Campylobacter Third Edition
Editors
Irving Nachamkin Department of Pathology and Laboratory Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania
Christine M. Szymanski Institute for Biological Sciences National Research Council Ottawa, Ontario Canada and
Martin J. Blaser Departments of Medicine and Microbiology New York University School o f Medicine and New York Harbor Veterans Affairs Medical Center New York, New York
ASM
-
PRESS
Washington, DC
CONTENTS
Contributors Preface
I.
ix xv
The Organism
1. Taxonomy of the Family Campylobacteraceae 3 Lies Debruyne, Dirk Gevers, and Peter Vandamme 2. Population Biology of Campylobacter jejuni and Related Organisms 27 Martin C. J. Maiden and Kate E. Dingle 3. Complexity and Versatility in the Physiology and Metabolism of Campylobacter 41 jejuni David J. Kelly 4. Comparative Genomics of Campylobacter 63 jejuni Olivia L. Champion, Suaad Al-Jaberi, Richard A. Stabler, and Brendan W. Wren
5. Comparative Genomics of Campylobacter Species Other Than Campylobacter 73 jejuni William G. Miller
7. Clinical Significance of Carnpylobacter and Related Species Other Than Campylobacter jejuni and Campylobacter coli 123 Albert J. Lastovica and Ban Mishu Allos
8. Burden of Illness of Campylobacteriosis and Sequelae 151 Kdre M d b a k and Arie Havelaar 9. Epidemiology of Carnpylobacter jejuni Infections in Industrialized Nations 163 Christine K. Olson, Steen Ethelberg, Wilfrid van Pelt, and Robert V. Tauxe 10. Molecular Epidemiology of Campylobacter Species 191 Stephen L. W. On, Noel McCarthy, William G. Miller, and Brent J. Gilpin 11. Isolation, Identification, Subspecies Differentiation, and Typing of Campylobacter 213 fetus Marcel A. P. van Bergen, Jos P. M. van Putten, Kate E. Dingle, Martin J. Blaser, and Jaap A. Wagenaar
12. Diagnosis and Antimicrobial Susceptibility of Campylobacter Species 227 Collette Fitzgerald, Jean Whichard, and Irving Nachamkin
Clinical and Epidemiologic Aspects of 11. Campylobacter Infections
13. Guillain-BarrC Syndrome and Campylobacter Infection 245 Bart C. Jacobs, Alex van Belkum, and Hubert P. Endtz
6. Clinical Aspects of Carnpylobacter jejuni and Campylobacter coli Infections 99 Martin J. Blaser and Jargen Engberg
14. Mechanisms of Antibiotic Resistance in Campylobacter 263 Qijing Zhang and Paul J. Plummer
vi
CONTENTS
15. National Molecular Subtyping Network for Food-Borne Bacterial Disease Surveillance in the United States 277 Peter Gerner-Smidt, Steven G. Stroika, and Collette Fitzgerald
111.
IV.
Glycobiology
25. N-Linked Protein Glycosylation in Campylobacter 447 Harald Nothaft, Saba Amber, Markus Aebi, and Christine M. Szymanski
Pathogenesis and Immunity 26. 0-Linked Flagellar Glycosylation in Campylobacter 471 Susan M. Logan, Ian C. Schoenhofen, and Patricia Guerry 9
16. Interaction of Campylobacter jejuni with Host Cells 289 Robert 0. Watson andlorge E. G a l h
Cell Biology of Human Host Cell Entry by 17. Campylobacter jejuni 297 Lan Hu and Dennis]. Kopecko 8
Campylobacter jejuni Secretes Proteins via 18. the Flagellar Type I11 Secretion System That Contribute to Host Cell Invasion and Gastroenteritis 3 15 Charles L. Larson, Jeffrey E. Christensen, Sophia A. Pacheco, Scott A. Minnich, and Michael E. Konkel
27. Campylobacter jejuni Lipooligosaccharides: Structures and Biosynthesis 483 Michel Gilbert, Craig T. Parker, and Anthony P. Moran 28. Campylobacter jejuni Capsular Polysaccharide 505 Andrey V. Karlyshev, Brendan W. Wren, and Anthony P. Moran 29. Campylobacter Metabolomics 523 Evelyn C. Soo, David J. McNally, lean-Robert Brisson, and Christopher W. Reid
19. Innate Immunity in Campylobacter Infections 333 Nicole M. Iovine
V.
20. Chemosensory Signal Transduction Pathway of Campylobacter jejuni 351 Victoria Korolik and Julian Ketley
Regulation of Flagellar Gene Expression 30. and Assembly 545 David R. Hendrixson
21. Animal Models of Campylobacter jejuni Infections 367 Linda S. Mansfield, David B. Schauer, andJames G. Fox
Natural Competence and Transformation in 3 1. Campylobacter 559 Rebecca S. Wiesner and Victor]. DiRita
Genes and Gene Expression
22. Rabbit Model of Guillain-BarrC 381 Syndrome Nobuhiro Yuki
Survival Strategies of Campylobacter jejuni: 32. Stress Responses, the Viable but Nonculturable State, and Biofilms 571 Sarah L. Svensson, Emilisa Frirdich, and Erin C. Gaynor
23. Pathogenesis of Campylobacter fetus 401 Martin 1.Blaser, Diane G. Newell, Stuart A. Thompson, and Ellen L. Zechner
33. Iron Metabolism, Transport, and Regulation 59 1 Alain Stintzi, Arnoud H. M. van Vliet, andlulian M. Ketley
24. Development of a Human 429 Vaccine David R. Tribble, Shahida Baqar, and Stuart A. Thompson
Regulation of Genes in Campylobacter 34. jejuni 611 Marc M. S. M. Wosten, Andries van Mourik, and 10s P. M. van Putten
8
CONTENTS
VI.
Food Safety and Intervention
35. Campylobacter in the Food 627 Supply Wilma Jacobs-Reitsma, Ulrike Lyhs, and Jaap Wagenaar Transmission of Antibiotic Resistance from 3 6. Food Animals to Humans 645 Frank M. Aarestrup, Patrick F. McDermott, and Henrik C. Wegener
37. Poultry Colonization with Campylobacter and Its Control at the Primary Production Level 667 Jaap A. Wagenaar, Wilma Jacobs-Reitsma, Merete Hofshagen, and Diane Newell 38. Bacteriophage Therapy and Campylobacter 679 Ian F. Connerton, Phillippa L. Connerton, Paul Barrow, Bruce S. Seal, and Robert J. Atterbuy Index
695
vii
CONTRIBUTORS
Frank M . Aarestrup National Food Institute, DK-1790 Copenhagen V, Denmark
Martin J. Blaser Department of Medicine, New York University School of Medicine, New York, NY 10016
Suaad Al-Jaberi Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WClE 7HT, United Kingdom
Jean-Robert Brisson NRC-Institute for Biological Sciences, Ottawa, Ontario K1A OR6, Canada Olivia L. Champion School of Biosciences, University of Exeter, Exeter EX4 4QD, United Kingdom
Ban Mishu Allos Division of Infectious Diseases, Vanderbilt University School of Medicine, Nashville, TN 37232
Jeffrey E. Christensen School of Molecular Biosciences, Washington State University, Pullman, WA 99 164-4234
Markus Aebi Department of Biology, Institute of Microbiology, Eidgenossische Technische Hochschule (ETH) Zurich, CH-8093 Zurich, Switzerland
Ian F. Connerton School of Biosciences, Division of Food Sciences, University of Nottingham, Loughborough, Leics LE12 SRD, United l n g d o m
Saba Amber Department of Biology, Institute of Microbiology, Eidgenossische Technische Hochschule (ETH) Zurich, CH-8093 Zurich, Switzerland
Phillippa L. Connerton School of Biosciences, Division of Food Sciences, University of Nottingham, Loughborough, Leics LE12 5RD, United Kingdom
Robert J. Atterbury Department of Clinical Veterinary Science, University of Bristol, Langford, Bristol BS40 ~ D U Y United Kingdom
Lies Debruyne Laboratory of Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium
Shahida Baqar Enteric Diseases Program, Naval Medical Research Center, Silver Spring, MD 20910-7500
Kate E. Dingle Nuffield Department of Clinical Laboratory Sciences, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, United Kingdom
Paul Barrow School of Veterinary Medicine and Science, Division of Food Sciences, University of Nottingham, Loughborough, Leics LE12 SRD, United Kingdom
Victor J. DiRita Department of Microbiology and Immunology and Unit for Laboratory Animal Medicine, University of Michigan, Ann Arbor, MI 48 103 ix
x
CONTRIBUTORS
Hubert P. Endtz Laboratory Sciences Division, International Centre for Diarrheal Disease Research Bangladesh, Dhaka 1212, Bangladesh
Patricia Guerry Naval Medical Research Center, Silver Spring, MD 20910
Jargen Engberg Department of Clinical Microbiology, Hvidovre Hospital, DK-2650 Hvidovre, Denmark
Arie Havelaar Laboratory for Zoonoses and Environmental Microbiology, National Institute for Public Health and the Environment, Bilthoven, The Netherlands
Steen Ethelberg Department of Epidemiology and Department of Bacteriology, Statens Serum Institut, DK-2300 Copenhagen S, Denmark
David R. Hendrixson Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, TX 75390
Collette Fitzgerald Enteric Diseases Laboratory Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333
Merete Hofshagen National Veterinary Institute, Oslo, Norway
James G. Fox Division of Comparative Medicine and Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139
Lan Hu Laboratory of Enteric and Sexually Transmitted Diseases, Center for Biologics Evaluation and Research, U.S. Food and Drug Administration, Bethesda, MD 20892
Emilisa Frirdicb Department of Microbiology and Immunology, University of British Columbia, Vancouver, British Columbia V6T 123, Canada
Nicole M . Iovine Department of Medicine, New York University School of Medicine, Veteran’s Administration Medical Center, New York, NY 10010
Jorge E. Gal& Section of Microbial Pathogenesis, Yale University School of Medicine, New Haven, CT 06536
Bart C . Jacobs Departments of Neurology and Immunology, Erasmus University Medical Center, Rotterdam 3015 CE, The Netherlands
Erin C. Gaynor Department of Microbiology and Immunology, University of British Columbia, Vancouver, British Columbia V6T 123, Canada Peter Gerner-Smidt Enteric Diseases Laboratory Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333 Dirk Gevers Laboratory of Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium Michel Gilbert Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario K1A OR6, Canada Brent J. Gilpin Water Management Programme, Institute of Environmental Science and Research Ltd., Christchurch, New Zealand
Wilma Jacobs-Reitsma RIKILT Institute of Food Safety, 6708 PD Wageningen, The Netherlands Andrey V. Karlysbev London School of Hygiene & Tropical Medicine, London WC1E 7HT, United Kingdom David J. Kelly Department of Molecular Biology and Biotechnology, University of Sheffield, Western Bank, Sheffield S10 2TN, United Kingdom Julian Ketley Department of Genetics, University of Leicester, Leicester LE1 7RH, United Kingdom Michael E. Konkel School of Molecular Biosciences, Washington State University, Pullman, WA 99 164-4234
CONTRIBUTORS
Dennis J. Kopeck0 Laboratory of Enteric and Sexually Transmitted Diseases, Center for Biologics Evaluation and Research, U.S. Food and Drug Administration, Bethesda, MD 20892 Victoria Korolik Institute for Glycomics, Griffith University, Gold Coast, QLD 4222, Australia Charles L. Larson School of Molecular Biosciences, Washington State University, Pullman, WA 99 164-4234 Albert J. Lastouica Department of Biotechnology, University of the Western Cape, Bellville 7535, South Africa
xi
Scott A. Minnich Department of Microbiology, Molecular Biology and Biochemistry, University of Idaho, MOSCOW, ID 83844-3 052 a r e Mslbak Department of Epidemiology, Statens Serum Institute, Copenhagen, Denmark Anthony P. Moran Department of Microbiology, National University of Ireland, Galway, Ireland, and Institute for Glycomics, Griffith University, Queensland, Australia Irving Nachamkin Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA 19104-4283
Susan M . Logan Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario K1A OR6, Canada
Diane G. Newel1 Veterinary Laboratories Agency, New Haw, Addlestone, Surrey KT15 3NB, United Kingdom
Ulrike Lyhs Seinajoki Unit, Ruralia Institute, University of Helsinki, FI-60320 Seinajoki, Finland
Harald Nothaft Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario K1A OR6, Canada
Martin C . J. Maiden Department of Zoology, University of Oxford, Oxford OX1 3PS, United Kingdom
Christine K. Olson Enteric Diseases Epidemiology Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333
Linda S. Mansfield Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI 48824
Stephen L. W . On Food Safety Programme, Institute of Environmental Science and Research Ltd., Christchurch, New Zealand
Noel McCarthy Department of Zoology, University of Oxford, Oxford OX1 3SY, United Kingdom
Sophia A. Pacheco School of Molecular Biosciences, Washington State University, Pullman, WA 99 164-4234
Patrick F. McDermott National Antimicrobial Resistance Monitoring System, Center for Veterinary Medicine, U.S. Food & Drug Administration, Laurel, MD 20708
Craig T. Parker Produce Safety and Microbiology Research, Agricultural Research Service, United States Department of Agriculture, Albany, CA 94710
David J. McNally NRC-Institute for Biological Sciences, Ottawa, Ontario K1A OR6, Canada
Paul J. Plummer Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, Ames, JA 50011
William G. Miller Produce Safety and Microbiology Research Unit, USDA, ARS, Albany, CA 94710
Christopher W. Reid NRC-Institute for Biological Sciences, Ottawa, Ontario K1A OR6, Canada
xii
CONTRIBUTORS
David B. Schauer Division of Comparative Medicine and Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139
David R. Tribble Preventive Medicine and Biometrics Department, Uniformed Services University of the Health Sciences, Bethesda, MD 208 14-5119
Ian C . Schoenhofen Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario K1A OR6, Canada
Alex van Belkum Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Center, Rotterdam 3015 CE, The Netherlands
Bruce S. Seal Poultry Microbiological Safety Research Unit, Russell Research Center, Agricultural Research Service, U.S. Department of Agriculture, Athens, GA 30605 Evelyn C . So0 NRC-Institute for Marine Biosciences, Halifax, Nova Scotia B3H 321, Canada Richard A. Stabler Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WClE 7HT, United Kingdom Alain Stintzi Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada Steven G. Stroika Enteric Diseases Laboratory Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333 Sarah L. Svensson Department of Microbiology and Immunology, University of British Columbia, Vancouver, British Columbia V6T 123, Canada Christine M . Szymanski Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario K1A OR6, Canada Robert V. Tauxe Division of Foodborne, Bacterial and Mycotic Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30333 Stuart A. Thompson Department of Biochemistry and Molecular Biology, Medical College of Georgia, Augusta, GA 30912
Marcel A. P. van Bergen Division of Infectious Diseases, Animal Sciences Group, OIE Reference Laboratory for Campylobacteriosis, 8200 AB Lelystad, The Netherlands Peter Vandamme Laboratory of Microbiology, Ghent University, B9000 Gent, Belgium Andries van Mourik Department of Infectious Diseases and Immunology, Utrecht University, 3584 CL Utrecht, The Netherlands Wilfrid van Pelt Department of Epidemiology and Surveillance, Rijksinsituut voor Volksgezondheid en Milieu, 3721 MA Bilthoven, The Netherlands Jos P. M . van Putten Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, University of Utrecht, 3508 TD Utrecht, The Netherlands
Arnoud H . M . van Vliet Institute of Food Research, Norwich NR4 7UA, United Kingdom Jaap A . Wagenaar Division of Infectious Diseases, Animal Sciences Group , OIE Reference Laboratory for Campylobacteriosis and WHO Collaboration Centre for Campylobacter, 8200 AB Lelystad, and Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, University of Utrecht, 3.508 TD Utrecht, The Netherlands Robert 0. Watson Section of Microbial Pathogenesis, Yale University School of Medicine, New Haven, CT 06536 Henrik C . Wegener National Food Institute, DK-1790 Copenhagen V, Denmark
CONTRIBUTORS
xiii
Jean Whichard Enteric Diseases Laboratory Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333
Nobuhiro Yuki GBS Laboratory, 3-9-5 Kamiikedai, Ota-ku, Tokyo 145-0064, Japan
Rebecca S. Wiesner University of Michigan, Ann Arbor, MI 48103
Ellen L. Zechner Institute of Molecular Biosciences, University of Graz, A-8010 Graz, Austria
Marc M . S. M . Wosten Department of Infectious Diseases and Immunology, Utrecht University, 3584 CL Utrecht, The Netherlands Brendan W. Wren Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WClE 7HT, United Kingdom
Qijing Zhang Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, Ames, IA 50011
PREFACE
late Gerald Aspinall, has provided a structural basis for important polysaccharides and glycolipids from the organism and great insight into glycosylation systems, which also are present in other prokaryotes and eukaryotes. This work, combined with an increased understanding of the molecular biology of gene expression in Campylobacter, is helping to form a more complete picture about the organism and its interaction with the host and environment. Ultimately, research will help improve human health through understanding the immunology of campylobacter infection and the pertinent host defenses. These will lead to the development of strategies to reduce infection, through either effective vaccines or improved food safety. The ecology of food safety also has seen a dramatic increase in research, with emphasis on understanding the extent of campylobacter in the food supply, transmission of antibiotic resistant campylobacters from food animals to humans, and control of campylobacter at the food source. The above advances have helped to form the nucleus for the 3'd edition of Campylobacter. We are grateful to the scientific groups from around the world that have contributed to the preparation of these outstanding chapters. These represent the generous sharing of our growing knowledge of these important zoonotic pathogens of humans. Finally, we especially thank Greg Payne at ASM Press for his help and guidance throughout the process of developing this volume. Irving Nachamkin Christine M. Szymanski Martin J. Blaser February 2008
Our understanding of the clinical aspects, epidemiology, and pathogenesis of Campylobacter infection has increased dramatically since publication of the second edition of Campylobacter in 2000. As the number of species within the family has expanded, so has our knowledge of this group of organisms in terms of their physiology, population biology, and diversity. The ability to understand genomic diversity in Campylobacter is due in great part to improvements in technology that have advanced comparative analyses. The sequencing of the first C. jejuni strain in 2000 was a milestone in Campylobacter genetics and brought to light aspects of biology that previously could not have been identified. Since then, additional complete Campylobacter genome sequences have been published and have provided new insights about the genomic diversity of these organisms, which now is covered extensively in this new edition. Campylobacter infections and their complications, such as Guillain-BarrC syndrome, cause significant morbidity in specific populations. There is growing recognition of Campylobacter among clinicians as well as by the lay public. Antimicrobial resistance also continues to increase and poses new issues regarding therapies. Of particular note are the impact that researchers in our field have had on government regulation of antimicrobial agent use in food animals and how effecting change will ultimately improve human health. Our understanding of the pathogenesis of Campylobacter infections has advanced greatly, especially in the area of signal transduction pathways and cell biology of the organism. The emerging field of Campylobacter glycobiology, which was pioneered by the
xv
I. THE ORGANISM
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 1
Taxonomy of the Family Campylobacteraceae LIESDEBRUYNE, DIRKGEVERS,AND PETER VANDAMME
fer, and species-specific core genomes and strainspecific accessory genomes have been described for several bacterial species (Dagan and Martin, 2006; Kunin et al., 2007). It would be unrealistic to expect that in the near future bacterial species will be described on the basis of whole-genome sequences only. Several recent studies (Goris et al., 2007; Konstantinidis et al., 2006), however, demonstrated that multiple parameters in whole-genome sequences reflect the traditional delineation of species at the 70% DNA-DNA hybridization level. In particular, multilocus sequence analysis (MLSA), a technique based on sequence analysis of fragments of multiple household genes present in the core genome, has been put forward as a novel approach for the classification of bacteria (Gevers et al., 2005). Konstantinidis et al. (2006) demonstrated that phylogenetic trees based on MLSA data can bear the same phylogenetic signal as trees that are based on sequence analysis of the entire common gene pool, implying that one does not need entire genome sequences for all strains under comparison to design taxonomic frameworks equivalent to frameworks that are based on whole-genome content. MLSA schemes have been developed for a growing number of bacteria, including some members of the family Campylobacteraceae (see below). Additional studies are needed to provide pragmatic criteria for determining how we can use MLSA data to define species and subdivisions within species for poorly studied or newly discovered taxa.
This chapter will present a traditional overview of the biological diversity of Campylobucter and Arcobacter but will also address the taxonomic information that has become available through whole-genome sequence analysis. Starting in the 1990s, genomic sequencing drastically changed our understanding of the relationships between microorganisms. Many new papers have been evaluating diversity within and among species by exploiting whole-genome information content, and the bacterial species concept emerged from obscurity into the spotlight (Doolittle and Papke, 2006; Gevers et al., 2005; Konstantinidis et al., 2006). Since the 1960s, the level of whole-genome DNA-DNA hybridization is the accepted standard for species delineation, with 70% DNA-DNA reassociation being the recommended species cutoff. Genomic DNA-DNA hybridization has been used as the key technology for the phylogenetic delineation of bacterial species because whole-genome sequences were assumed to provide a rational and objective basis for the description and delineation of bacterial species. While waiting for easy access to whole-genome sequences to emerge, the genomic DNA-DNA hybridizations were the best approach to comparing wholegenome sequences. Meanwhile, sequencing technology improved so dramatically that several hundreds of bacterial genomes have been-or are in the process of beingsequenced and annotated. It has become apparent that apart from the traditional nucleotide changes, other selective forces such as gene deletion, gene duplication, horizontal gene transfer, and chromosomal rearrangements shape the bacterial genome and that considerable fractions of the genome can be strain specific. Numerous studies, however, have stressed the dominance of vertical over horizontal gene trans-
TAXONOMIC HISTORY The early history of the genus Campylobacter, starting with the first description by Theodor Escher-
Lies Debruyne, Dirk Gevers, and Peter Vandamme Laboratory of Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Bioinformatics and Evolutionary Genomics, Ghent University/VIB, Technologiepark 927, B-9052 Gent, Belgium. Dirk Gevers Gent, Belgium.
3
4
DEBRUYNE ET AL.
ich in 1886, the first isolation of a Vibrio-like organism from aborted ovine fetuses by McFadyean and Stockman in 1913, the emergence of several novel “Vibrio” species in a range of animals with and without clinical symptoms, and the creation of the genus Campylobacter in 1963, has been outlined in detail in several previous reviews (Butzler, 2004; Skirrow, 2006). In 1963, Sebald and VCron transferred two of these Vibrio species, Vibrio fetus and Vibrio bubulus, into the new genus Campylobacter as Campylobacter fetus and Campylobacter bubulus, respectively. The former was the bacterium first isolated by McFadyean and Stockman (1913), which was subsequently also isolated from aborted bovine fetuses (Smith and Taylor, 1919) and from blood cultures of women who aborted (Vinzent et al., 1947); the latter was an organism isolated from the bovine vagina and semen (Florent, 1953). The reasons for transferring these two species into the novel genus Campylobacter included their low DNA base composition, their microaerophilic growth requirements, and their nonfermentative metabolism. Ten years later, VCron and Chatelain (1973) published a more comprehensive study on the taxonomy of the microaerophilic Vibrio-like organisms and considered four distinct species in the genus Campylobacter: C. fetus (the type species), Campylobacter coli (a vibrio isolated from feces of pigs with diarrhea [Doyle, 1948]), Campylobacter jejuni (a vibrio isolated from the feces of cattle with diarrhea Uones et al., 19311, blood cultures of humans with gastroenteritis [King, 1957; Levy, 19461 and aborted sheep fetuses [Bryans et al., 1960]), and Campylobacter sputorum. The latter comprised two subspecies, C. sputorum subsp. sputorum (a vibrio isolated from sputum of a patient with bronchitis [Prihot, 1940; Tunicliff, 19141) and C. sputorum subsp. bubulus (a vibrio isolated from the bovine vagina and semen [Florent, 19531). The application of a filtration technique used in veterinary microbiology allowed Butzler and coworkers in the early 1970s to exploit the small cell dimensions and the vigorous motility of Campylobacter cells to selectively isolate them from stools of humans with diarrhea (Butzler et al., 1973; Dekeyser et al., 1972). The main breakthrough, however, was provided a few years later by Skirrow, who described a selective supplement comprising a mixture of vancomycin, polymyxin B, and trimethoprim that was added to a basal medium (Skirrow, 1977). This simple isolation procedure thus enabled routine diagnostic microbiology laboratories to isolate campylobacters and to evaluate their clinical role. The availability of adequate isolation procedures led to a renewed interest in Campylobacter research
during the early 1980s. As a consequence, manifold Campylobacter-like organisms were isolated from a variety of human, animal, and environmental sources, and gradually, new species were described (Benjamin et al., 1983; Fox et al., 1989; Gebhart et al., 1985; Lawson et al., 1981; Marshall et al., 1984; McClung et al., 1983; Neil1 et al., 1985; Tanner et al., 1981). Classical biochemical tests routinely used for the identification of clinical bacteria often yielded negative or variable results within Campylobacter species. This poor biochemical reactivity and lack of clear-cut differential characters led to the wide application of vernacular names for many groups of Campylobacter-like organisms, and the term CLO (Campylobacter-like organism) became widely used. Some of these CLO groups were later classified as novel species, but several were identified as biochemical variants of well-known species (Vandamme, 2000). Still others remained unnamed: for example, the CLO-3 strain that was isolated in the early 1980s was not formally classified pending the isolation and characterization of additional strains belonging to the same taxon (Totten et al., 1985). An overview of these empirical names with their final classification has been given previously (Vandamme, 2000). Also in the 1980s, the study of bacterial phylogeny as imprinted in the degree of rRNA cistron similarity became more popular and was increasingly used to evaluate and revise bacterial classification schemes. Long-standing genera like Pseudomonas, Flavobacterium, Bacteroides, and many others were found to be extremely heterogeneous, and their classification was gradually revised to conform to the new phylogenetic (or “natural”) insights. The first rRNA-based phylogenetic studies of Campylobacter species were published in the late 1980s and demonstrated that Campylobacter too was heterogeneous, as it comprised three distinct phylogenetic lineages that also included members of the genus Wolinella and two Bacteroides species, Bacteroides gracilis and Bacteroides ureolyticus, which appeared phylogenetically distinct from the true Bacteroides species (Vandamme, 2000). In 1989, Goodwin et al. summarized genotypic and phenotypic arguments to exclude the gastric species Campylobacter pylori (humans) and Campylobacter mustelae (ferrets) from the genera Campylobacter and Wolinella and proposed a novel genus, Helicobacter, to accomodate both. Two years later, a complete revision of the taxonomy and nomenclature of the genus Campylobacter and related bacteria was proposed by Vandamme et al. (1991a). The results of an extensive DNA-rRNA hybridization study of over 60 strains representing all known Campylobacter species, CLO groups, and putative relatives such
CHAPTER 1
as Wolinella, Bacteroides, and “Flexispira” species corroborated and extended the 16s rRNA sequencing data. Carnpylobacter, together with Wolinella and “Flexispira” and B. gracilis and B. ureolyticus, was found to represent a separate, sixth rRNA superfamily sensu De Ley within the group of the gramnegative bacteria. This phylogenetic lineage is now known as the epsilon subdivision of the proteobacteria. Within this lineage, C. fetus, Campylobacter hyointestinalis (Gebhart et al., 1985), Campylobacter concisus (Tanner et al., 1981), Campylobacter mucosalis (Lawson et al., 1981; Roop et al., 1985a), C. sputorum, C. jejuni, C. coli, Campylobacter lari (Benjamin et al., 1983), Campylobacter upsaliensis (Sandstedt and Ursing, 1991), Wolinella curva (Tanner et al., 1984), Wolinella recta (Tanner et al., 1981), B. gracilis, and B. ureolyticus constituted the first rRNA homology cluster. Campylobacter nitrofigilis (McClung et al., 1983), Campylobacter cryaerophila (Neil1 et al., 1985), and several other aerotolerant CLO strains (which were later classified as Arcobacter butzleri and Arcobacter skirrowii vandamme et al., 1992]), constituted the second rRNA homology cluster. Helicobacter pylori (Goodwin et al., 1989; Marshall et al., 1984), H. mustelae (Fox et al., 1989; Goodwin et al., 1989), Campylobacter cinaedi (Totten et al., 1985), Campylobacter fennelliae (Totten et al., 1985), W. succinogenes (Wolin et al., 1961), “Flexispira rappini“ (Bryner et al., 1986), and the CLO-3 strain (Totten et al., 1985) constituted a third rRNA homology cluster. In addition, the free-living saprophytic Campylobacter strains of Laanbroek et al. (1977) and Wolfe and Pfennig (1977) (later reclassified as Sulfurospirillum species) belonged to the same phylogenetic lineage but did not belong to one of the three major rRNA homology clusters. A revised classification was proposed that took into account the phylogenetic results derived from the DNA-rRNA hybridization experiments and that published data on the genotype and phenotype of these organisms. The genus Carnpylobacter was restricted to those species belonging to the rRNA homology cluster containing C. fetus, the type species of the genus Campylobacter. The generically misnamed W. curva and W. recta, and in a subsequent study, B. gracilis (Vandamme et al., 1995), were included in the emended genus Campylobacter as Campylobacter curvus, Campylobacter rectus, and Campylobacter gracilis, respectively. B. ureolyticus was obviously a close relative of the emended genus Campylobacter, but its taxonomic status was not changed pending further study of this taxon and the isolation of similar bacteria (Vandamme et al., 1991a, 1995). The name Arcobacter (with Arcobacter nitrofigilis as the type species) was proposed for the organisms be-
-
TAXONOMY OF THE CAMPYLOBACTERACEAE
5
longing to the second rRNA homology cluster. Finally, C. cinaedi and C. fennelliae were transferred into the genus Helicobacter as H. cinaedi and H. fennelliae, respectively, and an emended genus description was proposed. Wolinella succinogenes remained the only species of the genus Wolinella. Several new Campylobacter, Arcobacter, and Helicobacter species have been reported since then. Campylobacter and Arcobacter species will be discussed below. A detailed overview of Helicobacter species is beyond the scope of the present chapter. Finally, the genera Campylobacter and Arcobacter were accomodated in the new bacterial family, the Campylobacteraceae (Vandamme and De Ley, 199l), because they shared several genotypic and phenotypic characteristics. The family Campylobacteraceae currently also comprises the generically misclassified species B. ureolyticus and the genus Sulfurospirillum (Stolz et al., 2005). In addition, a growing number of free-living Epsilonproteobacteria occupy phylogenetic positions among the Campylobacter, Arcobacter, Sulfurospirillum, and Wolinella-Helicobacter clades. The allocation of these bacteria to a bacterial family such as the Campylobacteraceae or others needs further study. Figure 1 represents a traditional phylogenetic tree of the family Campylobacteraceae that is based on percentage of 16s rRNA gene sequence similarity.
WHOLE-GENOME TAXONOMY OF CAMPYLOBACTER AND RELATED BACTERIA Species Tree of Campylobacter and Related Bacteria The progress in the whole-genome studies of Campylobacter and related bacteria presents an opportunity to produce a more robust species tree. Data sets consisting of a single gene or a small number of concatenated genes have a significant probability of supporting conflicting topologies. By contrast, a fully resolved species tree with maximum support can come from the concatenation of a large number of carefully selected genes, which whole-genome sequences provide in abundance. A total of 25 complete and incomplete genome sequences have been obtained for 15 species belonging to the Campylobacteraceae and related bacteria (A. butzleri, C. coli, C. concisus, C. curvus, C. fetus, C. jejuni, C. lari, C. upsaliensis, Helicobacter acinonychis, H. hepaticus, H. mustelae, H. pylori, Sulfurimonas denitrificans, Thiornicrospira crunogena, and W. succinogenes; http: //www.genomesonline.org/). These genomes were subjected to a comparative genome analysis to
6
DEBRUYNE ET AL.
-
-Sulfurospirillum cavolei Phe91T(AB246781) -Sulfurospirillum sp. DSM 806 (AF144694) Sulfurospirillum barnesii SES-3' (AF038843) ulfurospirillum deleyianum Spirillum 5 17.5' (Y13671) Sulfurospirillum multivorans K' (X82931) Sulfurospirillum halorespirans PCE-M2' (AF218076) "CandidatusArcobacter sulfidicus" (AY035822) Arcobacter butzleri ATCC 49616T(AY621116) obacter cibarius LMG 21996T(AJ607391) Arcobacter skirrowii CCUG 10374T(L14625) Arcobacter cryaerophilus CCUG 17801' (L14624) Arcobacter nitrofigilis CCUG 15893T(L14627) Arcobacter halophilus LA3 lBT(AF513455) Campylobacter hyointestinalis subsp. lawsonii CHY 5' (AF097685) -CCampylobacier lanienae NCTC 13004T(AF043425) Campylobacter hyointestinalis subsp. hyointestinalis NCTC 11608' (AF097689) mpylobacterfetus subsp.fetus ATCC 27374' (LO4314) Campylobacterfetus subsp. venerealis ATCC 19438T(L14633) Campylobacter canadensis L266T(EF621894) Campylobacter upsaliensis CCUG 14913' (L14628) -CCampylobacter helveticus NCTC 12470T(U03022) Campylobacter coli CCUG 11283' (LO4312) Campylobacter lari CCUG 23947' (LO4316) Campylobacter insulaenigrae NCTC 12927T(AJ620504) Campylobacterjejuni subsp. doylei CCUG 24567T(L14630)
: L
-
-I
-
c
Figure 1. Phylogenetic tree of the family Campylobacteruceae and related bacteria, based on percentage of 16s rRNA gene sequence similarity.
screen for protein families that together might reveal a major vertical component of species phylogeny (D. Gevers, unpublished data). Orthologous gene families were selected as phylogenetic markers if they met the following criteria: (i) had exactly one copy in each of
the 25 strains, (ii) had a minimum length of 100 amino acids after ambiguous positions were marked out, and (iii) showed no noisy phylogenetic signal from recombination or horizontal gene transfer. This resulted in a carefully selected set of 60 core genes
CHAPTER 1
that could be used to estimate the phylogeny of these bacteria by maximum likelihood (Fig. 2). In general, this species tree correlates nicely with the phylogenetic tree that is based on 16s rRNA gene sequences, but compared with the latter, it has a superior resolution to distinguish closely related bacteria such as C. jejuni subsp. jejuni, C. jejuni subsp. doylei, and C. coli. This tree also confirms and widens previous insights in the evolutionary relationships based on 12 concatenated protein data sets including only seven species (Fouts et al., 2005). By including more C. jejuni genomes, it is not only confirmed that C. jejuni RM1221 is more closely related to the C. jejuni subsp. jejuni strains than to C. coli, but it is also shown that it is closer to any C. jejuni subsp. jejuni than to C. jejuni subsp. doylei 269.97. Further, it confirms that W. succinogenes DSM 1740Tis more closely related to Helicobacter spp. than to any Campylobacter species. In addition, it was found that S. denitrificans ATCC 3 3 889= (reclassification from Thiomicrospira denitrificans) is more closely related to A. butzleri and Campylobacter than to T. crunogena.
0.2
I -
I
I
TAXONOMY OF T H E CAMPYLOBACTERACEAE
7
Whole-Genome Comparisons
As in other groups of bacteria, bioinformatic analyses of the growing number of whole-genome sequences of Campylobacter strains is explaining to which extent strains within species and strains representing different species are different. Goris et al. (2007) demonstrated that the recommended cutoff of 70% DNA-DNA reassociation for species delineation corresponds to 69% conserved DNA and 95% average nucleotide identity (ANI), a measure for evolutionary relatedness that is based on sequence similarity between orthologous genes. As shown in Fig. 3 , the AN1 values among C. jejuni strains do fall above this cutoff and range between 95.24 and 99.68%, with a significant partition between its two subspecies: intrasubspecies values for C. jejuni subsp. jejuni are on average 97.85% ( _ t 0.62%), and intersubspecies values for C. jejuni subsp. doylei versus C. jejuni subsp. jejuni are on average 95.80% (k0.28%). In comparison, Helicobacter pylori comprises a more heterogenous set of strains, with lower AN1 values ranging between 94.05 and 95.57%. The AN1 values
H . pylori 26695 H. p ylori H PAG 1 H . pylori399 H. acinon ychis str. Sheeba H. mustelae ATCC 43772 H . hepaticus ATCC 51449 Wolinella succinogenes DSM 1740 Sulfurimonas denitrificans ATCC 33889 Arcobacter butzleri 40 18 C. fetus subsp. fetus 82-40
HB93-13 81-176 260.94 C. j e j u n i subsp. j e j u n i CG8486 .C. j e j u n i RM1221 C. j e j u n i subsp. j e j u n i NCTC 11168 C. jejuni subsp. jejuni 84-25 CF93-6 Thiomicrospira crunogena X C L- 2
Figure 2. Whole-genome-based phylogeny of Campylobacter and related bacteria. Out of 365 Campylobacter core gene families, a concatenation of 60 carefully selected genes was used to estimate the phylogeny of these bacteria by maximum likelihood. Orthologous protein families were determined by OrthoMCL (Li et al., 2003). Protein sequence alignments were generated by Muscle (Edgar, 2004), and poorly conserved regions were trimmed by Gblocks (Castresana, 2000). Phylogenetic trees with 100 bootstrap replicates were. obtained by the PhyML algorithm (Guindon and Gascuel, 2003) with a JIT amino acid substitution matrix and a discrete gamma model. All nodes had 100% bootstrap support.
8
DEBRUYNE ET AL.
pylori-acinonychis 0
u 1
s!
