Whether costimulation of Opa-dependent and Opa-independent signal cascades differentially influences the cellular response to neisserial binding has not yet been determined.

4.

OPA-MEDIATED BINDING TO INTRACELLULAR PYRUVATE KINASE

Although intracellular gonococci are generally considered to remain inside a phagosomal compartment, occasional reports indicate that they may have the

12. NEISSERIA

585

capacity to escape into the cytoplasm [239]. Recently, gonococcal Opa proteins were reported to bind human pyruvate kinase (PK) subtype M2 in vitro, and this cytoplasmic enzyme appears to associate with intracellular Opa-expressing gonococci [240]. PK catalyzes the irreversible conversion of phosphoenolpyruvate to pyruvate with the resulting generation of ATP. Interestingly, a A^. gonorrhoeae mutant that is unable to use pyruvate or lactate is unaffected in its uptake into epithelial cells, but does not survive intracellularly [240]. It is thus possible that intracellular gonococci bind PK to gain an ample source of pyruvate, one of only three carbon sources known to be used by N. gonorrhoeae. Whether pyruvate does actually play a role in intracellular survival of Neisseria spp. in vivo, however, is still unknown. C. Opc-Mediated Interactions Ope is an outer membrane adhesin of pathogenic Neisseriae that is similar in size to the Opa proteins, but is structurally and antigenically distinct from them [241, 242]. Phase-variable expression of the meningococcal Ope is mediated by slip-strand errors in a homopolymeric cytidine tract residing within the promoter of the ope gene. The resulting changes in spacing between the -10 and -35 promoter elements affects the strength of transcription, resulting in either strong, low, or no expression of Ope (Fig. 3B) [243]. Although Ope has previously been considered to be a meningococci-specific protein, a gonococcal homolog has now been identified [242]. In contrast to the meningococcal ope gene, DNA sequence analysis suggests that expression of the gonococcal homolog is not to be subject to phase variation, but rather to a conventional transcriptional regulation. The gonococcal Ope protein is, therefore, a possible candidate adhesin for the contact-inducible invasion of HeelB cells [244]. However, the functional characterization of Ope has so far been reported only for the meningococcal homolog. Opc-producing meningococci interact with the serum glycoprotein vitronectin, and appear to use this molecule to attach to avP3 integrins that are present on the apical pole of endothelial cells [245]. This molecular bridging may result in bacterial uptake into endothelial cells [92]. Ope expression also promotes meningococcal binding to and entry into certain epithelial cell lines (i.e., Chang conjunctiva cells) in the absence of additional factors [91, 246]. The host-cell receptor for this process has recendy been identified as being syndecan-like heparan sulfate proteoglycans (HSPGs) [106]. The apparently analogous functions of the HSPGA^N-binding Opa proteins (e.g., Opa5(); see §IV.B.2) and Ope seen in vitro makes the reason for neisserial maintenance of these two otherwise unrelated adhesins an intriguing question. D. Interactions Mediated by a Novel Multiple Adhesin Family Screening of known host cell surface components for their capacity to bind gonococci has lead to the description of several lacto- and ganglio-series

586


glycolipids as being putative adhesion receptors for A^. gonorrhoeae [247, 248]. For one of these putative receptors (asialo-Gmi-Gral(pl-3)GalNAc(Pl-4)Gal(Pl4)Glc(pl-l)Cer), a corresponding bacterial Hgand has been identified. The heterologous expression of a plasmid encoding a 36-kDa neisserial protein in E. coli was suggested to confer bacterial adhesion to the immobilized glycolipid [249]. More recent experiments suggest that the 36-kDa lipoprotein exhibits a different as-yet-undefined binding function, and that a separate genetically linked adhesin instead possesses the above-cited glycolipid specificity [249a]. These adhesins are both encoded by multiple genes in the neisserial genome, and as members of a multiple adhesin family have now been termed MafA and MafB, respectively. P36/MafA is clearly different from the sialic acid-specific 27-kDa adhesin Sia-1, which is expressed by the commensal species Neisseria flava and recognizes the structure NeuAc(a2-3)Gal(pl-4)Glc on erythrocytes [250]. Although the ubiquitous expression of such glycolipids could potentially contribute to neisserial interactions with many host tissues, the current lack of described function for MafA and Maffi makes their role in neisserial infection uncertain.

V. PorB Porins are a major protein constituent of the neisserial outer membrane. They resemble a hydrophilic pore that allows the passage of small nutrients and waste products across the outer membrane of these Gram-negative bacteria. In other bacteria, such pores are typically formed by a trimeric arrangement of the porin monomer [251, 252]. The three-dimensional structure of neisserial porin protein has not yet been resolved; however, various topological membrane models have been proposed. Each of these predicts 16 membrane-spanning p-strands that come together to form a p-barrel with eight loops exposed at the bacterial surface [253, 254]. A^. meningitidis possesses two different porins: ForA (also called Class 1) and ForB (of either Class 2 or Class 3 serotype). While A^. gonorrhoeae does not express its ForA homolog [255], ForB can be antigenically separated into two serotypes (F.IA and FIB), reminiscent of the mutually exclusive class 2 and class 3 ForB proteins in meningococci [256]. Several unusual features of the gonococcal porin indicate that ForB may have additional functions aside from acting as a simple pore in the outer membrane. The ForB porins of pathogenic Neisseria species have been reported to translocate from live intact bacteria into artificial membranes, a property not observed with porins of commensal Neisseria species [257]. Furthermore, electron microscopic examination of epithelial cells that have been infected with N. gonorrhoeae has revealed that the neisserial ForB is present in the eukaryotic cytoplasmic and phagosomal membranes [190]. It is currently unknown how these integral outer membrane proteins translocate from the bacterial envelope into artificial membranes or into the membrane of target cells. It is, however, possible that differences in membrane fluidity might be the driving force, resulting in

12. NEISSERIA

587

movement of porin from the higher energy state of the relatively rigid bacterial membrane to the lower energy state of the more fluid cellular membrane [258]. Alternatively, the PorB-containing blebs that are released from the outer membrane of pathogenic Neisseria may fuse with the target cell, or there may be receptor-mediated integration of porin into the cellular membrane. This latter model is similar to that which described for the Staphylococcus spp. a-toxin. After binding to its receptor, multimers of the soluble a-toxin integrate into the target cell membrane to form amphipathic (i-sheet structures that are reminiscent of the porins [259]. Once inserted into the eukaryotic membrane, the neisserial porin forms a channel that is regulated by the cell [260]. This was demonstrated by patch-clamp experiments that showed the porin channel to remain closed in intact living cells but open immediately if the membrane fragment (patch) with inserted porin is excised from the cells. This approach also revealed that PorB has channel characteristics that are very similar to those of a class of eukaryotic porins known as the voltage-dependent anion channels (VDACs). Like neisserial PorB, the opening and closing of these channels is dependent on the applied voltage, with higher voltages (i.e., in the range of 40 to 80 mV), resulting in channel closure. In addition, PorB and VDACs both possess a binding site for nucleotides, which, once bound to the porin, also downregulate pore size and dramatically influence voltage dependence. Even at very low voltages (-10 mV), nucleotides bind to porin and close the channel. Both of these features are likely important for regulation of porin channels by the host cell, since high nucleotide concentrations and a significant membrane potential are present in the eukaryotic cell [260]. In bacteria, these properties appear to be restricted to the PorB porins of pathogenic Neisseria species, since, with the exception of Neisseria lactamica, they were not found in commensal neisserial strains or in other bacterial pathogens [260].

A. Influence of PorB on Cellular Interactions 1.

EFFECTS OF PORB ON THE PHAGOCYTIC RESPONSE

During the infection process, N. gonorrhoeae encounter the bactericidal action of PMNs. Purified gonococcal porin selectively interferes with the signaling machinery of PMNs, inhibiting degranulation, downregulating both actin polymerization and opsonin receptor expression, and reducing the level of bacterial phagocytosis [261, 262]. Consistent with this, recombinant gonococci in which PorB has been replaced by either the porin of the commensal species N. lactamica or by a PorB mutant lacking its amino-terminal extracellular loop are phagocytosed more efficiently by polymorphonuclear neutrophils (PMNs) than are the wild-type controls [263]. Preincubation of PMNs with purified porin also inhibits subsequent phagocytosis of wild-type A^. gonorrhoeae [262]. However, urethral

588


exudates from gonorrhea patients contain PMNs associated with both intracellular and extracellular gonococci, indicating that the phagocytic response of neutrophils is not completely abolished by PorB in vivo. Consistent with this, the activity of NADPH-oxidase, an enzyme necessary for the oxidative killing response of phagocytes, is unchanged by the presence of porin [261]. PorB does, however, also influence the maturation of phagosomes within the cell. This was most clearly demonstrated by comparing the protein profile of latex bead-containing phagosomes in cells incubated in the presence or absence of porin, since the composition and amount of proteins present within the phagosome do differ [264]. Several of the proteins enriched in PorB-containing phagosomes were identified as being typical markers of the early endosome (e.g., the Rab 5 protein). Consistent with this, phagosomes containing N. gonorrhoeae do not contain cathepsin G, a marker of mature phagosomes, while this protein is present in phagosomes that harbor phagocytosed E. coli (C. R. Hauck and T. F Meyer, unpublished observations). This is consistent with the fact that degranulation, the release of bactericidal enzymes at the cell surface or into the phagosome during its maturation into a lysosome, is strongly reduced in treated cells (Fig. 5) [261]. Taken together, there is a clear line of evidence that suggests a function of porin in helping A^. gonorrhoeae to survive the encounter with a hostile phagocyte by interfering with phagocytosis and/or processing of ingested bacteria.

2.