60
0 5
0
80
90
100 100 I
95
90
85
80
75
70
Average Nucleotide Identity (ANI)
Figure 3. Genetic diversity among genomes of Cumpylobacter and related bacteria. Every data point represents a wholegenome comparison between two genomes (both intra- and interspecies) and shows the percentage AN1 on the x axes plotted against the percentage of gene content similarity on the y axes. Only comparisons with AN1 values above 70% are shown because this is the range in which AN1 is considered to be a robust and sensitive method for measuring evolutionary relatedness (Konstantinidis and Tiedje, 2005). AN1 values were calculated as described previously (Konstantinidis and Tiedje, 2005), and gene content similarity values were calculated on the basis of reciprocal best BLASTp hits.
are useful for quantifying evolutionary relatedness between different species and correspond to wholegenome-based phylogenies. The closest related species of c. jejuni among the included genomes is c. coli (ANIs of 83.23 -t 0.29%), and the second closest is C. upsaliensis (ANIs of 72.89 0.10%). In comparison, the evolutionary distance between H. pylori and H. acinonychis is smaller than any two species of Campylobacter included in our analyses (ANIs of 89.25 ? 0.18). The gene content similarity values, shown in Fig. 3, were determined to reveal the effect of ecology on this group of organisms. It has been reported before that the environment is one of the main factors in shaping the evolution of the gene content of an organism (Konstantinidis et al., 2006). Larger intraspecific genomic differences could be the outcome of a differential evolution of those strains in different environments. The intraspecific values of the C. jejuni
*
strains (excluding the C. jejuni subsp. jejuni CG8486 strain) go as low as 74%, indicative of an ecologically versatile organism. These differences are likely to be determined by the broad host range of C. jejuni, which includes human, cattle, and poultry as well as a large variety of wild bird species (Schouls et al., 2003). Similar large genomic differences within a species have been reported for Escherichia coli (Konstantinidis et al., 2006). The unfinished C. jejuni subsp. jejuni CG8486 genome sequence has significantly fewer annotated genes in comparison with other strains of this species (1,421 versus 1,750 2 100, respectively), which explains why comparisons with this strain result in more extreme gene content similarity values, and thus its gene content similarity values should be considered with caution. The AN1 values do not require the same reservation because they are not affected by large genome size differences in the same way. In contrast to the intraspecific values
CHAPTER 1
of the C. jejuni strains, the H. pylori strains show substantially smaller genomic differences, ranging between 83 and 90%, even though their AN1 values indicate a larger evolutionary distance. This is a strong indication of its narrower ecological nichethat is, it is limited to the human host. It is well established that the phenotypes and genotypes of C. jejuni and C. coli are remarkably similar, resulting in an important taxonomic and diagnostic problem. Figure 3 provides us with a quantitative measure of this similarity. Although the AN1 values reflect a clear distinction between these two species, c. coli seems to have a large set of genes in common with some of the C. jejuni strains, likely explaining the confusing phenotypic test results. Which and How Many Genes Are Needed To Predict the Whole-Genome-Based Phylogeny? Because of the limitations of 16s rRNA gene analysis for accurate species-level identification of campylobacters (see below), several other genes were examined for taxonomic and identification purposes. Karenlampi et al. (2004) demonstrated that partial groEL gene sequences could be used to construct a phylogeny similar to that obtained by 16s rRNA gene sequences, but with lower interspecies and high intraspecies similarities. Korczak et al. (2006) used rpoB sequences and also obtained a good congruence between rpoB and 16s rRNA gene sequence-based trees. Again, resolution for the rpoB gene was much higher than for the 16s rRNA gene, enabling differentiation of even closely related Campylobacter species, as well as of most subspecies. In 2001, Dingle and coworkers (2001) reported a first multilocus sequence typing scheme that is based on seven housekeeping loci to study population biology in C. jejuni. Since then, the analysis of polymorphisms in multiple coding gene sequences has been used in several other Campylobacter species for typing purposes (Dingle et al., 2005; Litrup et al., 2007; van Bergen et al., 2005a). Miller et al. (2005) reported the first multispecies, multilocus sequence-based scheme that allowed analysis of multiple Campylobacter species (C. jejuni, C. coli, C. lari, C. upsaliensis, and C. helveticus) but used it for strain typing and not for studying the phylogeny and identification of these bacteria. Yet the analysis of multilocus sequences currently represents the most realistic approach for studying species diversity and evolution at large (Gevers et al., 2005). But how do a few genes correlate with whole-genome-based phylogeny in the case of the Campylobacteraceae, and what genes should preferentially be used to improve phylogenetic accuracy in a cost-effective way? Genes currently used in multi-
TAXONOMY OF THE CAMPYLOBACTERACEAE
9
locus sequencing approaches are often selected primarily on the basis of being used previously in other taxa, being essential, being unlinked in the genome, and having a successful primer design. However, their phylogenetic robustness is overlooked, and a selection based on their predictive power for the whole-genome-based phylogeny would be better. Taking such an approach, it has been shown that a more accurate phylogeny for the Escherichia coli group can be obtained from three best-performing genes compared with the concatenated alignment of eight commonly used markers (Konstantinidis et al., 2006). The availability of 17 Campylobacteraceae genomic sequences allows a similar analysis. We evaluated the correlation between all the 548 Campylobacteraceae core genes separately, and we assessed the whole-genome-based phylogeny by performing comparisons both in terms of tree topology (Robinson-Foulds symmetric difference distance) and branch length (branch score distance of Kuhner and Felsenstein) with the FTREEDIST (http: //emboss. sourceforge.net/) and TOPD (Puigbo et al., 2007) software tools. This methodology guided the selection toward a total of 60 best-performing genes with tree topologies and branch lengths highly similar to the whole-genome-based tree, i.e., the species tree. However, single-gene-based trees correlate poorly with the species tree at the shorter evolutionary scales, i.e., among the very closely related strains of C. jejuni, where only the concatenation of multiple loci provides enough resolving power. This can be explained by the finding that C. jejuni has a weakly clonal population structure as a result of high levels of intraspecific recombination (Suerbaum et al., 2001). Among the genes that are chosen as good phylogenetic markers, there appears to be no functional bias (Table 1). By use of the general functional classification from the COG (Cluster of Orthologous Genes) database (Tatusov et al., 1997), it was found that the genes are evenly distributed over the categories of information storage and processing (25%), cellular processes and signaling (28%), and metabolism (32%), with the remainder classified in the “poorly characterized” category or not classified in any COG at all. For a gene to be practically useful as a phylogenetic marker in an MLSA approach, it should allow the development of a set of conserved primers of limited degeneracy. To identify regions corresponding to the most highly conserved segments of each of the 60 best-performing genes, amino acid alignments were submitted to the BLOCKS Multiple Alignment Processor (Henikoff et al., 1995; http://
Table 1. Functional classification of 60 selected phylogenetic markers for the Campylobacteraceae" Gene name
Product description
P7fA miaA" miaB dusB
Peptide chain release factor 1 tRNA delta(2)-isopentenylpyrophosphatetransferase tRNA-i(6)A37 thiotransferase enzyme tRNA-dihydrouridine synthase B Ribosomal protein L25 GTP-binding protein Glycyl-tRNA synthetase beta subunit Methionyl-tRNA formyltransferase DNA gyrase, B subunit Exodeoxyribonuclease VII, large subunit Excinuclease ABC subunit C DNA polymerase type I Holliday junction DNA helicase RuvB DNA gyrase, A subunit DNA processing protein A PP-loop family protein Chromosome partitioning protein, ParA family Mur ligase family protein Membrane-associated zinc metalloprotease, putative Methyltransferase GidB Outer-membrane lipoprotein carrier protein precursor Permease, putative Peptidase, M23 /M37 family Lipopolysaccharide heptosyltransferase I Penicillin-binding protein DNA repair protein ATPase subunit of heat shock protein H s l W Heat shock protein 90 Mg chelatase-related protein Trigger factor Preprotein translocase, SecY subunit Signal recognition particle-docking protein Homoserine dehydrogenase Aspartate kinase Carbamoyl-phosphate synthase small chain Carbamoyl-phosphate synthase, large subunit CTP synthetase HIT family protein Phosphatase, Ppx/ GppA family Fructose-1,6-bisphosphatase ATP-NAD kinase, putative Dephospho-CoA kinase 4-Hydroxybenzoate octaprenyltransferase Glutamyl-tRNA reductase Dihydropteroate synthase Bifunctional 3,4-dihydroxy-2-butanone 4-phosphate Synthase/ GTP cyclohydrolase I1 protein Oxygen-independent coproporphyrinogen 111 oxidase Riboflavin synthase subunit alpha D-3-phosphoglycerate dehydrogenase Acetyl-CoA carboxylase beta subunit Phosphatidylserine decarboxylase Peptidase, putative GTP-binding protein, GTPl/ Obg family tRNA modification GTPase Conserved hypothetical protein Porphobilinogen deaminase Translation elongation factor Ts Multidrug resistance efflux transporter, putative Delta-aminolevulinic acid dehydratase Peptidase, U32 family
7p"'
ychF dYS fmt" gY7B
xseA" uvrc polA ruvB gY7A
dprA" b b
murF
gidB' lolA* z.
waaC radA hslU htpG
*
tig secY ftsY* hom lysC carA carB PYrG
*
fbP b
coaE* ubiA hemA folP ribB hemN ribE* serA accD psdb b
obg tmzE hemC tsf hemB
COG
J J J J J J J J
L L L L
L L LU D D M M M M M M M M
0 0 0 0 0 U U E E EF EF F FGR FP G G H H
H H H
H H HE I I R R R S
X X X X X
Genes for which automatic primer design was not successful are indicated with an asterisk. hFunctional annotation as the Cluster of Orthologous Genes (COG) database (Tatusov et al., 1997; http://www. ncbi.nlm.nih.gov/ COG).
a
10
Next Page CHAPTER 1
bioinformatics.weizmann.ac.il/blocks/).The output from BLOCKS was subsequently subjected to automatic PCR primer design by CODEHOP with parameters maintained at their default settings (Rose et al., 1998). For a total of 41 of 60 genes, a primer combination could be generated whose predicted products ranged from 400 to 1,500 bp (Table 1). The strength of an MLSA approach lies in the concatenation of multiple genes buffering against the distorting effects of recombination at a single locus and resulting in an increased resolving power. Knowing a list of best-performing genes raises the question of how many genes should be used to come up with a cost-effective consensus phylogeny. Figure 4 shows the results of an in silico MLSA analysis that is based on a random combination of 2 to 15 genes selected from the best-performing gene set. When we look at the averages of 100 repetitions per evaluated loci count, we find that each additional gene used improves the correlation with the whole-genome-based phylogeny. However, it does this in a more prominent way, by comparing the lower loci count, whereas the gain in the use of higher loci count is less pronounced. These results also show that combinations of two genes exist that are more congruent to the whole-genome-based phylogeny than some combinations of seven or more genes. However, when only a few genes are used, the buffering capac-
I
I
I
I
I
1
2
3
5
7
I
1
0
I
1
5
Number of genes
Figure 4. In silico MLSA analysis with variable loci count. For a variable number (n = 1-15) of concatenated loci (x axis) randomly selected from the set of best-performing genes (Table l), the differences in tree topology and branch lengths with the wholegenome-based phylogeny was determined and expressed as a tree distance value (y axis). For each number of loci, the procedure was repeated 100 times, and the distributions of the distances between the maximum likelihood trees are plotted as box and whisker plots.
TAXONOMY OF THE CAMPYLOBACTERACEAE
11
ity against unexpected conflicting signals in even a few lineages is too limited. Moreover, in interpreting this analysis, it should be kept in mind that this set of genes was defined by using the same set of strains whose phylogeny we are trying to estimate. Therefore, the analysis does not guarantee similar performance for an independent set of strains. Modern statistical approaches have the potential to make the choice of genes less critical by using sequence information to extract a phylogenetic signal while simultaneously identifying the genes that show discordant signals for particular stains and incorporating this information in the analysis (Didelot and Falush, 2007).
GENERAL DESCRIPTION The Family Campylobacteraceae The genera Campylobacter, Arcobacter, and Sulfurospirillum and the generically misclassified B. ureolyticus form a family of gram-negative, nonsaccharolytic bacteria with microaerobic growth requirements and a low G+C content. This definition is in complete agreement with the original criteria used by Sebald and V&on (1963) to separate several Vibrio species from the genuine vibrios, and to include the former into the new genus Campylobacter. Members of the family Campylobacteraceae occur as commensals or parasites in humans and domestic animals (Campylobacter and Arcobacter) or as free-living environmental bacteria (Arcobacter and Sulfurospirillum). The type genus is the genus Campylobacter Sebald and Vkron 1963. Members of the family Campylobacteraceae have following general characteristics. Cells are curved, S-shaped, or spiral rods that are 0.2 to 0.8 p m wide and 0.5 to 5 p m long. They are gram negative and do not form spores. Cells in old cultures may form spherical or coccoid bodies. They are typically motile, with a characteristic corkscrew-like motion performed by means of a single polar unsheathed flagellum at one or both ends of the cell. Campylobacter and Arcobacter cells grow under microaerobic conditions and have a respiratory and chemoorganotrophic type of metabolism. Some grow under aerobic or anaerobic conditions. Energy is obtained from amino acids or tricarboxylic acid cycle intermediates, not carbohydrates. Carbohydrates are neither fermented nor oxidized. The optimum growth temperature is 30 to 37°C. Typical biochemical characteristics are reduction of fumarate to succinate; negative methyl red reaction and acetoin and indole production; and for most species, reduction of nitrate, absence of hippurate hydrolysis, and presence of oxidase activity.
Previous Page 12
DEBRUYNE ET AL.
Menaquinones are the only respiratory quinones detected, with menaquinone-6 (three different structural types) and menaquinone-5 as major components. rRNA genes (16s and 23s) of some strains comprise internal transcribed spacers or intervening sequences. The presence of these internal transcribed spacers is strain specific, not species specific. The DNA base ratio ranges between 27 and 47 mol%. The Genus Campylobacter General characteristics
Campylobacter cells are mostly slender, spirally curved rods, 0.2 to 0.8 p m wide and 0.5 to 5 p m long. Cells of some species are predominantly curved or straight rods. Cells in old cultures may form coccoid bodies; these are considered degenerative forms rather than a dormant stage of the organism. Cells of most species are motile, with a characteristic corkscrew-like motion performed by means of a single polar unsheathed flagellum at one or both ends of the cell. Cells of some species are nonmotile (C. gracilis) or have multiple flagella (Campylobacter showae). In general, biochemical characteristics are as described for the family Campylobacteraceae. Several species grow in anaerobic conditions with fumarate or nitrate as electron acceptor; these species grow only in microaerobic conditions if hydrogen, formate, or succinate is supplemented as electron source. Gelatin, casein, starch, and tyrosine are not hydrolyzed. Oxidase activity is present in all species except C. gracilis. No lipase or lecithinase activity is present. Some species are pathogenic for humans and animals. This organism is found in the reproductive organs, intestinal tract, and oral cavity of humans and animals. The G + C of the DNA ranges from 29 to 47 mol%. At present, there are 17 validly named Campylobacter species, with Campylobacter fetus Sebald and VCron 1963 as the type species. There is no simple gold standard for the routine isolation of all Campylobacter species. Simultaneous application of a microaerobic atmosphere containing hydrogen with a filtration method and a selective base is methodologically the optimal solution. However, the predominant species in human infection can be readily grown in a microaerobic atmosphere on selective media without the necessity to use hydrogen. To evaluate the presence of less common species, appropriate culture conditions need to be applied. Menaquinone-6 (2-methyl-3-farnesyl-farnesyl1.4-naphthoquinone) and a methyl-substituted menaquinone-6 (2,[5 or 81-dimethyl-3-farnesyl-far-
nesyl-1.4-naphthoquinone)have been reported as major respiratory quinones in Campylobacter species. List of the species of the genus Campylobacter C . fetus. The first isolated Campylobacter was almost certainly C. fetus (McFadyean and Stockman, 1913). The species comprises two subspecies, C. fetus subsp. fetus and C. fetus subsp. venerealis, both of which are considered primary pathogens. C. fetus subsp. fetus has been associated with abortion in sheep and sporadic abortion in cattle, and it has been isolated from a wide range of sources, including ungulates, fowl, reptiles, and humans. It is considered an uncommon human pathogen. It causes systemic infection in immunodeficient patients, or, rarely, neonatal sepsis and septic abortion. C. fetus subsp. venerealis is the causative agent of bovine genital campylobacteriosis, an infectious venereal disease that can lead to infertility, abortion, and embryo death. It is mainly isolated from the bovine genital tract. Differentiation between the two subspecies is not straightforward. Biochemical differentiation is based on a single phenotypic test, i.e., tolerance to 1% glycine. It was suggested that, additionally, tolerance to 0.1% potassium permanganate, 0.005% basic fuchsin, and 64 mg literp1 cefoperazone may be useful for discrimination between the two subspecies (On and Harrington, 2001). Whole-cell protein electrophoresis, cellular fatty acid analysis, serotyping, and ribotyping do not separate the two taxa. Molecular methods used for discrimination between the two subspecies include numerical analysis of pulsedfield gel electrophoresis (PFGE) DNA profiles (On and Harrington, 2001), amplified fragment-length polymorphism (AFLP) (Duim et al., 2001), randomly amplified polymorphic DNA (RAPD) analysis (Tu et al., 2005), and several PCR tests (e.g., van Bergen et al., 2005b). C. fetus strains of reptile origin are divergent from C. fetus strains from mammalian origin, based on 16s rRNA, recA and sapD gene sequences, RAPD, PFGE, and DNA-DNA hybridization experiments. This suggests that they might represent a distinct taxon (Tu et al., 2005).
C. hyointestinalis. C. hyointestinalis was originally isolated from pigs with intestinal disorders and includes two subspecies: C. hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii. The former has been isolated from the intestines of pigs and hamsters, fecal samples from cattle, deer, and humans, and occasionally from the blood of patients with bacteremia. C. hyointestinalis subsp. hyointestinalis is associated with porcine proliferative
CHAPTER 1
enteritis and diarrhea in animals and humans, but its pathogenic role is unclear. C. hyointestinalis subsp. lawsonii has been isolated from the stomach of pigs. Its clinical significance is unknown. When grown on common blood agar bases, only some C. hyointestinalis strains grow microaerobically without hydrogen (all strains grow microaerobically in the presence of hydrogen). Both whole-cell protein and macrorestriction profiles, and AFLP analysis can be used for high-resolution identification and for determination of epidemiological relationships in this species (Duim et al., 2001; On and Harrington, 2000).
Campylobacter lanienae. During a routine screen of healthy abattoir workers, Campylobacterlike organisms were isolated from fecal samples of two individuals. Cells were slender, slightly curved rods, with single, bipolar unsheathed flagella (Logan et al., 2000). C. lanienae was also isolated from fecal samples from healthy pigs and has been detected by means of conventional and quantitative PCR in bovine fecal samples. C. sputorum. Bacteria currently classified in the species C. sputorum (On et al., 1998), were originally described as multiple Vibrio species. In 1914, Tunicliff observed a vibrio, later classified as Vibrio sputorum (PrCvot, 1940), in sputum of a patient with bronchitis. Similar bacteria isolated from the bovine vagina and semen were classified as Vibrio bubulus (Florent, 1953). These two species were thought to be related at the subspecies level on the basis of their extensive biochemical similarities and were renamed C. sputorum subsp. sputorum and C. sputorum subsp. bubulus (VCron and Chatelain, 1973). In addition, Firehammer (1965) described a fecal-type vibrio isolated from sheep that differed mainly from V. bubulus by its strong catalase activity, for which the name Vibrio fecalis was proposed. Subsequent DNA-DNA hybridization studies (Roop et al., 1985b) revealed a high level of DNA-DNA relatedness between C. sputorum subsp. sputorum, C. sputorum subsp. bubulus, and V. fecalis, and also indicated that C. sputorum subsp. mucosalis (see below) represented a distinct species. Limited biochemical variation between these taxa had been noted (catalase production, and growth on 3.5% NaCl and 1.0% ox-bile media), and therefore Roop et al. (1985b) proposed that these taxa be referred to as source-specific biovars of C. sputorum (C. sputorum bv. sputorum, C. sputorum bv. bubulus, and C. sputorum bv. fecalis). Subsequent studies by On et al. (1998) questioned the legitimacy of C. sputorurn bv. bubulus as a distinct taxon be-
TAXONOMY OF THE CAMPYLOBACTERACEAE
13
cause the tests used to distinguish this taxon from other biovars were found to be poorly reproducible. The identification of urease-positive variants of C. sputorum isolated from cattle feces (Atabay et al., 1997) resulted in the proposal of a new biovar structure, defined by reactions in two simple and reproducible tests: presence of catalase and urease activity. Strains of C. sputorum bv. sputorum produce neither catalase nor urease. They are found in the oral cavity, feces (normal and diarrheic), and abscesses and other skin lesions of humans; the genital tract of bulls; abortions in sheep; and the feces of sheep and pigs. Their pathogenicity is unknown. Strains of C. sputorum bv. faecalis produce catalase, but not urease. They have been isolated from the feces of sheep and cattle, and their pathogenicity is unknown. Finally, strains of C. sputorum bv. paraureolyticus produce urease, but not catalase. Strains are isolated from the feces of cattle and from human diarrhea; their pathogenicity is unknown.
C. mucosalis. The name C. sputorum subsp. mucosalis was originally proposed for a group of strains thought to be a causal agent of proliferative enteritis in pigs (Lawson et al., 1981). However, subsequent studies have identified Lawsonia intracellularis as the principal pathogen in this disease, and DNA-DNA hybridization demonstrated that this organism represented a distinct Campylobacter species (Roop et al., 1985a). C. mucosalis strains do not grow microaerobically on common agar bases in an atmosphere without hydrogen. Three distinct serovars, A, B, and C, have been described (Lawson et al., 1981), which were shown to have strikingly distinct whole-cell protein patterns (Costas et al., 1987; Vandamme et al., 1990). C. mucosalis strains have been isolated from the intestinal mucosa of pigs with porcine intestinal adenamatosis, necrotic enteritis, regional ileitis, and proliferative hemorrhagic enteropathy; and from the porcine oral cavity. Their clinical significance is unknown. Human infections supposedly caused by this organism were shown to be misidentified C. concisus infections (On, 1994). More recently, Tyrrell et al. (2003) reported the isolation of C. rnucosalis, identified by biochemical means, from the tongue dorsum of patients with oral malodor. C . concisus and C:curvus. Tanner et al. (1981, 1984) described two groups of strains isolated from gingival crevices of humans with gingivitis, periodontitis, and periodontosis, which were named C. concisus and C. curvus (the latter was originally described as w. cuwa; see above). Subsequent studies identified C. concisus in stool and blood samples of
14
DEBRUYNE ET AL.
children and adults with and without diarrhea, and more recently also in patients with Barrett’s esophagus, a complication of chronic gastroesophageal reflux disease. The high prevalence of C. concisus in clinical settings has been demonstrated when appropriate isolation conditions are used. Whole-cell protein electrophoresis and DNA-DNA hybridization experiments (Van Etterijck et al., 1996; Vandamme et al., 1989) revealed that C. concisus is a heterogeneous species, which has been confirmed by a range of molecular techniques, including RAPD (Engberg et al., 2005; Matsheka et al., 2006; Van Etterijck et al., 1996), ribotyping (Engberg et al., 2005), and PFGE (Matsheka et al., 2002). AFLP analysis revealed at least four distinct genomospecies within C. concisus (Aabenhus et al., 2005). At present, no clear differences have been observed between isolates from patients with diarrhea and isolates from healthy carriers; therefore, the role of C. concisus as a human enteric pathogen remains unclear. C. cuwus has been rarely isolated from human fecal samples, and it has been implicated in hepatic and bronchial abscesses. Strains of both species require hydrogen for growth in microaerobic conditions.
C. rectus, C . gracilis, and C. showae. C. rectus, C. gracilis, and C. showae are hydrogen-requiring campylobacters for which optimal growth is obtained in anaerobic conditions. Their cell morphology is most unusual. C. rectus (Tanner et al., 1981) is a rather plump, straight rod; its cellular surface is covered with a distinctive array of hexagonal, packed, macromolecular subunits. Strains have been isolated from gingival crevices of humans, in cases of appendicitis, in patients with Barrett’s esophagus, and from extraoral abscesses. It is considered a putative periodontal pathogen. It has been suggested that periodontal disease may contribute to preterm low birth weight, but because many confounding factors may influence preterm delivery, there is no conclusive evidence that periodontal pathogens like C. rectus are actually involved (Buduneli et al., 2005). C. gracilis cells are straight and nonmotile rods (Tanner et al., 1981). C. gracilis is the only oxidasenegative Campylobacter species, although the pattern of cytochromes found in C. gracilis resembled that reported for other campylobacters. Strains have been isolated from gingival crevices; from visceral, head, and neck infections; in soft tissue abscesses; in pneumonia and empyema; and in an ischial wound in humans. The association of C. gracilis with serious deep tissue infection, coupled with a high frequency of antibiotic resistance, suggests that it is an underesti-
mated pathogen, even though conclusive evidence for pathogenicity is lacking. C. showae cells are straight rods and have polar bundles of two to five unsheathed flagella (Etoh et al., 1993). Strains have been isolated from human dental plaque, from infected root canals, and from periodontitis lesions. Pathogenicity is unknown.
Campylobacter hominis. C. hominis was initially described as “Candidatus Campylobacter hominis,” an uncultivated species found in the gastrointestinal tract of healthy human individuals (Lawson et al., 1998). After modifications of the isolation methodology, the organism was subsequently isolated by anaerobic culture. Similar to C. gracilis, C. hominis cells are nonmotile straight rods without flagella (Lawson et al., 2001). C. horninis has been detected by speciesspecific PCR in both healthy and diarrheic human fecal samples. C. jejuni and C. coli. In 1927, Smith and Orcutt described a group of Vibrio-like bacteria from the feces of cattle with diarrhea. Jones and coworkers (193 1) showed a causal relationship between these microaerophilic vibrios and bovine dysentery and subsequently named the organism Vibrio jejuni. Similar organisms were detected in blood cultures of humans with gastroenteritis and aborted sheep fetuses. In 1944, Doyle isolated yet another vibrio from feces of pigs with diarrhea and classified it as Vibrio coli (Doyle, 1948). At present, C. jejuni and C. coli are by far the most important human enteropathogens among the campylobacters. C. jejuni infection is also a known antecedent of Guillain-BarrC syndrome. Within C. jejuni, two subspecies are recognized: C. jejuni subsp. jejuni and C. jejuni subsp. doylei. Strains of the latter species differ biochemically from the former by the absence of nitrate reduction, cephalothin susceptibility, and often catalase activity (a weak reaction may also be observed). Many common molecular-based methods are not suitable for subspeciation, but recently a multiplex PCR based on the nap locus was proposed for C. jejuni subspecies differentiation (Miller et al., 2007). The pathogenic role of C. jejuni subsp. doylei is unknown. Although it is isolated infrequently, it is obtained primarily from human clinical samples, and it is often associated with bacteremia, notably in infants. Biochemically, C. jejuni subsp. jejuni and C. coli differ only in their ability to hydrolyze hippurate, for which C. coli is negative. However, hippuratenegative C. jejuni subsp. jejuni strains are well recognized. Because of the limitations of biochemical differentiation, a wide range of alternative, mainly
CHAPTER 1
genotypic methods has been developed for differentiation between C. coli and C. jejuni subsp. jejuni (e.g., Jensen et al., 2005; Mandrel1 et al., 2005; Nayak et al., 2005; Sails et al., 2001). The complexity of this taxonomic area has been illustrated by data showing that strains first described as a distinct species, c. hyoilei, are in fact c. coli (Vandamme et al., 1997), despite a higher 16s rRNA sequence similarity to C. jejuni (Alderton et al., 1995). A multilocus sequence typing study of C. coli demonstrated that C. hyoilei strains belonged to the sequence type 1017 complex, which comprised the majority of the C. coli sequence types and included strains from Australia, Belgium, Canada, Denmark, Japan, and the United States and which were isolated from humans and a range of animals (Miller et al., 2005).
C. lari. C. luri was originally referred to as the group of nalidixic acid-resistant thermophilic campylobacters (the NARTC group). These strains were primarily differentiated from C. jejuni and C. coli by their resistance to nalidixic acid, their anaerobic growth in the presence of trimethylamine N-oxide hydrochloride, and later also by the absence of indoxy1 acetate hydrolysis. However, nalidixic acidsusceptible strains (NASC strains), urease-producing strains (UPTC strains), and urease-producing nalidixic acid-susceptible strains (Endtz et al., 1997) were also identified as C. luri variants by onedimensional whole-cell protein electrophoresis and semiquantitative DNA-DNA hybridization (Vandamme et al., 1991b). Endtz et al. (1997) reported a striking heterogeneity among and within the different groups of C. luri variants, which was confirmed by Duim and coworkers (2004), who clearly distinguished four subgroups by means of numerical analysis of AFLP profiles and of partial protein profiles. Preliminary DNA-DNA hybridization experiments revealed that some of these groups may represent distinct species, but at present, the taxonomic status of these groups remains unclear. C. luri strains have been isolated from intestinal contents of seagulls and other animals, river water, and shellfish. In humans, C. luri has been occasionally isolated from diarrheic feces and from cases of bacteremia and other extraintestinal infections in both immunocompetent and immunodeficient patients. Its pathogenicity is unknown. Campylobacter insulaenigae. Four Cumpylobacter-like strains were isolated from rectal swabs taken from common seals and the intestinal contents of a porpoise in northern Scotland and represented a distinct Campylobacter species, named C. insulaenigrae (Foster et al., 2004). C. insuluenigrue
TAXONOMY OF THE CAMPYLOBACTERACEAE
15
has also been isolated from free-living and stranded northern elephant seals in California. Although initially reported as nonthermotolerant, several of the Californian C. insuluenigrue strains were able to grow at 42"C, leading to an initial misidentification of some of these strains as C. lari (Stoddard et al., 2007). The pathogenic potential of these bacteria in marine mammals is unknown.
Campylobacter canadensis. During a study of the bacterial flora of the cloacae of whooping cranes (Grus umericuna), 10 atypical Campylobacter isolates were recovered on two separate occasions and were classified as C. canadensis. The cells were polymorphic in shape (sigmoid to coccoid) and possessed a single polar flagellum. The prevalence of C. cunudensis and its role are at present unknown (Inglis et al., 2007). C. upsaliensis and Campylobacter helveticus. Both C. upsaliensis and C. helveticus are catalasenegative or weakly positive thermotolerant campylobacters. C. upsuliensis strains have been isolated from feces of enteric and asymptomatic humans, dogs, cats, and meercats as well as from human blood samples. Strains associated with a human abortion, a breast abscess, hemolytic-uremic syndrome, and a prosthetic knee infection have been described. C. helveticus has been isolated from feline and canine fecal samples, but at present, no strains have been isolated from human sources. Reports have emerged worldwide implicating C. upsuliensis as a human enteric pathogen, and in some regions, it is preceded only by C. jejuni in being the most frequently isolated Cumpylobucter species in human stool samples. Accumulated clinical and epidemiological data support the importance of C. upsuliensis as a primary human pathogen, even though direct evidence is at present still lacking. Differentiation of Campylobacter species Apart from the recent single-locus, multilocus, or whole-genome-based analyses discussed above, a variety of approaches have been used to identify campylobacters to the species level. Some of these are discussed below. Classical phenotypic characteristics. Most routine laboratories use basic biochemical tests to identify campylobacters. The lack of application of highly standardized procedures and the well-known biochemical inertness of campylobacters often make biochemical identification of these bacteria difficult be-
16
DEBRUYNE ET AL.
cause the discrimination among species or subspecies may rely on one differential character, such as presence of hippuricase or urease activity (On, 1996). A computerized scheme of nearly 70 biochemical tests readily differentiated virtually all Campylobacter species (On et al., 1996; On and Holmes, 1995). Traditional taxonomic analyses. In taxonomic studies, the reference method for the delineation of bacterial species is determination of the level of DNA-DNA hybridization (Vandamme et al., 1996). A variety of DNA-DNA hybridization studies have been performed before the phylogenetic relationships of campylobacters were established by means of rRNA-directed studies and focused on the species known since the early 1980s (detailed overviews are given by Vandamme and Goossens [1992] and On [1996]). Chemotaxonomic techniques such as wholecell protein and fatty acid analysis have been used in a range of taxonomic studies for the identification of Campylobacter species. In particular, electrophoresis of whole-cell proteins by highly standardized sodium dodecyl sulfate-polyacrylamide gel electrophoresis has proven to be extremely reliable to identify Campylobacter and other strains at the species level because numerous studies have revealed a correlation between high similarity in whole-cell protein content and level of DNA-DNA hybridization (On, 1996; Vandamme et al., 1996). rRNA-based approaches. Sequence analysis of the rRNA genes became the ultimate tool to study bacterial phylogeny during the 1980s, and international databases have been constructed to collect sequences and to make them publicly available. Nowadays, it is common to identify unknown organisms that are of particular interest by means of 16s rRNA sequence analysis and comparison with rRNA sequences present in international databases (Gorkiewicz et al., 2003). Several important points should be addressed here. Although unsurpassed in its capacity to reveal the phylogenetic neighborhood of an unknown bacterium, comparison of entire 16s rRNA sequences is generally not adequate for the identification of strains to the species level. It has been reported that strains belonging to different species may have identical 16s rRNA sequences and that strains of one species may have 16s rRNA genes that differ up to 3% (Stackebrandt and Goebel, 1994). Harrington and On (1999) demonstrated differences of up to 4.3% of the total 16s rRNA sequence between C. hyointestinalis strains. There is definitely a lack of knowledge, not only of the strain-to-strain variation within a species, but also of the interoperon variation within a single
strain as reported by Clayton et al. (1995). Therefore, concluding that an unknown organism belongs to a particular species because it shares a high percentage of its 16s rRNA gene sequence with another strain of that species, or concluding that an unknown represents a novel species because it shares only 97% of its 16s rRNA gene sequence with its closest neighbor is premature in the absence of other data. Alternatively, sequence information derived from rRNA cistrons has been used successfully to design and validate many species-, group-, or genusspecific primers and probes (Liibeck et al., 2003; On, 1996). Applied in a PCR or a hybridization assay, these primers and probes offer valuable alternatives for identification of Campylobacter taxa. Finally, broad-spectrum molecular identification schemata that are based on restriction fragment analysis of PCR amplicons derived from 16s (Cardarelli-Leite et al., 1996; Marshall et al., 1999) and 23s (Gonzalez et al., 2006; Hurtado and Owen, 1997) rRNA genes have also been described. AFLP analysis. AFLP analysis is based on selective amplification of restriction fragments of chromosomal DNA and has been widely used in a variety of bacterial genera, including Campylobacter, both for identification and typing purposes. It is a reproducible and discriminatory technique, particularly well suited for computer-assisted data analysis. In Campylobacter, AFLP schemes were initially developed for typing purposes, incorporating only C. jejuni subsp. jejuni and C. coli, and two studies revealed clear differences in AFLP profiles between these two species (Duim et al., 1999; Kokotovic and On, 1999). These observations were extended to other Campylobacter species (Duim et al., 2001; On and Harrington, 2000) and demonstrated that numerical analysis of AFLP profiles is well suited for identification at the species, subspecies, and strain levels. Numerous subsequent studies have applied AFLP to explore intraspecies diversity (e.g., Aabenhus et al., 2005; Duim et al., 2004) and epidemiological relationships (e.g., Keller et al., 2007; Siemer et al., 2005) of Campylobacter strains. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS). In
MALDI-TOF MS, biological samples are mixed with a matrix that is chosen such that it specifically absorbs the energy of a laser beam. The high-energy impact is followed by the formation of ions that are extracted through an electric field; these are subsequently focused and detected as a m / z (mass/charge) spectrum. In a preliminary study, Winkler et al. (1999) analyzed a set of C. jejuni, C. coli, C. fetus,
CHAPTER 1
H. pylori, and H. mustelae strains by means of both whole cells and partially purified extracts from lysed cells, and they reported the detection of discriminatory biomarker ions in the 10- to 20-kDa range. Further research into the potential of MALDI-TOF MS for identification purposes was performed by Mandrell and coworkers (2005). They analyzed whole bacterial cells from a large collection of well-characterized reference strains from C. coli, C. jejuni, C. helveticus, C. lari, C. sputorum, and C. upsaliensis, and they observed species-identifying biomarker ions in the 9- to 14-kDa range for each of the six species. Most difficulty was observed for differentiating the closely related C. upsaliensis and C. helveticus. The Genus Arcobacter General characteristics
Arcobacter cells are slender, spirally curved rods 0.2 to 0.9 pm wide and 0.5 to 3 p m long; S-shaped or helical cells are often present. Cells in old cultures may form spherical or coccoid bodies and loose spiral filaments up to 20 p m long. Arcobacter is motile, with a characteristic corkscrew-like motion performed by means of a single polar unsheathed flagellum at one or both ends of the cell. In general, biochemical characteristics are as described for the family Campylobacteraceae. In addition, arcobacters are able to grow in aerobic and anaerobic conditions and at 15°C. Hydrogen is not required. Indoxyl acetate is hydrolyzed. No hydrolysis of gelatin, casein, starch, hippurate, and tyrosine is present. Habitat is extremely diverse. Arcobacter is unique among the Epsilonproteobacteria in that it combines plant and animal associated species and free-living environmental species within one genus. Some species are associated with humans and animals and are found in the reproductive organs and abortions of various animals and in the intestinal tract of humans and animals. Other sources include water reservoirs, river and surface water, sewage, oil field communities, marine and hypersaline environments, and deep-sea hydrothermal vents; one species is a plant-associated nitrogen fixer. The G+C of the DNA ranges from 27 to 35 mol%. Menaquinone-6 (2 -methyl-3 -farnesyl-farnesyl- lY4-naphthoquinone) and a second menaquinone-6, the detailed structure of which has not been determined, were reported as major respiratory quinones. The genus Arcobacter presently comprises six validly named species, with Arcobacter nitrofigilis Vandamme, Falsen, Rossau, Hoste, Segers, Tytgat, and De Ley 1991 as the type species.