EFFECT OF PORB ON NEISSERIAL INTERACTIONS WITH EPITHELIAAND ENDOTHELIA

Mutations in the gonococcal PorB also reduce the level of bacterial invasion into epithelial cells following Opa-mediated binding to the HSPG receptors expressed by epithelial and endothelial cells (see §IV.B.2). This is consistent with the previous observation that the expression of Opa5() in E. coli confers attachment to, but not uptake by, Chang epithelial cells [176, 265]. Thus, HSPG receptor-mediated invasion may need a factor in addition to the Opaô adhesin to trigger uptake. An interesting finding in this context was that the PorBj^ allele, one of the two major PorB isoforms found in N. gonorrhoeae, mediates invasion into Chang epithelial cells even in the absence of any Opa proteins. However, invasion in this case was strictly dependent on an absence of phosphate in the medium, which apparently bind to PorBj^ porin in a manner similar to nucleotides [266]. The reason for phosphate's inhibitory effect on PorBi^-mediated invasion is not yet clear, but it may close the pore, a function that was previously thought to be restricted to nucleotides [266]. If true, then an open, regulated channel may be the trigger for neisserial uptake into Chang cells. In fact, PorBi^ was recently shown to allow a Ca^"^ influx into target cells [267], an effect that could potentially enhance bacterial uptake by its influence on the activation state of protein kinase C (PKC) [197c].

12. NEISSERIA

589

Porin

-Ik

1 ^ ^ --- ™^

ATP T

Cellular

••

signals

\

y^

••

^ i , ^ i ^ ^

>-r^

±

•

151

\

t

HighCa^*

/

Azurophilic granules

>^: M Active ^ '^^ Active caspases^ coipain

Phogosome Apciptosis

F/g. 5 Model for the function of PorB in cell invasion, intracellular survival, and induction of apoptosis. A direct role of PorB for invasion of Neisseria into host cells has been shown under two experimental conditions: in the presence of phosphate, where invasion of epithelial cells requires expression of Opa5o protein that binds to the HSPG receptors, and under special in vitro conditions in the absence of phosphate, where gonococci invade epithelial cells without expression of any Opa protein. PorB translocates into the host cell cytoplasmic membrane, where it forms channels. PorB channels are closed by noncovalent binding of ATP immediately after translocation. PorB also interferes with phagosome maturation, probably by preventing fusion of the phagosome with azurophilic granules. Furthermore, by an unknown cellular signal (e.g., drop in the ATP level) an opening of the inserted PorB channels and an influx of Ca~^ into the host cell may occur. The increase in intracellular Ca~^, probably together with other signals, induces apoptotic death by active Ca^'^-dependent proteases like calpain and the central executioner proteases, the caspases.

B.

PorB Induction of Apoptosis

A^. gonorrhoeae induces apoptotic cell death of epithelial cells and phagocytes in vitro. Neisserial PorB is itself sufficient to induce this suicide program (Fig. 5) [267], which is present in all cells [268]. Cells treated with purified PorB respond with an immediate, transient uptake of extracellular Ca^^. Within 30 min, they start to form extensive membrane blebs, a hallmark of cells that have entered the apoptotic program. Then, the cells round up, detach from the monolayer, and eventually disintegrate into small particles known as apoptotic bodies. Interestingly, preincubation of porin with ATP closes the channel, thereby preventing both influx of Cd?^ and the subsequent apoptotic death of these cells [267]. This is consistent with the influx of Ca^"^ and the induction of apoptosis being linear processes. Normally, the process of apoptosis is controlled by a class of cysteinyl aspartate-specific proteinases (caspases). These enzymes are also activated in Neisseria- and porin-mediated apoptosis; however, the activity of the cysteine protease calpain is also elevated. Interestingly, the activity of calpain strictly

590


depends on Ca^"^, the level of which is significantly increased in porin-treated cells. Pretreatment of cells with inhibitors of caspases and calpain reduces porin-induced apoptotic cell death more efficiently than does inhibition of either caspases or calpain alone. This result suggests that porin induces cell death by at least two independent pathways. Dual activation of caspases and calpain is rarely seen during control of apoptotic cell death, making neisserial induction of this novel pathway a particularly intriguing topic of future study.

W. IgA 7 Protease There is a strict correlation between the virulence of neisserial species and their ability to synthesize and secrete a serine protease that was originally shown to display a high specificity for human IgAl antibodies. Because of this, the IgAl protease is presumed to have an important role in neisserial disease [269]. Numerous serologically distinct IgAl proteases exist in A^. gonorrhoeae and A^. meningitidis due to the horizontal exchange of genetic sequences [55, 270, 271]. In each case, the iga sequences encode an approximately 170-kDa precursor protein consisting of five distinct domains [272]. The amino-terminal signal peptide and carboxy-terminal IgAp domain are responsible for transporting the IgAl protease (IgAp) and its associated domains to the bacterial surface (Fig. 6). Ultimately, the soluble IgAp^^y preproteins are cleaved from the transporter domains, and are then cleaved in a stepwise fashion to generate mature IgAl protease, a-protein, and y-peptide products. Proper secretion of neisserial IgAl protease expressed in E, coli [273] implies that this "autotransporter" process is intrinsic in the precursor protein. The contribution of IgAl protease to the infection process is still unclear, since, for example, the levels of bacterial attachment, invasion, and damage seen during infection of human fallopian tube organ cultures are indistinguishable between an IgAl protease-deficient mutant and its parental strain [274]. Mucosal secretory immunity is primarily mediated by IgA antibodies. It is interesting that the IgAl subclass predominates in secretions at the primary surfaces colonized by the Neisseria (i.e., respiratory and urogenital mucosa), while IgA2, which is not cleaved by the neisserial protease, is present only in relatively small amounts. Cleavage of IgAl antibodies by the neisserial protease leads to separation of the F(ab) fragment, which is involved in antigen binding, from the Fc domain, which is involved in effector function. This should result in the pathogen becoming decorated by IgA F(ab) fragments, thus masking the immunodominant epitopes that could otherwise be targets for IgG-mediated complement activation. Although IgAl protease has a high cleavage specificity for IgAl antibodies, the elucidation of its cleavage consensus sequence prompted a search for other potential targets that may be encountered during infection. The human lyso-

12.

591

NEISSERIA

IgA proteose Igopra

.Slgnol peptide

Cytopiosm

Fig. 6 IgAl protease secretion. Following the N-terminal signal peptide-directed transport of the IgA 1 protease polyprotein (Iga) through the cytoplasmic membrane, the signal peptide is cleaved off the polyprotein. The C-terminal Igap domain then integrates into the outer membrane, forming an amphipathic P-barrel with a putative internal pore through which the remainder of the protein can translocate to the bacterial cell surface. This translocation step may also be necessary for correct folding of the IgAl protease into an enzymatically active conformation. Autoproteolytic cleavage at sites a, b, and c then leads to generation of mature IgAl protease, a-protein, and the small y-peptide. The functional significance of a-protein and y-peptide are still uncertain.

some/late endosome associated membrane protein 1 (h-lamp-1) contains an IgAl protease cleavage consensus sequence, and purified h-lamp-1 is cleaved in an in vitro reaction under conditions that are similar to those found in phagosomal compartments [275, 277]. h-lamp-1 is a major component of the lysosomal membrane, and is thought to protect the lysosomal membrane against the degradative enzymes present within the lysosome. This function is likely facilitated by the protein's high level of glycosylation, which should form a carbohydrate coat lining the luminal face of the lysosome [278, 279]. Consistent with the in vitro cleavage results, one report suggested a decrease of the intracellular levels of h-lamp-1 during neisserial infection of A431 endocervical epithehal cells, and the survival and growth of IgAl protease-deficient gonococcal mutants associated with these cells has been reported to be significantly reduced as compared to that of the isogenic parental strain [277]. A parallel study did not, however, show a similar effect of IgAl protease mutations on intracellular survival in other human epithelial cell types [275], suggesting that such a phenomenon may be cell-type dependent. This conclusion is consistent with the finding that h-lamp-1 produced

592


by phagocytic cells is not cleaved by the neisserial protease, likely due to the higher glycosylation of the phagocyte-expressed form of this protein [275]. Whether the potential effect of IgAl protease on phagosomal integrity might play a role in the suggested phagosomal escape of gonococci [239, 240] is an intriguing hypothesis. The release of the IgAl protease precursor into the cytosol of an infected cell would probably lead to targeting of IgAl protease and its associated a-protein to the nucleus owing to a functional nuclear location signal in the precursor sequence [276].

VIL Iron Acquisition in Vivo Iron is essential for life, and the sequestration of iron from invading microbes is an important mechanism of a host's defence against the establishment and propagation of bacterial infection. More than 99.9% of iron is held intracellularly in the storage protein ferritin or as a component of heme compounds. The rest is carried by the host's extracellular iron transport and sequestration proteins transferrin and lactoferrin, which are present in serum and mucosal secretions, respectively. In order to circumvent this normally effective means of nutritional immunity, the pathogenic Neisseria are capable of obtaining iron from transferrin and lactoferrin. Iron removal from transferrin and lactoferrin is mediated by two transferrin (TbpA and TbpB) or lactoferrin (LbpA and LbpB) binding proteins expressed at the bacterial surface (Fig. 7) [280]. TbpB and LbpB likely function in the initial binding of their respective host glycoproteins, while the integral outer membrane proteins TbpA and LbpA strengthen this interaction and function as gated pores through which iron can pass into the periplasmic space. This is an energy-dependent process that is thought to be powered by interactions between TbpA and the cytoplasmic membrane-bound TonB protein complex [281]. Once in the periplasm, the free iron is bound by a soluble ferric binding protein (FbpA), which then shutdes the iron to an ATP-driven cytoplasmic membrane permease complex for its delivery into the bacterial cytoplasm [282]. The importance of this process in vivo has been clearly demonstrated using experimental infection of human male volunteers, since Tbp mutants are greatly reduced in their ability to colonize the urogenital tract and are completely unable to cause urethritis [283]. Neisseria are also capable of acquiring iron from free heme, hemoglobin, or hemoglobin bound to its serum carrier protein haptoglobin [284]. Each of these molecules is bound by a different receptor [285], and each process likely leads to bacterial internalization of the intact haem molecule, which is comprised of a protoporphyrin ring complexed with iron. Uptake from hemoglobin or hemoglobin-haptoglobin complexes also requires TonB function, while the binding and uptake of free haem does not [286]. These processes have the obvious benefit that

o

ft

a.

£

• J -' I

•y-
^DQ ^ . t i -^ y Q .^

O

•^

i-i

c/5

OJ C l . >

P

^

>.^

.2 E

c >
.,

t>£)

eg '-S 8 :e b ^ g cd ^ 5 # H °^ i Ob o Cu O^ M-i o == ^

J^

"^ < ' «^ • •

S c ° i2 :r -p -o ,

X,

A

ON

»-

(U

CL Q C^ -^ ^

>>

;/5

D

C

k! S I H .S ^ I ;

13.

BORDETELLA

627

cies. A possible explanation for this observation is the occurrence of horizontal transfer from B. pertussis to a coinfecting human B. parapertussis strain that subsequently emerged as a highly successful clone. Although insertion sequences can be acquired by horizontal or vertical transmission, the pattern of distribution of IS elements within clusters of related multilocus genotypes is consistent with a predominantly vertical descent. In general, the distribution of ISIOOl, IS1002, and IS481 supports assignment of genetic relationships based on MEE. Several hypothesis can be advanced based on the data summarized in Figure 1 [45]. Although B. pertussis and B. parapertussis l^^^ both infect humans, they do not appear to have a recent common ancestor and may instead have evolved from different lineages at different times. It therefore appears that there have been two independent host-range adaptations to humans, the earliest by B. pertussis (cluster D) and the most recent by B. parapertussis l^^^ (ET 28 in Fig. 1). In both cases, B. bronchiseptica is likely to be the evolutionary progenitor, suggesting that B. pertussis and B. parapertussis û strains may be considered as host-adapted lineages of B. bronchiseptica. Sequence comparisons of genes encoding fimbrial subunits [48], adenylate cyclase toxin [49], pertactin [50], pertussis toxin [51], and BvgAS [52] confirm the very close relationship between B. parapertussis û and B. bronchiseptica and the more distant relationship between these subspecies and B. pertussis. The lack of diversity among B. parapertussis hu strains is also consistent with the hypothesis that human adaptation is a very recent event. Since human and ovine B. parapertussis isolates comprise genetically distinct populations [42, 45] (Fig. 1), it is unlikely that sheep provide a reservoir from which transmission to humans commonly occurs. Although speculative in nature, the framework advanced by van der Zee et al. [45] provides a valuable tool for investigating mechanisms of pathogenesis and host-range adaptation. For example, the close relationship between B. parapertussis hu strains and the B. bronchiseptica isolates in cluster J could potentially be exploited to identify genetic correlates of human adaptation. The complete genomic sequences of B. pertussis, B. parapertussis hu, and B. bronchiseptica, which are presently being determined, will undoubtedly facilitate this comparative analysis. There are several examples of differential expression of surface and secreted factors with potential roles in the infectious cycles of Bordetella subspecies. It is now possible to ''hang" various phenotypes on the evolutionary tree and determine if the presence or absence of a particular factor correlates with the proposed phylogenetic lineage and/or parameters of infection. For example, it is likely that the progenitor in Figure 1 expressed both the pertussis toxin operon (PT"^), which is activated by the BvgAS control system, as well as genes for motility (Mof^), which are repressed by BvgAS. Since B. bronchiseptica and B. parapertussis strains are uniformly PT~ [51], the ability to express the pertussis toxin operon must have been lost twice, and independently, early in the formation of clusters B and C. In contrast, the Mot"^ phenotype is observed exclusively and uniformly in B. bronchiseptica isolates [3, 53]. Motility expression might therefore have

628

PEGGY A. COTTER AND JEFF F. MILLER

been lost in three independent events, associated with the evolution ofB. pertussis (cluster D), B. parapertussisQ^, (cluster I) and B. parapertussis^^ (ET 28). This predicts that the specific lesion(s) responsible for the lack of motility gene expression should be conserved within these lineages, but divergent between them. It is also interesting to consider that loss of motility correlates with restricted adaptation to specific mammalian hosts. Finally, preliminary data suggest that the type III secretion system is functional in B. bronchiseptica and B. parapertussisQ^, but not in B. parapertussis \^^ or B. pertussis [2]. The lack of type III secretion, therefore, appears to correlate with human adaptation as well as with acute disease versus chronic infection. The simplest alternative hypothesis for differential expression of specific factors is that pertussis toxin genes, motility genes, or type III secretion loci were acquired at various points in evolution by horizontal gene transfer. This is inconsistent with the observation that these genes are present, but not necessarily expressed, in all of the subspecies under consideration. The challenge in understanding the evolution of Bordetella subspecies, as well as other bacterial pathogens is to make the transition from correlation to causation.

IV. Bordetella Virulence Factors Bordetellae express a large number of surface-exposed and secreted factors with postulated roles in pathogenesis (Fig. 2). Many of these were identified and characterized biochemically before genetic and molecular biological tools were available. However, even recent advances in genetic and molecular methodologies have in many cases not been sufficient to overcome problems stemming from the lack of a natural-host animal model for B. pertussis. Thus, while extensive in vitro characterization has greatly increased our appreciation of what many identified factors can do, we still know relatively litde about what they actually do during the course of respiratory infection.

A.

LPS

The lipopolysaccharides (LPSs) contained in the outer membranes of Bordetella species, like endotoxins from other Gram-negative bacteria, are pyrogenic, mitogenic, toxic, and can activate and induce tumor necrosis factor production in macrophages [54-56]. Bordetella LPS molecules differ in chemical structure from the well-known smooth-type LPS expressed by Enterobacteriaceae. B. pertussis LPS lacks a repetitive O-antigenic structure and is therefore more similar to

^

^

P i p ^ O ) CT3 ra

'o Q5

I o

:S^

^^^^^^ ^

0^ ;=^•^ n

en

I- >

^ c o

.CQ

0)

a

iQ

^ c o

tf) Zi

> o .^ 55 t: di Q.

2nj a

QD •D C CO CT3

a

Q) .V5

•C

5K

C

o QQ> CO

1 -?

O 5 o^ -Q ^ ^ ^

E 0 to

>^ C

g

"-^

CN 1

^

CO

— "o = O

Q) c^

CT3 O

^

o C/5

o ^ o > , ^

>J_ n

CD J5 t r -qif

•

Z

X

o Q LU 1-

1 c x o

o

o o

Q)

E

m o O

LU CO

z

o • z

CQ ^

^^ ^

£ S S S

:/:

^ i2^r

^

< .J^ ^

^

.T:

X)

Ci,

'5

'i;

3

c

C3

X

< ^ o

>, c o c (U

"O

cx

> 3^ >
_o ^ ^> "S >

00

c

a.

o ^ ~$ a. «2Q < 1 o > QQ < c

PQ

o 3

00

< S-

OH

tl}

D-

> ^ C y: 03 X CQ u. X CQ :2 ^ X H JB T3 , ^ CQ 3 t; .*« ^ DQ XQJ £O H k, ^

630


rough-type LPS. It resolves as two distinct bands (A and B) on silver-stained sodium dodecyl sulfate-polyacrylamide gels [57]. The faster migrating moiety, band B, consists of a Hpid A molecule linked via a single ketodeoxyoctulosonic acid (KDO) residue to a branched oligosaccharide core structure containing heptose, glucose, glucuronic acid, glucosamine, and galactosaminuronic acid (GalNAcA) [58-61]. The charged sugars—GalNAcA, glucuronic acid, and glucosamine—are not commonly found as core constituents in other LPS molecules. The slower-migrating moiety (band A) consists of band B plus a trisaccharide consisting of A^-acetyl-A^-methyl-fucosamine (FucNAcMe), 2,3dideoxy-2,3-di-A^-acetylmannosaminuronic acid (2,3-diNAcManA), and N-acetylglucosamine (GlcNAc) [58-61]. B. bwnchiseptica LPS is composed of band A and band B plus an 0-antigen structure consisting of a polymer of the single sugar 2,3-dideoxy-di-A^-acetyl-galactosaminuronic acid [62]. Human isolates of B. parapertussis contain LPS that lacks band A, has a truncated band B, and contains an 0-antigen that appears to consist of the same sugar polymer as in the B. bronchiseptica 0-antigen [62]. LPS from ovine B. parapertussis isolates appear to lack 0-antigen and contain band A- and B-like molecules that are distinct from those of the other Bordetella species [63]. Although a distinct role for LPS in Bordetella pathogenesis has not yet been demonstrated, monoclonal antibodies against band A can adoptively transfer immunity to B. pertussis in an infant mouse lung infection model [64]. The importance of LPS in pathogenesis is further suggested by the observation that changes in LPS structure in B. bronchiseptica are controlled by the BvgAS virulence regulatory system [63]. The LPS structures of ^. parapertussis^^ and B. parapertussis hu also vary in response to the same environmental signals to which BvgAS responds, suggesting they are controlled by BvgAS as well [63]. Allen and Maskell have recently identified, cloned, and sequenced genetic loci required for LPS biosynthesis in B. pertussis, B. parapertussis hu, and B. bronchiseptica, and have constructed strains with various mutant LPS phenotypes [65-67]. Compared with their wild-type parental strains, B. pertussis, B. parapertussis hu, and B. bronchiseptica strains that synthesize only band-B LPS show decreased colonization in a mouse model of respiratory infection [68]. For B. bronchiseptica and B. parapertussis hu, this difference may be attributable to differences in sensitivity to antibody-dependent serum killing [68]. Further characterization of these and other mutants with defined mutations affecting LPS structure will greatly facilitate deciphering the precise role(s) of LPS in Bordetella pathogenesis. B.