TAXONOMY OF THE CAMPYLOBACTERACEAE
17
Isolation procedures for Arcobacter spp. are often based on filtration methods or selective media developed for Campylobacter, Yersinia, or even Leptospira species (Vandamme, 2000). In addition, several isolation methods were developed specifically for Arcobacter spp. (deBoer et al., 1996; Houf et al., 2001a; Johnson and Murano, 1999a, 1999b) by using a variety of selective media, in some cases preceded by an enrichment step. However, several of these methods are suboptimal because arcobacters may be overgrown by campylobacters present in the same specimens, and they may be susceptible to some of the antimicrobial agents used in the selective media (Houf et al., 2001b). In 2001, an Arcobacter enrichment broth and Arcobacter selective agar were developed for the isolation of arcobacters from poultry products, which allowed the recovery of A. butzleri, Arcobacter cryaerophilus, and A. skirrowii and successfully inhibited the contaminating flora (Houf et al., 2001a). In recent years, this procedure was validated for several other matrices, including animal feces (Van Driessche et al., 2003), human feces (Houf and Stephan, 2007), and poultry intestinal content (Van Driessche and Houf, 2007). Species of the genus Arcobacter A. nitrofigilis. A. nitrofigilis strains occur in the roots and rhizosphere of the salt marsh plant Spartina alternifiora, where they fix nitrogen (McClung et al., 1983). They differ from most other members of the family Carnpylobacteraceae by their preference for a high salt concentration (optimum growth is in the presence of 10 to 40 g/liter NaCl), and by the presence of nitrogenase and urease activity. A. cryaerophilus. In 1985, Neil1 and coworkers performed an extensive phenotypic characterization of aerotolerant Campylobacter strains isolated from various animal sources. They concluded that the aerotolerant strains were only distantly related to strains of the other Campylobacter species examined and emphasized that they formed a heterogeneous group. Their findings were confirmed by integrated phenotypic, genotypic, and chemotaxonomic studies of, in part, the same isolates (Kiehlbauch et al., 1991; Vandamme et al., 1991a, 1992), which entailed in the description of a novel genus Arcobacter, with three separate animal-associated species: A. cryaerophilus, A. butzleri, and A. skirrowii. Within A. cryaerophilus, two subgroups, referred to as subgroup 1 or group 1A and subgroup 2 or group lB, have been described (Kiehlbauch et al., 1991; Vandamme et al., 1992). Strains of these subgroups differ in their whole-cell protein and fatty acid
18
DEBRUYNE ET AL.
patterns, and in the restriction fragment length polymorphisms of the rRNA genes. Moderate levels of DNA-DNA hybridization were detected between strains of the different subgroups, suggesting that they could be given a distinct taxonomic rank. However, at present, no formal nomenclatural modifications have been proposed. A. cryaerophilus strains have been isolated from cases of human bacteremia and diarrhea; from chicken, pork, and duck carcasses; from bovine, ovine, and porcine abortions; from fecal samples from cattle, pigs, and horses; and from mastitis in cattle. One study reported its isolation from a naturally infected rainbow trout. Their pathogenicity is unknown. A. butzleri. A. butzleri strains have been isolated from human blood and diarrheic feces; from feces of various animals with diarrhea including nonhuman primates, pigs, horses, cattle, an ostrich and a tortoise; from bovine and porcine abortions; from various food products including ground pork, chicken, and turkey; from surface and drinking water reservoirs and canal waters; from seawater; and from plankton. This species is associated with enteritis, abdominal cramps, bacteremia, and appendicitis in humans and with enteritis and abortion in animals. It was identified as the fourth most isolated Campylobacter-like organism from human diarrheal stool in Belgium and France (Prouzet-Mauleon et al., 2006; Vandenberg et al., 2004). Screening for arcobacters in asymptomatic humans yielded no positive results for A. butzleri (Houf and Stephan, 2007).
A. skirrowii. A. skirrowii strains have been isolated from preputial fluids of bulls; from bovine, ovine, and porcine abortions; from diarrheic feces of various animals including sheep and cattle; and from chicken carcasses. At present, there is one report on the isolation of A. skirrowii from a patient with chronic diarrhea, but it was unclear whether the strain was the etiological agent (Wybo et al., 2004). Their pathogenicity remains unknown. Arcobacter cibarius. A. cibarius strains were first isolated during a long-term study of Arcobacter contamination on poultry carcasses. Twenty isolates, recovered from 13 unrelated broiler carcasses, yielded no Arcobacter species-specific PCR amplicon, but they did yield a genus-specific PCR amplicon (Houf et al., 2005). Additional data demonstrated that the 20 isolates represented a novel Arcobacter species. The organism has also been isolated from porcine and ovine feces (K. Houf, unpublished data) and pig-
gery effluent (Chinivasagam et al., 2007). At present, their pathogenic potential is unknown.
Arcobacter halophilus. A. halophilus is an obligately halophilic Arcobacter species isolated from a hypersaline lagoon on Laysan Atoll in the northwestern Hawaiian Islands, and currently comprises a single strain. Phenotypic and genotypic data placed the type strain LA3 lBT within the genus Arcobacter, but it can be distinguished from all recognized arcobacters by its obligate halophily and fatty acid composition (Donachie et al., 2005). “Candidatus Arcobacter sulfidicus.” “Candidatus A. sulfidicus” is a highly motile sulfide-oxidizing bacterium, unique among prokaryotes in that it excretes elemental sulfur in the form of hydrophilic rigid filaments, as opposed to globular or amorphous sulfur found with other sulfur oxidizers (Wirsen et al., 2002). The organism currently remains a candidate species that is difficult to cultivate, not because of the nutritional aspects of cultivation, but because of the required growth apparatus and conditions (Wirsen et al., 2002). The formation of filamentous sulfur appears to be a specific adaptation permitting these organisms to colonize the specialized niche, characterized by dynamic fluid movement and high sulfide and low oxygen concentrations, required for their survival. The worldwide occurrence of microbial filamentous sulfur formation in marine ecosystems indicates that it is likely to play an important yet undefined role in global sulfur and carbon cycles (Sievert et al., 2007). As opposed to the chemoorganotrophic growth found in other Arcobacter spp., this organism has a chemoautotrophic mode of nutrition. Experiments demonstrated that it has a microaerobic growth preference and that it aggregates within the oxic-anoxic interface by exhibiting a chemotactic response (Sievert et al., 2007). Other Organisms On and coworkers (2002, 2003) reported the isolation of five strains associated with pig abortions and one from duck feces, which were believed to represent a novel Arcobacter species, as assessed by AFLP profiling, phenotype, and 16s rRNA gene sequence analysis. However, formal description was deferred until the identification of at least 10 isolates, in accordance with recommendations for minimal taxonomic standards for this group of bacteria (Ursing et al., 1994). Also, a number of culture-independent studies have documented the existence of additional
CHAPTER 1
-
TAXONOMY OF THE CAMPYLOBACTERACEAE
19
Arcobacter taxa, as assessed by 16s rRNA gene phylogeny, in environmental samples including oil field communities (Gevertz et al., 2000; Sette et al., 2007; Voordouw et al., 1996), cyanobacterial mats (Teske et al., 1996), activated sludge of a large municipal wastewater treatment plant (Snaidr et al., 1997), and deep-sea hydrothermal vents (Moussard et al., 2006). Other potentially novel species have been detected in the tunic matrix of the colonial ascidian Diplosoma migrans (Schuett et al., 2005) and as part of the Chilean oyster (Tiostrea chilensis) microbiota (Romero et al., 2002). Because none of these novel taxa was cultivated, they were not formally named.
PCR for identification of A. butzleri and A. cryaerophilus (Brightwell et al., 2007). As for the aforementioned PCR assays, A. cibarius was not included in the evaluation of these DNA-based identification methods. The nucleotide sequences of the gyrA gene for A. butzleri, A. cryaerophilus, A. cibarius, and A. skirrowii have been determined. Phylogenetic analysis of the generated gyrA sequences provided a result similar to 16s rRNA gene phylogeny. Phylogenetic analysis allowed a good separation of these species from each other and from related Campylobacter and Helicobacter species (Abdelbaqi et al., 2007).
Differentiation of Arcobacter Species
The Genus Sulfurospirillum
Basic biochemical tests are still routinely used to identify arcobacters to the species level, and, like for campylobacters, this is hampered by their biochemical inertness and intraspecies variability. In particular, the separation of A. cryaerophilus subgroup 2 and A. butzleri is tedious (Vandamme et al., 1992). As mentioned for Campylobacter, numerical analysis of a large number of phenotypic characteristics (On et al., 1996; On and Holmes, 1995) and whole-cell protein electrophoresis can also be used for Arcobacter identification. Whole-cell fatty acid analysis did not allow discrimination between A. butzleri from A. cryaerophilus subgroup 2, but this technique did permit differentiation of all other Arcobacter taxa (Vandamme and Goossens, 1992). Several PCR assays for the discrimination of animal-associated Arcobacter species have been developed in the last decade, often in the form of a multiplex reaction (Brightwell et al., 2007; Houf et al., 2000; Kabeya et al., 2003). At present, no PCR assay has been developed that included A. cibarius. Restriction fragment length polymorphism analysis of a PCR-amplified fragment of the gene coding for 16s rRNA (Cardarelli-Leite et al., 1996) or 23s rRNA (Hurtado and Owen, 1997) differentiated A. butzleri from other arcobacters but did not differentiate A. cryaerophilus from A. skirrowii. A PCR-restriction fragment length polymorphism analysis that used either a 16s or 23s rRNA amplicon in combination with only one restriction enzyme permitted genuslevel discrimination between Arcobacter, Campylobacter, and Helicobacter (Gonzalez et al., 2006). Numerical analysis of AFLP profiles allowed discrimination of four Arcobacter species (i.e., A. butzleri, A. cryaerophilus, A. skirrowii and A. nitrofigilis) but did not include A. cibarius or A. halophilus (On et al., 2003). Other DNA-based methods include 16s rRNA-based DNA probes (Wesley et al., 1995), ribotyping (Kiehlbauch et al., 1994), and real-time
The genus Sulfurospirillum was established by Schumacher et al. (1992) to accommodate two freeliving Campylobacter-like strains (Spirillum 5 17.ST and Campylobacter sp. “Veldkamp”), which formed a distinct clade within the rRNA superfamily VI (Vandamme et al., 1991a). At present, there are seven named Sulfurospirillum species, with Sulfurospirillum deleyianum (Wolfe and Pfennig 1977) Schumacher, Kroneck, and Pfennig 1993 as the type species. Strain Spirillum 5 175T, isolated from anoxic mud from a forest pond near Heiningen, Germany, was assigned as the type strain of the species Sulfurospirillum deleyianum. Sulfurospirillum arcachonense was isolated from oxidized surface sediment in an intertidal mud flat near Arcachon, France (Finster et al., 1997). Stolz et al. (1999) described Sulfurospirillum barnesii, formerly referred to as “Geospirillum barnesii,” and Sulfurospirillum arsenophilum. The type strain of S. barnesii, SES3T, was isolated from a selenate-contaminated freshwater lake in western Nevada, USA (Oremland et al., 1994), and is able to use selenate and arsenate as terminal electron acceptors. The type strain for S. arsenophilum, MIT-13T, was isolated from arsenic-contaminated watershed sediments in eastern Massachusetts, USA (Ahmann et al., 1994), and is able to reduce arsenate. Sulfurospirillum halorespirans PCE-M2T, isolated from a polluted anaerobic soil, and Sulfurospirillum multivorans KT (formerly Dehalospirillum multivorans), isolated from activated sludge, both have the ability to reduce the pollutant tetrachloroethylene (Luijten et al., 2004; Scholzmuramatsu et al., 1995). Sulfurospirillum cavolei Phe9 lTwas isolated from petroleum-contaminated groundwater in an underground crude oil storage cavity in Japan (Kodama et al., 2007). Sulfurospirilla are extremely fastidious, and very few strains have been isolated and studied. Speciation is primarily based on differences in 16s rRNA gene
20
DEBRUYNE ET AL.
sequences, although there are obvious differences in the metabolism of these organisms too. Cells of all of these species are slender, spirally curved rods 0.1 to 0.5 p m wide and 1.0 to 3 p m in length. They may form helical chains of two or more cells. As far as the data available indicate, their general biochemical profiles conform to that of the family Campylobacteraceae. All species show oxidase activity. Growth occurs between 8 and 36°C. The G+C of the DNA ranges from 32 to 42 mol%. Menaquinone-6, a methyl-substituted menaquinone-6 (referred to as thermoplasmaquinon-6), and menaquinone-5 have been reported as respiratory quinones in S. deleyianum; menaquinone-6 and thermoplasmaquinon-6 were found in S. arcachonense. Members of this genus are all sulfur reducers but exhibit metabolic versatility. Different isolation procedure have been described, all of which involve enrichment procedures (Finster et al., 1997; Oremland et al., 1994; Wolfe and Pfennig, 1977). Several other presently unnamed organisms also belong to the genus Sulfurospirillum. Coleman et al. (1993) isolated a marine sulfur-reducing bacterium, designated strain SM-5, implicated in the reduction of Fe(II1). A nitrate-reducing Sulfurospirillum sp., strain KW, was isolated from an oil field (Hubert and Voordouw, 2007). Strain “Veldkamp” (DSM 806), isolated from activated sludge (Laanbroek et al., 1977), was originally classified as a Sulfurospirillum sp. (Schumacher et al., 1992) and was later reclassified as S. deleyianum, but phylogenetic analysis of the 16s rRNA gene sequence of this strain indicates that it may represent a separate species (Fig. 1). Strain NP4, isolated from groundwater with high arsenic concentrations, is classified as a Sulfurospirillum species on the basis of 16s rRNA gene sequence analysis, and it differs from the other Sulfurospirillurn isolates in its carbon source and electron acceptor usage profiles (MacRae et al., 2007). This organism may adversely affect water quality by reducing arsenate to the more toxic arsenite.
Bacteroides ureolyticus B. ureolyticus (Jackson and Goodman, 1978) cells are nonmotile rods that grow microaerobically when hydrogen is provided. Representative strains of this species were included in a polyphasic taxonomic study to elucidate their taxonomic status (Vandamme et al., 1995). B. ureolyticus resembles campylobacters in respiratory quinone content, DNA base ratio, and most of its phenotypic characteristics, but differs from campylobacters in fatty acid composition and proteolytic metabolism. This organism was, however, not formally reclassified pending a thorough taxo-
nomic characterization of additional B. ureolyticuslike bacteria (Vandamme et al., 1995). Strains have been isolated from superficial ulcers and soft tissue infections, urethritis, vaginosis, and periodontal disease. A pathogenic role is suggested by its predominance in mixed infections and its strong proteolytic activity, which may enable tissue destruction. Acknowledgments. We are indebted to the Fund for Scientific Research-Flanders (Belgium) for research grants (P.V.) and for a position as a postdoctoral fellow (D.G.).
REFERENCES Aabenhus, R., S. L. On, B. L. Siemer, H. Permin, and L. P. Andersen. 2005. Delineation of Campylobacter concisus genomospecies by amplified fragment length polymorphism analysis and correlation of results with clinical data. J. Clin. Microbiol. 43: 5091-5096. Abdelbaqi, K., A. Menard, V. Prouzet-Mauleon, F. Bringaud, P. Lehours, and F. Megraud. 2007. Nucleotide sequence of the gyrA gene of Arcobacter species and characterization of human ciprofloxacin-resistant clinical isolates. FEMS Immunol. Med. Microbiol. 49:337-345. Ahmann, D., A. L. Roberts, L. R. Krumholz, and F. M. Morel. 1994. Microbe grows by reducing arsenic. Nature 371:750. Alderton, M. R., V. Korolik, P. J. Coloe, F. E. Dewhirst, and B. J. Paster. 1995. Campylobacter hyoilei sp. nov., associated with porcine proliferative enteritis. Int. J. Syst. Bacteriol. 45:61-66. Atabay, H. I., J. E. Corry, and S. L. On. 1997. Isolation and characterization of a novel catalase-negative, urease-positive Campylobacter from cattle faeces. Lett. Appl. Microbiol. 2459-64. Benjamin, J., S . Leaper, R. J. Owen, and M. B. Skirrow. 1983. Description of Campylobacter laridis, a new species comprising the nalidixic-acid resistant thermophilic Campylobacter (NARTC) group. Cuw. Microbiol. 8:231-238. Brightwell, G., E. Mowat, R. Clemens, J. Boerema, D. J. Pulford, and S. L. On. 2007. Development of a multiplex and real time PCR assay for the specific detection of Arcobacter butzleri and Arcobacter cryaerophilus.J. Microbiol. Methods 68:318-325. Bryans, J. T., A. G. Smith, and A. J. Baker. 1960. Ovine vibrionic abortion caused by a new variety of Vibrio. Cornell Vet. 505459. Bryner, J. H., J. Littleton, C. Gates, C. A. Kirkbride, and A. E. Ritchie. 1986. Flexispira rappini gen. nov., sp. nov., a Gramnegative rod from mammalian fetus and feces, p. 11-18. Microbe 86, Abstr. XIV Int. Cong. Microbiol., Manchester, England. Buduneli, N., H. Baylas, E. Buduneli, 0. Turkoglu, T. Kose, and G. Dahlen. 2005. Periodontal infections and pre-term low birth weight: a case-control study. J. Clin. Periodontol. 32:174-181. Butzler, J. P. 2004. Campylobacter, from obscurity to celebrity. Clin. Microbiol. Infect. 105368-876. Butzler, J. P., P. Dekeyser, M. Detrain, and F. Dehaen. 1973. Related vibrio in stools. J. Pediatr. 82:493-495. Cardarelli-Leite, P., K. Blom, C. M. Patton, M. A. Nicholson, A. G. Steigenvalt, S. B. Hunter, D. J. Brenner, T. J. Barrett, and B. Swaminathan. 1996. Rapid identification of Campylobacter species by restriction fragment length polymorphism analysis of a PCR-amplified fragment of the gene coding for 16s rRNA. J. Clin. Microbiol. 34:62-67.
CHAPTER 1
Castresana, J. 2000. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol. Biol. Evol. 17540-552. Chinivasagam, H. N., B. G. Corney, L. L. Wright, I. S. Diallo, and P. J. Blackall. 2007. Detection of Arcobacter spp. in piggery effluent and effluent-irrigated soils in southeast Queensland. J. Appl. Microbiol. 103:418-426. Clayton, R. A., G. Sutton, P. S. Hinkle, C. Bult, and C. Fields. 1995. Intraspecific variation in small-subunit ribosomal RNA sequences in GenBank: why single sequences may not adequately represent prokaryotic taxa. lnt. J. Syst. Bacteriol. 45595-599. Coleman, M. L., D. B. Hedrick, D. R. Lovley, D. C. White, and K. Pye. 1993. Reduction of Fe(II1) in sediments by sulphatereducing bacteria. Nature (London) 361:436-438. Costas, M., R. J. Owen, and P. J. H. Jackman. 1987. Classification of Campylobacter sputorum and allied campylobacters based on numerical analysis of electrophoretic protein patterns. Syst. Appl. Microbiol. 9: 125-13 1. Dagan, T., and W. Martin. 2006. The tree of one percent. Genome Biol. 7:118. deBoer, E., J. J. H. C. Tilburg, D. L. Woodward, H. Lior, and W. M. Johnson. 1996. A selective medium for the isolation of Arcobacter from meats. Lett. Appl. Microbiol. 23:64-66. Dekeyser, P., J. P. Butzler, J. Sternon, and M. Gossuind. 1972. Acute enteritis due to related Vibrio-first positive stool cultures. 1.Infect. Dis. 125:390-392. Didelot, X., and D. Falush. 2007. Inference of bacterial microevolution using multilocus sequence data. Genetics 175:12511266. Dingle, K. E., F. M. Colles, D. Falush, and M. C. Maiden. 2005. Sequence typing and comparison of population biology of Campylobacter coli and Campylobacter jejuni. J. Clin. Microbiol. 43: 340-347. Dingle, K. E., F. M. Colles, D. R. Wareing, R. Ure, A. J. Fox, F. E. Bolton, H. J. Bootsma, R. J. Willems, R. Urwin, and M. C. Maiden. 2001. Multilocus sequence typing system for Campylobacter jejuni. ]. Clin. Microbiol. 39:14-23. Donachie, S. P., J. P. Bowman, S. L. On, and M. Alam. 2005. Arcobacter halophilus sp. nov., the first obligate halophile in the genus Arcobacter. lnt. J. Syst. Evol. Microbiol. 55:1271-1277. Doolittle, W. F., and R. T. Papke. 2006. Genomics and the bacterial species problem. Genome Biol. 7: 116. Doyle, L. P. 1948. The etiology of swine dysentery. Am. J. Vet. Res. 9:50-51. Duim, B., P. A. Vandamme, A. Rigter, S. Laevens, J. R. Dijkstra, and J. A. Wagenaar. 2001. Differentiation of Campylobacter species by AFLP fingerprinting. Microbiology 147:2729-2737. Duim, B., J. A. Wagenaar, J. R Dijkstra, J. Goris, H. P. Endtz, and P. A. Vandamme. 2004. Identification of distinct Campylobacter lari genogroups by amplified fragment length polymorphism and protein electrophoretic profiles. Appl. Environ. Microbiol. 70:18-24. Duim, B., T. M. Wassenaar, A. Rigter, and J. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with amplified fragment length polymorphism fingerprinting. Appl. Environ. Microbiol. 65:236923 75. Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 3217921797. Endtz, H. P., J. S. Vliegenthart, P. Vandamme, H. W. Weverink, N. P. van den Braak, H. A. Verbrugh, and A. van Belkum. 1997. Genotypic diversity of Campylobacter lari isolated from mussels and oysters in The Netherlands. Int. J. Food Microbiol. 34:7988.
TAXONOMY OF THE CAMPYLOBACTERACEAE
21
Engberg, J., D. D. Bang, R. Aabenhus, F. M. Aarestrup, V. Fussing, and P. Gerner-Smidt. 2005. Campylobacter concisus: an evaluation of certain phenotypic and genotypic characteristics. Clin. Microbiol. Infect. 11:288-295. Escherich, T. 1886. Beitrage zur Kenntniss der Darmbacterien. 111. Uber das Vorkommen van Vibrionen im Darmcanal und den Stuhlgangen der Sauglinge. Miinch. med. Wschr. 33:815-817, 833-835. Etoh, Y., F. E. Dewhirst, B. J. Paster, A. Yamamoto, and N. Goto. 1993. Campylobacter showae sp. nov., isolated from the human oral cavity. lnt. J. Syst. Bacteriol. 43:631-639. Finster, K., W. Liesack, and B. J. Tindall. 1997. Sulfurospirillum arcachonense sp. nov., a new-microaerophilic sulfur-reducing bacterium. Int. J. Syst. Bacteriol. 47:1212-1217. Firehammer, B. D. 1965. The isolation of vibrios from ovine feces. Cornell Vet. 55:482-494. Florent, A. 1953. Isolement d’un vibrion saprophyte du sperme du taureau et du vagin de la vache (Vibrio bubulus). C. R. SOC. Biol. 147~2066-2069. Foster, G., B. Holmes, A. G. Steigerwalt, P. A. Lawson, P. Thorne, D. E. Byrer, H. M. Ross, J. Xerry, P. M. Thompson, and M. D. Collins. 2004. Campylobacter insulaenigrae sp. nov., isolated from marine mammals. lnt. J. Syst. Evol. Microbiol. 54:23692373. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Fox, J. G., T. Chilvers, C. S. Goodwin, N. S. Taylor, P. Edmonds, L. I. Sly, and D. J. Brenner. 1989. Campylobacter mustelae, a new species resulting from the elevation of Campylobacter pylori subsp. mustelae to species status. Int. J. Syst. Bacteriol. 39:301303. Gebhart, C. J., P. Edmonds, G. E. Ward, H. J. Kurtz, and D. J. Brenner. 1985. “Campylobacter hyointestinalis” sp. nov.: a new species of Campylobacter found in the intestines of pigs and other animals. J. Clin. Microbiol. 21:715-720. Gevers, D., F. M. Cohan, J. G. Lawrence, B. G. Spratt, T. Coenye, E. J. Feil, E. Stackebrandt, Y. Van de Peer, P. Vandamme, F. L. Thompson, and J. Swings. 2005. Opinion: re-evaluating prokaryotic species. Nut. Rev. Microbiol. 3:733-739. Gevertz, D., A. J. Telang, G. Voordouw, and G. E. Jenneman. 2000. Isolation and characterization of strains CVO and FWKOB, two novel nitrate-reducing, sulfide-oxidizing bacteria isolated from oil field brine. Appl. Environ. Microbiol. 66:24912501. Gonzalez, A., Y. Moreno, R. Gonzalez, J. Hernandez, and M. A. Ferrus. 2006. Development of a simple and rapid method based on polymerase chain reaction-based restriction fragment length polymorphism analysis to differentiate Helicobacter, Campylobacter, and Arcobacter species. Curr. Microbiol. 53:416-421. Goodwin, C. S., J. A. Armstrong, T. Chilvers, M. Peters, M. D. Collins, L. Sly, W. Mcconnell, and W. E. S. Harper. 1989. Transfer of Campylobacter pylori and Campylobacter mustelae to Helicobacter gen. nov. as Helicobacter pylori comb. nov. and Helicobacter mustelae comb. nov., respectively. lnt. J. Syst. Bacteriol. 39:397-405. Goris, J., K. T. Konstantinidis, J. A. Klappenbach, T. Coenye, P. Vandamme, and J. M. Tiedje. 2007. DNA-DNA hybridization values and their relationship to whole-genome sequence similarities. lnt. J. Syst. Evol. Microbiol. 57:81-91.
22
DEBRUYNE ET AL.
Gorkiewicz, G., G. Feierl, C. Schober, F. Dieber, J. Kofer, R. Zechner, and E. L. Zechner. 2003. Species-specific identification of campylobacters by partial 16s rRNA gene sequencing. J. Clin. Microbiol. 4 1:2537-2546. Guindon, S., and 0. Gascuel. 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 52:696-704. Harrington, C. S., and S. L. W. On. 1999. Extensive 16s rRNA gene sequence diversity in Campylobacter hyointestinalis strains: taxonomic and applied implications. Int. J. Syst. Bacteriol. 49: 1171-1 175. Henikoff, S., J. G. Henikoff, W. J. Alford, and S . Pietrokovski. 1995. Automated construction and graphical presentation of protein blocks from unaligned sequences. Gene 163:GC17-26. Houf, K., L. A. Devriese, L. De Zutter, J. Van Hoof, and P. Vandamme. 2001a. Development of a new protocol for the isolation and quantification of Arcobacter species from poultry products. Int. J. Food Microbiol. 71:189-196. Houf, K., L. A. Devriese, L. De Zutter, J. Van Hoof, and P. Vandamme. 2001b. Susceptibility of Arcobacter butzleri, Arcobacter cyaerophilus, and Arcobacter skirrowii to antimicrobial agents used in selective media. J. Clin. Microbiol. 39:1654-1656. Houf, K., S. L. On, T. Coenye, J. Mast, J. Van Hoof, and P. Vandamme. 2005. Arcobacter cibarius sp. nov., isolated from broiler carcasses. Int. J. Syst. Evol. Microbiol. 55:713-717. Houf, K., and R. Stephan. 2007. Isolation and characterization of the emerging foodborn pathogen Arcobacter from human stool. 1.Microbiol. Methods 68:408-413. Houf, K., A. Tutenel, L. De Zutter, J. Van Hoof, and P. Vandamme. 2000. Development of a multiplex PCR assay for the simultaneous detection and identification of Arcobacter butzleri, Arcobacter cryaerophilus and Arcobacter skirrowii. FEMS Microbiol. Lett. 193:89-94. Hubert, C., and G. Voordouw. 2007. Oil field souring control by nitrate-reducing Sulfurospirillum spp. that outcompete sulfatereducing bacteria for organic electron donors. Appl. Environ. Microbiol. 73:2644-2652. Hurtado, A., and R. J. Owen. 1997. A molecular scheme based on 23s rRNA gene polymorphisms for rapid identification of Campylobacter and Arcobacter species. J. Clin. Microbiol. 35:24012404. Inglis, G. D., B. M. Hoar, D. P. Whiteside, and D. W. Morck. 2007. Campylobacter canadensis sp. nov., from captive whooping cranes in Canada. Int. J. Syst. Evol. Microbiol. 57:26362644. Jackson, F. L., and Y. E. Goodman. 1978. Bacteroides ureolyticus, a new species to accommodate strains previously identified as Bacteroides corrodens, anaerobic. Int. J. Syst. Bacteriol. 28: 197200. Jensen, A. N., M. T. Andersen, A. Dalsgaard, D. L. Baggesen, and E. M. Nielsen. 2005. Development of real-time PCR and hybridization methods for detection and identification of thermophilic Campylobacter spp. in pig faecal samples.]. Appl. Microbiol. 99:292-300. Johnson, L. G., and E. A. Murano. 1999a. Comparison of three protocols for the isolation of Arcobacter from poultry. J. Food Prot. 62: 6 10-6 14. Johnson, L. G., and E. A. Murano. 1999b. Development of a new medium for the isolation of Arcobacter spp. J. Food Prot. 62: 456-462. Jones, F. S., M. Orcutt, and R B. Little. 1931. Vibrios (V. jejuni n. sp.) associated with intestinal disorders in cows and calves. J. Exp. Med. 53:853-863. Kabeya, H., Y. Kobayashi, S . Maruyama, and T. Mikami. 2003. One-step polymerase chain reaction-based typing of Arcobacter species. Int. J. Food Microbiol. 81:163-168.
Karenlampi, R. I., T. P. Tolvanen, and M. L. Hanninen. 2004. Phylogenetic analysis and PCR-restriction fragment length polymorphism identification of Campylobacter species based on partial groEL gene sequences. J. Clin. Microbiol. 425731-5738. Keller, J., B. Wieland, M. Wittwer, R. Stephan, and V. Perreten. 2007. Distribution and genetic variability among Campylobacter spp. isolates from different animal species and humans in Switzerland. Zoonoses Public Health 54:2-7. Kiehlbauch, J. A., D. J. Brenner, M. A. Nicholson, C. N. Baker, C. M. Patton, A. G. Steigerwalt, and I. K. Wachsmuth. 1991. Campylobacter butzleri sp. nov. isolated from humans and animals with diarrheal illness. J. Clin. Microbiol. 29:376-385. Kiehlbauch, J. A., D. N. Cameron, and I. K. Wachsmuth. 1994. Evaluation of ribotyping techniques as applied to Arcobacter, Campylobacter and Helicobacter. Mol. Cell. Probes 8:109-116. King, E. 0. 1957. Human infections with Vibrio fetus and a closely related vibrio. J. Infect. Dis. 101:119-128. Kodama, Y., T. Ha le, and K. Watanabe. 2007. Sulfurospirillum cauolei sp. nov., a facultatively anaerobic sulfur-reducing bacterium isolated from an underground crude oil storage cavity. Int. J. Syst. Evol. Microbiol. 57:827-831. Kokotovic, B., and S . L. On. 1999. High-resolution genomic fingerprinting of Campylobacter jejuni and Campylobacter coli by analysis of amplified fragment length polymorphisms. FEMS Microbiol. Lett. 173:77-84. Konstantinidis, K. T., A. Ramette, and J. M. Tiedje. 2006. The bacterial species definition in the genomic era. Phil. Trans. R. Soc. Lond. B Biol. Sci. 361:1929-1940. Konstantinidis, K. T., and J. M. Tiedje. 2005. Genomic insights that advance the species definition for prokaryotes. Proc. Natl. Acad. Sci. USA 102:2567-2572. Korczak, B. M., R. Stieber, S. Emler, A. P. Burnens, J. Frey, and P. Kuhnert. 2006. Genetic relatedness within the genus Campylobacter inferred from rpoB sequences. Int. ]. Syst. Evol. Microbiol. 56:937-945. Kunin, V., R. Sorek, and P. Hugenholtz. 2007. Evolutionary conservation of sequence and secondary structures in CRISPR repeats. Genome Biol. 8:R61. Laanbroek, H. J., W. Kingma, and H. Veldkamp. 1977. Isolation of an aspartate-fermenting, free-living Campylobacter species. FEMS Microbiol. Lett. 1:99-102. Lawson, A. J., D. Linton, and J. Stanley. 1998. 16s rRNA gene sequences of “Candidatus Campylobacter hominis,” a novel uncultivated species, are found in the gastrointestinal tract of healthy humans. Microbiology 144(Pt. 8):2063-2071. Lawson, A. J., S . L. On, J. M. Logan, and J. Stanley. 2001. Campylobacter hominis sp. nov., from the human gastrointestinal tract. Int. J. Syst. Evol. Microbiol. 51:651-660. Lawson, G. H. K., J. L. Leaver, G. W. Pettigrew, and A. C. Rowland. 1981. Some features of Campylobacter sputorum subsp. mucosalis subsp. nov., nom. rev. and their taxonomic significance. Int. 1.Syst. Bacteriol. 31:385-391. Levy, A. J. 1946. A gastroenteritis outbreak probably due to a bovine strain of vibrio. Yale]. Biol. Med. 18:243. Li, L., C. J. Stoeckert, Jr., and D. S . Roos. 2003. OrthoMCL: identification of ortholog groups for eukaryotic genomes. Genome Res. 13:2178-2189. Litrup, E., M. Torpdahl, and E. M. Nielsen. 2007. Multilocus sequence typing performed on Campylobacter coli isolates from humans, broilers, pigs and cattle originating in Denmark. J. Appl. Microbiol. 103:2 10-21 8. Logan, J. M., A. Burnens, D. Linton, A. J. Lawson, and J. Stanley. 2000. Campylobacter lanienae sp. nov., a new species isolated from workers in an abattoir. Int. J. Syst. Evol. Microbiol. 50(Pt. 2):865-872.
CHAPTER 1
Liibeck, P. S., P. Wolffs, S. L. On, P. Ahrens, P. Radstrom, and J. Hoorfar. 2003. Toward an international standard for PCR-based detection of food-borne thermotolerant campylobacters: assay development and analytical validation. Appl. Environ. Microbiol. 695664-5669. Luijten, M., S. A. B. Weelink, B. Godschalk, A. A.M. Langenhoff, M. H. A. van Eekert, G . Schraa, and A. J. M. Stams. 2004. Anaerobic reduction and oxidation of quinone moieties and the reduction of oxidized metals by halorespiring and related organisms. FEMS Microbiol. Ecol. 49:145-150. MacRae, J. D., I. N. Lavine, K. A. McCaffery, and K. Ricupero. 2007. Isolation and characterization of NP4, arsenate-reducing Sulfurospirillum, from Maine groundwater. J. Environ. Eng. 133:81-88. Mandrell, R. E., L. A. Harden, A. Bates, W. G . Miller, W. F. Haddon, and C. K. Fagerquist. 2005. Speciation of Campylobacter coli, C. jejuni, C. helveticus, C. lari, C. sputorum, and C. upsaliensis by matrix-assisted laser desorption ionization-time of flight mass spectrometry. Appl. Environ. Microbiol. 71:62926307. Marshall, B. J., H. Royce, D. I. Annear, C. S. Goodwin, J. W. Pearman, J. R. Warren, and J. A. Armstrong. 1984. Original isolation of Campylobacter pyloridis from human gastric mucosa. Microbios Lett. 25:83-88. Marshall, S. M., P. L. Melito, D. L. Woodward, W. M. Johnson, F. G. Rodgers, and M. R. Mulvey. 1999. Rapid identification of Campylobacter, Arcobacter, and Helicobacter isolates by PCRrestriction fragment length polymorphism analysis of the 16s rRNA gene. J. Clin. Microbiol. 37:4158-4160. Matsheka, M. I., B. G. Elisha, A. L. Lastovica, and S. L. On. 2002. Genetic heterogeneity of Campylobacter concisus determined by pulsed field gel electrophoresis-based macrorestriction profiling. FEMS Microbiol. Lett. 211:17-22. Matsheka, M. I., A. J. Lastovica, H. Zappe, and B. G. Elisha. 2006. The use of (GTG), oligonucleotide as an RAPD primer to type Campylobacter concisus. Lett. Appl. Microbiol. 42:600605. McClung, C. R., D. G. Patriquin, and R. E. Davis. 1983. Campylobacter nitrofigilis sp. nov., a nitrogen-fixing bacterium associated with roots of Spartina alterniflora Loisel. Int. ]. Syst. Bacteriol. 33:605-612. McFadyean, J., and S. Stockman. 1913. Report of the Departemental Committee Appointed by the Board of Agriculture and Fisheries to Enquire into Epizootic Abortion. His Majesty's Stationery Office, London. Miller, W. G., S. L. On, G. Wang, S. Fontanoz, A. J. Lastovica, and R. E. Mandrell. 2005. Extended multilocus sequence typing system for Campylobacter coli, C. lari, C. upsaliensis, and C. helveticus. J. Clin. Microbiol. 43:23 15-2329. Miller, W. G., C. T. Parker, S. Heath, and A. J. Lastovica. 2007. Identification of genomic differences between Campylobacter jejuni subsp. jejuni and C. jejuni subsp. doylei at the nap locus leads to the development of a C. jejuni subspeciation multiplex PCR method. BMC Microbiol. 7:ll. Moussard, H., E. Corre, M. A. Cambon-Bonavita, Y. Fouquet, and C. Jeanthon. 2006. Novel uncultured Epsilonproteobacteria dominate a filamentous sulphur mat from the 13 degrees N hydrothermal vent field, East Pacific Rise. FEMS Microbiol. Ecol. 58:449-463. Nayak, R., T. M. Stewart, and M. S. Nawaz. 2005. PCR identification of Campylobacter coli and Campylobacter jejuni by partial sequencing of virulence genes. Mol. Cell. Probes 19:187193. Neill, S. D., J. N. Campbell, J. J. Obrien, S. T. C. Weatherup, and W. A. Ellis. 1985. Taxonomic position of Campylobacter cryaerophila sp. nov. Int. ]. Syst. Bacteriol. 35:342-356.
TAXONOMY OF THE CAMPYLOBACTERACEAE
23
On, S. L. 1994. Confirmation of human Campylobacter concisus isolates misidentified as Campylobacter mucosalis and suggestions for improved differentiation between the two species. J. Clin. Microbiol. 32:2305-2306. On, S. L. 1996. Identification methods for campylobacters, helicobacters, and related organisms. Clin. Microbiol. Rev. 9:405422. On, S. L., H. I. Atabay, J. E. Corry, C. S. Harrington, and P. Vandamme. 1998. Emended description of Campylobacter sputorum and revision of its infrasubspecific (biovar) divisions, including C. sputorum biovar paraureolyticus, a urease-producing variant from cattle and humans. Int. J. Syst. Bacteriol. 48(Pt. 1): 195-206. On, S. L., and C. S. Harrington. 2001. Evaluation of numerical analysis of PFGE-DNA profiles for differentiating Campylobacter fetus subspecies by comparison with phenotypic, PCR and 16s rDNA sequencing methods. J. Appl. Microbiol. 90:285-293. On, S. L., and C. S. Harrington. 2000. Identification of taxonomic and epidemiological relationships among Campylobacter species by numerical analysis of AFLP profiles. FEMS Microbiol. Lett. 193~161-169. On, S. L., C. S. Harrington, and H. I. Atabay. 2003. Differentiation of Arcobacter species by numerical analysis of AFLP profiles and description of a novel Arcobacter from pig abortions and turkey faeces. J. Appl. Microbiol. 95:1096-1105. On, S. L., B. Holmes, and M. J. Sackin. 1996. A probability matrix for the identification of campylobacters, helicobacters and allied taxa. J. Appl. Bacteriol. 81:425-432. On, S. L., T. K. Jensen, V. Bille-Hansen, S. E. Jorsal, and P. Vandamme. 2002. Prevalence and diversity of Arcobacter spp. isolated from the internal organs of spontaneous porcine abortions in Denmark. Vet. Microbiol. 85:159-167. On, S. L. W., and B. Holmes. 1995. Classification and identification of campylobacters, helicobacters and allied taxa by numerical analysis of phenotypic characters. Syst. Appl. Microbiol. 18: 374-390. Oremland, R. S., J. S. Blum, C. W. Culbertson, P. T. Visscher, L. G. Miller, P. Dowdle, and F. E. Strohmaier. 1994. Isolation, growth, and metabolism of an obligately anaerobic, selenaterespiring bacterium, strain SES-3. Appl. Environ. Microbiol. 60: 3011-3019. PrCvot, A. R. 1940. Etudes de systimatique bactirienne. V. Essai de classification des vibrions anaerobies. Ann. Inst. Pasteur 64: 117-1 25. Prouzet-Mauleon, V., L. Labadi, N. Bouges, A. Menard, and F. Megraud. 2006. Arcobacter butzleri: underestimated enteropathogen. Emerg. Infect. Dis. 12:307-309. Puigbo, P., S. Garcia-Vallve, and J. 0. Mdnerney. 2007. TOPD/ FMTS: a new software to compare phylogenetic trees. Bioinformatics 23:1556-1558. Romero, J., M. Garcia-Varela, J. P. Laclette, and R T. Espejo. 2002. Bacterial 16s rRNA gene analysis revealed that bacteria related to Arcobacter spp. constitute an abundant and common component of the oyster microbiota (Tiostrea chilensis). Microb. Ecol. 44:365-371. Roop, R. M., R. M. Smibert, J. L. Johnson, and N. R. Krieg. 1985a. Campylobacter mucosalis (Lawson, Leaver, Pettigrew, and Rowland 1981) comb. nov.: emended description. Int. J. Syst. Bacteriol. 35:189-192. Roop, R. M., R. M. Smibert, J. L. Johnson, and N. R. Krieg. 1985b. DNA homology studies of the catalase-negative campylobacters and Campylobacter fecalis, an emended description of Campylobacter sputorum, and proposal of the neotype strain of Campylobacter sputorum. Can. J. Microbiol. 31:823-83 1. Rose, T. M., E. R. Schultz, J. G. Henikoff, S. Pietrokovski, C. M. McCallum, and S. Henikoff. 1998. Consensus-degenerate hybrid
24
DEBRUYNE ET AL.
oligonucleotide primers for amplification of distantly related sequences. Nucleic Acids Res. 26:1628-1635. Sails, A. D., A. J. Fox, F. J. Bolton, D. R. Wareing, D. L. Greenway, and R. Borrow. 2001. Development of a PCR ELISA assay for the identification of Campylobacter jejuni and Campylobacter coli. Mol. Cell. Probes 15:291-300. Sandstedt, K., and J. Ursing. 1991. Description of Campylobacter upsaliensis sp. nov. previously known as the CNW group. Syst. Appl. Microbiol. 14:39-45. Scholzmuramatsu, H., A. Neumann, M. Messmer, E. Moore, and G. Diekert. 1995. Isolation and characterization of Dehalospirillum multivorans gen. nov., sp. nov., a tetrachloroetheneutilizing, strictly anaerobic bacterium. Arch. Microbiol. 163:4856. Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41:15-26. Schuett, C., H. Doepke, W. Groepler, and A. Wichels. 2005. Diversity of intratunical bacteria in the tunic matrix of the colonial ascidian Diplosoma migrans. Helgoland Mar. Res. 59:136-140. Schumacher, W., P. M. H. Kroneck, and N. Pfennig. 1992. Comparative systematic study on Spirillum 5 175, Campylobacter and Wolinella species: description of Spirillum 5 175 as Sulfurospirillum deleyianum gen. nov., spec. nov. Arch. Microbiol. 158:287293. Sebald, M., and M. V6ron. 1963. Teneur en bases de I’ADN et classification des vibrions. Ann. lnst. Pasteur 105:897-910. Sette, L. D., K. C. Simioni, S. P. Vasconcellos, L. J. Dussan, E. V. Neto, and V. M. Oliveira. 2007. Analysis of the composition of bacterial communities in oil reservoirs from a southern offshore Brazilian basin. Antonie Leeuwenhoek 91:253-266. Siemer, B. L., E. M. Nielsen, and S. L. On. 2005. Identification and molecular epidemiology of Campylobacter coli isolates from human gastroenteritis, food, and animal sources by amplified fragment length polymorphism analysis and Penner serotyping. Appl. Environ. Microbiol. 71:1953-1958. Sievert, S. M., E. B. Wieringa, C. 0. Wirsen, and C. D. Taylor. 2007. Growth and mechanism of filamentous-sulfur formation by “Candidatus Arcobacter sulfidicus” in opposing oxygensulfide gradients. Environ. Microbiol. 9:271-276. Skirrow, M. B. 1977. Campylobacter enteritis: a “new” disease. Br. Med. J. 2:9-11. Skirrow, M. B. 2006. John McFadyean and the centenary of the first isolation of Campylobacter species. Clin. Infect. Dis. 43: 1213-1217. Smith, T., and M. Orcutt. 1927. Vibrios from calves and their serological relation to Vibrio fetus. J. Exp. Med. 45:391-397. Smith, T., and M. S. Taylor. 1919. Some morphological and biochemical characters of the spirilla (Vibrio fetus n. sp.) associated with disease of the fetal membranes in cattle. J. Exp. Med. 310: 299-312. Snaidr, J., R. Amann, I. Huber, W. Ludwig, and K. H. Schleifer. 1997. Phylogenetic analysis and in situ identification of bacteria in activated sludge. Appl. Environ. Microbiol. 63:2884-2896. Stackebrandt, E., and B. M. Goebel. 1994. A place for DNA-DNA reassociation and 16s ribosomal RNA sequence analysis in the present species definition in bacteriology. lnt. J. Syst. Bacteriol. 44~846-849. Stoddard, R. A., W. G. Miller, J. E. Foley, J. Lawrence, F. M. Gulland, P. A. Conrad, and B. A. Byrne. 2007. Campylobacter insulaenigrae isolated from Northern Elephant Seals (Mirounga angustirostris) in California. Appl. Environ. Microbiol. 73:17291735.
Stolz, J. F., D. J. Ellis, J. S. Blum, D. Ahmann, D. R. Lovley, and R. S. Oremland. 1999. Sulfurospirillum barnesii sp. nov. and Sulfurospirillum arsenophilum sp. nov., new members of the Sulfurospirillum clade of the epsilon Proteobacteria. lnt. J. Syst. Bacteriol. 49:1177-1180. Stolz, J. F., R. S. Oremland, B. J. Paster, F. E. Dewhirst, and P. Vandamme. 2005. Genus 111. Sulfurospirillum Schumacher, Kroneck, and Pfennig 1993, 188VP (Effective publication: Schumacher, Kroneck, and Pfennig 1992, 291), emend. Fienster, Liesack and Tindall 1997d, 1216, p. 1165-1168. In D. J. Brenner, N. R. Krieg, J. T. Staley, and G. M. Garrity (ed.), BergeySManual of Systematic Bacteriology, vol. 2. Springer-Verlag, New York. Suerbaum, S., M. Lohrengel, A. Sonnevend, F. Ruberg, and M. Kist. 2001. Allelic diversity and recombination in Campylobacter jejuni. J. Bacteriol. 183:2553-2559. Tanner, A. C. R., S. Badger, C. H. Lai, M. A. Listgarten, R. A. Visconti, and S. S. Socransky. 1981. Wolinella gen. nov., Wolinella succinogenes (Vibriosuccinogenes Wolin et. al.) comb. nov., and description of Bacteroides gracilis sp. nov., Wolinella recta sp. nov., Campylobacter concisus sp. nov., and Eikenella corrodens from humans with periodontal disease. lnt. J. Syst. Bacteriol. 31:432-445. Tanner, A. C. R., M. A. Listgarten, and J. L. Ebersole. 1984. Wolinella curva sp. nov.: “Vibriosuccinogenes” of human origin. Int. J. Syst. Bacteriol. 34:275-282. Tatusov, R. L., E. V. Koonin, and D. J. Lipman. 1997. A genomic perspective on protein families. Science 278:63 1-637. Teske, A., P. Sigalevich, Y. Cohen, and G. Muyzer. 1996. Molecular identification of bacteria from a coculture by denaturing gradient gel electrophoresis of 16s ribosomal DNA fragments as a tool for isolation in pure cultures. Appl. Environ. Microbiol. 62:4210-4215. Totten, P. A., C. L. Fennell, F. C. Tenover, J. M. Wezenberg, P. L. Perine, W. E. Stamm, and K. K. Holmes. 1985. Campylobacter cinaedi (sp. nov.) and Campylobacter fennelliae (sp. nov.): two new Campylobacter species associated with enteric disease in homosexual men. J. Infect. Dis. 151:131-139. Tu, Z. C., W. Eisner, B. N. Kreiswirth, and M. J. Blaser. 2005. Genetic divergence of Campylobacter fetus strains of mammal and reptile origins. J. Clin. Microbiol. 43:3334-3340. Tunicliff, R. 1914. An anaerobic vibrio isolated from a case of acute bronchitis. J. Infect. Dis. 15:350-351. Tyrrell, K. L., D. M. Citron, Y. A. Warren, S. Nachnani, and E. J. Goldstein. 2003. Anaerobic bacteria cultured from the tongue dorsum of subjects with oral malodor. Anaerobe 9:243-246. Ursing, J. B., H. Lior, and R. J. Owen. 1994. Proposal of minimal standards for describing new species of the family Campylobacteraceae. Int. J. Syst. Bacteriol. 445342445. van Bergen, M. A., K. E. Dingle, M. C. Maiden, D. G. Newell, L. van der Graaf-Van Bloois, J. P. van Putten, and J. A. Wagenaar. 2005a. Clonal nature of Campylobacter fetus as defined by multilocus sequence typing. J. Clin. Microbiol. 435888-5898. van Bergen, M. A., G. Simons, L. van der Graaf-van Bloois, J. P. van Putten, J. Rombout, I. Wesley, and J. A. Wagenaar. 2005b. Amplified fragment length polymorphism based identification of genetic markers and novel PCR assay for differentiation of Campylobacter fetus subspecies. J. Med. Microbiol. 54: 1217-1224. Vandamme, P. 2000. Taxonomy of the family Campylobacteraceae, p. 3-26. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Vandamme, P., M. I. Daneshvar, F. E. Dewhirst, B. J. Paster, K. Kersters, H. Goossens, and C. W. Moss. 1995. Chemotaxonomic analyses of Bacteroides gracilis and Bacteroides ureolyticus and reclassification of B. gracilis as Campylobacter gracilis comb. nov. lnt. J. Syst. Bacteriol. 45:145-152.
CHAPTER 1
Vandamme, P., and J. De Ley. 1991. Proposal for a new family, Campylobacteraceae.Int. J. Syst. Bacteriol. 41:45 1-455. Vandamme, P., E. Falsen, B. Pot, B. Hoste, K. Kersters, and J. De Ley. 1989. Identification of EF group 22 campylobacters from gastroenteritis cases as Campylobacter concisus. J. Clin. Microbiol. 27: 1775-1781. Vandamme, P., E. Falsen, R. Rossau, B. Hoste, P. Segers, R. Tytgat, and J. De Ley. 1991a. Revision of Campylobacter, Helicobacter, and Wolinella taxonomy: emendation of generic descriptions and proposal of Arcobacter gen. nov. Int. J , Syst. Bacteriol. 41~88-103. Vandamme, P., and H. Goossens. 1992. Taxonomy of Campylobacter, Arcobacter, and Helicobacter: a review. Zentralbl. Bakteriol. 276:447-472. Vandamme, P., B. Pot, E. Falsen, K. Kersters, and J. De Ley. 1990. Intra- and interspecific relationships of veterinary campylobacters revealed by numerical analysis of electrophoretic protein profiles and DNA:DNA hybridizations. Syst. Appl. Microbiol. 13~295-303. Vandamme, P., B. Pot, M. Gillis, P. de Vos, K. Kersters, and J. Swings. 1996. Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol. Rev. 60:407-438. Vandamme, P., B. Pot, and K. Kersters. 1991b. Differentiation of campylobacters and Campylobacter-like organisms by numerical analysis of one-dimensional electrophoretic protein patterns. Syst. Appl. Microbiol. 1457-66. Vandamme, P., L. J. Van Doorn, S. T. a1 Rashid, W. G. Quint, J. van der Plas, V. L. Chan, and S. L. On. 1997. Campylobacter hyoilei Alderton et al. 1995 and Campylobacter coli Veron and Chatelain 1973 are subjective synonyms. Int. J. Syst. Bacteriol. 47~1055-1060. Vandamme, P., M. Vancanneyt, B. Pot, L. Mels, B. Hoste, D. Dewettinck, L. Vlaes, C. van den Borre, R. Higgins, and J. Hommez. 1992. Polyphasic taxonomic study of the emended genus Arcobacter with Arcobacter butzleri comb. nov. and Arcobacter skirrowii sp. nov., an aerotolerant bacterium isolated from veterinary specimens. Int. J. Syst. Bacteriol. 42:344-356. Van Driessche, E., and K. Houf. 2007. Discrepancy between the occurrence of Arcobacter in chickens and broiler carcass contamination. Poult. Sci. 86:744-751. Van Driessche, E., K. Houf, J. Van Hoof, L. De Zutter, and P. Vandamme. 2003. Isolation of Arcobacter species from animal feces. FEMS Microbiol. Lett. 229:243-248.
TAXONOMY OF THE CAMPYLOBACTERACEAE
25
Van Etterijck, R., J. Breynaert, H. Revets, T. Devreker, Y. Vandenplas, P. Vandamme, and s. Lauwers. 1996. Isolation of Campylobacter concisus from feces of children with and without diarrhea. J. Clin. Microbiol. 34:2304-2306. Vandenberg, O., A. Dediste, K. Houf, S. Ibekwem, H. Souayah, S. Cadranel, N. Douat, G. Zissis, J. P. Butzler, and P. Vandamme. 2004. Arcobacter species in humans. Emerg. Infect. Dis. 10:1863-1867. Viron, M., and R. Chatelain. 1973. Taxonomic study of the genus Campylobacter Sebald and V6ron and designation of the neotype strain for the type species Campylobacter fetus (Smith and Taylor) Sebald and VCron. lnt. J. Syst. Bacteriol. 23:122-134. Vinzent, R., J. Dumas, and N. Picard. 1947. Septicemie grave eu cours de la grossesse, due iun vibrion. Avortement consecutif. C. R Acad. Med. 131:90. Voordouw, G., S. M. Armstrong, M. F. Reimer, B. Fouts, A. J. Telang, Y. Shen, and D. Gevertz. 1996. Characterization of 16s rRNA genes from oil field microbial communities indicates the presence of a variety of sulfate-reducing, fermentative, and sulfide-oxidizing bacteria. Appl. Environ. Microbiol. 62: 16231629. Wesley, I. V., L. Schroeder-Tucker, A. L. Baetz, F. E. Dewhirst, and B. J. Paster. 1995. Arcobacter-specific and Arcobacter butzleri-specific 16s rRNA-based DNA probes. J. Clin. Microbiol. 33:1691-1698. Winkler, M. A., J. Uher, and S. Cepa. 1999. Direct analysis and identification of Helicobacter and Campylobacter species by MALDI-TOF mass spectrometry. Anal. Chem. 71:3416-3419. Wirsen, C. O., S. M. Sievert, C . M. Cavanaugh, S. J. Molyneaux, A. Ahmad, L. T. Taylor, E. F. DeLong, and C. D. Taylor. 2002. Characterization of an autotrophic sulfide-oxidizing marine Arcobacter sp. that produces filamentous sulfur. Appl. Environ. Microbiol. 68:3 16-325. Wolfe, R. S., and N. Pfennig. 1977. Reduction of sulfur by Spirillum 5175 and syntrophism with Chlorobium. Appl. Environ. Microbiol. 33 :427-43 3. Wolin, M. J., E. A. Wolin, and N. J. Jacobs. 1961. Cytochromeproducing anaerobic vibrio, Vibrio succinogenes, sp. n. J. Bacteriol. 81:911-917. Wybo, I., J. Breynaert, S. Lauwers, F. Lindenburg, and K. Houf. 2004. Isolation of Arcobacter skirrowii from a patient with chronic diarrhea. J. Clin. Microbiol. 42:1851-1852.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Biaser 6 2008 ASM Press, Washington, DC
Chapter 2
Population Biology of Campylobacter jejuni and Related Organisms MARTIN
c. J. MAIDEN AND KATE E. DINGLE
must take account of population biology and evolution. Population studies of bacteria comprise three elements: (i) sampling the population, (ii) measuring variation in the population, and (iii) interpreting the variation. Sampling the natural world is a nontrivial problem, and it is impossible to achieve the ideal of a truly random or even wholly representative population sample. In microbiology there are additional technical questions about the reliability of culture methods-or indeed other methods-for accurately capturing all of the variation present in a given sample. For widely distributed multihost bacteria, such as the members of the genus Campylobacter, these issues are particularly relevant (Bolton and Robertson, 1982) and have to be mitigated by the adoption of appropriately designed structured sampling schemes that are driven by the particular question being investigated. Molecular methods can go some way to mitigating problems with culture (Abulreesh et al., 2006), but even these can be subject to sensitivity issues, and the lack of bacterial isolates limits the extent to which biological variation can be accurately characterized. Once bacterial isolates are available, exhaustively measuring the variation present among genomes is possible by the appropriate application of highthroughput nucleotide sequence determination technology (Maiden, 2000). Two approaches can be used. The existence of a complete genome from one example of the population being analyzed enables the nucleotide sequence of genes in many isolates to be determined with PCR-based direct nucleotide sequence determination; this has the advantage that it can also be done on specimens containing noncultur-
Studies of the ecology and evolution of bacterial populations, which require measurement of variation, theory development, and testing, are often carried out independently from the investigation of microbial function that are usually conducted within a reductionist framework by means of controlled experiments. There is a potentially creative tension between these two approaches: population studies require the collection and analysis of large numbers of bacterial isolates, while on the other hand, functional studies commonly involve the detailed examination of one, or at most a few, example isolates, with little or no reference to population variation. The integration of these approaches is necessary if recent technical advances in data collection are to be fully exploited. The availability of whole-genome sequences for many bacteria, including the principal human and animal pathogens (Bentley and Parkhill, 2004), has highlighted the need for better interaction of these approaches. A whole-genome sequence is in many ways the apotheosis of a single organism study, and in principle, the determination of the complete genetic information for any organism enables all aspects of its biology to be deduced. In practice, however, current and anticipated knowledge of biochemistry and genetics is too limited to permit such deductions from genome sequences alone. Biochemistry is still not a high-throughput science in the sense that genome sequence determination is, and detailed investigations of all cistrons from all genomes are impractical. Consequently, most of the information in annotated genomes comes from comparative studies, the quality of which is dependent on the quality of the comparisons (Brenner, 1999). Once a comparative element is introduced into a study, it perforce
Martin C. J. Maiden Department of Zoology, University of Oxford, South Parks Rd., Oxford OX1 3PS, United Kingdom. Kate E. Dingle Nuffield Department of Clinical Laboratory Sciences, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, United Kingdom.
27
28
MAIDEN AND DINGLE
able organisms. Conversely, the complete genome of many samples can be compared with microarray, or better, rapid whole-genome resequencing approaches, although this generally requires the availability of isolates to prepare genomic DNA. The former approach has the advantage of producing defined data rapidly, whereas the latter provides the intriguing prospect of comprehensive comparative analysis. At the present time, the whole-genome sequence determination approach remains impractical as a result of cost constraints. Even if such data were to become available, new analysis algorithms would be required to cope with the quantity of data generated. The use of an evolutionary framework is essential to the interpretation of biological variation, and a number of analytical approaches have been developed. These allow the variation present in a given population to be explored with a view to establishing the biological forces that have generated the variation present. In the case of the bacteria, which are fast growing, have small genomes, and live in intimate contact with their environment, genetic variation can be especially informative in understanding their ecology. The most effective analysis methods are parametric and model based, and in this respect, recently developed analytical tools that are based on coalescent theory are especially valuable (Balding, 2003; Didelot and Falush, 2007; Falush et al., 2003). Informal techniques can also be valuable in the primary analysis of data (Feil et al., 2004). Here we explore the contribution that population studies have made to our understanding of the biology of the Campylobacter, and we argue that such studies have a central role to play in understanding the epidemiology and pathogenesis of this important group of gramnegative bacteria.
ANALYSIS OF POPULATIONS OF CAMPYLOBACTER The factors that make members of the genus Campylobacter difficult to study with conventional single-isolate approaches-in particular their ubiquity and diversity-also make them tractable to, and indeed attractive subjects of, population studies. The biology underlying the diversity of members of the genus Campylobacter, including factors such as host range and specialization, as well as being interesting biological problems in their own right, are also relevant to the main human interaction with these bacteria, namely their propensity to cause gastroenteritis. Carnpylobacter jejuni and Campylobacter coli cause the majority of human cases of Campylobacter-associated gastroenteritis; these two organisms are asso-
ciated with approximately 90 and 10% of cases, respectively (Gillespie et al., 2002). Campylobacter upsaliensis is frequently identified in domestic cats and dogs, and it causes occasional cases of gastroenteritis in humans. Campylobacter helveticus is also found in domestic animals, but it appears not to cause human disease. Campylobacter lari can cause gastroenteritis in humans. It has been isolated from shellfish (Endtz et al., 1997), but it is most commonly found among wild birds. Campylobacter fetus is primarily associated with infections in livestock, which result in abortion and infertility, particularly in sheep and cattle. C. fetus can also be isolated from goats, pigs, horses, fowl, and reptiles and causes occasional human infections. Transmission to humans generally occurs orally through contaminated food or water (Garcia et al., 1983). A range of techniques have been used to study variation within the genus Campylobacter. Indeed, the plethora of isolate characterization techniques available could be said to have obscured rather than illuminated research into populations of these organisms (Wassenaar and Newell, 2000). In this respect, it is necessary to distinguish those bacterial characterization schemes that can be sensibly used for population studies from those that cannot. It is relatively trivial to develop a typing scheme that can classify isolates as distinct or indistinguishable (Achtman, 1996), but it is much more problematic to develop schemes that generate reproducible data that can be analyzed to investigate the biology of the organism in question. Because virtually all variation in bacterial populations is genetic or ultimately genetically determined, the availability of high-throughput nucleotide sequence determination has provided influential advances in isolate characterization. Thanks to the combination of the availability of complete genome sequences of C. jejuni, for example that of isolate NCTC 11168 (Parkhill et al., 2000), and PCR amplification technology, it is now possible to establish definitively the genetic variation present at any part of the genome by PCR-direct nucleotide sequence determination. Other methods, such as multilocus enzyme electrophoresis (Aeschbacher and Piffaretti, 1989), pulsed-field gel electrophoresis analysis (Yan et al., 1991), and amplified fragment-length polymorphism (Duim et al., 1999), are at best proxies for nucleotide sequence diversity and therefore generally provide less information (Sails et al., 2003a, 2003b). In addition, the data produced by these techniques are not as tractable to phylogenetic and statistical analysis as nucleotide sequences themselves. The application of high-throughput nucleotide sequence-based approaches (Maiden, 2000) to large
CHAPTER 2
isolate collections and the analysis of these data have enabled the diversity of C. jejuni populations to be defined (Dingle et al., 2002) and the relationship of this population to other members of the genus compared (Dingle et al., 2005; Miller et al., 2005; van Bergen et al., 2005). At present, the major technique used is multilocus sequence typing (MLST) (Maiden et al., 1998). MLST is a generic approach for the characterization of bacterial isolates that aims to identify lineages in bacterial populations by indexing the variation present in several (commonly seven) housekeeping genes located at various parts of the chromosome (Maiden, 2006). The advantages of using housekeeping genes is their presence in all, or nearly all, isolates, combined with the fact that although they contain measurable variation, they are subject to stabilizing selection for conservation of metabolic function. Given that horizontal genetic exchange is common in the microbial world (Narra and Ochman, 2006), the use of multiple loci is important because single loci can be misleading indicators of relationship by descent. MLST is in effect a nucleotide sequence-based version of multilocus enzyme electrophoresis (Selander et al., 1986) in that it indexes of genetic variation at housekeeping loci with nucleotide sequencing. This approach has a number of advantages in addition to generating definitive, reproducible, and transportable data. For MLST, a viable isolate is not required, and the process is scalable from one isolate to many hundreds or thousands by the exploitation of liquid handling technology and automated DNA analyzers. For a group of organisms as diverse as those making up the genus Campylobacter, this is a particular advantage because it facilitates the analysis of very large numbers of isolates. MLST also introduced the innovation of sharing data via the Internet (Chan et al., 2001; Jolley et al., 2004) and promoted the development of novel analysis tools (Didelot and Falush, 2007; Feil et al., 2004, 1999; Jolley et al., 2001). MLST approaches have the advantage that they can simultaneously address large-scale or more local population questions, especially in combination with other sequence-based typing approaches (Maiden, 2006). In the Campylobacter MLST scheme, the sequences of seven allele fragments are determined after amplification from genomic DNA or a killed cell suspension. In common with other schemes, short gene fragments, generally 400 to 550 bp, are used because these are easily sequenced on both DNA strands by using only one extension primer pair, ensuring rapidity, cost effectiveness, and accuracy. Each unique sequence of an allele fragment is assigned an arbitrary number in order of discovery. These unique
POPULATION BIOLOGY OF C. lElUN1
29
sequences are stored electronically in the “profiles database,” which is essentially a dictionary that catalogs variation and that is used to assign typing data as they are collected (Jolley et al., 2004). The allele numbers are combined into allelic profiles, or sequence types (STs), which are also arbitrarily assigned. Allele numbers and STs are in effect shorthand codes for bacterial variation, with each ST representing a unique combination of nucleotide sequence of, in the case of C. jejuni, 3,309 bp in length and including the variation present at seven separate loci. STs can be grouped into “clonal complexes,” surrogates of genetic lineages that are described in detail for C. jejuni below. The definitions of these clonal complexes are also stored in the profiles database. Analysis can be conducted either at the level of clonal complex, ST, allele, or nucleotide sequences, including all or a subset of the MLST data, depending on the questions being addressed (Maiden, 2006).
VARIATION WITHIN THE GENUS CAMPYLOBACTER MLST schemes are now available for many members of the genus Campylobacter. The first to be developed was for C. jejuni (Dingle et al., 2001a); schemes are also available for C. coli (Dingle et al., 2005), C. lari, C. upsaliensis, C. helveticus (Miller et al., 2005), and C. fetus (van Bergen et al., 2005). These schemes have established the basic population structure of the members of the genus and are beginning to demonstrate associations of genotypes with particular animal hosts (Dingle et al., 2002; French et al., 2005). The relationships of these different Campylobacter microbiological species with human and veterinary disease are being more precisely defined on the basis of the insights available from MLST data. The level of genetic divergence among Campylobacter housekeeping genes is as high as 80.9% (Miller et al., 2005). Consequently, only C. jejuni can be MLST typed reliably by using the oligonucleotide primers originally described (Dingle et al., 2001b), which were designed by using the first draft of the C. jejuni genome sequence (Parkhill et al., 2000). Two approaches have been used to adapt this MLST scheme to other Campylobacter species. First, the scheme was extended to C. coli and C. fetus by using a large collection of different C. jejuni oligonucleotide primers in PCR amplifications of DNA from C. coli and C. fetus (Dingle et al., 2005; van Bergen et al., 2005). The amplification reactions were performed under conditions of reduced stringency to
30
MAIDEN AND DINGLE
permit the amplification of corresponding housekeeping gene sequences from heterologous species. The sequences of these amplicons were determined, and new primers were designed that were specific to C. coli or C. fetus. The same seven housekeeping genes were used as the C. jejuni scheme to allow comparisons of diversity, population structure, and epidemiology among the different species. The C. coli MLST scheme was validated by using 53 isolates from humans, chickens, and pigs, together with 15 Penner serotype reference isolates (Dingle et al., 2005). The C. fetus MLST scheme was validated by using a diverse collection of 140 geographically diverse C. fetus isolates from a variety of animal hosts and humans (van Bergen et al., 2005). A second approach to extending the C. jejuni MLST scheme to other species was made possible by the availability of the draft genome sequences of C. coli, C. upsaliensis, and C. lari (Fouts et al., ZOOS), enabling the design of degenerate oligonucleotide primers, so that individual primer pairs could be used to amplify MLST loci from up to five different Campylobacter species (Miller et al., 2005). This had the advantage of easily detecting interspecies genetic recombinants, genetic hybrids, within the genus Campylobacter and detecting mixed samples. In addition to C. jejuni, the species targeted were C. coli and C. helveticus, for which the same loci are used as the C. jejuni scheme. Schemes have also been developed for C. lari, for which adk and pgi were substituted for aspA and gltA, and C. upsaliensis, for which adk and pgi were substituted for gltA and pgm. Multiple C. coli (n = 57), C. lari (n = 20), C. upsaliensis (n = 78), and C. helveticus (n = 9) isolates, representing both clinical and environmental sources, were typed, demonstrating the utility of the approach for distinguishing among these bacteria. More than 80% of the isolates had unique STs. Genetic recombinants have since been described between C. jejuni and either C. coli or C. lari and between C. upsaliensis and C. helveticus (D’lima et al., 2007; Kinana et al., 2007; Miller et al., 2005). The MLST data for all of the above schemes are available at the pubMLST website (http: //pubmlst. org/campylobacter/), stored in web-accessible customized databases (Fig. 1)that permit online analysis and comparison (Jolley et al., 2004; Jolley and Maiden, 2006). The MLST scheme used for C. jejuni and C. coli is represented by a single PubMLST isolate database, which is integrated with typing data from the flaA, flaB, and porA genes used for antigen gene sequencing by AgdbNet software (Jolley and Maiden, 2006) (Fig. lb). Data from the other MLST schemes are also available at pubMLST. Although MLST was originally developed with a view to identifying clones or lineages within bacterial
populations, especially those of pathogens (Maiden et al., 1998), it has many other applications, including addressing the vexed question of bacterial “speciation” (Stackebrandt et al., 2002). Where the same genes have been chosen for different members of the genus, their interrelationships can be investigated in a number of ways, the simplest being a phylogeny based on the nucleotide sequences. A simple phylogeny drawn for those members of the genus Campylobacter for which MLST data are available shows that C. fetus is the most divergent member of the group, with C. jejuni and C. coli sharing a common ancestor, as do C. lari and C. insulaenigrae, and C. upsaliensis and C. helveticus (Fig. 2). Phylogenic trees represent genealogies, with the branches illustrating pathways of descent. Conventional algorithms use genetic distances, calculated in various ways, to reconstruct these phylogenies, by comparing nucleotide sequences on the assumption that the sequence changes occur according to a more or less constant clock rate. Such a representation is unproblematic if there is no, or at least very little, genetic exchange among the taxa being analyzed, as in the case of the named Campylobacter species. However, as the frequency of gene transfer increases among the organisms placed on divergent branches of the tree, the reliability of a conventionally generated treelike phylogeny decreases. This is because a single genetic transfer event, which is common in some bacterial populations, may introduce a large number of sequence changes, resulting in a long branch, while a single mutation, a comparatively rare event, would generate a short branch. The generation of mosaic genes by intragenic recombination further complicates the interpretation of conventional phylogenies (Maiden, 1993). It is now known that genetic exchange (sometimes referred to as lateral gene transfer or “localized sex” [Maynard Smith et al., 19911) is quite common in the bacteria and needs to be taken into account when reconstructing bacterial genealogies (Maynard Smith et al., 1993). In the case of the relationships among the different members of the genus, a conventional tree, such as that shown in Fig. 2, is more or less reliable, as there is relatively little genetic exchange among the bacteria classified as microbiological species. Indeed, bacterial “species” are probably best defined as populations of bacteria that recombine infrequently, rather than the more traditional definitions that are based on DNA homology (Stackebrandt et al., 2002). The problem of such a definition is that it is difficult to define the levels of gene flow that permit the emergence of bacterial species, and in any case, estimating these parameters is complex and can be sensitive to the sampling frame used. Further, the different levels
CHAPTER 2
POPULATION BIOLOGY OF C. JETUNI
31
a
N
N
O 0
s
0
0
N
N
0
% 0
0 0 0
0
N
P
N O 0
tn
N
0 O
0-4
Year
b
Data access1
Figure 1. Carnpylobacter MLST databases. (a) The growth of the Carnpylobacterjejuni and Campylobacter coli database since its inception in 2001. Open squares represent isolates submitted; closed squares, sequence types. (b) The structure of the pubMLST database network, which is used for all of the Campylobacter databases. It allows the integration of data from various sources in a variety of public or private databases, ensuring that a common typing nomenclature is used.
of horizontal genetic exchange may have different effects on diverse bacterial populations. Wholly asexual populations, i.e., those in which genetic exchange is negligible, will conform to the clonal model of population structure. Clonality is an inevitable consequence of reproduction by binary fission because any mutations that occur in the population are limited to the descendants of the mother cell in which they arose. Diversity reduction events such as bottlenecking (Achtman, 1997) or periodic selection (Levin, 1981) purge the population of diversity, resulting in a population that is dominated by lineages with distinct genotypes and therefore phenotypes. Clonal populations will conform to a num-
ber of predictions: (i) alleles at different loci will exhibit linkage disequilibrium; that is, they will occur in nonrandom combinations; (ii) they will exhibit relationships that conform to a treelike phylogeny, with branches generated by mutation events; and (iii) they will exhibit congruence; that is, the same phylogeny will be recorded at different loci. The population structure of wholly sexual populations is characterized by a lack of lineage structure, with genetic variation that exhibits (i) linkage equilibrium, (ii) a netlike phylogeny, and (iii) incongruence, with each locus recording a unique evolutionary history (Maynard Smith, 1989). In practice, few bacteria occupy the extreme of an entirely clonal or
32
MAIDEN AND DINGLE
C.insulaenigrae \
C. lari \
,
C. jejuni C. upsaliensis
I
w
C. helveticus
C. fetus
Figure 2. Relationships among the different Campylobacter microbiological species. The phylogeny was reconstructed with concatenated sequences from the alleles common to all of the schemes (gln, gly, and tkt).
wholly nonclonal population, and the process of binary fission imposes some element of clonality on all bacterial populations (Feil et al., 1999; Spratt and Maiden, 1999). Understanding bacterial population structure therefore becomes an issue of determining the rate of gene flow and applying an appropriate model, i.e. clonal, partially clonal, or nonclonal. Within the genus Campylobacter, different types of population structure have been observed within bacteria currently assigned to different microbiological species. By far the most studied to date are C. jejuni and C. coli, reflecting their importance in human disease.