TCT

Of the various toxins and virulence-associated factors synthesized by Bordetellae, only tracheal cytotoxin (TCT) has been shown to reproduce the specific epithelial cytopathology characteristic of pertussis. TCT corresponds to a disaccharide-

13.

BORDETELIA

631

tetrapeptide monomer of peptidoglycan. Its structure is A^-acetylglucosaminyll,6-anhydro-A^-acetylmuramyl-(L)-alanyl-Y-(D)-glutamyl-m^5odiaminopimelyl-(D )-alanine [69]. Although this peptidoglycan fragment is produced by all Gramnegative bacteria as they break down and rebuild their cell wall during growth, only Bordetella spp. [70] and Neisseria gonorrhoeae [71] have been shown to release it into the environment. Other bacteria, such as E. coli, recycle this peptidoglycan fragment by transporting it back into the cytoplasm via an integral cytoplasmic membrane protein called AmpG [72]. It appears that Bordetella do not express a functional AmpG, although sequences resembling ampG can be identified in the recendy released B. pertussis genome database (R. Lyon and W. E. Goldman, personal communication) [73]. Expression of £". coli ampG in B. pertussis results in a substantial decrease in the amount of TCT produced (R. Lyon and W. E. Goldman, personal communication). The activities of TCT have been studied in vitro using hamster tracheal organ culture and cultured hamster tracheal epithelial (HTE) cells. TCT causes mitochondrial bloating, disruption of tight junctions, and extrusion of ciliated cells, with little or no damage to nonciliated cells, in hamster tracheal ring cultures and a dose-dependent inhibition of DNA synthesis in HTE cells [74, 75]. TCT has also been shown to cause loss of ciliated cells, cell blebbing and mitochondrial damage in human nasal epithelial biopsies [76]. There is strong evidence that this cytopathology is due to a TCT-dependent increase in nitric oxide (NO). Inducible nitric oxide synthase (iNOS) expression is positively controlled by interleukin-la (IL-la), and TCT has been shown to trigger IL-1 a production in HTE cells [77]. Both TCT and IL-la result in increased NO production when added to HTE cells [78]. It is hypothesized that TCT stimulates IL-la production in noncihated mucus-secreting cells, which activates expression of iNOS, resulting in the formation of large amounts of NO. This NO then diffuses to neighboring ciliated cells, which are much more susceptible to its damaging effects [79]. The ability to construct TCT-deficient mutants by expressing a heterologous ampG gene in Bordetella will allow this hypothesis to be tested using in vivo models. C. Pertactin and Other Auto-Exporters Bordetellae express a number of related surface-associated proteins that appear to direct their own export to the outer membrane, where they undergo autoproteolytic processing of their C termini in a manner similar to the IgA proteases of Neisseria gonorrhoeae [80] and Haemophilus influenzae [81] and the elastase of Pseudomonas aeruginosa [82]. The first member of this family to be identified in Bordetella, and the best characterized, is pertactin. Mature pertactin is a 68-kDa protein in B. bronchiseptica, the subspecies in which it was first discovered [83], a 69-kDa protein in B. pertussis [84], and a 70-kDa protein in B. parapertussis i^^^ [85]. All three pertactin proteins contain an Arg-Gly-Asp (RGD) tripeptide motif as well as several proline-rich regions and leucine-rich repeats, motifs commonly

632


present in molecules that form protein-protein interactions involved in eukaryotic cell binding [86]. The B. pertussis, B. bronchiseptica, and B. parapertussis pertactins differ primarily in the number of proline-rich regions they contain [50]. The X-ray crystal structure of B. pertussis pertactin was recently determined. It consists of a 16-stranded parallel p-helix with a V-shaped cross-section and is the largest (i-helix known to date [87]. Other Bordetella proteins with predicted autoexport ability include TcfA [88], BrkA [89], and Vag8 [90]. All of these proteins show significant amino-acid sequence similarity in their C termini and contain one or more RGD tripeptide motifs. Based on predicted amino-acid sequence similarity with all of these proteins, the B. pertussis genome appears to encode at least three additional members of this autoexporter family. It is hypothesized that the C termini of the autoexporting precursor proteins form a pore in the outer membrane through which the N-terminal portions are threaded and then cleaved, but which remain cell-associated by an uncharacterized mechanism. In support of this model, Charles et al have shown that deletion of the y region of prn^^ prevents surface exposure of the molecule [91]. In the case of Vag8, however, cleavage may not occur since the predicted size of the entire protein encoded by vagS corresponds to the size seen by SDS-PAGE [90]. The presence of RGD and other sequence motifs suggests that pertactin and the other autoexporters may function as adhesins. In vitro studies demonstrated that purified pertactin could promote binding of CHO cells to tissue culture wells and that expression of prn in Salmonella or E. coli could increase the adherence and/or invasiveness of these bacteria to various mammalian cell lines [92]. In contrast, a Pm~ strain of B. pertussis did not differ from its wild-type parent in its ability to adhere to or invade HEp2 cells in vitro or to colonize the respiratory tracts of mice in vivo [93]. Similarly, we have constructed a B. bronchiseptica strain with an in-frame deletion mutation in prn and found it to be indistinguishable from wild-type B. bronchiseptica in its ability to establish a persistent respiratory tract infection in rats (our unpublished data). Thus, although pertactin appears to be a strong and potentially protective immunogen [83, 84, 94], its role in pathogenesis remains unknown. Potential adhesive functions for TcfA, BrkA, and Vag8 have not been investigated directly, although TcfA" B. pertussis show a decreased ability to colonize the murine trachea compared to wild-type B. pertussis [88]. BrkA has been proposed to play a role in serum resistance [89]. D.

Fimbriae

Like most Gram-negative pathogenic bacteria, Bordetellae express fimbriae. The major fimbrial subunits that form the two predominant Bordetella fimbrial serotypes—Fim2 and Fim3—are encoded by unlinked chromosomal loci fim2 and fimS, respectively [95, 96]. A third unlinked locus,//mX, is expressed only at very low levels, if at all [97]. In addition to positive regulation by BvgAS, the//m genes are subject to fimbrial phase variation by a mechanism that appears to

13.

BORDETELLA

633

involve slipped-strand mispairing within a stretch of ctyosine residues located between the -10 and -35 elements of ihefim2,fim3, and fimX promoters [98]. Since transcription of the individual genes is affected independently by this mechanism, bacteria may express Fim2, Fim3, and FimX, or any combination at any given time. In all cases, the minor fimbrial subunit that most likely forms the tip adhesin appears to be FimD [99]. The/zmD gene is located within the fimbrial biogenesis operon adjacent to fimB and fimC [100, 101], Interestingly, this operon is positioned between///(^^ and fhaC, genes required for synthesis and processing of another putative adhesin, filamentous hemagglutinin (FHA). The proposed functions of FimB and FimC as chaperone and usher, respectively, are based on the similarity of their predicted amino-acid sequences with those of the E. coli PapD and PapC proteins [100, 101]. Mutation of any one of the genes in the fimBCD locus results in a complete lack of fimbrial structures on the bacterial cell surface, suggesting/FmBCD is the only functional fimbrial biogenesis locus on the Bordetella chromosome [102]. In B. pertussis, a truncated major fimbrial subunit gene,/?mA, is located at the 5' end of the fimBCD gene cluster [101]. It was recently shown that in B. bronchiseptica and B. parapertussis, fimA is intact and capable of encoding a fourth fimbrial subunit type, FimA [103]. The putative promoter region of fimA does not contain a "C stretch" and therefore probably does not undergo phase variation. As a critical early step in bacterial pathogenesis, attachment to host epithelium is often mediated by fimbriae. Although it is likely that Bordetella fimbriae function as adhesins in establishment of respiratory tract colonization, definitive evidence that they in fact perform this function has been difficult to obtain for several reasons. First, the ability to construct strains devoid of fimbriae, but unaltered with respect to other putative adhesins, has been hampered by the unlinked locations of multiple major fimbrial subunits encoding genes as well as the complex genomic organization and transcriptional and translational couphng of the fimbrial biogenesis operon with the flia operon. Second, the presence of several other putative adhesins with potentially redundant functions has, in many cases, obscured the ability to detect clear phenotypes for Fim~ mutants. Finally, since the interactions between bacterial adhesins and host receptor molecules are expected to be highly specific, the use of heterologous hosts for studies with B. pertussis could severely limit the ability to detect in vivo roles for putative adhesins. Nonetheless, a number of studies by van Furth, Mooi, and colleagues (e.g., [105, 106, 108]) suggest fimbriae may mediate binding of Bordetella to respiratory epithelium via the major fimbrial subunits and to monocytes via FimD. Sulfated sugars are ubiquitous in the mammalian respiratory tract and Geuijen et al. have shown that purified B. pertussis fimbria, with or without FimD, were able to bind to heparan sulfate, chondroitin sulfate, and dextran sulfate [104]. Heparin-binding domains within the Fim2 subunit were identified and found to be similar to those of the eukaryotic extracellular matrix protein, fibronectin [104]. FimD-mediated adherence to monocytes/macrophages was suggested by a

634

PEGGY A. COTTER AND JEFF R MILLER

series of in vitro studies by Hazenbos et al. The resulting model suggests FimD mediates binding of nonopsonized B. pertussis to the very late antigen 5 (VLA-5) on the surface of monocytes, which then causes activation of complement receptor type 3 (CR3), which enhances its ability to bind FHA [105, 106]. As discussed below, there is evidence that FHA-mediated binding to monocytes, via the leukocyte response integrin/integrin associated protein (LRI/IAP) complex, also stimulates CR3 binding activity [107]. In vivo studies have shown that Fim~ B. pertussis strains are defective in their ability to multiply in the nasopharynx and trachea of mice [99, 108]. Using a B. bronchiseptica strain devoid of fimbriae but unaltered in its expression of FHA and other putative adhesins, we have recently shown that fimbriae contribute to the efficiency of establishment of tracheal colonization and are absolutely required for persistence in the trachea using both rat and mouse models [108a]. Moreover, the serum antibody profiles of animals infected with Fim~ bacteria differ qualitatively and quantitatively from those of animals infected with wild-type B. bronchiseptica. Taken together, these results suggest that Fim-mediated interactions with epithelial cells and/or monocyte/macrophages may play important roles, not only in adherence, but also in the nature and magnitude of the host immune response to Bordetella infection.