Campylobacter jejuni The MLST data collected to date show that C. jejuni is highly diverse with a total of 2,243 distinct STs from some 3,704 isolates deposited in the pubMLST database (http://pubmlst.org/ campylobacter/); many more isolates, perhaps in excess of 10,000, have been characterized by MLST in a variety of studies. In addition to the most widely used scheme, two other C. jejuni MLST schemes have been developed. As expected, given the nature of MLST systems, all three schemes provide similar perspectives on C. jejuni biology (Dingle et al., 2001a; Manning et al., 2003; Suerbaum et al., 2001). Analyses of subsets of the MLST data by maximum likelihood methods show that the phylogenies generated from the different loci are incongruent,
violating the predictions of clonality (Dingle and Maiden, 2005). In addition, there is little evidence for treelike structure in nucleotide sequences of the individual alleles (Dingle et al., 2001a; Suerbaum et al., 2001). There is, however, some evidence of linkage disequilibrium in these data sets (Dingle and Maiden, 2005; Schouls et al., 2003). These features, which are seen in a number of other bacteria analyzed by MLST (Maiden, 2006), are consistent with a partially clonal population, that is, one that contains distinct clusters of related isolates that do not share clonal relationships with each other. In other words, although each member of each cluster shares a common ancestor with other members of the same cluster, cluster recombination has reassorted genetic variation wholesale among clusters. Clusters of this type are recognized in MLST data as clonal complexes, which are defined informally within the MLST databases on the basis of a number of biological properties (Maiden, 2006). Examination of data sets from partially clonal organisms reveals differences in the frequency distribution of STs in that some STs are much more frequent in population samples than others. Typically these STs are also recovered in isolates obtained over prolonged periods of time and from diverse geographic locations. In addition, such STs are often characterized by having many low-frequency relatives that vary from them at one, two, or three loci. These STs correspond to “central genotypes” (sometimes referred to as “founders”) that define and give their name to the clonal complex. The clonal complex then includes all isolates with the central ST plus those with STs sharing up to four identical loci. Thus, the ST45 clonal complex includes ST-45 itself and all of its single, double, and triple locus variants (Fig. 3). In most cases, the variants are generated by recombination rather than point mutation. The influence of recombination on the population structure of C. jejuni is further indicated by the relationship of alleles to STs in the database. Although there were a total of 2,243 distinct STs, organized into 42 clonal complexes at the time of writing, the number of alleles at each locus was much smaller, ranging from 134 (for aspA) to 256 (for pgm). This relationship shows that most genotypes are assembled by reassortment of preexisting alleles, not by the generation of novel alleles by mutation or intra-allele recombination. This in turn implies that the recombination fragment sizes are rather larger than the average size of the MLST alleles (402 to 507 bp in C. jejuni) and, given that these are substantial fractions of the average size of C. jejuni cistrons (948 bp; Parkhill et al., 2000), the genes themselves. This observation is somewhat at odds with some estimates
I
34
MAIDEN AND DINGLE
of average recombination fragment sizes of 225 to 750 bp obtained by approximate likelihood methods (Fearnhead et al., 2005) but consistent with the estimate of 3.3 kbp obtained with apparently less precise methods (Schouls et al., 2003). However, these observations further underline the importance of genetic exchange in the population biology and evolution of Campylobacter. Clonal complexes are surrogates for lineages, and the utility of the clonal complex as a unit of analysis presumably reflects the strength of the lineage structure within populations of C. jejuni. The recent development of the ClonalFrame algorithm (Didelot and Falush, 2007) permits the formal analysis of these lineages for the first time. This program identifies clonal lineages while estimating the likelihood that particular changes have occurred by either recombination or single base-pair mutation, avoiding the problems associated with distance-based methods. The principal disadvantage of this approach is that it is computationally intensive to the extent that it is impractical to apply to all samples in the MLST databases and has to be rerun each time a new ST is added. It is, however, invaluable for testing the robustness of the informal clonal complex designations. Analysis of sample data sets with ClonalFrame shows that C. jejuni clonal complexes correspond to genetic lineages, as predicted. It also confirms the observation that particular clonal complexes are associated with different isolation sources. This probably reflects the fact that clonal complexes represent lineages that are adapted to particular niches that, at least in some cases, correspond to different host species. For example, ST-45 complex C. jejuni are isolated primarily from chickens, environmental sources, or human disease, ST-61 complex mostly from bovine or ovine sources, the environment, or human disease, and ST-42 complex mostly from bovine or ovine sources and human disease (Dingle et al., 2002) (Fig. 4). Some relationships among clonal complexes are apparent from the ClonalFrame analysis; for example, ST-283 complex and ST-45 complex appear to be related, as do ST-48 complex, ST-353 complex, and ST-682 complex (Fig. 4). Most branches, however, converge in a “star phylogeny,” showing that at least for data from seven loci, ClonalFrame analysis cannot resolve deeper phylogenies. This may be a consequence of extensive horizontal genetic exchange disrupting all clonal signal in the population, or it may be a limitation of the numbers of loci used in the analysis. The ST-21 complex is also poorly resolved in this analysis, also appearing as a star phylogeny; this also might be resolved by the inclusion of more loci in future studies. Because the
ST-21 complex is found in a variety of host sources, there may be sublineages within this clonal complex adapted to particular niches. Further support for these ideas comes from the assignment of host types on the basis of MLST data. By use of the structure algorithm (Falush et al., 2003) and a reference set of isolates, the origins of variant alleles in a set of ST-21 complex isolates were assigned and compared with the host origin of the isolate itself. In 18 of 22 cases, the assignment of the variant alleles corresponded with the isolation source, in terms of host animal of the variant strain, indicating that the strain (i) had been resident in the host, and (ii) had preferentially acquired alleles associated with that particular host (McCarthy et al., 2007). Clonal complex, a surrogate for lineage, is therefore a fundamental unit of C. jejuni biology, and much of the confusion present in the field was a consequence of the fact that earlier typing schemes, especially serotyping, did not accurately reflect the clonal complexes (Dingle and Maiden, 2005). The lineages are sufficiently stable to be recovered over periods of time and to be associated with particular hosts. Given that most sampling effort is focused on human disease cases and food animals, it is likely that clonal complexes associated by nonfarm animal sources are poorly represented in current isolate collections, and it is likely that many additional lineages associated with wild animal sources exist (French et al., 2005). In one case a lineage had been identified and named as a discrete subspecies. Isolates characterized as C. jejuni subspecies doylei have been classified as distinct from other C. jejuni isolates on the basis that they are primarily isolated from humans and have a propensity to cause bacteremia. Analysis of MLST data with conventional phylogenetic techniques and microarray-based comparative genome indexing suggested that C. jejuni doylei was distinct from “C. jejuni jejuni” (Parker et al., 2007). This division into two clades is not supported by ClonalFrame analysis, which suggests that the isolates belonging to C. jejuni doylei are just another lineage. This analysis also provides the following explanation for the discrepancy. The C. jejuni doylei isolates are located in a sparsely sampled part of the C. jejuni tree generated by ClonalFrame, which contains isolates mostly of environmental origin. Moreover, the well-sampled part of the tree, which contains mostly isolates derived from humans or food animals, contains a number of major lineages that are related to each other as well as to the ST-21 complex. Thus, a collection of reference isolates that reflected current isolate collections would have very few of the samples of nonhuman and nonhuman food origin, and very many of those from humans and their food animals. There-
CHAPTER 2
206 -
0.02
-4
1275
POPULATION BIOLOGY OF C. lElUN1
35
\ C.doylei
Figure 4. Correspondence of clonal complexes with genealogies. Clonal complex designations were used to annotate a phylogeny generated by ClonalFrame. The size of the pie charts represents the number of isolates in each clonal complex, with the color indicating the relative contribution of each source: black, human disease; white, chickens and chicken meat; dark gray, ovines and ovine meat; light gray, environmental sources. Data from Dingle et al. (2002) and Parker et al. (2007).
fore, analyses comprising C. jejuni doylei and other human disease isolates, however performed, will show C. jejuni doylei to be distinct because other isolates with a similar relationship to the humancentered isolates will be absent. An interpretation that is consistent with both sets of data is that C. jejuni doylei isolates represent a lineage associated with an environmental source that very rarely comes into contact with humans, but when it does, it tends to cause more severe disease. The observations on the potentially misleading characterization of the C. jejuni doylei isolates as a subspecies distinct from all other C. jejuni lineages introduces another implication of the population
structure of C. jejuni. Comparisons of the genomes of multiple isolates, especially by microarray methods, are popular and considered to be important (Champion et al., 2005). However, the inferences that can be gained from such studies are dependent on the genes included in the microarrays used and the isolates included in the analysis. It is known that not all C. jejuni isolates contain all of the genes present in the population, and it is likely that different lineages typically have distinct gene complements. A consequence of this is that any comparison of gene complement that is based on a microarray assembled with only one genome sequence can only provide resolution of the C. jejuni population into two groups:
36
MAIDEN AND DINGLE
“like the reference isolate” and “unlike the reference isolate.” Thus, microarray-based “genomotyping” with microarrays that are based on C. jejuni 11168 can only divide C. jejuni collections into two groups: like NTCC1168 (which happens to be an ST-21 complex isolate) and unlike NTCC1168 (Champion et al., 2005). Adding data from more genomes will help in this process, but the MLST data assembled so far indicate that data from a minimum of 42 isolates (one from each clonal complex identified to date) should be used to generate the microarrays if the association of particular clonal complexes with host sources are to be informatively defined. In general, clonal complex structure should be taken into account when choosing isolates for inclusion in functional analyses or multiple genome sequencing. Although human infection is one of the most important practical applications of studies of Campylobacter populations, in terms of Campylobacter population dynamics and evolution, infection is probably irrelevant. Human infection is almost certainly incidental, and little if any importance should be attached in terms of onward evolutionary spread of C. jejuni. Understanding the population biology of Carnpylobacter is, however, crucial in understanding the transmission to humans and developing means for its control. MLST has permitted direct comparisons to be made among studies of C. jejuni epidemiology (Gillespie et al., 2002). The identification of outbreaks and potential sources of infection relies on the precise characterization of isolates and knowledge of their interrelationships. The clonal complex is a valuable unit of analysis for epidemiological investigation, allowing the comparison of data from diverse sources worldwide. This has demonstrated that different clonal complexes are important in different geographical regions of the world, probably reflecting different food animal sources of human infection (Dingle et al., 2002; Duim et al., 2003; Mickan et al., 2007; Sopwith et al., 2006) (Fig. 4). MLST has also been of value in investigating the link of C. jejuni genotypes with specific disease syndromes (Dingle et al., 2001b). Antigen sequence typing generates high levels of discrimination among bacterial isolates by measuring the variation present in antigen genes that are subject to diversifying selection, presumably by virtue of their exposure to the host immune system. The inclusion of these data are therefore useful in identifying very closely related bacterial isolates, for example in an outbreak situation, and in studies of the immune response to bacteria, but their high rates of change make them less useful in unambiguously identifying lineages. Combining MLST and antigen gene sequence data can be a highly effective means of
achieving high-resolution characterization data that also allow long-range comparisons. For the identification of disease outbreaks, clonal complex designation alone is insufficiently precise to reliably detect outbreaks (Clark et al., 2003; Fitzgerald et al., 2001; Sails et al., 2003b). Sequence typing data also provide the resolution and reproducibility necessary to investigate the diversity of Campylobacter isolates obtained from farms and the environment (Colles et al., 2003; French et al., 2005; Manning et al., 2003). These studies have shown that genotypes most like those from human disease are recovered from farm animals (Colles et al., 2003; Manning et al., 2003), while the genotypes from environmental sources seem to be distinct (Fig. 4) (French et al., 2005). Certain clonal complexes are associated with particular food animal sources, with the ST-61 complex especially associated with cattle (Dingle et al., 2002; French et al., 2005) and the ST-403 with pigs (Manning et al., 2003). Large-scale studies of isolates from a wide variety of sources provide the prospect that in the near future, it will be possible to trace the movement of Campylobacter genotypes through the food chain (Allen et al., 2007) and attribute cases of human infection to particular host sources with reasonable precision (McCarthy et al., 2007). MLST has also been used to examine antibiotic-resistant isolates obtained from poultry (Kinana et al., 2006, 2007).
Campylobacter coli C. coli has a similar host range to C. jejuni and is associated with up to 10% of human Campylobacter infections. At the time of writing, the MLST allelic profile database contained 708 C. coli STs. Preliminary analyses of MLST data from C. coli (Miller et al., 2005) identified two putative clonal complexes among a small collection of 37 STs. The first was designated ST-1017 complex (31STs) and the second ST-1052 complex (4 STs). Now that more data are available, ST-1017 has been shown to be part of the ST-828 complex, and ST-1052 is currently unassigned. Two large C. coli clonal complexes have now been confirmed, the largest being ST-828 complex, with 425 STs, and the other ST-1150 complex, with 25 STs (http: //pubmlst.org/campylobacter/). This indicates the difficulty of identifying the central genotype of a clonal complex when only few data are available. Of the 708 STs currently in the PubMLST database, 258 are unassigned to a clonal complex. These STs were analyzed for the presence of clonal complexes by the BURST algorithm (Feil et al., 2004), and a new large lineage was identified, ST-
CHAPTER 2
1172 complex, containing 130 STs. Two smaller possible clonal complexes, ST-2013 (eight STs) and ST2461 (seven STs), were also identified. Five groups of 4 isolates, two groups of 3 isolates, and 10 paired isolates were identified, with 67 singletons remaining. This suggests that the C. coli population structure is similar to but not identical to C. jejuni, C. coli apparently comprising a smaller number of genetic lineages. The two large clonal complexes identified so far appear to resemble ST-21 complex in C. jejuni in that they are multihost and could potentially be better understood by the sequencing of additional loci. Five studies performed in four different countries have used MLST to study C. coli isolated from food animals and human disease (D’lima et al., 2007; Kinana et al., 2007; Litrup et al., 2007; Thakur and Gebreyes, 2005; Thakur et al., 2006). Pigs are considered a common reservoir for C. coli infections in humans; therefore, several studies have used MLST to examine the genotypes associated with pigs. A MLST-based comparison of 151 C. coli isolates from pigs on the farm and at slaughter demonstrated a high level of genetic diversity, with a total of 74 STs identified among the isolates. Most STs found on the farm were not found at slaughter, but different STs were, suggesting an important role for crosscontamination at slaughter (Thakur and Gebreyes, 2005; Thakur et al., 2006). They also compared isolates from animals reared with and without the use of antibiotics. ST-828 complex was described as containing isolates from both the production systems examined. A study of 36 C. coli isolates from chickens in Senegal described a low level of genetic diversity, with 13 STs, and no clear genetic lineages, probably due to the small size of the data set (Kinana et al., 2007); however, 10 of the STs identified were members of ST-828 complex. There was no apparent link between quinolone resistance and ST. In a study of 160 Danish C. coli isolates from humans, broilers, pigs, and cattle, high genetic diversity was observed, with 84 STs being identified. Ten percent of the isolates possessed STs that were found in human, poultry, and pig isolates. Only 10% of the isolates from pigs shared STs with isolates from humans, and these shared STs were found in poultry isolates as well. Therefore, C. coli isolates from pigs were presumed not to be a significant source of human infections in Denmark (Litrup et al., 2007). A study of 59 multidrug-resistant C. coli strains from turkeys demonstrated that only 3 of the 14 STs accounted for 69% of the isolates, and these STs were turkey specific, implying that multidrug-resistant C. coli strains from turkeys in North Carolina, USA, are characterized by a relatively small number of clonal groupings (D’Lima et al., 2007).
POPULATION BIOLOGY OF C. 7EIUNI
37
The studies by Litrup et al. and D’Lima et al. illustrate the potential of C. coli MLST for tracking the origins of human infection. They also illustrate the requirement for large data sets before host associations start to become apparent. Miller et al. (2006) have made major advances in this respect by typing 488 C. coli strains from four different food animal sources (cattle, chickens, swine, and turkeys) collected at different times over a 6-year period from different geographical locations in the United States. A total of 149 STs were identified. The 185 swine strains were the most diverse, possessing 82 STs. The cattle strains were the most clonal; 52 (83%) of 63 strains possessed a single ST (ST-1068). Most STs and alleles were host associated, i.e., found primarily in strains from a single food-animal source. Only 12 (8%) of 149 STs were found in multiple sources, indicating the potential for MLST to help trace human infections (Miller et al., 2006). C . fetus Population Structure and Molecular Epidemiology
A large and diverse collection of 140 C. fetus isolates, including a wide range of hosts and geographic locations, was used to validate the C. fetus MLST scheme (van Bergen et al., 2005). Fourteen different STs were identified, and these exhibited low levels of inter-ST genetic diversity, with only 22 variable sites in 3,312 nucleotides. The C. fetus MLST data indicated that this species is genetically homogeneous compared with the heterogeneity of other Campylobacter species for which MLST data are available (C. jejuni, C. coli, C. lari, C. upsaliensis, and C. helveticus). A total of 91% of C. fetus isolates was assigned to only four STs. C. fetus is divided into two subspecies: C. fetus subspecies fetus (associated with abortions most often in sheep, and a wide range of other hosts) and C. fetus subspecies venerealis (isolated mainly from cattle). The two C. fetus subspecies were extremely closely related genetically, but ST-4 was associated only with C. fetus subsp. venerealis, which appears to represent a bovine clone. Congruence was observed among C. fetus subspecies, sap types, and ST. MLST therefore confirms that mammalian C. fetus is genetically stable, probably as result of the introduction of a single ancestral clone into a mammalian niche. No associations among the STs and host or geographic region were identified. Other Species The extended MLST method (Miller et al., 2005) was used to type 20 C. lari isolates, 78 C. upsaliensis, and nine C. helveticus. Each species was ge-
38
MAIDEN AND DINGLE
netically diverse, the majority of the STs being unique (http://pubmlst.org/campylobacter/).A total of 15 C. lari STs, 66 C. upsaliensis STs, and 8 C. helveticus STs were identified.
CONCLUSIONS Since the development of MLST 10 years ago (Maiden et al., 1998), major progress has been made in understanding the population biology of C. jejuni and its relatives. The application of high-throughput nucleotide sequence-based methods, and particularly MLST, has not only improved isolate characterization but also illuminated areas of the biology of these highly diverse organisms that are relevant to understanding and controlling disease. It is instructive to reflect that the first study of C. jejuni and C. coli population structure by multilocus enzyme electrophoresis (Aeschbacher and Piffaretti, 1989) provided many insights that have proved to be correct and that have been extended and deepened by MLST studies. The clonal complex has proved to be a powerful and adaptable unit for analysis of population structure that also enables long-term epidemiological monitoring. Including data from antigen-encoding genes provides a highly discriminatory system for tracking genotypes through the food chain and in epidemiological monitoring. The association of particular genotypes with particular host species provides the prospect of accurately determining the relative importance of different sources of human infection. Ongoing nucleotide sequence-based studies involving large numbers of isolates and improved genealogical analysis tools provide the highly attractive prospect that well within the next 10 years, the population biology of these organisms, at least insofar as it relates to human infection, will be effectively resolved. Acknowledgments. M.C.J.M. is a Wellcome Trust Senior Research Fellow in Basic Biomedical Sciences. The United Kingdom Department of Environment, Food and Rural Affairs supported our work (contract numbers 020604 and 020611). We thank Samuel Sheppard for assistance in the preparation of the figures.
REFERENCES Abulreesh, H. H., T. A. Paget, and R. Goulder. 2006. Campylobacter in waterfowl and aquatic environments: incidence and methods of detection. Enuiron. Sci. Technol. 40:7122-7131. Achtman, M. 1996. A surfeit of YATMs? J. Clin. Microbiol. 34: 1870. Achtman, M. 1997. Microevolution and epidemic spread of serogroup A Neisseria meningitidis-a review. Gene 192:135-140. Aeschbacher, M., and J.-C. Piffaretti. 1989. Population genetics of human and animal enteric Campylobacter strains. Infect. lmmun. 57~1432-1437.
Allen, V. M., S. A. Bull, J. E. Corry, G. Domingue, F. Jorgensen, J. A. Frost, R. Whyte, A. Gonzalez, N. Elviss, and T. J. Humphrey. 2007. Campylobacter spp. contamination of chicken carcasses during processing in relation to flock colonisation. Int. ]. Food Microbiol. 11354-61. Balding, D. J., M. J. Bishop, and C. Cannings. 2003. Handbook of Statistical Genetics. John Wiley and Sons Ltd., Chichester, United Kingdom. Bentley, S. D., and J. Parkhill. 2004. Comparative genomic structure of prokaryotes. Annu. Rev. Genet. 38:771-792. Bolton, F. J., and L. Robertson. 1982. A selective medium for isolating Campylobacter jejuniicoli. J. Clin. Pathol. 35 :462-467. Brenner, S. E. 1999. Errors in genome annotation. Trends Genet. 15:132-133. Champion, 0. L., M. W. Gaunt, 0. Gundogdu, A. Elmi, A. A. Witney, J. Hinds, N. Dorrell, and B. W. Wren. 2005. Comparative phylogenomics of the food-borne pathogen Campylobacter jejuni reveals genetic markers predictive of infection source. Proc. Natl. Acad. Sci. USA 102:16043-16048. Chan, M. S., M. C. Maiden, and B. G. Spratt. 2001. Databasedriven multi locus sequence typing (MLST) of bacterial pathogens. Bioinformatics 17: 1077-1 083. Clark, C. G., L. Price, R. Ahmed, D. L. Woodward, P. L. Melito, F. G. Rodgers, F. Jamieson, B. Ciebin, A. Li, and A. Ellis. 2003. Characterization of waterborne outbreak-associated Campylobacter jejuni, Walkerton, Ontario. Emerg. Infect. Dis. 9: 12321241. Colles, F. M., K. Jones, R. M. Harding, and M. C. Maiden. 2003. Genetic diversity of Campylobacter jejuni isolates from farm animals and the farm environment. Appl. Environ. Microbiol. 69: 7409-74 13. Didelot, X., and D. Falush. 2007. Inference of bacterial microevolution using multilocus sequence data. Genetics 175:125 11266. Dingle, K. E., F. M. Colles, D. Falush, and M. C. Maiden. 2005. Sequence typing and comparison of population biology of Campylobacter coli and Campylobacter jejuni. ]. Clin. Microbiol. 43: 340-347. Dingle, K. E., F. M. Colles, R. Ure, J. Wagenaar, B. Duim, F. J. Bolton, A. J. Fox, D. R. A. Wareing, and M. C. J. Maiden. 2002. Molecular characterisation of Campylobacter jejuni clones: a rational basis for epidemiological investigations. Emerg. Infect. Dis. 8:949-955. Dingle, K. E., F. M. Colles, D. R. A. Wareing, R. Ure, A. J. Fox, F. J. Bolton, H. J. Bootsma, R. J. L. Willems, R. Urwin, and M. C. J. Maiden. 2001a. Multilocus sequence typing system for Campylobacter jejuni. 1. Clin. Microbiol. 39: 14-23. Dingle, K. E., and M. C. Maiden. 2005. Population genetics of Campylobacter jejuni, p. 43-58. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter Molecular and Cellular Biology. Horizon Bioscience, Wymondham, United Kingdom. Dingle, K. E., N. van den Braak, F. M. Colles, L. J. Price, D. L. Woodward, F. G. Rodgers, H. P. Endtz, A. van Belkum, and M. C. J. Maiden. 2001b. Sequence typing confirms that Campylobacter jejuni strains associated with Guillain Barrt and Miller Fisher syndromes are of diverse genetic lineage, serotype and flagella type. 1. Clin. Microbiol. 39:3346-3349. D’lima, C. B., W. G. Miller, R. E. Mandrel, S. L. Wright, R. M. Siletzky, D. K. Carver, and S. Kathariou. 2007. Clonal population structure and specific genotypes of multidrug-resistant Campylobacter coli from turkeys. Appl. Environ. Microbiol. 73: 2 156-2 164. Duim, B., P. C. Godschalk, N. van den Braak, K. E. Dingle, J. R. Dijkstra, E. Leyde, J. van der Plas, F. M. Colles, H. P. Endtz, J. A. Wagenaar, M. C. Maiden, and A. van Belkum. 2003. Molecular evidence for dissemination of unique Campylobacter je-
CHAPTER 2
juni clones in Curacao, Netherlands Antilles. 1.Clin. Microbiol. 415593-5597. Duim, B., T. M. Wassenaar, A. Rigter, and J. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with amplified fragment length polymorphism fingerprinting. Appl. Environ. Microbiol. 65:236923 75. Endtz, H. P., J. S. Vliegenthart, P. Vandamme, H. W. Weverink, N. P. van den Braak, H. A. Verbrugh, and A. van Belkum. 1997. Genotypic diversity of Campylobacter lari isolated from mussels and oysters in The Netherlands. Int. J. Food Microbiol. 34:7988. Falush, D., M. Stephens, and J. K. Pritchard. 2003. Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164: 1567-1587. Fearnhead, P., N. G. Smith, M. Barrigas, A. Fox, and N. French. 2005. Analysis of recombination in Campylobacter jejuni from MLST population data. J. Mol. Evol. 61:333-340. Feil, E. J., B. C. Li, D. M. Aanensen, W. P. Hanage, and B. G. Spratt. 2004. eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J. Bacteriol. 186:1518-1530. Feil, E. J., M. C. J. Maiden, M. Achtman, and B. G. Spratt. 1999. The relative contributions of recombination and mutation to the divergence of clones of Neisseria meningitidis. Mol. Biol. Evol. 16:1496-1502. Fitzgerald, C., L. 0. Helsel, M. A. Nicholson, S . J. Olsen, D. L. Swerdlow, R. Flahart, J. Sexton, and P. I. Fields. 2001. Evaluation of methods for subtyping Campylobacter jejuni during an outbreak involving a food handler. J. Clin. Microbiol. 39:23862390. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S . C. Daugherty, R. J. Dodson, A. S . Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple campylobacter species. PLoS Biol. 3:e15. French, N., M. Barrigas, P. Brown, P. Ribiero, N. Williams, H. Leatherbarrow, R. Birtles, E. Bolton, P. Fearnhead, and A. Fox. 2005. Spatial epidemiology and natural population structure of Campylobacter jejuni colonizing a farmland ecosystem. Environ. Microbiol. 7:1116-1126. Garcia, M. M., G. M. Ruckerbauer, M. D. Eaglesome, and W. E. Boisclair. 1983. Detection of Campylobacter fetus in artificial insemination bulls with a transport enrichment medium. Can. J. Comp. Med. 47:336-340. Gillespie, I. A., S . J. O’Brien, J. A. Frost, G. K. Adak, P. Horby, A. V. Swan, M. J. Painter, K. R. Neal, and C. S. S. S. Collaborators. 2002. A case-case comparison of Campylobacter coli and Campylobacter jejuni infection: a tool for generating hypotheses. Emerg. Infect. Dis. 8:937-942. Joky, K. A., M. S . Chan, and M. C. Maiden. 2004. mlstdbNetdistributed multi-locus sequence typing (MLST) databases. BMC Bioinfomzatics 5:86. Joky, K. A., E. J. Feil, M. S . Chan, and M. C. Maiden. 2001. Sequence type analysis and recombinational tests (START). Bioinfomzatics 17: 1230-123 1. Joky, K. A., and M. C. Maiden. 2006. AgdbNet-antigen sequence database software for bacterial typing. BMC Bioinformatics 7:314. Kinana, A. D., E. Cardinale, I. Bahsoun, F. Tall, J. M. Sire, S. Breurec, B. Garin, C. Saad-Bouh Boye, and J. D. Perrier-GrosClaude. 2007. Campylobacter coli isolates derived from chickens
-
POPULATION BIOLOGY OF C. TETUNZ
39
in Senegal: diversity, genetic exchange with Campylobacter jejuni and quinolone resistance. Res. Microbiol. 158:138-142. Kinana, A. D., E. Cardinale, F. Tall, I. Bahsoun, J. M. Sire, B. Garin, S. Breurec, C. S . Boye, and J. D. Perrier-Gros-Claude. 2006. Genetic diversity and quinolone resistance in Campylobacter jejuni isolates from poultry in Senegal. Appl. Environ. Microbiol. 72:3309-3313. Levin, B. R. 1981. Periodic selection, infectious gene exchange and the genetic structure of E. coli populations. Genetics 99:l-23. Litrup, E., M. Torpdahl, and E. M. Nielsen. 2007. Multilocus sequence typing performed on Campylobacter coli isolates from humans, broilers, pigs and cattle originating in Denmark.1. Appl. Microbiol. 103:210-218. Maiden, M. C. 2006. Multilocus sequence typing of bacteria. Annu. Rev. Microbiol. 60561-588. Maiden, M. C. J. 1993. Population genetics of a transformable bacterium: the influence of horizontal genetical exchange on the biology of Neisseria meningitidis. FEMS Microbiol. Lett. 112: 243-250. Maiden, M. C. J. 2000. High-throughput sequencing in the population analysis of bacterial pathogens. Int. J. Med. Microbiol. 290~183-190. Maiden, M. C. J., J. A. Bygraves, E. Feil, G. Morelli, J. E. Russell, R. Urwin, Q. Zhang, J. Zhou, K. Zurth, D. A. Caugant, I. M. Feavers, M. Achtman, and B. G. Spratt. 1998. Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc. Natl. Acad. Sci. USA 95:3140-3145. Manning, G., C. G . Dowson, M. C. Bagnall, I. H. Ahmed, M. West, and D. G. Newell. 2003. Multilocus sequence typing for comparison of veterinary and human isolates of Campylobacter jejuni. Appl. Environ. Microbiol. 69:6370-6379. Maynard Smith, J. 1989. Trees, bundles or nets. Trends Ecol. Evol. 4 :3 02-3 04. Maynard Smith, J., C. G. Dowson, and B. G. Spratt. 1991. Localized sex in bacteria. Nature 349:29-31. Maynard Smith, J., N. H. Smith, M. O’Rourke, and B. G. Spratt. 1993. How clonal are bacteria? Proc. Natl. Acad. Sci. USA 90: 4384-4388. McCarthy, N. D., F. M. Colles, K. E. Dingle, M. C. Bagnall, G. Manning, M. C. Maiden, and D. Falush. 2007. Host-associated genetic import in Campylobacter jejuni. Emerg. Infect. Dis. 13: 267-272. Mickan, L., R. Doyle, M. Valcanis, K. E. Dingle, L. Unicomb, and J. Lanser. 2007. Multilocus sequence typing of Campylobacter jejuni isolates from New South Wales, Australia. J. Appl. Microbiol. 102:144-152. Miller, W. G., M. D. Englen, S . Kathariou, I. V. Wesley, G . Wang, L. Pittenger-Alley, R. M. Siletz, W. Muraoka, P. J. FedorkaCray, and R. E. Mandrell. 2006. Identification of host-associated alleles by multilocus sequence typing of Campylobacter coli strains from food animals. Microbiology 152:(Pt. 1):245-255. Miller, W. G., S . L. On, G. Wang, S . Fontanoz, A. J. Lastovica, and R. E. Mandrell. 2005. Extended multilocus sequence typing system for Campylobacter coli, C. lari, C. upsaliensis, and C. helveticus. J. Clin. Microbiol. 43:2315-2329. Narra, H. P., and H. Ochman. 2006. Of what use is sex to bacteria? Cum Biol. 16:R705-R710. Parker, C. T., W. G. Miller, S . T. Horn, and A. J. Lastovica. 2007. Common genomic features of Campylobacter jejuni subsp. doylei strains distinguish them from C. jejuni subsp. jejuni. BMC Microbiol. 7 5 0 . Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S . Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H.
40
MAIDEN AND DINGLE
van Vliet, S . Whitehead, and B. G. Barrel]. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Sails, A. D., B. Swaminathan, and P. I. Fields. 2003a. Clonal complexes of Campylobacter jejuni identified by multilocus sequence typing correlate with strain associations identified by multilocus enzyme electrophoresis. J. Clin. Microbiol. 41:4058-4067. Sails, A. D., B. Swaminathan, and P. I. Fields. 2003b. Utility of multilocus sequence typing as an epidemiological tool for investigation of outbreaks of gastroenteritis caused by Campylobacter jejuni. J. Clin. Microbiol. 41:4733-4739. Schouls, L. M., S . Reulen, B. Duim, J. A. Wagenaar, R. J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41:15-26. Selander, R. K., D. A. Caugant, H. Ochman, J. M. Musser, M. N. Gilmour, and T. S . Whittam. 1986. Methods of multilocus enzyme electrophoresis for bacterial population genetics and systematics. Appl. Environ. Microbiol. 515337-884. Sopwith, W., A. Birtles, M. Matthews, A. Fox, S . Gee, M. Painter, M. Regan, Q. Syed, and E. Bolton. 2006. Campylobacter jejuni multilocus sequence types in humans, northwest England, 20032004. Emerg. Infect. Dis. 12:1500-1507. Spratt, B. G., and M. C. J. Maiden. 1999. Bacterial population genetics, evolution and epidemiology. Proc. R. SOC. Lond. B Biol. Sci. 354:701-710.
Stackebrandt, E., W. Frederiksen, G. M. Garrity, P. A. Grimont, P. Kampfer, M. C. Maiden, X. Nesme, R. Rossello-Mora, J. Swings, H. G. Truper, L. Vauterin, A. C. Ward, and W. B. Whitman. 2002. Report of the ad hoc committee for the reevaluation of the species definition in bacteriology. Int. /. Syst. Bacteriol. 52(Pt. 3):1043-1047. Suerbaum, S., M. Lohrengel, A. Sonneveld, F. Ruberg, and M. Kist. 2001. Allelic diversity and recombination in Campylobacter jejuni. J. Bacteriol. 183:2553-2559. Thakur, S., and W. A. Gebreyes. 2005. Campylobacter coli in swine production: antimicrobial resistance mechanisms and molecular epidemiology. J. Clin. Microbiol. 435705-5714. Thakur, S., W. E. Morrow, J. A. Funk, P. B. Bahnson, and W. A. Gebreyes. 2006. Molecular epidemiologic investigation of Campylobacter coli in swine production systems, using multilocus sequence typing. Appl. Environ. Microbiol. 725666-5669. van Bergen, M. A., K. E. Dingle, M. C. Maiden, D. G. Newell, L. van der Graaf-Van Bloois, J. P. van Putten, and J. A. Wagenaar. 2005. Clonal nature of Campylobacter fetus as defined by multilocus sequence typing. J. Clin. Microbiol. 435888-5898. Wassenaar, T. M., and D. G. Newell. 2000. Genotyping of Campylobacter species. Appl. Environ. Microbiol. 66: 1-9. Yan, W., N. Chang, and D. E. Taylor. 1991. Pulsed-field gel electrophoresis of Campylobacter jejuni and Campylobacter coli genomic DNA and its epidemiologic application. J. Infect. Dis. 163:1068-1072.
Cumpylobucter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 3
Complexity and Versatility in the Physiology and Metabolism of
Campylobacter jejuni DAVIDJ. KELLY
UNDERSTANDING PHYSIOLOGY AND METABOLISM HELPS TO UNDERSTAND THE COLONIZATION AND DISEASE PROCESS
comparative genomics. The challenge now is to use this information to discover which metabolic pathways or enzyme activities are essential in the different environments that C. jejuni is capable of surviving or growing in. This will give us insights into the strategies it uses for successful colonization, and it will suggest how C. jejuni can be a commensal in the avian gastrointestinal tract, a pathogen in the human gastrointestinal tract, and a persistent survivor in the environment and the food chain. In these different situations, the metabolism of the bacteria may be significantly altered, yet these crucial changes are poorly understood at present. In this chapter, the current knowledge on some of the metabolic aspects of C. jejuni physiology, with emphasis on those features of carbon, nitrogen, and electron flow that are likely to be of importance in understanding growth in the environment and in vivo, will be reviewed. The chapter will focus on catabolic pathways, i.e., those involved in the breakdown of extracellular solutes, yielding energy and key intracellular intermediates that are the building blocks for new cell growth. Since I last reviewed this area (Kelly, 2005), a number of important papers have been published. I have endeavored to include most of these, but because of space limitations, the material is necessarily selective.
The recent availability of the genome sequences of several campylobacters, particularly Campylobacter jejuni, has refocused attention on the physiological and metabolic properties of this medically and economically important group of the Epsilonproteobacteria. For many years, such studies have taken second place to work on virulence and colonization, but it is now becoming clear that the particular metabolic properties of any pathogen are intimately related to its ability to cause disease, and a greater understanding of these relationships is crucial to gain a full picture of pathogenicity. The campylobacters can be difficult organisms to culture, often have complex nutritional requirements, and are microaerophilic, and this combination of features has undoubtedly hindered progress in elucidating many aspects of their physiology and metabolism. Nevertheless, such studies are central to understanding growth in animal and human hosts, survival in the environment and the food chain, and mechanisms of pathogenicity, as well as what carbon sources are being used in vivo. Physiological studies may also provide leads in identifying potential targets for controlling or preventing the growth of campylobacters. The determination of the complete genome sequence of C. jejzkni NCTC 11168 (Parkhill et al., 2000) provided many new insights into its biology, not least in terms of metabolism. Complete genome sequences are now available for additional diverse strains of C. jejuni, for example 81-176 (Hofreuter et al., 2006), RM1221 (Fouts et al., 2005), and CG8486, a recent clinical isolate from Thailand (Poly et al., 2007), which provide a wealth of material for
TRANSPORTING SOLUTES IN AND OUT OF CAMPYLOBACTER JEJUNI Overview
Solute transport is a fundamental prerequisite for metabolism in the cytoplasm of hydrophilic and polar growth substrates. Active solute transport sys-
David J. Kelly Department of Molecular Biology and Biotechnology, University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, United Kingdom.
41
42
KELLY
tems can be broadly classified into two major groups on the basis of their energy-coupling mechanisms. Primary systems are often multicomponent and depend on direct hydrolysis of chemical bonds to release free energy for solute transport, often ATP hydrolysis. In contrast, secondary systems are usually simpler (sometimes a single transport protein) and use a preexisting electrochemical ion-gradient across the cytoplasmic membrane to power solute uptake, often the proton motive force or a sodium ion gradient. C. jejuni contains many types of both classes of transport systems. As an example, the NCTC 11168 genome contains about 92 distinct transport proteins, 37% of which form primary systems and 57% of which form secondary systems. Most primary systems are accounted for by ABC-type transporters (about 30 complete systems, both importers and exporters), with the remaining primary systems comprising three P-type ATPases (the cation transporters Cjll5.S and Cj1161, and the Kdp potassium transporter) and one F-type ATPase (the proton-translocating ATP synthase). Of the secondary transporters, most fall into the major facilitator superfamily group, consisting of a single transport protein, but there are one or more representatives of about 20 other distinct families of secondary transporters, each of which will have specialized roles in solute movements. It is often difficult to accurately predict the substrates used by transporters from sequence homologies. As a general observation, we can conclude that C. jejuni NCTC 11168, for example, has few carbohydrate transporters, but many systems for amino acid and organic acid transport. This reflects the asaccharolytic nature of its metabolism and its reliance on amino acids as major carbon sources (discussed below). Here, there is only space to describe a selection of the transport systems present that relate to the major metabolic pathways in C. jejuni for which some functional data are available. Amino Acid Transport As described below, amino acids are thought to be the most important carbon and energy sources for C. jejuni, and data are now accumulating on the transport of these substrates, which suggests a number of interesting connections between pathogenicity and metabolism. This is particularly evident in the dual roles of several amino acid binding proteins, which are also immunogenic surface protein antigens and adhesins. PEBla and the PEBl transport system: aspartate and glutamate transport In a study designed to identify proteins suitable for use in serology and as vaccine candidates, Pei et
al. (1991) identified four proteins of 28, 29, 30, and 31 kDa in acidic glycine extracts of whole cells of strain 81-176. These were termed PEBl to PEB4, respectively. The structural gene for PEBl (pebla)was cloned by Pei and Blaser (1993) by using polyclonal antibody raised against purified PEBla protein in a screen of a hgtl 1 library of genomic DNA from strain 81-176. The gene was mutated, and the null mutant strain showed 50- to 100-fold less adherence to and 15-fold less invasion of epithelial cells in culture (Pei et al., 1998). Moreover, mouse challenge studies indicated that the rate and duration of intestinal colonization was significantly lower and shorter, respectively, compared with the wild-type strain. These data have established the PEBla protein as an important colonization and virulence factor in C. jejuni. Western blot analysis showed that PEBla is present in all C. jejuni and C. coli strains examined, but it appears to be absent from C. fetus, C. lari, and Helicobacter pylori (Pei et al., 1991). From database searching that used the deduced amino acid sequence of PEBla, Pei and Blaser (1993) showed that the protein is homologous to the periplasmic-binding protein component of amino acid ABC transporters, for example GlnH (27.8% identity) and HisJ (28.9% identity). A linked open reading frame encodes a protein with homology with the membrane proteins GlnQ and HisP, and the genome sequence of strain NCTC 11168 clearly reveals that pebla in this strain (CjO921c) is part of a four-gene operon encoding a typical ABC transporter. A potential dual role for PEBla as both a surface-exposed adhesin and a periplasmic solute binding protein can be rationalized because of the unusual existence of two predicted processing sites in the signal sequence for signal peptidase I and 11, which may be responsible for the localization of the protein both in the periplasm and on the cell surface (Pei and Blaser, 1993). PEBla is encoded by CjO921c in the C. jejuni NCTC 11168 genome (Parkhill et al., 2000). Overexpression and purification of the recombinant form of Cj092 1 followed by steady-state fluorescence titrations with all 20 amino acids (Leon-Kempis et al., 2006) showed that the protein specifically bound Laspartate and L-glutamate with low (micromolar) Kd values but had a low affinity for L-asparagine and Lglutamine. The other amino acids did not bind. Significantly, a CjO921c mutant was completely deficient in glutamate transport, showed a -20-fold reduction in aspartate transport compared with the parent strain, and was unable to grow on either aspartate or glutamate as a carbon source in minimal media (Leon-Kempis et al., 2006). It is thus clear that PEBla is the periplasmic component of an aspartate/gluta-
CHAPTER 3
mate ABC transporter, essential for optimal microaerobic growth on these dicarboxylic amino acids. The vast majority of PEBla in C. jejuni was found to be periplasmic (Leon-Kempis et al., 2006), consistent with its solute binding function, but some must be surface exposed in order for the protein to act as an adhesin. The presence of a signal peptidase I1 recognition site might suggest that the protein could be anchored in the outer membrane, but no evidence for this was found from fractionation and localization studies (Leon-Kempis et al., 2006). Rather, the presence of PEBla in acidic glycine extracts and in concentrated culture supernatant samples supports the suggestion of Pei and Blaser (1993) that the protein can be exported across the outer membrane. Although the mechanism for this is unclear, the two-component signal peptide of PEBla is extremely similar to that of endoglucanase (Egl) from Pseudomonas solanacearum (Huang and Schell, 1992; Pei and Blaser, 1993), the two-step export of which has been studied in detail. Egl is first directed across the inner membrane to the periplasm by the lipoprotein signal peptide, then is further cleaved at the signal peptidase I site during export across the outer membrane (Huang and Schell, 1992). A similar process may apply to PEBla, although in this case, most of the fully mature protein appears to remain in the periplasm. Recently, the crystal structure of PEBla at 1.5 8, resolution has been obtained (Miiller et al., 2007), allowing insights into the mechanism of solute binding. The protein has a two-domain a / p structure, characteristic of periplasmic extracytoplasmic solute receptors and a chain topology related to the type I1 subfamily. An aspartate ligand, clearly defined by electron density in the interdomain cleft, forms extensive polar interactions with the protein, the majority of which are made with the larger domain. Arg89 and Asp174 form ion-pairing interactions with the main chain a-carboxyl and a-amino groups, respectively, of the ligand, while Arg67, Thr82, Lysl9, and Tyrl56 coordinate the ligand side chain carboxyl. Lysl9 and Arg67 line a positively charged groove, which favors binding of Asp over the neutral Asn. The ligand-binding cleft is also of sufficient depth to accommodate a glutamate. This is the first structure to be determined of an ABC-type aspartate binding protein and explains the high affinity of the protein for aspartate and glutamate and its much weaker binding of asparagine and glutamine. Stopped-flow fluorescence spectroscopy indicates a simple bimolecular mechanism of ligand binding, with high association-rate constants (Miiller et al., 2007). Interestingly, sequence alignments and phylogenetic analyses revealed PEBla homologues to be
PHYSIOLOGY AND METABOLISM OF C. TETUNI
43
present in some gram-positive bacteria. The alignments also suggested a more distant homology with GltI from Escherichia coli, a known glutamate and aspartate binding protein, but Lysl9 and Tyrl56 are not conserved in GltI. These results provide a structural basis for understanding both the solute transport and adhesin/virulence functions of PEBla.
CjaA: cysteine transport Interestingly, PEBla is not the only differentially localized solute binding protein in C. jejuni with an unusual signal peptide. Another C. jejuni immunodominant surface antigen and vaccine candidate, CjaA (Cj0982), has been shown to be a cysteine binding protein component of an ABC system, and the three-dimensional structure of this protein has been determined (Miiller et al., 2005). The signal peptide of CjaA contains a strongly predicted recognition sequence for signal peptidase I1 but a much less convincing atypical signal peptidase I site, so the sorting mechanism may not be the same as for PEBla. Indeed, preliminary data suggest that CjaA is mainly located on the periplasmic side of the inner membrane (Godlewska et al., 2005). CjaA is localized in an operon with CjaB, which encodes a predicted solute:proton symporter (Wyszynska et al., 2006), but there is no evidence that CjaB has a functional role in cysteine transport. Instead, the predicted membrane permease and ABC protein subunits that could make up a functional ABC-type cysteine transporter along with CjaA appear to be localized in an unlinked region of the chromosome, and they have been detected by bioinformatic comparisons with other bacteria (Miiller et al., 2005). Serine transport As described below, serine is a good carbon source for the growth of C. jejuni. The genome sequence of NCTC 11168 reveals a two-gene operon (Cj1624c and Cj1625c) encoding homologues of the E. coli serine dqhydratase (SdaA) and serine transporter (SdaC) for L-serine transport and catabolism. The hydropathy profile of SdaC shows that it contains 10 or 11 potential membrane-spanning segments, and its overall topology is similar to the E. coli SdaC protein and other members of the hydroxyl aromatic amino acid permease family. Biochemical evidence suggests that serine transport in C. jejuni is actually carried out by at least two systems, a lowaffinity L-serine-specific transporter encoded by sdaC, and a high-affinity transporter, which was revealed by analysis of the residual serine transport kinetics in an sdaC mutant (Velayudhan et al., 2004). The iden-
44
KELLY
tity of the high-affinity transporter is unknown, but competition studies with other amino acids suggest it could be a dual serine-threonine transporter. This type of system is known in other bacteria-for example the Na+-coupled serine-threonine uptake system in E. coli (Hama et al., 1987). Organic acid transport systems There appears to be a multiplicity of transporters for organic acids encoded in the C. jejuni genome; at least nine distinct secondary systems can be deduced in strain NCTC 11168. The physiological roles for many of these are not clear, but various strains of C. jejuni have been shown to grow on pyruvate (Velayudhan and Kelly, 2002), citrate, malate, fumarate, succinate, and lactate (Hinton, 2006), all of which will require active transport across the cytoplasmic membrane. C4-dicarboxylate transport (malate, succinate, fumarate, and possibly also aspartate) seems to be particularly important in C. jejuni. These substrates can feed directly into the citric acid cycle (CAC), malate and succinate can act as direct electron donors for aerobic respiration, and fumarate is an alternative electron acceptor under oxygen-limiting conditions. There are three genes (CjOO25c, Cj1097, and Cj 1192) that encode members of the dicarboxylate/ amino acid: cation symporter family. Cj1192 is a likely DctA homologue, well known in other bacteria to transport malate/succinate/fumarate for catabolism under aerobic conditions. DcuA (Cj088) and DcuB (Cj0671) are paralogous and partially redundant C4-dicarboxylate secondary uptake systems that are known to function in fumarate:succinate antiport during anaerobic fumarate respiration in enteric bacteria like E. coli (Unden and Kleefeld, 2004). Because C. jejuni is known to carry out fumarate respiration (see below) and the active site of fumarate reduction by the fumarate reductase catalytic subunit FrdA is on the cytoplasmic side of the membrane, the Dcu proteins are essential for this process. Moreover, by acting as electroneutral antiporters, by using the outwardly directed succinate concentration gradient to power the uptake of fumarate, they provide a bioenergetically efficient solution to the need to deliver fumarate to the inside of the cell. Significantly, microarray and reverse transcriptase PCR studies have shown that both dcuA and dcuB are upregulated in expression in C. jejuni grown in the chick gut compared with laboratory media (Woodall et al., ZOOS), suggesting that fumarate respiration may be of increased importance in this oxygen-limited in vivo environment (see below). E. coli also contains another Dcu protein, DcuC. The role of this is thought to be
as a dedicated succinate efflux system during fermentative growth (Unden and Kleefeld, 2004). Interestingly, strains NCTC 11168 and RM1221 each contain a dcuC pseudogene, while in strain 81-176, there are apparently two functional full-length dcuC genes in addition to dcuA and dcuB homologues (Hofreuter et al., 2006). This suggests an increased role for dicarboxylate metabolism in 8 1-176. Whether this might relate to the increased pathogenicity of this strain is not yet clear.
CENTRAL CARBON METABOLISM IN C. JEJUNI Gluconeogenesis and Anaplerotic Reactions Genes encoding enzymes of the Embden-Meyerhof (EM) and pentose phosphate pathways are present in the genome of the currently sequenced strains of C. jejuni, but these strains lack the key enzymes of the Entner-Doudoroff pathway. Most of the reactions of the EM pathway are reversible in vivo, with the notable exceptions of the 6-phosphofructokinase and pyruvate kinase (PYK) reactions, both of which constitute key control points in many organisms. The C. jejuni strains sequenced thus far lack a gene that could encode a 6-phosphofructokinase, which would account for the well-known inability of this species to catabolize glucose (Velayudhan and Kelly, 2002; Kelly, 2005). However, a gene encoding fructose-1,6bisphosphatase is present, which suggests that the EM pathway functions solely in gluconeogenesis. Surprisingly, a homologue of PYK (Cj0392 in strain NCTC 11168) is present in the C. jejuni genome. This is unexpected because PYK catalyzes the physiologically irreversible conversion of phosphoenolpyruvate (PEP) and ADP to pyruvate and ATP at the final stage of the glycolytic pathway (Fig. 1).PYK is thus a catabolic enzyme that should not be required if the only function of the EM pathway in C. jejuni is in gluconeogenesis. A p y k mutant displayed a slightly decreased growth rate in complex media but was able to grow with pyruvate, lactate, and malate as carbon sources in defined media (Velayudhan and Kelly, 2002). PYK is present in cell extracts at a high specific activity (>800 nmol min-' mg-' protein), is activated by fructose-l,6-bisphosphate,and displays other regulatory properties strongly indicative of a catabolic role. Therefore, PYK may function in the catabolism of unidentified substrates, which are metabolized through PEP. Thus, the lower part of the EM pathway in C. jejuni may have a catabolic role in some situations (e.g., during in vivo growth) even though the absence of 6-phosphofructokinase does not allow the full catabolism of hexose sugars. In this
CHAPTER 3
PHYSIOLOGY AND METABOLISM OF C. 7E7UNI
45
Serine I
1
+Acetate A
proline
PutA
\ glutama&e
cytoplasm
i GSSG --GG-T Glu
Proline
i
Glutamate Aspartate
Glutathione (GSSG) Figure 1. Major pathways of central carbon metabolism and amino acid utilization in C. jejuni. The cell is represented with an outer membrane (OM) and inner membrane (IM), enclosing the periplasm, and a cytoplasm within which the major pathways of carbon metabolism are shown. Key amino acid transport systems are shown as black rectangles in the inner membrane. The black circle emphasizes the transamination reaction that converts glutamate to aspartate. The major enzymes are shown next to the reactions catalyzed. SdaC, serine transporter; SdaA, serine dehydratase; Pyk, pyruvate kinase; Pyc, pyruvate carboxylase; Pck, PEP carboxykinase; Por, pyruvate:acceptor oxidoreductase; Acs, acetyl-CoA synthetase; Pta, phosphotransacetylase; AckA, acetate kinase; GltA, citrate synthase; Acn, aconitase; Icd, isocitrate dehydrogenase; Oor, 2oxog1utarate:acceptor oxidoreductase; SUC,succinyl-Cod synthetase; Sdh, succinate dehydrogenase; Fum, fumarase; Mqo, ma1ate:quinone oxidoreductase; Mdh, malate dehydrogenase (NAD linked); AspA, aspartase; Aat, aspartate:glutamate aminotransferase; GlnA, gluramine synthase; GltBD, glutamate synthase; PutA, proline dehydrogenase; PutP, proline transporter; GGT, y-glutarnyl transpeptidase. The conversion of glutathione to glutamate ocurrs in the periplasm of some strains only (dotted arrow). The Pebl system is an ABC transporter containing the periplasmic aspartateightamate binding protein Pebla. Fld, flavodoxin; Fd, ferredoxin. OAA, oxaloacetate; PEP, phosphoenol pyruvate; AcP, acetyl-phosphate.
context, it is interesting to note that although a pseudogene in NCTC 11168, strain 81-176 contains a glpT gene encoding a glycerol-3-phosphate transporter homologue (Hofreuter et al., 2006), which may allow the latter strain to grow on this substrate, which can be catabolized by the lower part of the glycolytic pathway. Net synthesis of oxaloacetate (OM) from pyruvate or PEP is required to permit the CAC to fulfill
both biosynthetic and energy-conserving roles. C. jejuni possesses homologues of anaplerotic enzymes that may function in this capacity (Fig. l),the activities of some of which have been detected by nuclear magnetic resonance (Mendz et al., 1997) and direct assay (Velayudhan and Kelly, 2002). Pyruvate carboxylase (PYC) is a key anaplerotic enzyme encoded by pycA (Cj1037c) and pycB (CjO933c) homologues, which correspond to the biotin carboxylase and bi-
46
KELLY
otin carboxyl carrier subunits, respectively, of the two-subunit type of carboxylase (Goss et al., 1981). PYC fulfills an important function in many organisms by catalyzing the ATP-dependent carboxylation of pyruvate to yield OAA. Phosphoenolpyruvate carboxykinase (PCK) provides the PEP required for gluconeogenesis and catalyzes the ATP-dependent decarboxylation of OAA to PEP. C. jejuni has a pckA gene (CjO932c), and interestingly, it is contiguous with the pycB homologue. In addition to PYC and PCK, another predicted decarboxylating/carboxylating enzyme located at the metabolic node around pyruvate is malate oxidoreductase (malic enzyme [MEZ]). MEZ catalyzes the oxidative decarboxylation of malate to pyruvate coupled with NAD(P) reduction. Cj 1287c encodes a probable malate oxidoreductase homologous to that of Bacillus stearothermophilus (Parkhill et al., 2000). Insertion mutants in pycA, pycB, and mez have been generated (Velayudhan and Kelly, 2002). However, this could not be achieved for pckA, indicating that PCK could be an essential enzyme in C. jejuni. The lack of PEP synthase and pyruvate orthophosphate dikinase activities in cell-free extracts suggest a unique role for PCK in PEP synthesis. A pycA mutant was able to grow normally in complex media but was unable to grow in defined media with pyruvate or lactate as major carbon sources, thus confirming an important role for PYC in anaplerosis. In view of the high K, of MEZ for malate (-9 mM) and the lack of a growth phenotype of the rnez mutant, MEZ seems to have only a minor anaplerotic role in C. jejuni (Velayudhan and Kelly, 2002). A Complete CAC with Some Characteristics of Anaerobes The genome sequences of the strains of C. jejuni so far sequenced indicate the presence of a complete oxidative CAC (Fig. l),with all of the key enzymes present including homologues of succinate dehydrogenase or a type B fumarate reductase (which could also act in reverse as a succinate dehydrogenase), the a and /3 subunits of succinyl-CoA synthetase (SCS; sucAB genes) and an NAD-linked malate dehydrogenase and a ma1ate:quinone oxidoreductase. Importantly however, genes for the pyruvate and 2-oxoglutarate dehydrogenase complexes usually found in respiratory aerobes are absent. In fact, C. jejuni uses unrelated enzymes more commonly found in obligately anaerobic bacteria to perform the same reactions. This has consequences for understanding the bioenergetics of campylobacters and so will be discussed in some detail.
The entry of carbon from pyruvate into the cycle requires its oxidative decarboxylation to acetyl-CoA, which, in most aerobic bacteria, is carried out by the pyruvate dehydrogenase multienzyme complex, with NAD used as an electron acceptor. However, in many obligate anaerobes, a flavodoxin- or ferredoxin-dependent pyruvate:acceptor oxidoreductase (POR) catalyzes this reaction. Daucher and Kreig (1995) showed that many species and strains of Campylobacter possess this latter type of enzyme, which can be detected by using benzyl or methyl viologen as an artificial electron acceptor in cell-free extracts. The significance of this is that such acceptor: oxidoreductases are iron-sulfur-cluster-containing enzymes, which are usually sensitive to inactivation by molecular oxygen or reactive oxygen species. This has been clearly shown in the related H. pylori by the oxygen lability of the purified pyruvate:flavodoxin oxidoreductase (Hughes et al., 1995). In addition, both H. pylori and C. jejuni contain a related oxygen labile CAC enzyme, 2-oxog1utarate:acceptor oxidoreductase (OOR; Hughes et al., 1998; Kelly, 2001), which replaces the function of the 2-oxoglutarate dehydrogenase multienzyme complex found in conventional aerobes. Campylobacter jejuni is a true microaerophile that exhibits oxygen-dependent growth but is unable to grow at normal atmospheric oxygen tensions, and it has been proposed that the presence of these proteins might contribute to the microaerophilic phenotype of both Helicobacter and Campylobacter (Kelly, 1998, 2001, 2005). However, there is an interesting difference between the POR of C. jejuni and H. pylori in that the former enzyme is a large single-subunit protein containing four separate functional domains that are present as separate gene products in H. pylori. In contrast, the OORs of both bacteria are similar four-subunit enzymes. The situation in C. jejuni and the fact that there is only a rather low overall sequence similarity between the POR and OOR subunit sequences in H. pylori indicate that these two enzymes have evolved independently. An important question is how the reduced flavodoxin or ferredoxin produced in the POR/OOR reactions is reoxidized and thus how the substrate derived electrons are used by the bacteria for energy conservation. In H. pylori, it was shown that in cellfree extracts in the presence of CoA and flavodoxin, pyruvate could support the reduction of NADP, even though NADP is not a substrate of the purified POR enzyme (Hughes et al., 1998). This suggested that an additional redox protein might mediate the transfer of electrons from POR-reduced flavodoxin to NADP, which might then be used as a respiratory substrate. A likely candidate for this enzyme has recently been identified in H. pylori as HP1164, a flavodoxin:qui-
CHAPTER 3
none reductase that can catalyze flavodoxin-dependent NADP reduction in vitro and that appears to be unique to and highly conserved in many Epsilonproteobacteria, including C. jejuni (St Maurice et al., 2007). In NCTC 11168, the flavodoxin:quinone reductase homologue is Cj0559 (St Maurice et al., 2007), but it has not yet been shown whether this might mediate the same reaction. An alternative route for the electrons derived from pyruvate or 2-0x0glutarate via ferredoxin/ flavodoxin is a more direct interaction with the respiratory chain. As discussed below, an interesting possibility is via the unusual NDH-1 complex, which does not contain NAD(P)Hbinding subunits in C. jejuni and related bacteria.
AMINO ACID CATABOLISM AND NITROGEN ASSIMILATION The main carbon sources used by C. jejuni in vivo are likely to be amino acids because the bacterium is unable to metabolize exogenous sugars but does have the transport and enzymatic capacity for amino acid catabolism. There are essentially four enzymatic amino acid catabolic mechanisms that might be operative, depending on the particular chemistry of the amino acid side chain: (i) deamination (oxidative or nonoxidative), (ii) transamination via an oxo-acid acceptor like 2-oxoglutarate or OAA, (iii) p elimination of ammonia via dehydratases or amino acid:ammonia lyases, and (iv) decarboxylation. These four mechanisms must provide intermediates that can readily feed into central metabolic pathways, and some amino acids might require a combination of enzymes to satisfy this requirement. In C. jejuni NCTC 11168, on the basis of the available genome sequence and a consideration of the most likely catabolic pathways in which these enzymes participate, there are enzymes that would be predicted to allow the complete catabolism of only aspartate, asparagine, glutamate, glutamine, serine, and proline (Fig. 1). This is a narrow range of amino acids, but it accords exactly with the amino acid utilization data of Leach et al. (1997), and interestingly, aspartate, glutamate, proline, and serine are the most abundant amino acids in chicken excreta (Parsons et al., 1982). CjOO21c encodes a potential fumarylacetoacetate hydrolase that is the final enzyme in tyrosine catabolism in some bacteria, but the other enzymes of this complex pathway appear to be absent. No amino acid catabolic decarboxylases are present, but the three other principal mechanistic routes are represented (Fig. 1).Asparagine might be deaminated to aspartate by Cj0029 (a predicted asparaginase), and there are two possible aspartate/glutamate transaminases (CjOl50 and
PHYSIOLOGY AND METABOLISM OF C. TETUNI
47
Cj0762). A key aspartate catabolic enzyme is likely be encoded by Cj0087, a predicted aspartate:ammonia lyase (aspartase) that would directly form fumarate by p-elimination of ammonia. Proline could be oxidized to glutamate via a predicted bifunctional proline dehydrogenase PutA (Cj1503) with A'pyrroline-5-carboxylate as an intermediate. The electrons from this reaction may feed into the respiratory chain at the level of menaquinone (Fig. 2). Glutamate catabolism is likely to proceed via transamination to aspartate. Glutamate and glutamine metabolism is also dominated by the glutamine synthase and glutamine:2-oxoglutarate amino-transferase system, which is the major route by which nitrogen (in the form of ammonium ions) is incorporated into cellular amino acids. 2-Oxoglutarate is the primary amino group acceptor in the interconversion of glutamine and glutamate, and it is a crucial intermediate in C. jejuni amino acid metabolism. Some biochemical studies have been carried out on amino acid utilization in both batch and continuous cultures (Karmali et al., 1986; Westfall et al., 1986; Leach et al., 1997; Mendz et al., 1997). In complex media, the most heavily depleted amino acids were serine, aspartate, glutamate, and proline (Leach et al., 1997). In continuous culture, there was a strong growth rate effect on the utilization of different amino acids, with aspartate and serine being metabolized preferentially at high dilution (growth) rates, with a switch to glutamine and proline at lower dilution rates (Leach et al., 1997). L-Serine seems to be a particularly favored amino acid for catabolism. C. jejuni possesses an active serine dehydratase (deaminase) that converts serine to pyruvate and ammonia. The pyruvate can be rapidly and efficiently metabolized via POR and the CAC, which may explain why this amino acid is so well utilized (Fig. 1).Most serine dehydratases from aerobic bacteria are pyridoxal 5'-phosphate (PLP)dependent enzymes. Results (Velayudhan et al., 2004) have shown that the sdaA gene in C. jejuni encodes an L-serine dehydratase that is devoid of PLP; instead, it is an oxygen-labile iron-sulfur enzyme. The enzyme lacks the consensus sequence centered around a lysyl residue to which PLP binds via a Schiff base (Ogawa et al., 1989), in addition to lacking the glycine-rich region reported for all PLP-dependent enzymes (Marceau et al., 1988). Moreover, the PLP-reactive carbonyl reagents phenylhydrazine and hydroxylamine did not significantly affect enzyme activity. The absorption spectrum of the purified protein was, however, typical of that for an iron-sulfur protein, with broad peaks at 300 to 350 nm and at 420 nm (Velayudhan et al., 2004). Consistent with this, the C. jejuni SdaA se-
48
KELLY
FldA/Fd
NO3-
NapAB
(Cj078 11782)
Cj0020 and Cj0358
NrfH --* (Cj1358)
NAD(P)H
SorAB Malate Proline
? e Gluconate
Tor/Dor Cj026
(CjO408-410)
f
Succinate 4
TMAO/DMSO
TMADMS
Figure 2. Major electron transport pathways in C. jejuni. Integral membrane oxidoreductases on the electron donor side of the menaquinone (MK) pool include an NDH-1-like complex (Cj1566-1579), the electron donor to which is unknown, hydrogenase, formate dehydrogenase, and succinate dehydrogenase. Peripherally associated oxidoreductases include (among several others) ma1ate:quinone oxidoreductase, proline dehydrogenase, and a lactate dehydrogenase. Reducing equivalents are transferred to menaquinone in the lipid bilayer of the inner membrane. Menaquinol reduces the cytochrome bc, complex, which in turn reduces periplasmic cytochrome c. Cytochrome c is reoxidized by one of the terminal oxidases, a cb-type cytochrome c oxidase. A separate quinol oxidase (CioAB) is also present. Cytochrome c may also be reoxidized by hydrogen peroxide in the periplasm through the activity of two separate CCPs. Several alternative reductases are present in C. jejuni. Fumarate reductase (FrdCAB) catalyzes electron transfer from menaquinol to fumarate as terminal acceptor. Periplasmic nitrate (Nap), nitrite (Nrf), and TMAOiDMSO reductases (ToriDor) are also present. An additional type of DMSO reductase (DmsABC) is present in strain 81-176 (not shown; see text and Fig. 3 for details). Cj0378/379 are homologues of the E. coli YedZ/YedY proteins, which are a b-type cytochrome and a molybdoprotein reductase, respectively, but the substrate reduced is unknown. Solid lines indicate experimentally established or highly likely routes of electron transport; dotted lines indicate uncertainty as to the exact route, possibly with the participation of unidentified additional redox proteins. Figure modified and updated from Kelly (2005).
quence contains four conserved cysteine residues, which could coordinate a [4Fe-4S] cluster. Moreover, both the spectral features and enzyme activity were lost on exposure to air, but activity could be restored by treatment with ferrous iron and a reducing agent (Velayudhan et al., 2004). These properties are similar to those exhibited by the L-serine dehydratases of many anaerobes, which are PLP-independent, iron-sulfur-containing enzymes (Grabowski et al., 1993). They form a family of enzymes that share mechanistic similarities with the dehydration of citrate by aconitase (Grabowski et al., 1993), but they are generally highly specific for L-serine deamination. The use of an oxygen-labile dehydratase for serine catabolism is another example of the use by C. jejuni of a type of enzyme more commonly found in anaerobic bacteria than aerobes.
Mutation of sduA resulted in a 10-fold reduction in the activity of serine dehydratase in cell-free extracts. The residual activity was found to be due to the PLP-dependent biosynthetic L-threonine dehydratase (IlvA; CjO828c), but this activity was unable to support growth of the mutant on L-serine in a minimal medium, and there was no evidence from 'H-nuclear magnetic resonance spectroscopy for significant utilization of L-serine by intact cells (Velayudhan et al., 2004). However, growth of the sduA mutant in a complex medium was indistinguishable from that of the parent strain. This indicates that an inability to deaminate serine does not result in a growth disadvantage when a larger array of carbon sources in addition to amino acids are available. Most significantly, however, it was found that there was a defect in colonization of 2-week-old chickens by the
CHAPTER 3
isogenic sdaA mutant compared with its wild-type colonization-proficient parent strain (Velayudhan et al., 2004). The data suggest that a source of serine is readily available to the bacterium and that its catabolism via SdaA is a determinant of its ability to grow in the chicken cecum. These results may also indicate a high degree of selectivity in the types of amino acid utilized by C. jejuni in this niche. Role for y-Glutamyl Transpeptidase in Amino Acid Catabolism and Colonization in C. j e j m i Although all of the key enzymes discussed above are present in each of the sequenced C. jejuni strains, there is an interesting situation regarding one enzyme, y-glutamyl transpeptidase (GGT), which is present in some strains, e.g., 81-176 (Hofreuter et al., 2006) and 81116 (Barnes et al., 2007), but not in NCTC 11168 or RM1221. Colonization studies with isogenic ggt mutants have clearly shown that this gene is required for persistent colonization of l-day-old chicks (Barnes et al., 2007) and also a mouse model of infection (Hofreuter et al., 2006). GGT catalyzes the transpeptidation of the y-glutamyl group of glutathione, and it is an essential enzyme in glutathione metabolism in most animal tissues. Its role in bacteria, in which it is also widespread, has been less clear because other enzymes of glutathione metabolism and the y-glutamyl cycle are often absent, and also because in gram-negative bacteria, GGT is a periplasmic enzyme. However, GGT also catalyzes the hydrolysis of glutamine to form glutamate and ammonia, and glutathione to form glutamate and cysteinylglycine. The hydrolysis reaction of GGT has been largely overlooked in assessing the function of the enzyme in prokaryotes, but a recent elegant study in H. pylori by Shibayama et al. (2007) has revealed that the kinetics of the hydrolysis reaction are far more favorable than the y-glutamyl transferase reaction, and that the likely role of GGT in H. pylori is to supply the bacteria with glutamate for catabolism by the hydrolysis of extracellular glutathione or glutamine. These substrates would be hydrolyzed in the periplasm and then the glutamate transported inside the cell (Shibayama et al., 2007). Given the importance of amino acid catabolism for the growth of C. jejuni, as emphasized above, and the known abundance of glutathione (and almost certainly glutamine) in avian and mammalian tissues, it is tempting to speculate that in those strains that possess a ggt gene, a similar physiological role in supplying glutamate applies (Fig. 1). Other Sources of Nitrogen Low concentrations of environmental ammonium ions and those produced as a result of amino
PHYSIOLOGY AND METABOLISM OF C. TETUNI
49
acid catabolism by deamination or @elimination reactions as described above can be assimilated via the glutamine synthase and glutamine:2-oxoglutarate amino-transferase system. Other sources of nitrogen are also likely to be available to C. jejuni in the host-for example, urea and uric acid in mammalian and avian hosts, respectively. Unlike H. pylori and some thermophilic campylobacters, C. jejuni does not possess a urease gene to allow the direct hydrolysis of urea to ammonia and carbon dioxide. An alternative route of urea catabolism is via urea carboxylase to allophanate (urea-l-carboxylate), which can then be hydrolyzed to yield ammonia (Kanamori et al., 2004). These reactions may be carried out by Cj1541 and Cj1542 in NCTC 11168, with the cognate genes also being conserved in RM1221 and 81176, although biochemical data are lacking. Degradation of the purine ring system of uric acid is more complex and requires a set of enzymes to oxidize and hydrolyze uric acid to allantoin, allantoate, and then on to urea. Clear homologues of uricase (urate oxidase) and allantoinase appear to be lacking in C. jejuni, but a possible enzyme to hydrolyze the product of the uricase reaction (5-hydroxyisourate) is present in all the genome-sequenced strains (Cj0715 in NCTC 11168). This is hydroxyisourate hydrolase, a protein containing a eukaryotic-like transthyretin domain (Lee et al., 2005).
PHYSIOLOGY OF MICROAEROBIC GROWTH Campylobacters and many related Epsilonproteobacteria are well known as microaerophiles. The term microaerophile refers to microorganisms that, although requiring oxygen for growth, are unable to grow at normal atmospheric oxygen tensions; these organisms are thought to be adapted to particular niches that contain low oxygen concentrations (Kreig and Hoffman, 1986; Kelly et al., 2001). Most strains of C. jejuni are routinely cultured in gas atmospheres containing 3 to 10% (vol/vol) oxygen and 5 to 10% (vol/vol) carbon dioxide. However, during its life cycle, C. jejuni must be exposed to highly variable oxygen concentrations, and it thus presents an interesting paradox; although it is oxygen sensitive, it nevertheless must be able to survive high environmental oxygen tensions, resist the oxidative stresses encountered in vivo, and adapt to the severe oxygen limitation of the gut. The relationship between C. jejuni and oxygen is therefore one of the major defining features of the biology of this food-borne pathogen, yet it is also one of the least studied and understood. Indeed, although many genera of freeliving and pathogenic microaerophilic bacteria are
50
KELLY
known, it is still the case that the molecular and physiological bases of microaerophily are not fully understood for any bacterium. A facile explanation would be a deficiency of oxidative stress defenses. All aerobic bacteria are at risk from damage by molecular oxygen, as the stepwise one-electron reduction of 0, results in the formation of the superoxide radical (02-) and hydrogen peroxide (H,O,), which can be removed by the well-characterized superoxide dismutase and catalaselperoxidase enzymes. Significantly, although some microaerophiles do lack these activities (Kreig and Hoffman, 1986) there are many species-including C. jejuni-that contain them, so additional, more complex explanations need to be invoked. As is apparent from the discussion above, C. jejuni, and indeed related microaerophiles, contain several key enzymes that are oxygen sensitive and that may be major targets for inactivation by both molecular oxygen and reactive oxygen species. This may be one reason for the oxygen sensitivity of growth of these bacteria. Sensing of oxygen is clearly important in microaerophiles both for oxygen avoidance at supraoptimal levels and for maximizing the use of alternative pathways of electron transport (and thus energy conservation) if oxygen concentrations become too limiting for aerobic growth. Mechanisms of oxygen sensing are unclear in C. jejuni; from the genome sequence, it is obvious that many of the global regulatory systems present in “model” aerobic bacteria are absent. There are few transcriptional regulators, only seven two-component systems and no RpoStype stationary-phase sigma factor. Significantly, the SoxRS and OxyR oxidative stress regulators are absent, and there are no Arc or Fnr-type systems (Parkhill et al., 2000). Nevertheless, previous microarray studies have provided evidence that C. jejuni upregulates several alternative electron transport pathways in response to oxygen limitation, both in vitro (Gaynor et al., 2004) and in the chick gut (Woodall et al., 2005). The data from the latter study suggest that there may be a coordinated response to limiting oxygen in the sense that metabolically related sets of genes show upregulation in vivo, i.e., the chick gut, compared with in vitro.