E. FHA

FHA is an unusually large and highly immunogenic hairpin-shaped molecule that has been included as a primary component in acellular pertussis vaccines [109]. It is synthesized as a 367-kDa precursor, FhaB, which is modified at its N terminus [110] and cleaved at its C terminus [111] to form the mature 220-kDa FHA protein. Although efficiently secreted via a process requiring the outer membrane protein FhaC, a significant amount of FHA remains associated with the cell surface by an unknown mechanism [111]. In vitro studies using a variety of mammalian cell types suggest FHA possesses at least four distinct attachment activities, and four separate FHA-binding domains have been proposed. The Arg-Gly-Asp (RGD) triplet [112], situated in the middle of FHA and locaHzed to one end of the proposed hairpin structure [113], stimulates adherence to monocyte/macrophages and possibly other leukocytes via LRI/IAP and CR3 [107, 112,114]. The CR3-recognition domain in FHA has yet to be identified. FHA also possesses a carbohydrate-recognition domain (CRD), which mediates attachment to ciliated respiratory epithelial cells as well as to macrophages in vitro [114,115]. Finally, a lectin-like activity for heparin and other sulfated carbohydrates, which can mediate adherence to nonciliated epithelial cell lines, has been identified [116]. This heparin-binding site is distinct from the CRD and RGD sites and is required for FHA-mediated hemagglutination [116].

13.

BORDETELLA

635

Evidence for FHA-dependent phenotypes in vivo has been more difficult to obtain. Using a rabbit model, Saukkonen et al found fewer FHA mutants than wild-type B. pertussis in the lungs 24 hours after intratracheal inoculation [114]. Based on in v/rro-determined binding characteristics of the various mutants used in their study, they inferred that wild-type B. pertussis were adhering both to ciliated epithelial cells and macrophages, and competition experiments with lactose and anti-CR3 antibody suggested both CRD- and RGD-dependent binding was involved [114]. Using mouse models, however, others have found FHA mutants to be indistinguishable from wild-type B. pertussis in their ability to persist in the lungs, but defective for tracheal colonization [99, 108, 117]. Still others [118-122], also using mouse models, have observed no difference between FHA mutants and wild-type B. pertussis. The difficulty in achieving a complete and detailed understanding of the role of FHA in the anatomic localization of B. pertussis during infection probably reflects the absence of a natural animal host (other than humans), as well as the complexity of this molecule and its associated biological activities. We have recently explored the role of FHA in pathogenesis by constructing two types of FHA mutant derivatives oiB. bronchiseptica: one containing an in-frame deletion in jhaB, the FHA structural gene, and one in which FHA is expressed ectopically in the Bvg~ phase, in the absence of the array of Bvg"^ phase virulence factors with which it is normally expressed [123]. Comparison of these mutants with wild-type B. bronchiseptica showed that FHA is both necessary and sufficient to mediate adherence to rat lung epithelial cells in vitro. Using a rat model of respiratory infection, we showed that FHA is absolutely required, but not sufficient, for tracheal colonization in healthy, unanesthetized animals. FHA was not required for initial tracheal colonization in anesthetized animals, however, suggesting its role in establishment may be dedicated to overcoming the clearance activity of the mucociliary escalator. The use of anesthetized rats also revealed a role for FHA in persistence. B. pertussis has recently been shown to inhibit T-cell proliferation to exogenous antigens in vitro in an FHA-dependent manner [124]. Taken together, these data suggest FHA may also perform immunomodulatory functions in vivo.

F. Dermonecrotic Toxin Although initially misidentified as endotoxin, dermonecrotic toxin (Dnt) was one of the first B. pertussis virulence factors to be described [125]. This heatlabile toxin induces locaUzed necrotic lesions in mice and other laboratory animals when injected intradermally, and is lethal for mice at low doses when administered intravenously [125-128]. The Dnts ofB. pertussis, B. parapertussis, and B. bronchiseptica are much more highly related genetically and biologically to each other than to the Dnt of B. avium [128a]. The Dnts of B. pertussis and B.

636


bwnchiseptica are cytoplasmically located single polypeptide chains of about 140 kDa [129-132]. In vitro studies have shown that purified Dnt from B. bwnchiseptica induces dramatic morphological changes, stimulates DNA replication, and impairs differentiation and proliferation in osteoblastic clone MC3T3 cells [133, 134]. Recent evidence indicates that these effects are due to Dnt-mediated activation of the small GTP-binding protein rho [135], which results in tyrosine phosphorylation of focal adhesion kinase (pi25^''^) and paxillin [136]. pi25^"^ and paxillin are involved in embryonic development and cell locomotion [137], and their activation leads to profound alterations in the actin cytoskeleton and assembly of focal adhesions [138-143]. Lacerda et al also showed that Dnt stimulates DNA synthesis without activation of p42'""/'^ and p44'"^'^'^, providing evidence for a novel p2P^'^^-dependent signaling pathway that leads to entry into the S phase of the cell cycle in Swiss 3T3 cells [136]. If and how these effects of Dnt contribute to Bordetella pathogenesis is not known. Although B. bwnchiseptica strains with decreased dermonecrotic toxin activity have been associated with decreased turbinate atrophy in infected pigs [144, 145], transposon mutants of B. pertussis lacking dermonecrotic toxin are no less virulent than wild-type bacteria in mice [121]. G. Adenylate Cyclase All of the Bordetella species that infect mammals secrete CyaA, a bifunctional calmodulin-sensitive adenylate cyclase/hemolysin. CyaA is synthesized as a protoxin monomer of 1706 amino acids. Its adenylate cyclase catalytic activity is located within the N-terminal 400 amino acids [146, 147]. The 1300-aa C-terminal domain mediates delivery of the catalytic domain into the cytoplasm of eukaryotic cells and possesses low but detectable hemolytic activity for sheep red blood cells [147-149]. Amino-acid sequence similarity among the C-terminal domain of CyaA, the hemolysins of Escherichia coli (HlyA) and Actinobacillus pleuropneumoniae (HppA), and the leukotoxins of Pasteurella hemolytica (LktA) 2ind Actinobacillus actinomycetemcomitans (AaLtA), places CyaA within a family of calcium-dependent pore-forming cytotoxins known as RTX (repeats in toxin) toxins [150]. Each of these toxins contains a tandem array of a 9-aa repeat (L-X-G-G-X-G-(N/D)-D-X) that is thought to be involved in calcium binding [150]. Before the CyaA protoxin can intoxicate host cells, it must be activated by the product of the cyaC gene, which is located adjacent to, and transcribed divergendy from, the cyaABDE operon [151]. CyaC activates the CyaA protoxin by catalyzing the palmitoylation of an internal lysine residue (Lys-983) [152], The E. coli HlyA protoxin is also activated by fatty acyl group modification [153155]. CyaA can enter a variety of eukaryotic cell types [156]. Once inside, CyaA is activated by calmodulin [157] and catalyzes the production of supraphysiologic

13.

BORDETELLA

637

amounts of adenosine 3',5'-monophosphate (cAMP) from adenosine triphosphate (ATP) [158-161]. Purified CyaA inhibits chemiluminescence, chemotaxis, and superoxide anion generation by peripheral blood monocytes [ 162] and PMNs in vitro with little effect on phagocytic activity [163]. CyaA has also been shown to induce apoptosis in cultured murine macrophages [119]. In vivo studies have shown that, compared with wild-type B. pertussis, CyaA~ mutants are defective in their ability to cause lethal infections in infant mice [121, 164] and grow in the lungs of older mice [118, 164]. Moreover, significandy fewer inflammatory cells, particularly PMNs, are recruited to the lungs in response to murine respiratory infection with CyaA" mutants compared with wild-type B. pertussis [165]. Taken together, these results suggest CyaA functions primarily as an antiinflammatory factor during infection. By comparing wild-type and CyaA~ B. bronchiseptica strains in wild-type and immunodeficient mice, we have recendy confirmed this suggestion and shown that phagocytic cells are indeed a primary in vivo target of the Bordetella adenylate cyclase toxin [166].

H. Pertussis Toxin

Among the Bordetella species identified so far, only B. pertussis synthesizes and secretes the ADP-ribosylating toxin known as pertussis toxin (Ptx). Ptx is composed of six polypeptides, designated S1-S5, which are present in a ratio of 1:1:1:2:1. Each subunit is synthesized with an N-terminal signal sequence, suggesting that transport into the periplasmic space occurs via a general export pathway analogous to the sec system of E. coli. Secretion across the outer membrane requires a specialized transport apparatus composed of nine Pd (pertussis toxin Hberadon) proteins [167, 168]. Extensive similarity between the ptl locus and the Agrobacterium tumefaciens virB operon, which encodes a secretion system involved in exporting single-stranded "T-DNA," suggests these systems function by a common mechanism [ 169-171 ]. Both appear to be involved in transporting large protein complexes since T-DNA is exported as a proteincoated DNA complex [172]. Furthermore, there is evidence that only the fully assembled Ptx holotoxin is efficiendy secreted [173, 174]. The ptl genes are located direcdy 3' to, and within the same transcripdonal unit as, \htptxA-E genes that encode the Ptx subunits [168, 175]. While the chromosomes of 5. parapertussis and B. bronchiseptica also contain ptx-ptl loci that encode functional polypeptides, these genes are transcriptionally silent due to nucleodde differences in the promoter regions [4-7]. In both B. parapertussis and B. bronchiseptica, replacement of native ptx-ptl promoter sequences with those of B. pertussis results in secredon of biologically acdve Ptx [176]. The biological relevance of differential Ptx expression among Bordetellae is not known.