PATHWAYS OF ELECTRON TRANSPORT IN C. JEJUNI Organization and Assembly of Electron Transport Chains The complexity of electron transport systems are a key element in the ability of bacteria to grow under a variety of environmental conditions. This is because
the composition and degree of branching of the electron transport chain determines the metabolic flexibility a bacterium possesses in terms of the variety of electron donors and acceptors that can be used to support growth. The simplest types of electron transport chain consist of a primary dehydrogenase to extract electrons from a substrate electron donor, a membrane soluble quinone as a mobile carrier of the electrons (in the form of hydrogen atoms), and a terminal reductase to reduce an appropriate electron acceptor. Key properties are (i) the degree of branching at both “dehydrogenase” and “reductase” ends of the chain; (ii) the ability to use alternative electron acceptors to molecular oxygen, with suitably positive midpoint redox potentials (EM7);(iii) the presence of a variety of types of cytochromes as additional electron carriers and often more than one type of quinone; (iv) interactions between electron transport pathways, optimizing the possibility of each reductant being paired with a wide choice of oxidants; and (v) the degree to which each electron transport chain contributes to proton translocation and energy transduction. The early detailed studies of Hoffman and Goodman (1982) and Carlone and Lascelles (1982) indicated that the respiratory chain of C. jejuni was complex. This work can now be interpreted in terms of the available genome sequences of C. jejuni. Figure 2 shows a partial reconstruction of the major predicted electron transport pathways in strain NCTC 11168 as deduced from the genome sequence and biochemical evidence. A variety of primary dehydrogenases can be identified that feed electrons to a menaquinone pool. C. jejuni contains a proton-translocating cytochrome bc, complex feeding electrons to a periplasmic c-type cytochrome (probably Cjll53 in NCTC 11168) and then to a high-affinity cb-type oxidase, which allows efficient energy conservation when oxygen is used as electron acceptor. Some periplasmic dehydrogenases ( e g , for sulfite and gluconate; see below) probably feed electrons to the same periplasmic cytochrome c (Fig. 1).C. jejuni also contains a number of terminal reductases that allow respiration with a variety of alternative electron acceptors to oxygen. These are discussed in more detail below. The respiratory chain in C. jejuni is clearly highly branched and is significantly more complex than might be expected for such a small-genome pathogen. It should be noted, however, that this complexity is currently underestimated because although a large majority of the components of the respiratory chains can be fitted into diagrams such the one in Fig. 2, there are still a number of uncharacterized cytochromes and other redox proteins. For example, the
CHAPTER 3
molybdoenzyme encoded by Cj0379 is a homologue of the E. coli YedY protein, the three-dimensional structure of which is known (Loschi et al., 2004), but the physiological substrate has not been identified (Fig. 1).Strain 81-176 contains an additional c-type cytochrome and an associated biogenesis system not present in the other strains (Hofreuter et al., 2006). All the key electron transport enzymes contain redox components like heme, iron-sulfur clusters, or molybdenum, nickel, copper, or other cofactors, which have their own pathways of biosynthesis and incorporation into the relevant apoproteins. The biogenesis of the electron transport chains of C. jejuni is thus critically dependent on metal-ion transport and complex cofactor biosynthesis enzymes. In addition, many of the proteins involved in the various pathways of respiration described below must be exported to the periplasm for their function. In the case of the c-type cytochromes, this is achieved via the Sec translocation system, with covalent heme attachment to the apoprotein occurring posttranslationally in the periplasm. In the case of complex cofactor containing proteins like molybdoenzymes (see below), cofactor attachment occurs in the cytoplasm and the mature proteins are translocated across the cytoplasmic membrane via the “twin arginine” translocation (TAT) system. TAT signal peptides can be easily recognized in the N-terminal regions of such proteins in C. jejuni, which also contains homologues of the tatAIE (Cj1176c), tatB (CjOS79c), and tatC (CjOS78c) translocase genes. Electron Donors Hydrogen and formate may be key energy sources in vivo C. jejuni contains a set of structural genes for a NiFe-type “uptake” hydrogenase, as well as accessory genes for nickel incorporation, which could mediate the transfer of electrons from gaseous hydrogen to the quinone pool (Fig. 2). C. jejuni has been shown to possess hydrogenase activity in membrane fractions (Carlone and Lascelles, 1982; Hoffman and Goodman, 1982). Hydrogen could be an important source of electrons for the growth of campylobacters in the gut because many obligate anaerobes in the intestine produce hydrogen from redox reactions associated with fermentation. It would be of great interest to know whether hydrogenase activity is essential for or contributes to the colonization of the avian gut by C. jejuni, as studies with H. pylori have shown that hydrogenase mutants do not colonize a mouse model of infection, and microelectrode measurements have shown that hydrogen is present in the
*
PHYSIOLOGY AND METABOLISM OF C. TETUNI
51
mouse stomach at concentrations much greater than the K , value of the hydrogenase enzyme (Olson and Maier, 2002). In addition to hydrogen, formate is also a good electron donor because the formate/bicarbonate couple has a highly negative redox potential (EM, -420 mV). Formate is also produced by anaerobes in the gut from mixed-acid-type fermentation reactions, so it is available to an organism like C. jejuni, which has the enzymatic capacity for formate oxidation. Respiratory activities in membrane vesicles were SO to100 times greater with formate or hydrogen as substrates when compared with the rates achieved with other electron donors such as succinate, lactate, malate, or NADH, suggesting that the former are indeed excellent electron donors (Hoffman and Goodman, 1982) and indicating the presence of an active formate dehydrogenase. The genome sequence of C. jejuni 11168 reveals an operon encoding putative formate dehydrogenase subunits IfdhA-D/CjlSllc-CjlS08c).C. jejuni fdhA (Cjl51lc) encodes a large 104-kDa selenocysteine containing molybdoprotein equivalent to the E. coli 110 kDa FdnG (a)subunit. fdhB (Cj1510c) encodes a 24 kDa iron-sulfur subunit equivalent to the E. coli 32-kDa FdnH (p) subunit. fdhC (CjlS09c) encodes a 3.5-kDa cytochrome b subunit equivalent to the E. coli 20-kDa FdnI ( y ) subunit. In addition to these three subunits, both E. coli and C. jejuni encode an FdhD protein (29 kDa) required for activity of the formate dehydrogenase enzyme complex (Berg et al., 1991). Molecular and functional studies are lacking on the C. jejuni enzyme. In particular, it would be of great interest to know whether fdh mutants are deficient in colonization of chickens or other animal models because this would provide evidence for the utilization and importance of formate as an in vivo electron donor. The puzzle of NAD(P)H oxidation and the function of complex I In most organisms, NADH is the major electron donor for oxidative phosphorylation, via interaction with a proton-translocating quinone oxidoreductase (often referred to as complex I, Nuo, NQO, or NDH-l), whereas the usual role of NADPH is as a source of electrons for cytoplasmic biosynthetic reactions. NDH-1 is a complex enzyme with a multitude of redox centers and is most commonly made up of 14 different subunits in bacteria. Electrons are passed from NADH via one FMN and nine Fe centers in NDH-1, to ubiquinone/menaquinone in the respiratory chain. Four protons are translocated
52
KELLY
across the membrane for every two electrons transferred. The genome sequences of C. jejuni show that the bacterium contains a cluster of genes (Cj1566 to Cj1579 in NCTC 11168) encoding a potential NADH:quinone oxidoreductase of the NDH-1 type (Fig. 2). However, examination of the deduced proteins encoded by this gene cluster suggests that the complex may not actually oxidize NADH because of the lack of the N Q O l (NuoE) and N Q 0 2 (NuoF) subunits (Smith et al., 2000; Kelly, 2005). These subunits are thought to be essential components for the function of the NDH-1 complex. NuoF binds NADH and also possesses a bound FMN and an Fe-S center. NuoE, G, By and I have cysteine residues that coordinate Fe-S clusters. In place of the expected nuoE and nuoF genes, C. jejuni (and also H. pylori and related Epsilonproteobacteria) has open reading frames of unknown function. Significantly, these open reading frames encode unique small proteins that do not contain any obvious NAD(P)H-binding motif. There are two possibilities regarding the function of this complex in C. jejuni. The first is that electrons from NAD(P)H are transferred to it via an intermediate protein that interacts with the N Q O l and N Q 0 2 replacements (Finely 1998; Smith et al., 2000). Alternatively, coupling of NAD(P)H with the respiratory chain may not occur via the NDH-1 homologue at all, but through an alternative quinone reductase, which may or may not be proton translocating. This implies that an as yet unidentified electron donor interacts with the Cj1566-1579 complex, and that its function is different from most bacterial NDH-1 systems. The nature of the electron donor and whether this type of enzyme is proton translocating are important unresolved questions for understanding the bioenergetics of C. jejuni. One possibility, discussed above, is that flavodoxin/ ferredoxin reduced by the POR/OOR complexes in the CAC donates electrons directly to the NDH-1-like enzyme. Organic acids as electron donors Several types of organic acids (in addition to formate) appear to be able to act as direct electron donors to the respiratory chain in c. jejuni (Fig. 2). Hoffman and Goodman (1982) demonstrated the activity of a lactate dehydrogenase in oxygen-linked respiration, although analysis of the NCTC 11168 genome sequence suggests no obvious locus for a membrane associated lactate dehydrogenase enzyme in C. jejuni. An annotated L-lactate dehydrogenase (Cj1167) is probably a fermentative enzyme. Malate can act as a direct electron donor to the quinone pool
due to the presence of a flavoprotein-type malate oxidoreductase (Mqo). Analysis of the C. jejuni genome reveals a putative oxidoreductase (Cj0393) with sequence similarity to the E. coli ma1ate:quinone oxidoreductase and to the Mqo of H. pylori, indicating that these bacteria use a similar enzyme for malate oxidation (Kather et al., 2000). This is also consistent with the data of Hoffman and Goodman (1982), which showed that malate stimulated respiration in C. jejuni, producing a H'/O ratio of 2.0. Although Mqo and thus malate oxidation is predicted to be localized on the cytoplasmic side of the inner membrane, an interesting additional potential periplasmic electron donor in each of the sequenced C. jejuni strains is gluconate (Fig. 2), which is based on the prediction of a flavin-containing gluconate dehydrogenase encoded by the Cj0414 and Cj0415 genes in strain NCTC 11168, the former with a TAT signal sequence. It is likely that gluconate-derived electrons will enter the chain at the level of the periplasmic cytochrome c (Fig. 2). In enteric bacteria, gluconate has been shown to be a significant in vivo growth substrate (Chang et al., 2004) and is an abundant component of intestinal mucus. It is highly likely that this substrate will be encountered by C. jejuni in vivo, but it is unlikely to be used as a growth substrate and carbon source because C. jejuni lacks the Entner-Doudoroff pathway for its intracellular catabolism. The interconversion of fumarate and succinate can be carried out by two related enzymes: by succinate dehydrogenase, which is expressed under aerobic conditions, or by fumarate reductase, which is induced under anaerobiosis. Succinate dehydrogenase (succinate:quinone oxidoreductase) catalyzes electron transfer from succinate to quinone, where succinate is oxidized to fumarate as part of the tricarboxylic acid cycle. As noted above, C. jejuni appears to have a complete CAC with genes encoding clearly identifiable homologues of all of the conventional enzymes (Parkhill et al., 2000), including separate sdh and frd operons. Sulfite: an unexpected electron donor in C . jejuni The ability to use sulfite as a respiratory electron donor is usually associated with free-living chemolithotrophic sulfur-oxidizing bacteria. However, it has been shown that C. jejuni has the ability to respire sulfite, with oxygen as the electron acceptor (Myers and Kelly, 2005). The C. jejuni NCTC 11168 CjOOO4c and CjOOOSc genes encode a monoheme cytochrome c and molybdopterin oxidoreductase, respectively, homologous to the su1fite:cytochrome c
CHAPTER 3
oxidoreductase of Starkeya novella (Fig. 2). Western blot tests of C. jejuni periplasm probed with a SorA antibody demonstrated cross-reaction of a 45-kDa band, consistent with the size of Cj0005, and a CjOOO4c mutant was unable to respire with sulfite or metabisulfite as electron donors but showed wildtype rates of formate dependent respiration (Myers and Kelly, 2005). Evidence from inhibitor studies suggested that electrons from sulfite enter the respiratory chain after the bc, complex at the level of cytochrome c (Fig. 2), and a fractionation study confirmed that the majority of the su1fite:cytochrome c oxidoreductase activity is located in the periplasm (Myers and Kelly, 2005). Sulfite is an extremely widespread inorganic anion in many environments, particularly low-oxygen niches in soil or water, where it is more stable than in aerobic conditions. The possession of a sulfite respiration system may thus contribute to the survival of C. jejuni in such environments, particularly because this bacterium has a high-affinity cb-type cytochrome c oxidase, which could allow sulfite respiration at extremely low oxygen concentrations. Sulfite respiration may also provide a detoxification mechanism for this normally growth-inhibiting compound. The presence of a sulfite respiration system in C. jejuni is another example of the surprising diversity of the electron transport chain in this small genome pathogen. Oxygen-Dependent Respiration in C. jejuni Most bacteria possess at least two terminal oxidases, often a quinol oxidase and a cytochrome c oxidase (Poole and Cook, 2000), with distinct catalytic properties and regulation of expression. Whereas only one terminal oxidase, a cb-type cytochrome c oxidase, has been found to terminate the respiratory chain of Helicobacter pylori (Nagata et al., 1996), all sequenced strains of Campylobacter jejuni have two terminal oxidases (Parkhill et al., 2000; Fouts et al., 2005; Hofreuter et al., 2006). This confirms the earlier spectroscopic work of Carlone and Lascelles (1982) and Hoffman and Goodman (1982). One oxidase is a cb-type cytochrome c oxidase, similar to that described in H. pylori, and the other resembles a cytochrome bd-type quinol oxidase. Analysis of the NCTC 11168 genome sequence, for example, reveals the presence of a cydAB-like operon encoding subunits I and I1 of such a quinol oxidase. However, in a recent detailed study of the terminal oxidases of NCTC 11168 (Jackson et al., 2007), it is clear from the lack of optical spectral signals associated with the high-spin hemes b and d that this
. PHYSIOLOGY AND METABOLISMOF c. mmr
53
oxidase is not of the conventional cytochrome bd type. Mutation of the cydAB operon was without significant effect on growth, but it did reduce formatesupported respiratory activity. However, the cydABencoded oxidase enhanced survival of cells cultured in a gaseous environment containing 5% (vol/vol) oxygen and was relatively cyanide resistant and has now been renamed CioAB (cyanide-insensitive oxidase). The oxygen affinity of each oxidaie was determined by means of a highly sensitive assay that exploits globin deoxygenation during respirationcatalyzed oxygen uptake (Jackson et al., 2007). The CioAB-type oxidase exhibited a relatively low affinity for oxygen (K, = 0.8 pM) and a V,, of >20 nmol/ mg/s. The alternative ccoNOQP-encoded cyanidesensitive oxidase, which encodes a cytochrome cbtype enzyme, plays a major role in the aerobic respiratory metabolism of C. jejuni because it appeared to be essential for viability, as judged by the failure of repeated attempts to construct a knockout mutation. The cytochrome cb-oxidase exhibited a high oxygen affinity, with a K,,, value of 40 nM and a V, of 6 to 9 nmol/mg/s. Therefore, C. jejuni has two oxidases with different structural and kinetic properties. The marked differences in affinity for oxygen probably reflect distinct physiological roles in response to, and survival under, various degrees of oxygen provision. In particular, the possession of a cb-type oxidase is likely to endow the bacterium with the ability to continue respiration at extremely low oxygen concentrations, as might be found in the niches that it occupies in the gut of animals or humans. Hydrogen Peroxide as an Electron Acceptor Reduction of oxygen to water is a four-electron reaction, but toxic oxygen intermediates formed by incomplete reduction include the superoxide anion (O,-), hydrogen peroxide (H,O,), and the hydroxyl radical (HO'). Hydrogen peroxide can be degraded to H,O and 0, by the cytoplasmic enzyme catalase, but in the periplasm, H,O, can be broken down to water alone by a periplasmic cytochrome c peroxidase (CCP), with the requirement of reduced cytochrome c as an electron donor (Atack and Kelly, 2007). Proton translocation and energy conservation may thus accompany electron transport from a primary electron donor to hydrogen peroxide via the electron transport chain. Although the periplasmic location of the peroxidase means that the enzyme itself is not energy conserving, the reduction of hydrogen peroxide could allow proton translocation to occur at the level of the primary dehydrogenases and also,
54
KELLY
as in C. jejuni, through the operation of the cytochrome bc, complex. The importance of the energyconserving (as opposed to the peroxide detoxification) function of CCPs is difficult to define, but Goodhew et al. (1988), working with the catalasenegative microaerophile Campylobacter mucosalis, were able to show that addition of hydrogen peroxide resulted in uncoupler-sensitive proton extrusion in classical proton-pulse type experiments, albeit with a low stoichiometry (about 0.6 H+/H202).It was suggested that this may be important in the bioenergetics of the bacterium during growth on formate where periplasmic hydrogen peroxide may be generated as a side reaction of the formate dehydrogenase. The true physiological role of bacterial CCPs is actually not well understood (Atack and Kelly, 2007). It is well known that exposure of bacterial cells to excess oxygen increases the rate of H,O, production (Seaver and Imlay, 2004). Paradoxically, however, bacterial ccp genes appear to be upregulated preferentially under low oxygen (microaerobic) or even anaerobic conditions. This may be relevant in pathogenic bacterialike C. jejuni because a drop in oxygen levels may be a signal of entry into the host and would allow the deployment of a periplasmic first line of defense against the attack of hydrogen peroxide-producing phagocytes, for example. In prokaryotes, CCPs are usually diheme proteins, containing two hemes C, one high potential and one low potential, although triheme CCPs have recently been recognized (Atack and Kelly, 2007). The high-potential heme is the source of the second electron for H,O, reduction, and the low-potential heme acts as a peroxidatic center. C. jejuni strains NCTC 11168, 81-176, and RM1221 each have two separate genes encoding putative CCP homologues. In NCTC strain 11168, Cj0020 and Cj03.58 are 34 kDa and 37 kDa, respectively, with both proteins containing two cytochrome c heme-binding site signatures. Cj0358 appears to be a typical CCP, which clusters phylogenetically with other similar enzymes from the bacteria in the epsilon group of the proteobacteria (e.g., H. pylori). These enzymes are all related to the “classical” Pseudomonas aeruginosa type of CCP, which has been studied in detail mechanistically (Atack and Kelly, 2007). However, Cj0020 appears not to be a closely related paralogue of Cj0358, and it is not contained within any of the four major groups that can be identified by phylogenetic analysis (Atack and Kelly, 2007). It does appear to be distantly related to the triheme enzymes from Zymononas mobilis and Gluconobacter oxydans, even though Cj0020 contains only two heme-binding mo-
tifs. Mutants in the CjOO2Oc gene show increased M. sensitivity to killing by hydrogen peroxide Atack and D. J. Kelly, unpublished data), and mutants in the 81-176 homologue are unable to effectively colonize chicks (Hendrixson and DiRita, 2004). It is not yet known whether this putative CCP also plays a role in virulence. The second CCP in C. jejuni, Cj0358, is also needed for resistance to hydrogen peroxide (J. M. Atack and D. J. Kelly, unpublished data), and its expression is increased significantly in the chick cecum (Woodall et al., 200.5), which is thought to be a low-oxygen niche. Cj0358 is also negatively regulated under iron-limited conditions by the iron response regulator Fur, by its homologue PerR (the peroxide response regulator [Holmes et al., 200.5]), and by the temperature-responsive regulator, RacR (Bras et al., 1999). Taken together, these studies suggest a complex regulatory pattern, which may well be related to colonization and possibly pathogenicity.
u.
Growth of C. jejuni under Oxygen Limitation: Respiration by Alternative Electron Acceptors In addition to oxygen-dependent electron transport, it is also clear that the genome of C. jejuni encodes a number of reductases that would be predicted to allow the bacterium to carry out respiration with several alternative electron acceptors to oxygen, including fumarate, nitrate, nitrite, trimethy1amine-Noxide (TMAO), and dimethyl sulfoxide (DMSO) (Kelly, 2001; Sellars et al., 2002; Myers and Kelly, 2005; Pittman et al., 2007). Nevertheless, Sellars et al. (2002) found that although growth with these alternative electron acceptors (see below) was insignificant under strictly anaerobic conditions, such growth was possible under severely oxygen-limited conditions. Their results indicated that some oxygen requiring metabolic reaction or reactions prevented anaerobic growth. Because C. jejuni only contains genes for an oxygen-requiring class I type of ribonucleotide reductase (RNR), it was suggested that the inability to synthesize DNA anaerobically is the most likely explanation. Consistent with this, cells incubated anaerobically with electron acceptors did not divide properly but formed filaments analogous to those seen after treatment of aerobic cells with the RNR inhibitor hydroxyurea (Sellars et al., 2002). Thus, C. jejuni can use alternative electron acceptors in energy-conserving reactions, but only if some oxygen is present to satisfy the requirement for deoxyribonucleotide production. This conclusion is directly relevant for understanding growth of C. jejuni in the gut because it implies that in the niches occupied by
CHAPTER 3
the bacterium, small amounts of oxygen must be present for continued viability. Whether in these severely oxygen-limited niches alternative electron acceptors are being used in addition to or even instead of molecular oxygen will require studies with the relevant mutants in animal models. Fumarate as electron acceptor In the absence of oxygen, fumarate (EM7+30 mV) can be used as an alternative terminal electron acceptor for an electron transport chain by using menaquinol (EM7-74 mV), and fumarate respiration is important in energy conservation in many anaerobic bacteria. C. jejuni strains have an frdCAB operon, encoding subunits highly similar to the thoroughly investigated fumarate reductase of Wolinella succinogenes (Lancaster, 2002). This enzyme is a type B quino1:fumarate reductase, which is reversible and has two b-type hemes bound to the quinol-oxidizing FrdC subunit, which spans the membrane. FrdB is an iron-sulfur cluster containing electron transfer protein, which passes electrons on to the peripheral FrdA subunit, which contains a flavin adenine dinucleotide (FAD) cofactor, the site of fumarate reduction. Because the site of fumarate reduction (which consumes 2H+ per mole of fumarate reduced to succinate) is on the cytoplasmic side of the membrane, and quinol oxidation releases protons into the periplasm, this type of fumarate reductase should be an electrogenic (and energy conserving) enzyme. However, experimental evidence from W. succinogenes has shown that the overall reaction is electroneutral (Kroger et al., 2002). To explain this, the E-pathway hypothesis has been advanced (Lancaster, 2002), which proposes a counterbalancing proton-translocating pathway in the enzyme from periplasm to cytoplasm via a conserved glutamate residue that is coupled to electron transfer between the two b hemes. The operation of this energy-dissipating pathway is essential for the function of the enzyme, even though the overall redox reaction of fumarate reduction by menaquinol is thermodynamically favorable. It is highly likely that similar considerations apply to the homologous C. jejuni Frd complex and that, like the remaining periplasmic reductases described below, net energy conservation occurs only at the level of the primary dehydrogenases. Nitrate and nitrite as electron acceptors
C. jejuni NCTC 11168 has both nitrate (Nap) and nitrite (Nrf) reductases located in the periplasm (Sellars et al., 2002; Pittman et al., 2007). The Nap type of nitrate reductase is a two-subunit enzyme consisting of NapA, a -90-kDa catalytic subunit that
PHYSIOLOGY AND METABOLISM OF C. 7E7UNI
55
binds a bis-molybdenum guanosine dinucleoside cofactor and a [4Fe-4S] cluster, and NapB, a -16-kDa diheme c-type cytochrome. NapA is a substrate for the TAT system and requires an additional protein, NapD, as a “proofreading” chaperone in the export process (Palmer et al., 2005). In C. jejuni, the electron transport pathway to nitrate via NapAB is not immediately obvious because nitrate reduction in other Nap systems is usually coupled to quinol oxidation by a membrane-anchored tetraheme cytochrome, NapC. However, a napAGHBLD operon is present in C. jejuni, which does not contain napC. Analysis of the NCTC strain 11168 genome shows that it does encode a napC homologue (Cj1358c), which is directly upstream of the nrfA nitrite reductase gene (Cj13.57~)~ implying that it is part of the nitrite reductase system. This is the case in Wolinella succinogenes, which encodes a NapC-like subunit (NrfH) similar to the product of Cj1358c (Simon et al., 2000). How, then, is the NapAB complex coupled to menaquinol oxidation in C. jejuni? Pittman et al. (2007) showed that the iron-sulfur protein NapG has a major role in electron transfer to the NapAB complex, and an additional protein, NapH, probably acts together with NapG as a quinol dehydrogenase, as in E. coli (Brondijk et al., 2004). It was found, however, that a napG mutant still showed slow nitratedependent growth under oxygen-limited conditions, which was abolished in a nrfH/napG double mutant, suggesting that a low rate of electron transfer from NrfH to NapAB can occur (Pittman et al., 2007). Interestingly, NapG/H are specifically involved in electron transfer from ubiquinol to nitrate in the E. coli Nap system (Brondijk et al., 2002) and are not required for menaquinol-dependent nitrate reduction (Brondijk et al., 2002), indicating a fundamentally different quinol specificity in C. jejuni. The epsilon group of proteobacteria (including W. succinogenes and C. jejuni) are unusual in containing an additional gene of unknown function, designated napL, within their nap operons (Simon et al., 2003). NapL is predicted to be a soluble periplasmic protein with no obvious cofactor binding motifs. A napL mutant possessed -50% lower NapA activity than the wild type but showed normal growth with nitrate as the electron acceptor (Pittman et al., 2007). It may possibly be involved in the assembly or maturation of NapA. There is evidence that nitrate respiration may play a significant role in the growth of human and animal pathogens in vivo. In Mycobacterium bovis, membrane-bound nitrate reductase (Nar) activity has been shown to contribute to virulence (Weber et al., 2000). As yet, it is not known if the C. jejuni Nap has any in vivo role; it may also have other functions,
56
KELLY
such as in survival in foods or the environment where nitrate is present, or in nitrogen assimilation in combination with nitrite reductase, which produces ammonia. It should be noted that because of the periplasmic location of the Nap and Nrf enzymes, a proton-motive force can only be generated at the level of the primary dehydrogenases and not from quinol oxidation during electron transfer. A role for the nitrite reductase NrfA in nitric oxide detoxification Cj1357c in strain NCTC 111168 encodes the periplasmic pentaheme cytochrome c nitrite reductase (NrfA), which is the terminal enzyme in the sixelectron dissimilatory reduction of nitrite to ammonia (Simon, 2002; Pittman et al., 2007). High specific activities of nitrite reduction can be detected in intact cells of C. jejuni by using methyl viologen as the electron donor (Sellars et al., 2002); a nrfA mutant lacks this activity and shows no oxygen-limited growth with nitrite (Pittman et al., 2007). NrfA is a wellstudied enzyme, which in most bacteria contains an unusual CXXCK motif at the heme 1 binding site, instead of the CXXCH motif found in the vast majority of heme c-containing enzymes. This requires a dedicated lyase for covalent attachment of the active site heme (Pisa et al., 2002). The C. jejuni NrfA contains a novel ligation site at heme 1, with a “conventional” CXXCH motif instead of CXXCK. This would explain the absence of genes encoding Nrfspecific heme lyases that are found in other bacteria (Einsle et al., 2000), but a site-directed His-for-Lys substitution in the W. succinogenes NrfA results in low catalytic activity (Pisa et al., 2002), so how the enzyme achieves high activity in C. jejuni is unknown. Interestingly, the predicted NrfAs of the closely related C. coli, C. lari, and C. upsuliensis (Fouts et al., 2005) also contain the heme 1 CXXCH motif rather than CXXCK. Cytochrome c nitrite reductases can also carry out the five electron reduction of the nitric oxide radical ( N O ) to NH4+ (Costa et al., 1990). Work in E. coli has shown that this reaction may be physiologically relevant under microaerobic conditions, and it has been proposed that NrfA plays a significant role in nitric oxide detoxification in addition to flavohemoglobin (Hmp) and flavorubredoxin (NorV), the other major E. coli N O detoxification systems (Poock et al., 2002). In C. jejuni, Hmp and NorV are absent, but there is a single domain globin (Cgb), which has been shown to protect against nitrosative stress (Elvers et al., 2004). In keeping with this, Cgb is inducible by nitrosative stress via the regulatory protein NssR (Elvers et al., 2005). nrfA mutants were found
to be hypersensitive to various forms of nitrosative stress, suggesting that NrfA does indeed play a role in nitric oxide metabolism (Pittman et al., 2007). However, nitrite also induced cgb expression in an NssR-dependent manner, suggesting that growth of C. jejuni with nitrite itself causes nitrosative stress. This was confirmed by lack of growth of cgb and nssR mutants, and slow growth of a nrfA mutant, in media containing nitrite (Pittman et al., 2007). Thus, NrfA and Cgb together provide C. jejuni with constitutive and inducible components of a robust defense against nitrosative stress.
S- or N-oxides as electron acceptors TMAO and DMSO are structurally similar compounds that are widely distributed in many environments, particularly water; TMAO is an excretory product of some fish, and DMSO is produced by some algae (McCrindle et al., 2005). It is therefore likely that campylobacters will come into contact with these electron acceptors in aquatic environments, soil, etc. In fact, a wide range of bacteria possess enzyme systems that catalyze the reduction of TMAO and DMSO (often by the same enzyme), which is a two electron transfer process, yielding trimethylamine or dimethyl sulfide, respectively: (CH,),SO
+ 2H+ + 2e-
-+ -+
(CH,),S (CH,),NO
+ 2H+ + 2e-
(CH,),N
H,O EM, = +160 mV
H,O EM, = +130 mV
As can be seen, the midpoint redox potentials of the DMSO/ dimethyl sulfide and TMAO/ trimethylamine couples are not highly positive, making them thermodynamically poorer electron acceptors compared with, say, oxygen or nitrate, but better than fumarate. Both TMAO and DMSO reductase activities have been demonstrated in intact cells of C. jejuni NCTC 11168 by using methyl viologen as an artificial electron donor, and TMAO- and DMSOdependent growth under oxygen-limited conditions has also been demonstrated (Sellars et al., 2002). Analysis of the C. jejuni NCTC 11168 genome reveals a gene encoding a 93-kDa molybdoprotein TMAO/DMSO reductase homologue (Cj0264) containing conserved residues for the binding of a molybdenum cofactor and an N-terminal “twin arginine” translocase recognition motif for transport to the periplasm. The CjO264c gene was mutated, and the mutant was found to be deficient in both TMAO and DMSO reductase activity, indicating that a single enzyme is responsible for both activities (Sellars et al.,
PHYSIOLOGY AND METABOLISM OF C. TETUNl
CHAPTER 3
2002). In addition, there is an upstream gene encoding a 22-kDa monoheme c type cytochrome (CjO265c) with similarity to the C terminus of the membrane-anchored pentaheme c type cytochrome TorC of E. coli, which transfers electrons to the catalytic TMAO reductase subunit, TorA (Gon et al., 2001; Sellars et al., 2002). An interesting unresolved issue here is how electrons get from the quinol pool to the CjO265/CjO264 complex (Fig. 3 ) . In other bacteria, a family of tetraheme cytochromes (the NapC family; see above) are often involved in connecting terminal reductases in the periplasm with the quinol pool. The additional four hemes of the E. coli TorC protein have this role, but the only NapC homologue in C. jejuni is NrfH (described above), the electron donor to the nitrite reductase. Growth studies with a nrfH mutant clearly showed that it was still capable of oxygen-limited growth with both TMAO and DMSO (Pittman et al., 2007). Because Cj0265 alone is unlikely to act as a quinol oxidase, there appears to be a novel route of electron transfer to these electron acceptors in strain NCTC 11168, which merits further investigation. Of the other genome-sequenced strains of C. jejuni, RM1221 and 81-176 contain a Cj0264 homologue and are thus likely to have a similar type
TMAOiDMSO + 2H+
HCOO - + H.0
He',-+
TMA/DMS
of TMAO/DMSO reductase to that found in NCTC 11168. Given the presence of TMAO and DMSO in water and soil, the physiological role of this enzyme might be assumed to aid survival outside the host. However, strain 81-176 contains, in addition to a Cj0264 homologue, a structurally distinct type of DMSO reductase, which has proven to be interesting with regard to a potential role in vivo (Hofreuter et al., 2006). This system is similar to the DmsABC DMSO reductase complex in E. coli (McCrindle et al., 2005) consisting of a large extrinsic membrane bound molybdoprotein (DmsA), a smaller electron transferring iron-sulfur subunit (DmsB), and an intrinsic membrane anchor and quinol binding subunit (DmsC). Electrons would be transferred from the quinol pool via DmsC then DmsB to DmsA in the periplasm, where the reduction of DMSO would take place (Fig. 3 ) . Like the periplasmic Cj0264/ Cj0265 system, the arrangement of this enzyme across the membrane means that it is not energy conserving, but electron flow through either system can allow energy to be conserved at the level of the primary dehydrogenases (Fig. 3 ) , as with the Nap and Nrf systems described above. Significantly, Hofreuter et al. (2006) showed that a dmsA mutant was less able to colonize a mouse
+ 2H20
TMAODMSO
TMA/DMS
2H+
'&
Cj026
membrane I
2H+
57
cytoplasm
I
2H+
Figure 3. Likely topological organization of two types of TMAO/DMSO reductases found in strains of C. jejuni. In each case, formate is depicted as a typical electron donor to the menaquinone pool (MK), through the action of formate dehydrogenase (Fdh). (a) The Tor/Dor type of reductase system that has been characterized in strain 11168 (Sellars et al., 2002) is illustrated, which is also present in strains RM1221 and 81-176. There is uncertainty over the mechanism of electron transfer from the MK to the monoheme Cj0265 cytochrome (dashed rectangular hypothetical quinol oxidoreductase). In (b), the additional DmsABC-type system shown in strain 81-176 to be needed for optimal colonization of a mouse infection model (Hofreuter et al., 2006) is illustrated. Note that because both of these TMAO/DMSO reductases are predicted to be periplasmic, reduction of the electron acceptor (which consumes protons-in bold) and quinol oxidation (which releases protons-in bold) occur on the same (periplasmic) side of the membrane. Thus, a proton-motive force is only generated at the level of the primary dehydrogenase (binding and releasing protons on opposite sides of the membrane). MGD, molybdenum guanosine dinucleoside cofactor.
58
KELLY
model of infection than the isogenic wild-type parent strain, implying that this enzyme has an important role in the colonization process. It seems unlikely that DMSO is the substrate of this enzyme in an animal host, and it is possible that in vivo, it is reducing a different S oxide-for example, methionine sulfoxide-because it is known that this class of enzyme often has a broad substrate specificity. It is intriguing that this type of DmsABC system has also been shown to be important for virulence in another bacterial pathogen, Actinobacillus pleuropneumoniae (Baltes et al., 2003). The fact that some strains of C. jejuni, a small-genome pathogen, can possess two distinct functional electron transport chains to TMAO/ DMSO underlines its surprising metabolic versatility and indicates a possible role for these electron acceptors in aiding survival outside and inside the host.
Leon-Kempis, John Atack, Marie Thomas, Jon Smart, Chat Phansopa, and Andy Hitchcock.
ADDENDUM IN PROOF Since this chapter was written, several papers have been published which clarify some aspects of the topics covered, particularly growth in vivo. D. Weerakoon and J. Olson (1.Bacteriol, 190: 915-925, 2008) have shown experimentally that reduced flavodoxin is indeed the electron donor to the unusual complex I in C. jejuni. Weingarten et al. (Appl. Environ. Microbiol., in press, 2008) have shown that mutants in napA and nrfA colonize chicks less well than the wild-type parent strain, while cydA and CjO264c mutants show wild-type levels of colonization. This indicates a possible role for nitrateinitrite reduction in growth of C. jejuni in chickens. A ccoN::cm (cb-oxidase) mutant showed no colonization of chickens, although this strain was very oxygen sensitive and had a significant microaerobic growth defect. Bingham-Ramos and Hendrixson (Infect. Immun., in press, 2008) have characterized the cytochrome c peroxidases in strain 81-176 and have shown a potential role in vivo for the Cj0020 homologue but not for the Cj0358 homologue.
CONCLUSIONS REFERENCES
Understanding the metabolism and physiology of C. jejuni and related bacteria has been greatly facilitated by the availability in recent years of several genome sequences and by better experimental techniques for their culture and biochemical analysis. One conclusion that can be drawn is that many campylobacters show an unexpected metabolic versatility, which is particularly reflected in the complexity of the electron transport chains in C. jejuni, as reviewed here. Despite a genome size of only 1.7 Mb, it is clear that this bacterium has retained complex enzymes with a variety of redox cofactors that enable it to be flexible with respect to the variety of electron donors and acceptors it can use. This is surely a function of its ability to occupy so many niches. Published and ongoing colonization studies with metabolic mutants are crucial and are beginning to answer questions about the nature of C. jejuni metabolism during growth in vivo. This information could be useful in identifying a means to control the growth of C. jejuni in animal reservoirs, for example. It is now also becoming increasingly clear that the application of both conventional biochemical techniques and postgenomic technologies like metabolomics and systems biology will be required to fully elucidate the metabolism of this important pathogen. Acknowledgments. Work in my laboratory on Campylobacter jejuni has been supported by grants and studentships from the United Kingdom Biotechnology and Biological Sciences Research Council and by Don Whitley Scientific Ltd., Shipley, West Yorkshire, United Kingdom. I gratefully acknowledge the hard work of past and current students and postdocs on Campylobacter in my laboratory: Jyoti Velayudhan, Mike Sellars, Mark Gidley, Stephen Hall, Marc Pittman, Jon Myers, Ed Guccione, Maria del Rocio
Atack, J. M., and D. J. Kelly. 2007. Structure, mechanism and physiological roles of bacterial cytochrome c peroxidases. Adv. Microb. Pbysiol. 52:73-106. Baltes, N., I. Hennig-Pauka, I. Jacobsen, A. D. Gruber, and G . F. Gerlach. 2003, Identification of dimethyl sulphoxide reductase in Actinobacillus pleuropneumoniae and its role in infection. Infect. Immun. 71:6784-6792. Barnes, H. A., M. C.Bagnall., D. D. Browning, S. A. Thompson, G . Manning, and D. G. Newell. 2007. y-Glutamayl transpeptidase has a role in persistent colonisation of the avian gut by Campylobacter jejuni. Microb. Pathog. 43: 198-207. Berg, B. L., J. Li, J. Heider, and V. Stewart. 1991. Nitrateinducible formate dehydrogenase in Escherichia coli K-12. I. Nucletide sequence of the fdnGHI operon and evidence that opal (UGA) encodes selenocysteine. J. Biol. Chem. 266:2238022385. Bras, A. M., S . Chatterjee, B. W. Wren, D. G. Newell and J. M. Ketley. 1999. A novel Campylobacter jejuni two-component regulatory system important for temperature-dependent growth and colonization. J. Bacteriol. 181:3298-3302. Brondijk, T. H. C., D. Fiegen, D. J. Richardson, and J. A. Cole. 2002. Roles of NapF, NapG, and NapH, subunits of the Escherichiu coli periplasmic nitrate reductase, in ubiquinol oxidation. Mol. Microbiol. 44:245-255. Brondijk, T. H. C., A. Nilavongse, N. Filenko, D. J. Richardson, and J. A. Cole. 2004. NapGH components of the periplasmic nitrate reductase of Eschericbia coli K-12: location, topology, and physiological roles in quinol oxidation and redox balancing. Biochem. J. 379:47-55. Carlone, G . M., and J. Lascelles, J. 1982. Aerobic and anaerobic respiratory systems in Carnpylobacter fetus subsp. jejuni grown in atmospheres containing hydrogen. J. Bacteriol. 152:306-3 14. Chang, D. E., D. J. Smalley, D. L.Tucker, M. P. Leatham, W. E. Norris, S. J. Stevenson, A. B. Anderson, J. E. Grissom, D. C. Laux, P. S. Cohen, and T. Conway. 2004. Carbon nutrition of Escherichia coli in the mouse intestine. Proc. Natl. Acad. Sci. USA 101:7427-7432. Costa, C., A. Macedo, I. Moura, J. Moura, J. LeGall, Y. Berlier, M. Y. Liu, and W. Payne. 1990. Regulation of the hexa-heme
CHAPTER 3
nitrite: nitric oxide reductase of Desulfovibrio desulfuricans, Wolinella succinogenes, and Escherichia coli: a mass spectrometric study. FEBS Lett. 276:67-70. Daucher, J. A., and N. R. Kreig. 1995. Pyruvate-ferredoxin oxidoreductase in Campylobacter species. Can. J. Microbiol. 41: 198-201. Einsle, O., P. Stach, A. Messerschmidt, J. Simon, A. Kroger, R. Huber, and M. H. Kroneck. 2000. Cytochrome c nitrite reductase from Wolinella succinogenes. J. Biol. Chem. 275:3960839616. Elvers, K. T., S. M. Turner, L. M. Wainwright, G. Marsden, J. Hinds, J. A. Cole, R. K. Poole, C. W. Penn, and S. F. Park. 2005. NssR, a member of the Crp-Fnr superfamily from Campylobacter jejuni, regulates a nitrosative stress-responsive reguIon that includes both a single-domain and a truncated haemoglobin. Mol. Microbiol. 57:735-750. Elvers, K. T., G. Wu, N. J. Gilberthorpe, R. K. Poole, and S. F. Park. 2004. Role of an inducible single-domain hemoglobin in mediating resistance to nitric oxide and nitrosative stress in Camplylobacter jejuni and Campylobacter coli. J. Bacteriol. 186: 5332-534 1. Finel, M. 1998. Does NADH play a central role in energy metabolism in Helicobacter pylori? Trends Biochem. Sci. 23:412-414. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. Scott-Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms form the genomes of multiple Campylobacter species. PLoS Biol. 3:72-85. Gaynor, E. C., S. Cawthraw, G. Manning, J. K. MacKichan, S. Falkow, and D. G. Newell. 2004. The genome-sequenced variant of Campylobacter jejuni NCTC 11168 and the original clonal clinical isolate differ markedly in colonization, gene expression, and virulence-associated phenotypes. J. Bacteriol. 186: 503-5 17. Godlewska, R., M. Pawowski, A. Wyszynska, J. Bujnicki, and E. Jagusztyn-Krynicka. 2005. Sorting of Campylobacter jejuni lipoproteins, abstr. F17, p. 82. Abstr. 13th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Gold Coast, Australia, 4 to 8 September 2005. Gon, S., M. T. Guidici-Orticoni, V. Mejean, and C. Iobbi-Nivol. 2001. Electron transfer and binding of the c-type cytochrome TorC to the trimethylamine N-oxide reductase in Escherichia coli.J. Biol. Chem. 276:11545-11551. Goodhew, C. F., A. B. EIKurdi, and G. W. Pettigrew. 1988. The microaerophilic respiration of Campylobacter mucosalis. Biochim. Biophys. Acta 933:114-123. Goss, J. A., N. D. Cohen, N. D., and M. F. Utter. 1981. Characterisation of the subunit structure of pyruvate carboxylase from Pseudomonas citronellolis. J. Biol. Chem. 256: 1181911825. Grabowski, R., A. E. M. Hofmeister, and W. Buckel. 1993. Bacterial L-serine dehydratases: a new family of enzymes containing iron-sulfur clusters. Trends Biochem. Sci. 18:297-300. Hama, H., T. Shimamoto, M. Tsuda, and T. Tsuchiya. 1987. Properties of a Na+-coupled serine-threonine transport system in Escherichia coli. Biochim. Biophys. Acta. 905:23 1-239. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Hinton, A. 2006. Growth of Campylobacter in media supplemented with organic acids. J. Food Prot. 69:34-38. Hoffman, P. S., and T. G. Goodman. 1982. Respiratory physiology and energy conservation efficiency of Campylobacter jejuni. J. Bacteriol. 150:319-326.