638


The Ptx subunits are held together by noncovalent interactions and arranged in an A-B architecture typical of many bacterial toxins. The A protomer, consisting of the enzymatically active S1 subunit, sits atop a ring, the B oligomer, formed by the remaining S2-S5 subunits [177-179]. The B oligomer binds to eukaryotic cell membranes and dramatically increases the efficiency with which the SI subunit gains entry into host cells [178]. Unlike diphtheria toxin, Ptx does not require an acidic environment for entry into eukaryotic cells [180]. It has therefore been suggested that Ptx may traverse the membrane directly without need for endocytosis. Within the host cell cytosol, binding of ATP to the B oligomer that has intercalated into the cytoplasmic membrane causes release of SI subunit, which becomes active upon reduction of its disulfide bond [181]. The biochemical properties and biological effects of Ptx have been extensively characterized in vitro. In its reduced form, the SI subunit catalyzes the transfer of ADP-ribose from NAD to the a subunit of guanine nucleotide-binding proteins (G proteins) in eukaryotic cells [178, 182, 183]. G proteins that Ptx ADP-ribosylates, and hence inactivates, include Gj, G^ (transducin), and GQ. When active, Gj inhibits adenylyl cyclase and activates K"^ channels, G^ activates cyclic GMP phosphodiesterase in specific photoreceptors, and GQ activates K+ channels, inactivates Ca^"^ channels, and activates phospholipase C-p [184]. Biological effects attributed to disruption of these signaling pathways include histamine sensitization, enhancement of insulin secretion in response to regulatory signals, and both suppressive and stimulatory immunologic effects [24, 185]. Using in vitro assays, Ptx has been shown to inhibit chemotaxis, oxidative responses, and lysosomal enzyme release in neutrophils and macrophages [182, 186-193]. Using mouse and rat models, Ptx has been shown to inhibit chemotaxis and migration of neutrophils, monocyte/macrophages, and lymphocytes [194-196]. Ptx has also been suggested to function as an adhesin involved in adherence of B. pertussis to human macrophages and ciliated respiratory epithelial cells [112, 197]. Ptx is commonly cited as the major virulence factor expressed by B. pertussis, and pertussis has been proposed to be a toxin-mediated disease, with Ptx being responsible for many, if not all, of the disease's typical symptoms [24]. However, despite a plethora of experimental evidence demonstrating what Ptx can do in vitro and in animal models, clear evidence for an in vivo role for Ptx in human disease is lacking. One approach has been to compare symptomatology in children infected with either B. pertussis or B. parapertussis hu since these organisms differ primarily in the absence of Ptx expression by B. parapertussis hy. Such studies have indicated that the only significant difference between the two is increased leukocytosis in B. pertussis-mftciQd children [23, 198]. These observations suggest Ptx may not play a decisive role in causing the paroxysmal coughing, whooping, and vomiting characteristic of pertussis. The exact role of Ptx in establishment of infection, disease, and/or transmission of pertussis remains a mystery.

13.

BORDETELLA

639

I. The Bordetella Type III Secretion System A functional type III secretion system has recently been discovered in B. bronchiseptica [2]. Type III, or contact-dependent, secretion systems have been identified in several pathogenic Gram-negative bacteria (for a review see [199]). A common theme is the use of these systems to deliver bacterial effector proteins directly into the cytoplasm of host cells, leading to induction of signal transduction cascades [200]. While it has not yet been determined if any of the proteins secreted by the B. bronchiseptica type III system enter host cells, B. bronchiseptica can induce cytotoxicity in several cultured cell lines [2, 201], dephosphorylation of specific host cell proteins [2], apoptosis in fibroblast and macrophage cell lines, and nuclear translocation of NFKB [201a]. All of these phenotypes require a functional type III system [2, 201a]. In vivo, the B. bronchiseptica type III system contributes to persistent colonization of the trachea in both rat and mouse models of respiratory infection [2, 201a]. Additionally, animals infected with a mutant defective in type III secretion are completely protected against superinfection by wild-type B. bronchiseptica (P. A. Cotter and J. F. Miller, unpublished data). Taken together, these data suggest the B. bronchiseptica type III secretion system may be involved in modulating the host immune response and hence may contribute to the typically chronic nature of B. bronchiseptica infections. Interestingly, while the chromosomes of all Bordetella subspecies shown in Figure 1 contain bsc {Bordetella secretion) loci, only B. parapertussis^y isolates, an atypical B. pertussis strain (18323), and B. bronchiseptica strains express these genes in vitro [2]. Although it is possible that, for B. parapertussis hu and most B. pertussis strains, the requirements for induction of the type III system in vitro are more stringent than for B. bronchiseptica. Western blot analysis using sera from children recovering from pertussis suggests the type III system of B. pertussis is not expressed in vivo either [201a]. Additional phylogenetic analyses will help clarify the relationship between expression of bsc loci, host range, and/or disease.

V. The Bordetella-Hosf Interaction Bordetella species interact with their mammalian hosts primarily, and perhaps exclusively, at respiratory surfaces. Several scanning-electron-micrographic studies have demonstrated the predilection of these organisms for the cilia of respiratory epithelia [202-205]. In the nasal cavity, requirements for colonization appear to be few; B. bronchiseptica strains multiply deficient in the expression of

640


FHA, Fim, Pm, and CyaA are capable of persisting in the nasal cavities of rats for at least 60 days (S. Mattoo, P. A. Cotter, and J. F. Miller, unpublished data). Establishment of infection in the trachea, however, requires that bacteria be able to resist or overcome the clearance action of the mucociliary escalator as well as the killing effects of defensins, complement, and other antimicrobial factors. FHA, presumably by functioning as an adhesin, appears to be essential for overcoming mucociliary clearance [123], and LPS may be important for resistance to complement [205a]. TCT, released by Bordetella growing among the cilia, is proposed to stimulate IL-la production in neighboring nonciliated mucus-secreting cells, inducing expression of iNOS. The large amount of NO that is produced then diffuses into the ciliated epithelial cells, causing mitochondrial bloating, membrane blebbing, and extrusion from the mucosal surface. Damage and loss of tracheal epithelial cells containing adherent bacteria probably contributes to respiratory disease symptoms and possibly also to transmission by the aerosol route. Damage to respiratory epithelia undoubtedly also results in release of inflammatory cytokines. Although the specificity and magnitude of cytokine release has not been measured, recruitment of inflammatory cells, predominantly neutrophils, into the lungs of mice within 3 days following intranasal inoculation with either B, pertussis or B. bronchiseptica has been demonstrated [34, 165, 166]. This inflammatory response is significantly decreased in animals infected with ptx or cyaA mutants [34, 165, 166]. Both Ptx and CyaA have been shown to inhibit the microbicidal activities of neutrophils and macrophages in vitro. By doing so in vivo, these toxins may serve as anti-defense mechanisms, allowing Bordetella to resist the killing action of phagocytic cells. Without CyaA and Ptx, Bordetella mutants are efficiently eliminated by fewer numbers of neutrophils and macrophages, and hence less inflammation occurs. The importance of CyaA in resisting constitutive host defense mechanisms was recently demonstrated using mice that lack the ability to mount an adaptive immune response. SCID/Beige mice, defective in the development of T and B cells and NK-cell activity, are dependent on constitutive nonadaptive defense mechanisms for protection against microbial pathogens. Wild-type B. bronchiseptica killed these immunodeficient mice within about 50 days, while CyaA-deficient mutants did not even cause noticeable disease in these animals [166]. SCID mice are also killed by B. bronchiseptica, but with a slightly longer time to death [166]. This result indicates a role for CyaA in overcoming the defense mechanisms that are retained in SCID and SCID/Beige mice, which includes neutrophils. Mice rendered neutropenic, either by administration of cyclophosphamide or by a homozygous null mutation of the gene encoding granulocyte colonystimulating factor (G-CSF), were killed by both wild-type and cyaA mutant B. bronchiseptica, demonstrating that in the absence of neutrophils cyaA is not

13.

BORDETELLA

641

required [166]. The conclusion drawn from these experiments is that neutrophils are in vivo targets for the action of CyaA. Experiments using immunodeficient mice also revealed the importance of adaptive immunity in controlling Bordetella infections. B. bronchiseptica is confined to the upper respiratory tract in immunocompetent hosts but causes a lethal systemic infection in SCID or SCID/Beige mice. For B. pertussis, the ability of mice to mount an adaptive immune response determines whether the infection persists indefinitely in the lungs or is cleared entirely [205a]. While these observations demonstrate a crucial role for adaptive immunity, they do not reveal the nature of the immune response that is involved. It has been assumed that B. pertussis, as a noninvasive respiratory pathogen, is controlled primarily by a humoral immune response. The generation of ^nii-Bordetella antibodies in response to Bordetella infection is not only well documented but is considered diagnostic for pertussis in the absence of positive nasopharyngeal swab cultures. Using mice that do not express the interferon-y receptor, interleukin-4, or immunoglobulin heavy chain genes, Mills et al have demonstrated an absolute requirement for B cells, or their products, in clearance of B. pertussis infection in mice [206] and we have demonstrated that serum containing dinii-Bordetella antibodies can rescue SCID/Beige mice from lethal infection by B. bronchiseptica [206a]. However, there is increasing evidence that cell-mediated immunity also plays a significant role in controlling Bordetella infections. CD4"^ T cell clones that proliferate when stimulated with Ptx or FHA have been identified in peripheral blood mononuclear cells from children and adults following recovery from pertussis [207-210]. Secretion of interferon-y by these cells suggests they are Th-1 T cells [211], which are crucial for activation of phagocytic cells such as macrophages. Spleens of mice recovering from respiratory infection with B. pertussis were also shown to contain T cells that proliferate in response to heat-killed B. pertussis and secrete cytokines indicative of a Th-1 response [212, 213]. These spleen cells, upon adoptive transfer, were capable of inducing clearance of ^. pertussis from the lungs of nu/nu mice that lack functional T cells [214]. It therefore appears that both humoral and cell-mediated immune responses are important for the control and/or clearance of Bordetella infections. The relative contributions of each, and the mechanisms involved, await further investigation. Perhaps even less is known about the relative contributions of humoral and cell-mediated immunity in protection against reinfection. Recovery from pertussis in humans corresponds to the development of long-lasting protection against subsequent disease [215]. It has been assumed that this immunity is antibody-mediated, and vaccination strategies have been focused on the induction of anii-Bordetella antibodies. More recent evidence suggests, however, that, while vaccination against B. pertussis in humans and B. bronchiseptica in lower animals confers protection against disease, it is less effective at preventing colonization [216218]. The ability to detect Bordetella-spec'ific CD4"^ T cell clones in humans

642

PEGGY A. COTTER AND JEFF R MILLER

recovering from pertussis suggests the possibility that, while antibody may suffice to neutralize the effects of secreted toxins, activation of phagocytic cells may be required for eliminating bacteria from the respiratory tract.

VL The BvgAS Sensory Transduction System The genes and operons encoding all of the protein virulence factors described above are positively regulated by the products of bvgAS (originally called vir). This locus was first identified by Weiss and Falkow as the site of a Tn5 insertion in B. pertussis, which had simultaneously lost the ability to synthesize Ptx, CyaA, and FHA [219]. The locus was proposed, and subsequently proven, to encode trans-aciing regulatory factors required for coordinate activation of virulence gene expression. Inactivation of bvgAS by mutation, or of its products (BvgAS) by the presence of modulating signals, results in loss of virulence gene expression. Modulating signals that inactivate BvgAS in the laboratory include millimolar concentrations of nicotinic acid or MgS04, or growth at low temperatures (

I

@

h-

H

H

h-

^

©h X

© :^

ka L^^—1

XCLOf-• J ^

1^—'

X Q . o | - •pi^

4^

CL

I To O

^

h-

CE y

C/D

OH

o

3 03

OJ

o

Oi

ID

x:

< x y: cd

>>

a.

Q x: X:)

3 (U

c

^ c2^

(U X

OH

JO

o xy :j

OJ

in

c

Q c3

a

(Nl

^ r-

3

bX)

CD

'^ w & c3 X

o

y3

a, o E _c c a

DQ T3

O

> >

o x:

x: cu o a,

_c o 3, (N

o D 3

7^

O

^

D

o o c x:

OH

x:

_CJ

O H

-o o c x: a

o c

V2

_>> 15 B M r2 1 o>^ CL

o o ^3 00

>s ^ > >

^ 3

cd C CD

o c CO c o (U t»JO c/} c o (U o 'S o o

_c 'c3

E o

yj

T3

o

X)

>

P .

CJ

c

X)

C3

c -a

3 O &X)

O

C C3

OS

^

o ,o x: a.

OH

o 3 x: O OH o t>X) x: a o a. y) x: C o o x: x:

;5

c

iH Id o o X < yj too y^ c Z > ao c/) «D CO c '5i) cd PQ CQ O , ( N O 2^ O x: o a.

c

C3

y '^ B

TD C cd

p

>

g

3C H

Cu D

o x: p. o x:

a

i sX UD

O

O

r-) yj

t/)

> C3

OQ

O

x: a. o a.

>

x: a. i2+ . . B OXj yj (N x : r-- O c3 OQ "c5 ZH x :

-a

C

>1 03

fn

k:

.fei

13.

BORDETELLA

645

respond to signal intensity rather than signal diversity so that, rather than functioning like a switch that is responsive to many different signals, BvgAS may function like a rheostat that is adjusted in response to variations in intensity of a limited number of signals. As discussed below, there is evidence that BvgAS controls expression of a spectrum of phenotypic phases in response to subde yet distinct quantitative differences in environmental cues [37]. There is also evidence that the different patterns of gene expression required to produce these various phenotypic phases occur in response to variations in BvgA~P levels. The ability of the BvgS receiver to mediate phosphorylation and dephosphorylation of the HPD as well as dephosphorylation of the transmitter suggests it plays a pivotal role in the phosphorelay and may be responsible for regulating the flow of phosphate to the HPD and hence to BvgA [226]. Further genetic and biochemical analyses will be required to determine the validity of this hypothesized mechanism for adjusting BvgA~P levels. Although, as mentioned above and below, signals to which BvgS responds in the laboratory have been identified and extensively characterized, the true signals that are sensed in nature are unknown. The large BvgS periplasmic domain is thought to be involved in signal recognition since mutations in this region alter or abrogate signal sensitivity [38, 240]. Two regions with similarity to E. coli solute-binding proteins involved in glutamine and histidine transport are present in the BvgS periplasmic domain; however, their relevance is currently unknown. Also unknown is the relevance of PAS/PAC domains that are present in the BvgS linker, a region where many mutations that render BvgS insensitive to modulating signals occur. PAS/PAC domains were originally associated with light reception, light regulation, and clock proteins [241]. A broad, comprehensive search has revealed a large family of PAS/PAC-containing signal-transducing proteins, many of which mediate responses to changes in concentration of oxygen, redox carriers and carbon sources [242]. Although it has not been determined if BvgAS responds to oxygen concentration or redox potential, BvgAS reciprocally controls the expression of cytochrome d [243] and c oxidases (M. Liu, P. A. Cotter, and J. F. Miller, unpublished data), which presumably facilitate respiration under low and high oxygen environments, respectively. In addition to contributing to our understanding of how the BvgAS signal transduction system functions mechanistically, mutational analyses ofbvgAS have also produced a number of valuable tools for deciphering the structure of the BvgAS regulon and for investigating the role of Bvg-mediated signal transduction in vivo. Most notable are mutations that alter, rather than abrogate, the signal transduction pathway. One class, the bvgS constitutive mutations, is exemplified by the bvgS-C3 allele (Fig. 3) [240]. These mutations result in single amino-acid substitutions in the BvgS linker that lock BvgS in its active form, rendering it insensitive to modulating conditions. Strains that are isogenic with wild-type Bordetella except for these single nucleotide differences constitutively express all known Bvg-activated virulence factors [30, 240]. A second class, exemplified by

646


the bvgS-ll mutation (Fig. 3), appears to decrease the overall activity of the system [37]. The bvgS-ll mutation results in a methionine-to-threonine substitution four amino acids away from the primary site of phosphorylation in the BvgS transmitter. A third class, which includes the bvgAW60 and bvgA1056 mutations, maps to the extreme 3' end of bvgA [244]. These mutations abrogate the ability of BvgA to activate transcription of a subset of Bvg-activated genes and were instrumental in deciphering how BvgAS differentially controls virulence gene expression.

VIL Phenotypic Modulation The phenomenon of phenotypic modulation was first recognized early in the twentieth century [245] and was rigorously characterized by Lacey in 1960 [246]. It was originally defined as the reversible loss of virulence-associated phenotypes by Bordetella species that occurs in response to certain growth conditions [246]. Using colony morphology, hemolysin production, and antigenicity as indicators, Lacey described distinct phenotypic phases and designated them X mode (virulent phase), C mode (avirulent phase), and I mode (intermediate phase). High temperature (37°C), and certain ions such as sodium, potassium, halides, formate, and nitrate were shown to favor the virulent phase, while low temperature (25°C) and ions such as S04~^, and mono- and dicarboxylic acids favor avirulent phase growth. Further investigation showed that chlorate anions and nicotinic acid derivatives also resulted in downregulation of virulence factors [247]. It is now known that these changes are controlled by BvgAS. Lacey's X mode apparently corresponds to the Bvg"^ phase, C mode to the Bvg" phase, and I mode to the more recently characterized Bvg-intermediate (BvgO phase. A. The Bvg Phase Although Lacey's work provided evidence for C mode (Bvg~ phase)-specific antigens, BvgAS was traditionally considered to function merely as an ON/OFF switch for virulence gene expression. It was the discovery of Bvg-repressed genes, vrgs in B, pertussis [248], and genes required for motility in B. bronchiseptica [53] that proved that BvgAS also controls phenotypes expressed exclusively in the Bvg" phase. These discoveries necessitated a redefinition of phenotypic modulation to include the induction of Bvg~ phase factors and demonstrated that BvgAS actually functions to mediate a transition between at least two distinct phases, each with unique characteristics and unique patterns of gene expression (Fig. 4).

13.

BORDETELLA

647

In B. bronchiseptica, the Bvg~ phase is distinguished by the prominent phenotype of motihty. Two motihty loci, flaA (the flagelUn structural gene) and frlAB (the motility master regulatory locus, similar io flhDC of E. coli), have been characterized [3, 53] (Fig. 2). Both are expressed only in the Bvg~ phase [3, 53]. Based on complementation studies, and by analogy with the E. coli motility regulon, iht frlAB gene products are proposed to function together as a transcriptional activator positioned at the top of a regulatory cascade required for synthesis and assembly of flagella as well as proteins required for chemotaxis [3]. Substitution of the frlAB promoter with the Bvg-activated/Txfl^ promoter, so that frlAB transcription is activated by BvgAS, results in ectopic synthesis of flagella and motility in the Bvg"*" phase [31]. This result confirmed that activation offrlAB is sufficient to induce the entire motility regulatory cascade and demonstrated that repression offrlAB expression is the point at which BvgAS controls motility./r/A5 represents the first intermediate regulatory locus in the BvgAS regulon to be identified. Although B. pertussis, B. parapertussis hu, and B. parapertussis^y strains are nonmotile, their chromosomes contain sequences that hybridize io flaA [53] (U. Heininger, P. A. Cotter, and J. F. Miller, unpublished data). B. pertussis also contains anfrlAB locus that is functional when expressed in B. hronchiseptica (B. J. Akerley and J. F. Miller, unpublished data); however, its role in B. pertussis is unknown. Urease activity has been used as a diagnostic feature to distinguish B. hronchiseptica and B. parapertussis (Urease^) from B. pertussis (Urease") for many years. It was shown to be a Bvg-repressed phenotype in B. hronchiseptica [249]. Interestingly, urease expression is also repressed by BvgAS in B. parapertussis hu, while B. parapertussis^^ strains appear to have lost BvgAS control of urease (Fig. 1) (U. Henninger, unpublished data). The nucleotide sequence of the urease gene cluster in B. hronchiseptica was recendy determined revealing putative structural and accessory genes (ureA-G, 7, /), as well as a postulated regulatory locus (hhuR) with potential to encode a LysR transcriptional activator homolog [250]. B. pertussis and B. hronchiseptica produce an iron-chelating siderophore called alcaligin [251], and it is likely that the siderophore produced by B. parapertussis strains is also alcaligin [252]. Genes required for alcaligin biosynthesis (alcABC) and a potentially Fur-responsive AraC-type transcriptional regulator (alcR) have been identified in B. pertussis and B. hronchiseptica [252, 253]. Alcaligin biosynthesis is negatively regulated by BvgAS in some B. hronchiseptica strains [254] but appears not to be regulated by BvgAS in B. pertussis [255]. BvgAS regulation of alcaligin production in B. parapertussis strains has not been examined. Using transposon mutagenesis, we have recently identified additional Bvg-repressed loci in B. hronchiseptica (M. Liu, P. A. Cotter, and J. F. Miller, unpublished data). These genes include ccoC, which encodes a cytochrome c oxidase subunit 3 homolog, and prpE, a propionyl-coA synthetase homolog. The discovery that genes involved in electron transport-mediated oxidative phosphorylation and carbon utilization are BvgAS regulated indicates that, in addition to controlling the expression of virulence genes and other "accessory" factors, BvgAS


648

starvation survival (B.b.)

respiratory tract colonization

aerosol transmission?

' 4 Bvg- h

• ( Bvgi } . .

^4-*- t

• n ^ ' ,

-•-•

u^.

Bvg- )

B. p.

(Bvg-

T ^ ^ ^ B . b .

^ 1 !»l !»^'

"late" Bvg-activated genes (e.g. cyaA)

100

"early" Bvg-activated genes (e.g. fhaB, bvgAS) 2 Bvg' phase genes (e.g. bip1) A Bvg-repressed genes ^ (e.g. frIAB. flaA)

mM nic. Q

0.4

t

AbvgS

Bvgi

05

Bvgwt 0

f

t

Bvgc

mMnic. 0

4.0

>

CD

0.1 0.2 0.4 0.8 1.6 2.0 4.0 0

n ^

t

O)

CD

0

{f

(

BvgWt

>

0

0.1 0.2 0.4 0.8 1.6 2.0 4.0 0

M -84 -63

uis

I serum:

anti-Bvgwt B.

bronchiseptica

-36 -26 anti-Bvg' B. bronchiseptica

13.

BORDETELLA

649

controls basic physiologic processes. This observation highlights the importance of this global regulatory system in the life cycle of the organism. In comparison to B. bronchiseptica, little is known about the Bvg" phase of B. pertussis. Although Knapp and Mekalanos discovered five Bvg~ phase-specific B. pertussis genes {vrg6, vrgl8, vrg24, vrg53, and vrg73) in 1988 [248], the functions of the gene products are still unknown. Their regulation, however, has been studied in some detail and involves bvgR, the second intermediate regulatory locus in the BvgAS regulon to be identified [256]. Additional Bvg-repressed factors were discovered using a biochemical approach. Using two-dimensional gel electrophoresis and a hybridoma bank of monoclonal antibodies, Stenson and Peppier demonstrated the existence of several Bvg~ phase-specific outer membrane proteins in B. pertussis [257]. Two that were further characterized, vra-a and vra-b, appear not to be encoded by any of the previously identified vrgs and are not expressed in B. bronchiseptica or B. parapertussis [257]. Roles for these factors have also not yet been determined. Repression of vra-a and vra-b has been shown to involve BvgR [258, 259].

Fig. 4 (opposite) The three phases of Bordetella. BvgAS control at least three distinct phenotypic phases in response to environmental conditions. (A) The Bvg^ phases of B. pertussis (B.p.) and B. bronchiseptica {B.b.) are nearly identical. Both subspecies express a variety of surface molecules (solid and dashed lines) and secreted factors (solid circles). Both subspecies appear to also express a similar Bvg' phase, characterized by the expression of a subset of Bvg-activated factors (dashed lines) as well as factors expressed maximally in this phase (shaded triangles). The Bvg" phases of B. pertussis and B. bronchiseptica differ, vrgs and vras (whose products are represented by solid ovals), are expressed only by B. pertussis, while motility (flagella are represented by curved lines) and other coregulated factors such as urease (represented by solid circles) are expressed by B. bronchiseptica but not by B. pertussis. The Bvg"^ phases of both B. bronchiseptica and B. pertussis have been shown to be necessary and sufficient for respiratory tract colonization, while the Bvg' phase is hypothesized to be involved in aerosol transmission. The Bvg~ phase of B. bronchiseptica is necessary and sufficient for survival under nutrient poor conditions and may contribute to transmission of this organism by allowing it to survive in an environmental reservoir. (B) BvgAS controls expression of at least four classes of genes. "Late" Bvg-activated genes (1), such as cyaA, are expressed only under Bvg"'" phase conditions (in the absence of modulators such as nicotinic acid), while "early" Bvg-activated genes (2), such asfhaB and bvgAS), are expressed under both Bvg' and Bvg^ phase conditions. Bvg' phase genes (3), such as bipl, are expressed maximally under Bvg' phase conditions (e.g., in the presence of low levels of nicotinic acid). Expression of Bvg-repressed genes (4) is maximal under Bvg~ phase (fully modulating) conditions. (C) Western blots showing three classes of Bvg-regulated antigens. Whole cell lysates of Bvg' or Bvg~ phase-locked B. bronchiseptica, or wild-type B. bronchiseptica grown in the presence of various concentrations of nicotinic acid (mM nic), were separated by SDS-PAGE, transferred to PVDF, then probed with serum from a rat infected with wild-type B. bronchiseptica (left panel) or a rat infected with a Bvg' mutant (right panel). Polypeptides encoded by "late" Bvg-activated genes (1) are present only in wild-type B. bronchiseptica grown in the presence of

Principles of Bacterial Pathogenesis

Bacterial Pathogenesis

Bacterial Pathogenesis Methods and Protocols

Bacterial Pathogenesis Methods and Protocols

Bacterial Pathogenesis: A Molecular Approach

Bacterial Pathogenesis (Methods in Molecular Biology, v431)

Fungal Pathogenesis: Principles and Clinical Applications (Mycology)

Fungal Pathogenesis: Principles and Clinical Applications (Mycology)

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Pathogenesis of Neurodegenerative Disorders

Genetics of Bacterial Polysaccharides

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Biochemistry of bacterial growth

Bacterial Chromatin

Bacterial Systematics

Bacterial Biofilms

Bacterial Signaling

Bacterial Systematics

Methods in Enzymology, Volume 235: Bacterial Pathogenesis, Part A: Identification and Regulation of Virulence Factors

Mims' Pathogenesis of Infectious Disease

Mims' Pathogenesis of Infectious Disease

Genomics of Foodborne Bacterial Pathogens

Viral Pathogenesis and Immunity

Bacterial Resistance to Antimicrobials

Acinetobacter: Biology and Pathogenesis (Infectious Agents and Pathogenesis)

Principles of Bacterial Pathogenesis

Bacterial Pathogenesis

Bacterial Pathogenesis Methods and Protocols

Bacterial Pathogenesis Methods and Protocols

Bacterial Pathogenesis: A Molecular Approach

Bacterial Pathogenesis (Methods in Molecular Biology, v431)

Fungal Pathogenesis: Principles and Clinical Applications (Mycology)

Fungal Pathogenesis: Principles and Clinical Applications (Mycology)

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Pathogenesis of Neurodegenerative Disorders

Genetics of Bacterial Polysaccharides

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Handbook of Bacterial Adhesion: Principles, Methods, and Applications

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Biochemistry of bacterial growth

Bacterial Chromatin

Bacterial Systematics

Bacterial Biofilms

Bacterial Signaling

Bacterial Systematics

Methods in Enzymology, Volume 235: Bacterial Pathogenesis, Part A: Identification and Regulation of Virulence Factors

Mims' Pathogenesis of Infectious Disease

Mims' Pathogenesis of Infectious Disease

Genomics of Foodborne Bacterial Pathogens

Viral Pathogenesis and Immunity

Bacterial Resistance to Antimicrobials

Acinetobacter: Biology and Pathogenesis (Infectious Agents and Pathogenesis)

Recommend Documents