PHYSIOLOGY AND METABOLISM OF c. r E r w
59
Hofreuter, D., J. Tsai, R 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J, E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Holmes, K., F. Mulholland, B. M. Pearson, C. Pin, J. McNichollKennedy, J. M. Ketley, and J. M. Wells. 2005. Campylobacter jejuni gene expression in response to iron limitation and the role of Fur. Microbiology 151:243-257. Huang, J., and M. A. Schell. 1992. Role of the two-component leader sequence and mature amino-acid sequences in extracelMar export of endoglucanase EGL from Pseudomonas solanacearum. J. Bacteriol. 174:1314-1323. Hughes, N. J., P. A. Chalk, C. L. Clayton, and D. J. Kelly. 1995. Identification of carboxylation enzymes and characterization of a novel four-subunit pyruvate:flavodoxin oxidoreductase from Helicobacter p y lori. J. Bacteriol. 177: 3 953-3 959. Hughes, N. J., C. L. Clayton, P. A. Chalk, and D. J. Kelly. 1998. Helicobacter pylori porCDAB and oorDABC genes encode distinct pyruvate:flavodoxin and 2-oxoglutarate:acceptor oxidoreductases which mediate electron transport to NADP. J. Bacteriol. 180:1119-1128. Jackson, R J., K. T. Elvers, L. J. Lee, M. D. Gidley, L. M. Wainwright, J. Lightfoot, S. F. Park, and R K. Poole. 2007. Oxygen reactivity of both respiratory oxidases in Campylobacter jejuni: the cydAB genes encode a cyanide-resistant, low-affinityoxidase that is not of the cytochrome bd type. J. Bacteriol. 189:16041615. Kanamori, T., K, Norihisa, H. Atomi, and T. Imanaka. 2004. Enzymatic characterisation of a prokaryotic urea carboxylase. J. Bacteriol. 186: 253 2-25 39. Karmali, M. A., M. Roscoe, and P. C. Fleming. 1986. Modified ammonia electrode method to investigate D-asparagine breakdown by Campylobacter strains. ],Clin. Microbiol. 23:743-747. Kather, B., K. Sting], M. E. van der Rest, K. Altendorf, and D. Molenaar. 2000. Another unusual type of citric acid cycle enzyme in Helicobacter pylori: ma1ate:quinone oxidoreductase. J. Bacteriol. 182:3204-3209. Kelly, D. J. 1998. The physiology and metabolism of the human gastric pathogen Helicobacter pylori. Adv. Microb. Physiol. 40: 137-189. Kelly, D. J. 2001. The physiology and metabolism of Campylobacter jejuni and Helicobacter pylori. J. Appl. Microbiol. 90:16S24s. Kelly, D. J. 2005. Metabolism, electron transport and bioenergetics of Campylobacter jejuni: implications for understanding life in the gut and survival in the environment, p. 275-292. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter jejuni: Molecular and Cellular Biology. Horizon Press, Wymondham, United Kingdom. Kelly, D. J., N. J. Hughes, N. J., and R. K. Poole. 2001. Microaerobic physiology: aerobic respiration, anaerobic respiration, and carbon dioxide metabolism, p. 113-124. In H. L. T. Mobley, G . L. Mendz, and S. L. Hazel1 (ed.), Helicobacter pylori: Physiology and Genetics. ASM Press, Washington, DC. Kreig, N. R. and P. S. Hoffman. 1986. Microaerophily and oxygen toxicity. Ann. Rev. Microbiol. 40:107-130. Kroger, A., S. Biel., J. Simon, R. Gross, G. Unden, and C. R. Lancaster. 2002. Fumarate respiration of Wolinella succinogenes: enzymology, energetics and coupling mechanism. Biochim. Biophys. Acta 1553:23-38. Lancaster, C. R. D. 2002. Wolinella succinogenes quinol: fumarate reductase-2.2 resolution crystal structure and the E-pathway hypothesis of coupled transmembrane proton and electron transfer. Biochim. Biophys. Acta 1565:215-231. Leach, S., P. Harvey, and R. Wait. 1997. Changes with growth rate in the membrane lipid composition of and amino-acid util-
a
60
KELLY
isation by continuous cultures of Campylobacter jejuni. J. Appl. Microbiol. 82:631-640. Lee, Y., D. H. Lee, C. W. Kho., A. Y. Lee, M. Jang., S. Cho, C. H. Lee, J. S . Lee, P. K. Myung, B. C. Park, and S. G. Park. 2005. Transthyretin-related proteins function to facilitate the hydrolysis of 5-hydroxyisourate, the end product of the uricase reaction. FEBS Lett. 579:4769-4774. Leon-Kempis, M del R., E. Guccione, F. Mulholland, M. P. Williamson, and D. J. Kelly. 2006. The Campylobacter jejuni PEBla adhesin is an aspartate/glutamate-binding protein of an ABC transporter essential for microaerobic growth on dicarboxylic amino acids. Mol. Microbiol. 60:1262-1275. Loschi, L., S. J. Brokx, T. L. Hills, G. Zhang, M. G. Botero, A. L. Lovering, J. H. Weiner, and N. C. Strynadka. 2004. Structural and biochemical identification of a novel bacterial oxidoreductase. J. Biol. Chem. 27950391-503400. Marceau, M., S. D. Lewis, and J. A. Shafer. 1988. The glycinerich region of Escherichia coli D-Serine dehydratase. Altered interaction with pyridoxal 5’-phosphate produced by substitution of aspartic acid for glycine. J. Biol. Chem. 263:16934-16941. McCrindle, S. L., U. Kappler and A. G. McEwan. 2005. Microbial dimethylsulfoxide and trimethylamine-N-oxide respiration. Adv. Microb. Phys. 50:147-198. Mendz, G. L., G. E. Ball, and D. J. Meek. 1997. Pyruvate metabolism in Campylobacter spp. Biochim. Biophys. Acta. 1334:291302. Miiller, A., M. del R. Leon-Kempis, E. Dodson, K. S. Wilson, A. J. Wilkinson, and D. J. Kelly. 2007. A bacterial virulence factor with a dual role as an adhesin and a solute-binding protein: the crystal structure at 1.5 A resolution of the PEBla protein from the food-borne human pathogen Campylobacter jejuni. J. Mol. Biol. 372:160-171. Miiller, A., G. H. Thomas, R. Horler, J. A. Brannigan, E. Blagova, V. M. Levdikov, M. J. Fogg, K. S. Wilson, and A. J. Wilkinson. 2005. An ATP-binding cassette-type cysteine transporter in Campylobacter jejuni inferred from the structure of an extracytoplasmic solute receptor protein. Mol. Microbiol. 57:143155. Myers, J. D., and D. J. Kelly. 2005. A sulphite respiration system in the chemoheterotrophic human pathogen Campylobacter jejuni. Microbiology 15 1:233-242. Nagata, K., S. Tsukita, T. Tamura and N. Sone. 1996. A cb-type cytochrome-c oxidase terminates the respiratory chain in Helicobacter pylori. Microbiology 142: 1757-1763. Ogawa, H., K. Konishi, K., and M. Fujioka. 1989. The peptide sequences near the bound pyridoxal phosphate are conserved in serine dehydratase from rat liver and threonine dehydratases from yeast and Escherichia coli. Biochim. Biophys. Acta. 139: 139-141. Olson, J. W., and R. J. Maier. 2002. Molecular hydrogen as an energy source for Helicobacter pylori. Science 298:1788-1790. Palmer, T., F. Sargent, and B. C. Berks. 2005. Export of complex cofactor- containing proteins by the bacterial Tat pathway. Trends Microbiol. 13:175-1 80. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M-A. Rajandream, K. M. Rutherford, A. H. M. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Parsons, C. M., L. M. Potter, and R. D. Brown Jr. 1982. Effects of dietary protein and intestinal microflora on excretion of amino-acids in poultry. Poult. Sci. 61:939-946. Pei, Z., and M. J. Blaser. 1993. PEB1, the major cell-binding factor of Campylobacter jejuni, is a homolog of the binding component
in gram-negative nutrient transport systems. J. Biol. Chem. 268: 18717-18725. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. 2. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943. Pei, 2. H., R. T. Ellison 111, and M. J. Blaser. 1991. Identification, purification, and characterization of major antigenic proteins of Campylobacter jejuni. J. Biol. Chem. 266:16363-16369. Pisa, R, T. Stein, R. Eichler, R. Gross, and J. Simon. 2002. The nrfr gene is essential for the attachment of the active site haem group of Wolinella succinogenes cytochrome c nitrite reductase. Mol. Microbiol. 43:763-770. Pittman, M. S., K. T. Elvers, L. Lee, M. Jones, R K. Poole, S . F. Park, and D. J. Kelly. 2007. Growth of Campylobacter jejuni on nitrate and nitrite: electron transport to NapA and NrfA via an atypical NrfH and distinct roles for NrfA and the globin Cgb in protection against nitrosative stress. Mol. Microbiol. 63575590. Poly, F., T. Read., D. R. Tribble, S. Baqar, M. Loenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. Immun. 75:3425-3433. Poock, S. R., E. R. Leach, J. W. B. Moir, J. A. Cole, and D. J. Richardson. 2002. Respiratory detoxification of nitric oxide by the cytochrome c nitrite reductase of Escherichia coli. J. Biol. Chem. 277:23 664-23 669. Poole, R. K., and G. M. Cook. 2000. Redundancy of aerobic respiratory chains in bacteria: routes, reasons and regulation. Adv. Microb. Physiol. 42: 165-224. Seaver, L. C., and J. A. Imlay. 2004. Are respiratory enzymes the primary sources of intracellular hydrogen peroxide? J. Biol. Chem. 279:48742-48750. Sellars, M. J., S. J. Hall, and D. J. Kelly. 2002. Growth of Campylobacter jejuni supported by respiration of fumarate, nitrate, nitrite, trimethylamine-N-oxide or dimethylsulfoxide requires oxygen. J. Bacteriol. 184:4187-4196. Shibayama, K., J-I. Wachino, Y. Arakawa, M. Saidijam, N. G. Rutherford, and P. J. F. Henderson. 2007. Metabolism of glutamine and glutathione via y-glutamyl transpeptidase and glutamate transport in Helicobacter pylori: possible significance in the pathophysiology of the organism. Mol. Microbiol. 64:396406. Simon, J. 2002. Enzymology and bioenergetics of respiratory nitrite ammonification. FEMS Microbiol. Rev. 26:285-309. Simon, J., R. Gross, 0. Einsle, P. M. H. Kroneck, A. Kroger, and 0. Klimmek. 2000. A NapCiNirT-type cytochrome c (NrfH) is the mediator between the quinone pool and the cytochrome c nitrite reductase of Wolinella succinogenes. Mol. Microbiol. 35: 686-696. Simon J., M. Sanger, S. C. Schuster, and R. Gross. 2003. Electron transport to the periplasmic nitrate reductase (NapA) of Wolinella succinogenes is independent of a NapC protein. Mol. Microbiol. 49:69-79. Smith, M. A., M. Finel, V. Korolik, and G. L. Mendz. 2000. Characteristics of the aerobic respiratory chains of the microaerophiles Campylobacter jejuni and Helicobacter pylori. Arch. Microbiol. 174: 1-10. St Maurice, M., N. Cremades, M. A. Croxen, G. Sisson, J. Sancho, and P. S. Hoffman. 2007. F1avodoxin:quinone reductase (FqrB): a redox partner of pyruvate:ferredoxin oxidoreductase that reversibly couples pyruvate oxidation to NADPH production in Helicobacter pylori and Campylobacter jejuni. J. Bacteriol. 189: 4764-4773. Unden, G., and A. Kleefeld. 2004. Chapter 3.4.5. C,-dicarboxylate degradation in aerobic and anaerobic growth. In R. Curtiss 111
CHAPTER 3
et al. (ed.), EcoSal-Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, DC. http: // www.ecosal.org/. Velayudhan, J., M. A. Jones, P. A. Barrow, and D. J. Kelly. 2004. L-Serine catabolism via an oxygen-labile L-serine dehydratase (SdaA) is essential for colonisation of the avian gut by Campylobacter jejuni. Infect. lmmun. 72:260-268. Velayudhan, J., and D. J. Kelly. 2002. Analysis of gluconeogenic and anaplerotic enzymes in Campylobacter jejuni: an essential role for phosphoenolpyruvate carboxykinase. Microbiology. 148~685-694. Weber, I., C. Fritz, S. Ruttkowski, A. Kreft, and F. C. Bange. 2000. Anaerobic nitrate reductase (narGHJI)activity of Mycobacterium
PHYSIOLOGYAND METABOLISM OF c. rErmrI
61
bovis BCG in vitro and its contribution to virulence in immunodeficient mice. Mol. Microbiol. 35:1017-1025. Westfall, H. N., D. M. Rollins, and E. Weiss. 1986. Substrate utilisation by Campylobacter jejuni and Campylobacter coli. Appl. Environ. Microbiol. 52:700-705. Woodall, C. A., M. A. Jones, P. A. Barrow, J. Hinds, G. L. Marsden, D. J. Kelly, N. Dorrell, B. W. Wren, and D. J. Maskell. 2005. Campylobacter jejuni gene expression in the chick cecum: evidence for adaptation to a low-oxygen environment. Infect. lmmun. 735278-5285. Wyszynska, A., M. Pawlowski, J. Bujnicki, D. Pawelec., J. P. van Putten., E. Brzuszkiewicz, and E. K. Jagusztyn-Krynicka. 2006. Genetic characterisation of the cj& operon of Campylobacter coli. Pol. J. Microbiol. 55:85-94.
Cumpylobucter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 4
Comparative Genomics of Campylobacter jejuni OLIVIA L. CHAMPION, SUAAD AL-JABERI,RICHARDA. STABLER,AND BRENDANW. WREN
The sequencing of the first Campylobacter jejuni strain in 2000 was a watershed in Campylobacter genetics and brought to light unanticipated aspects of biology of this important gastrointestinal pathogen. Since this milestone, three further complete C. jejuni genome sequences have been published that provide definitive glimpses of the genome diversity of this species. However, for the genome comparison of larger sets of C. jejuni strains, microarray-based technology has been widely applied and is currently the method of choice. Such studies have identified key genes that could be used to differentiate strains from different sources. The use of comparative genomics combined with robust methods for data analysis will continue and will form the basis for the development of rational intervention strategies to reduce C. jejuni in the food chain. In this chapter, we review the salient comparative features of the four fully sequenced genomes and reveal highlights from selected wholegenome microarray studies.
the basis of amino acid similarity to proteins of known function (Parkhill et al., 2000). After the completion of further C. jejuni genome sequences, these programs could be used for both inter- and intraspecies genomic comparisons. Furthermore, the availability of multiple C. jejuni genome sequences resulted in a shift away from the notion that any single strain could typify the entire species. More recently, the original annotation of NCTC 11168 has been updated in light of new studies on Campylobacter species and other microorganisms. The c. jejuni NCTC 11168 reannotation reduced the total number of coding sequences from 1,654 to 1,643, of which 90.0% have additional information regarding the motif identification and/or new relevant literature (Gundogdu et al., 2007). The reannotation led to 18.2% of coding sequence product functions being revised. This provides a useful framework for future C. jejuni studies and is a testament to the rapid pace of Campylobacter research in the postgenomic era.
CAMPYLOBACTER JEJUnrI GENOME SEQUENCES
GENOME DIVERSITY The relative genome diversity of bacterial species varies from clonal (genetically uniform) to genetically highly variable. The genomes of the three human clinical isolates share 1,474 core genes, with 35 genes unique to 81-176, 38 genes unique to CG8486, and 8 genes unique to NCTC 11168 (Fig. 1). All four sequenced C. jejuni genomes have revealed evidence of chromosomal loci that appear to be hot spots for intra- and intergenomic combination. For example, the genomes of NCTC 11168, RM1221, and 81-176 all have DNA unique to each strain in the loci cjO.564 and cj0.570. Similarly, identical homologues of cj1687 and cjl688 flank two
At the time of writing, four fully sequenced and annotated C. jejuni genomes have been published, including three human and one chicken isolate. Details of the strains and their basic genome statistics are summarized in Table 1. The publication of the first C. jejuni genome (Parkhill et al., 2000) paved the way for comparative genomics of this species. Initially the NCTC 11168 genome was compared with other publicly available sequences by computer programs such as Blast and the Artemis comparison tool. Such comparisons provide a means for attributing putative phenotypes to C. jejuni predicted coding sequences on
Olivia L. Champion School of Biosciences, University of Exeter, Geoffrey Pope Building, Exeter EX4 4QD, United KingSuaad Al-Jaberi, Richard A. Stabler, and Brendan W. Wren * Department of Infectious & Tropical Diseases, London School dom. of Hygiene & Tropical Medicine, Keppel St., London WClE 7HT, United Kingdom.
63
64
CHAMPION ET AL.
Table 1. Summary of C. jejuni genomes sequenced to date and comparison to C. doylei Strain 11168" 81-176 CG8486 Fa41221 C. doylei
G+C content Size (Mb) No. of genes % coding 30.55 30.61 30.43 30.31 31.02
1.64 1.59 1.61 1.78 1.85
1,643 1,707 1,588 1,940 2,037
92 93 ? 90 82
No. of No. of structural RNAs pseudogenes 54 54 ?b
56 55
20 0 ?b
46 25 1
Source
Origin
Human diarrhea Human diarrhea Human diarrhea Chicken skin
United Kingdom United States Thailand United States
Reannotated genome (Gundogdu et al., 2007). *None reported.
a
unique 81-176 open reading frames (ORFs), although no DNA inserts were found between these two ORFs in the NCTC 11168 and RM1221 genomes. A PCR screening of C. jejuni clinical isolates revealed that 6 of 16 possessed inserts between cj1687 and cj1688 (Hofreuter et al., 2006). C. jejuni demonstrates diversity in the DNA restriction modification systems, which is thought to influence its ability to acquire genetic information by horizontal gene transfer (Miller et al., 2005). Strain 81-176 encodes a complete type I DNA restriction modification system absent in 11168 and RM1221. This restriction modification has a G+C content of 28%, varying from the genome average of 30.61% G+C, suggesting that it may have been acquired by horizontal gene transfer. Likewise, strain CG8486 encodes two complete type I restriction modification systems, the first of which is positioned between homologues of
n NCTC 11168
1
8
Figure 1. Venn diagram of the genome content of the three sequenced C. jejuni human isolates (NCTC 11168, CG8486, and 81-176) (adapted from Poly et al., 2007). The gene content of 81-176 is based on the results of Hofreuter et al. (2006). These estimations exclude the capsule, LOS, and flagellin posttranslational modification loci.
cj0759 and cj0760 (as is the novel restriction modification type I system of 81-176). The second novel type I restriction modification system is located between cj1448 and cjl555, replacing the corresponding NCTC 11168 ORFs that are absent from the CG8486 genome. Strain CG8486 also contains one novel type I11 restriction modification system. GENE GAZING AND CLUES T O SURVIVAL AND PATHOGENESIS Although isolated from a human with campylobacteriosis and thus clearly pathogenic, NCTC 11168 contains none of the usual suspects in terms of pathogenesis. The genome of strain NCTC 11168, like those of RM1221 and CG8486, does not contain plasmids. However, strain 81-176 possesses two previously sequenced plasmids: pVir and pTet (Bacon et al., 2000, 2002; Batchelor et al., 2004). This suggests that pVir may be important for pathogenesis of a subset of C. jejuni strains. However, no clear mechanism has yet been demonstrated. Conflicting reports have been published regarding the correlation of pVir in C. jejuni strains and the clinical presentation of bloody diarrhea. A study carried out in Canada revealed that pVir was present in 17 of 104 clinical isolates of C. jejuni and was significantly associated with bloody diarrhea, a marker for invasiveness (Tracz et al., 2005). However, a later study of 125 C. jejuni isolates carried out in Holland revealed no correlation between the presence of pVir and bloody diarrhea (Louwen et al., 2006). The 81-176 genome sequence also revealed a 6-kb element integrated at a leu tRNA with the characteristics of an integrated plasmid. This integrated plasmid encodes TraN and TraG homologues that function in type IV secretion systems. The integrated plasmid also contains a CJEll55 homologue contained on the CJIE3 genomic island found in the genome of RM1221. The integrated plasmid in strain 81-176 is positioned between homologues of cj0936 and cj0937. The integrated plasmids identified in all C. jejuni strains
CHAPTER 4
sequenced to date have not been linked with pathogenesis. In addition to lacking plasmids, NCTC 11168 revealed virtually no insertion sequences, no prophages or pathogenicity islands, and few repeat sequences. Similarly, C. jejuni CG8486 (Poly et al., 2007) also lacks these genetic elements. However, this is not true of all sequenced C. jejuni strains. Although the genome of RM1221 shows synteny with that of NCTC 11168, the genome of RM1221 is disrupted with four large integrated elements (Fouts et al., 2005). A Campylobacter Mu-like phage (CMLP1) encoding MuA and MuB transposase homologues is located upstream of argC (CjE0275). CMLP1 contains 5 '-TG-3' dinucleotides flanked by a 5-bp direct repeat (TATGC), consistent with those found in Mulike prophages (Morgan et al., 2002). Bacteriophages can be vehicles of horizontal gene transfer that can increase bacterial fitness (Desiere et al., 2001; Hendrix, 2002) and can influence many aspects of virulence, from adhesion and invasion to host evasion and toxin production (Wagner and Waldor, 2002). Although CMLPl does not encode any characterized virulence determinants, its presence may influence phenotypes by insertional inactivation (Fouts et al., 2005). Pathogenicity islands are often associated with tRNA loci, and unlike CMLP1, integrated elements 2 and 4 (CJIE2 and CJIE4) in the RM1221 genome have integrated into the 3' end of arginyl and methionyl tRNA genes. Phage-related endonucleases, methylases, or repressors are predicted to be encoded by ORFs found in these elements. However, RM1221 CJIE3 is located within an arginyl tRNA yet possesses no phage-related ORFs, suggesting that it is of plasmid origin or a genomic island. The majority of genes on CJIE3 are of unknown function, although 23% share homology with Helicobacter hepaticus ATCC 51449 genomic island (HHGI1) (Suerbaum et al., 2003). CMLPl is inducible with mitomycin C, but it is unknown whether the remaining three integrated elements in strain RM1221 can be excised. Because of the unknown function of the majority of genes sequenced in these integrated elements, putative roles cannot be predicted. However, functions affecting virulence or bacterial fitness cannot be ruled out. Although strain 81-176 exhibits high levels of invasion in vitro (Hu and Kopecko, 1999; Oelschlaeger et al., 1993) and is highly pathogenic in vivo (Black et al., 1988; Prendergast et al., 2004; Russell et al., 1989), the genome sequence lacks prophages. The genome of strain 81-176 is syntenic with both the NCTC 11168 and RM1221 genomes, with a few disruptions where genomic islands have been in-
COMPARATIVEGENOMICS OF c. rEruNr
65
serted. The C. jejuni 81-176 genome contains 37 novel genes that are absent or pseudogenes in strains 11168 and RM1221. These are involved with electron pathways, energy metabolism, and respiratory pathways and include ORFs encoding predicted proteins with homology to DmsA, DmsB, DmsC, DmsD, and TorD, components of a Wollinella succinogenes putative anaerobic dimethyl sulfoxide reducatase (Hofreuter et al., 2006). Genome 81-176 also contains genes encoding proteins with homology to cytochrome c and several cytochrome c biosynthesis proteins. These findings suggest that strain 81-176 may have expanded respiratory pathways (Hofreuter et al., 2006). Similarly, although present as pseudogenes and thus redundant in both NCTC 11168 and RM1221, genes kdpA-C involved with potassium uptake are apparently functional in 8 1-176. Although all three genomes possess the KtrA/KtrB potassium uptake system (Tholema et al., 2005), the presence of an additional system in 81-176 may enhance its fitness in inside phagocytic vacuoles where potassium concentrations are low (Wagner et al., 2005). Strain 81-176 also contains an ORF, ggt, encoding a predicted protein with significant homology to a gammaglutamyl transpeptidase found in Helicobacter pylori. Although the mechanism is not characterized, this enzyme has been shown to have a role in H. pylori colonization in vivo (McGovern et al., 2001).
GENE SEQUENCING AND POTENTIAL PHASE VARIATION Scouring the genomes has revealed few definitive clues to genetic elements that may be important in virulence. However, scrutiny of raw nucleotide data can reveal repeat sequences or polymeric tracts (e.g., homopolymeric tracts of repeat As, Cs, Gs, and Ts) that can provide clues to genes that slip in and out of frame (Parkhill et al., 2000). Such phase-variable genes can provide a simple primordial genetic mechanism for switching key determinants on and off depending on the environment's selective pressures (van Belkum et al., 1998). The presence of hypervariable sequences in genes encoding the biosynthesis or modification of surface structures such as capsular polysaccharide, flagellum, and lipooligosaccharide (LOS) has been a common theme in each of the C. jejuni genomes sequenced to date. The combinatorial mathematics of phase variation means that a bacterium with just 20 phase-variable loci can potentially exist in over a million different states. Several dozen phase-variable genes in surface antigen biosynthesis loci have been identified in the strains sequenced to date. The rep-
66
CHAMPION ET AL.
ertoire of envelope-associated structures provides C. jejuni with astonishing diversity when interacting with its environment, including immune evasion in the human host but also during its survival in nonhuman environmental reservoirs.
GENE SEQUENCING AND CELL SURFACE GLYCOSTRUCTURES Surface structures, particularly glycan-modified surface structures, often play important roles for bacterial pathogens in interacting with host cells and the host immune system. The most rapid advance in the biology of C. jejuni postgenomic era has been our basic understanding of cell surface glycoconjugates fueled by the importance of these structures in disease pathogenesis, including immune-mediated neuropathies. C. jejuni has been termed a “hyperglycemic bug” in that it has dedicated a considerable part of its relatively small genome to the biosynthesis of several glycostructures. These include LOS, capsular polysaccharide, and both 0- and N-linked glycosylation pathways (the only bacterium proven to have both pathways). The C. jejuni LOS, capsule, and 0linked glycosylation system of flagellin have been found to be highly variable between strains, with multiple genetic mechanisms used to affect the observed structural differences (Gilbert et al., 2002; Karlyshev et al., 2005). One of the key discoveries resulting from the NCTC 11168 genome sequence was the presence of a highly variable C. jejuni capsular polysaccharide (Karlyshev and Wren, 2001; Karlyshev et al., 2005). Genomic comparisons of C. jejuni strains of different Penner serotype by means of microarrays have revealed that capsule is the likely Penner serodeterminant (Dorrell et al., 2001). In addition, all C. jejuni strains sequenced to date contain highly variable LOS biosynthesis genes and flagellin modification loci. The C. jejuni cell surface polysaccharides are discussed in more detail elsewhere (chapters 25 to 29).
C . JEJUnrr GENOMICS TO PHENOMICS Genome sequence is not an end in its own right, but a means for hypothesis generation to be tested in the laboratory. Genomic comparisons must be coupled with assays in the laboratory to unravel gene function and understand differences in strain phenomes. In strain 8 1-176, loss-of-function mutations were constructed in dmsA, cytC, and ggt to assess their contribution to virulence. These genes are thought to be involved in respiratory pathways of
strain 81-176 and so were tested under low 0, conditions, as would be experienced by C. jejuni in the gut during colonization. Interestingly, in vitro, no growth defect was demonstrated for all dmsA, cytC, and ggt mutants in BHI 10% CO, or under anaerobic conditions. Furthermore, mutants showed no reduction in invasiveness and survival in intestinal epithelial cells compared with wild type. However, a significant colonization defect was observed for dmsA and cytC in vivo after 3 weeks (Hofreuter et al., 2006). The ggt knockout also showed a colonization defect compared with wild-type 81-176 during competitive index experiments. By week 7 after infection, shedding of ggt knockout strains was significantly reduced compared with wild type, and by week 8, significantly fewer ggt knockout strains were recovered from mouse tissue compared with wild-type 8 1-176. These differences led the authors to suggest that 81176 specific genes dmsA, cytC, and ggt may contribute to the increased virulence of strain 8 1-176 in vivo (ferret model) compared with strain 11168. However, during in vivo experiments in the ferret model, strain CG8486 was equally as pathogenic as strain 81-176 (five of seven animals developed diarrhea after infection with both CG8486 and 81-176). Strain 8 1-176 genes dmsA, cytC, and ggt were not found in strain CG8486, although the same level of invasiveness in vivo was observed. Previous studies that used strain 81-176 have indicated that invasiveness in vitro correlates with disease in the ferret model (Bacon et al., 2001; Goon et al., 2006; Yao et al., 1997). However, despite displaying equal levels of pathogenicity in vivo, during in vitro comparisons of invasiveness in INT-407 cells, CG8486 invaded 1,000-fold less than strain 81-176. The disparity between in vitro and in vivo phenotypes of loss-of-function mutants dmsA, cytC, and ggt in strain 81-176 is not understood. Multiple C. jejuni genome sequences have revealed that the presence of specific virulence genes in the genome does not appear to correlate with the level of pathogenicity in vitro and in vivo of a strain. Loci encoding C. jejuni surface antigens possess the highest levels of genetic variation. Understanding the variation in these loci may hold the key to understanding strain differences in pathogenicity. Further research underpinning environmental stimuli, including host-pathogen interactions, implicit in surface antigen variability is required.
COMPARATIVE GENOMICS BY MICROARRAY Traditional phylogenetic classification of bacteria to study evolutionary relatedness is based on the
CHAPTER 4
characterization of a limited number of genes and of rRNA or signature sequences. However, because of the acquisition of DNA through lateral gene transfer, the differences between closely related bacterial strains can be vast. By contrast, whole-genome sequencing comparisons allow a multitude of genes to be compared. Nevertheless, whole-scale genome sequencing remains an expensive endeavor, and such comparisons are limited to only a handful of strains. DNA microarrays represent an alternative technology for whole-genome comparisons, enabling a bird’s-eye view of all the genes absent or present in a given genome as compared with the reference genome on the microarray. Harnessing DNA microarray information through interrogative and robust algorithms has enabled a true “comparative phylogenomics” approach to be developed. Recent comparative phylogenomics studies have been undertaken on increasingly large collections of strains from defined origins. A common feature from many of these studies has been the unexpectedly large genetic diversity between strains within the same species, blurring our definition of species boundaries. Whole-genome comparisons typically identify sets of core genes shared by all strains in a species and accessory genes present in one of more strains in a species that often result from gene acquisition. It is these differences that can often be used to identify genes/genetic islands related to gainof-function traits in pathogenic strains. Uncovering the mechanisms behind this variability is fundamental in understanding and ultimately counteracting infection. Moreover, given the range of diseases associated with some bacterial pathogens and the diverse genotypic and phenotypic properties of clinical and environmental isolates, microarrays have proved to be particularly useful for determining correlates of pathogenicity (Hinchliffe et al., 2003; Howard et al., 2006) and in deciphering the epidemiology and ecological niches of the organism (Champion et al., 2005; Stabler et al., 2006).
CASE STUDIES OF C. JEjUNI GENOMIC COMPARISONS BY MICROARRAY The first C. jejuni microarray, or clone array, which used clone vectors as a by-product from the genome sequencing project, was constructed in 2001 (Dorrell et al., 2001). This clone array demonstrated a proof of principle for microarray technology for C. jejuni and provided a platform for comparison of multiple strains at a whole-genome level. Microarray comparisons revealed an unexpectedly large level
COMPARATIVE GENOMICS OF C. lETUNI
67
of genetic diversity among strains of C. jejuni. A set of core genes shared by all strains compared to date have been identified. In addition, accessory genes have also been identified that are present in one or more strains in the species and are often the result of gene acquisition. These differences can be particularly useful for retrospective analysis of bacterial evolution and population genetics, both recent and in the long term. The initial observation on genetic variability in the gastrointestinal pathogen C. jejuni (Dorrell et al., 2001) has been expanded by several recent studies (Leonard et al., 2003; Pearson et al., 2003; Taboada et al., 2004), suggesting C. jejuni exhibits high levels of genome diversity and low levels of genome plasticity (Taboada et al., 2004). The original estimate of C. jejuni accessory genes of ~ 2 0 %(Dorrell et al., 2001) has now been increased to ~ 3 7 % (Taboada et al., 2004). However, many of these accessory genes mapped to previously defined variable loci, suggesting that large regions of the C. jejuni genome are genetically stable (Taboada et al., 2004). Unfortunately, it has proved difficult to use microarray data to identify genetic markers predictive of source or outcome of infection. A comparison of isolates from patients who developed Guillain-BarrC syndrome (GBS) with isolates from patients with uncomplicated gastrointestinal infection failed to identify specific GBS genes or regions (Leonard et al., 2004). A whole-genome microarray was used to determine the activities of cytolethal distending toxin (CDT) and hemolysin along with survival characteristics under aerobic conditions at room temperature (On et al., 2006). Numerical analysis of microarray data clearly delineated strains into two clusters, genotype cluster l and 2 (GC1 and GC2). CDT and hemolysin activities of GC1 strains were not statistically different from those of cluster GC2 strains. However, viability during aerobic incubation of cluster GC1 strains was statistically lower than corresponding estimates of cluster GC2 strains. The number of missing or highly divergent genes in cluster GC1 strains with respect to NCTC 11168 was also statistically significantly greater compared with those of cluster GC2 strains. Out of the genes present in NCTC 11168, 67 genes were characteristically missing or divergent among cluster GC1 strains. Of these, 53 genes were localized within 11 major gene clusters, of which 8 were associated with surface structures, including flagellar, lipooligosaccharide, and membrane transport proteins. This study indicates a correlation between C. jejuni genomic content, particularly in surface-coding regions, and its capacity for environmental survival, and it may explain why
68
CHAMPION ET AL.
genetics, which plays an important role in the investigation of C. jejuni pathogenicity.
certain serotypes are more commonly reported in human diseases (On et al., 2006). Studies that use an ORF-specific C. jejuni DNA microarray to assess clinical isolates associated with five independent clusters of infection were compared with data from random amplified polymeric DNA and Penner serotyping (Leonard et al., 2003). Microarray analysis provided a method for comparison of isolates that is more gene specific than are other existing genotyping methods and that can reveal genomic differences between isolates that randomly amplified polymorphic DNA analysis may not identify (Leonard et al., 2003). By means of comparative genomic hybridization, many genes were found to be divergent across multiple strains and were uniquely variable in single strains. Despite the large increase in the collective number of variable genes, nearly half of them were mapped to previously defined variable loci, and it therefore appears that large regions of the C. jejuni genome are genetically stable (Taboada et al., 2004). On the whole bacterial chromosome, regions of variable genes often are present in large clusters, suggesting that they were acquired or lost from the genome in groups during evolution. The data generated by microarray technology are highly accurate and readily comparable among laboratories. Studies carried out by microarray technology promise a clearer understanding of the population biology of this important pathogen. This in turn will lead to a better understanding of population
COMPARATIVE PHYLOGENOMICS Although the whole-genome comparisons of C. jejuni by DNA microarrays have been useful in identifying candidate genes that may be involved in pathogenesis and survival, they do not provide any clue to the evolutionary origins and phylogeny of C. jejuni. Therefore, we developed comparative phylogenomics, whole-genome comparisons of microbes via DNA microarrays combined with Bayesian-based algorithms to model the phylogeny of a given microorganism (Fig. 2). We applied this initially to 111 C. jejuni isolates from a spectrum of sources, including humans (n = 70), chickens (n = 17), bovines (n = 13), ovines (n = 5), and the environment (n = 6). Remarkably, from this data, the Bayesian phylogeny of the isolates revealed two distinct clades unequivocally supported by Bayesian probabilities (P = 1) (Champion et al., 2005). On further analysis, these two clades appeared to correlate with the origin of the strains and included a livestock clade comprising 3 1 (88.6%) of 35 of the livestock isolates and a nonlivestock clade containing all environmental isolates. Several genes were identified as characteristic of strains in the livestock clade. The most prominent
MrBayes
C. jejuni Strain Collection
I
1
gDNA
Microarray
-
K '
Molecular Biology Analysis
A Genespring
Test Vs Common Reference (C. jejuni 11168) Quality Contrc
/
\
I-
Phylogenetic Tree production
I
GACK
Gene Calling
Figure 2. Comparative phylogenomics pipeline. Graphical representation of steps involved in production of a phylogenetic tree. Genomic DNA from each member of the C. jejuni strain collection was hybridized to a BpGS C. jejuni 11168 microarray with C. jejuni 11168 gDNA as control. Florescent intensities were calculated by BlueFuse. GeneSpring calculated intensity ratios and removed low-quality data points. GACK was used to convert ratio data to binary present/absent data. MrBayes used the binary data to construct a putative phylogeny by means of a Bayesian algorithm. The resulting tree was statistically tested for robustness. Further C. jejuni strains can be fed into the pipeline, and selection may be influenced by current phylogeny prediction.
CHAPTER 4
was a cluster of six genes (cj1321 to cj1326) within the genetically variable flagellin glycosylation locus. In subsequent studies, this locus has been confirmed as being present in approximately 50% of strains, and its precise role in these C. jejuni strains is the subject of intense study. Surprisingly, the initial comparative phylogenomics study showed that 39 (55.7%) of 70 C. jejuni human isolates were found in the nonlivestock clade, suggesting that many C. jejuni infections may be from nonlivestock (and possibly nonagricultural) sources. More recently, we have studied over 230 C. jejuni strains from diverse origin and have further demonstrated the split of strains into two clades. Approximately half of the human isolates from this study are not associated with the livestock clade (S. Al-Jaberi, and B. W. Wren, unpublished data). These studies may provide insight into a previously unidentified reservoir of C. jejuni infection that may have implications in disease control strategies. The comparative phylogenomics approach provides a robust methodological prototype and has been applied to other bacterial pathogens (Howard et al., 2006; Stabler et al., 2006).
CONCLUSION AND FUTURE PERSPECTIVES Until the year 2000, the genetics of C. jejuni was investigated in a piecemeal gene-by-gene approach, with no clear overview. The availability of whole-genome sequences has facilitated research into genotypes and phenotypes in their entirety, heralding a golden era in Campylobacter microbiology. Robust methods of data analysis are required to maximize the new information gained from both genome sequences and comparative genome analyses via microarrays. DNA microarrays represent a powerful enabling technology for the whole-scale comparison of bacterial genomes. This, coupled with new methods to model DNA microarray data, is facilitating the development of robust comparative phylogenomics analyses. Such studies have increased our ability to differentiate between C. jejuni, highlighting previously undetected genetic differences and population structures, and providing new insight into virulence and evolution of C. jejuni. Hypothesis trawling is possible through comparison of whole genomes of strains with different characteristics by both genome sequencing and DNA microarrays. Subsets of genes that differ between the strains can be identified and their roles hypothesized on the basis of strain phenotypes. Gene homology searches with other sequenced bacteria provide clues
*
COMPARATIVE GENOMICS OF C. lElUNI
69
to the function of many of the genes in the genome. However, the function of many C. jejuni genes remains unknown, and learning their roles will aid understanding of the unique functioning of this problematic pathogen. Advances in genetic manipulation of C. jejuni, including the development of speciesspecific promoters and complementation techniques, have fueled research into gene function. However, the lack of a small, portable animal model to study C. jejuni pathogenesis is still a stumbling block for studying gene function in vivo. The next challenge for microbiologists in this postgenomic era is to correlate C. jejuni genome to phenome. This will provide a clear and more comprehensive understanding of the biology of C. jejuni.
REFERENCES Bacon, D. J., R. A. Alm, D. H. Burr, L. Hu, D. J. Kopecko, C. P. Ewing, T. J. Trust, and P. Guerry. 2000. Involvement of a plasmid in virulence of Campylobacterjejuni 81-176. lnfect. Immun. 68~4384-4390. Bacon, D. J., R. A. Alm, L. Hu, T. E. Hickey, C. P. Ewing, R. A. Batchelor, T. J. Trust, and P. Guerry. 2002. DNA sequence and mutational analyses of the pVir plasmid of Campylobacterjejuni 81-176. Infect. lmmun. 70:6242-6250. Bacon, D. J., C. M. Szymanski, D. H. Burr, R. P. Silver, R. A. Alm, and P. Guerry. 2001. A phase-variable capsule is involved in virulence of Campylobacter jejuni 8 1-176. Mol. Microbiol. 40~769-777. Batchelor, R. A., B. M. Pearson, L. M. Friis, P. Guerry, and J. M. Wells. 2004. Nucleotide sequences and comparison of two large conjugative plasmids from different Campylobacter species. Microbiology 150:3507-3517. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacterjejuni infection in humans. 1.lnfect. Dis. 157:472-479. Champion, 0. L., M. W. Gaunt, 0. Gundogdu, A. Elmi, A. A. Witney, J. Hinds, N. Dorrell, and B. W. Wren. 2005. Cornparative phylogenomics of the food-borne pathogen Campylobacter jejuni reveals genetic markers predictive of infection source. Proc. Natl. Acad. Sci. USA 102:16043-16048. Desiere, F., W. M. McShan, D. van Sinderen, J. J. Ferretti, and H. Brussow. 2001. Comparative genornics reveals close genetic relationships between phages from dairy bacteria and pathogenic Streptococci: evolutionary implications for prophage-host interactions. Virology 288:325-34 1. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. Al-Ghusein, B. G. Barrell, J. Parkhill, N. G . Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Campylobacterjejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome. Res. 11:1706-1715. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS. Biol. 3:e15.
70
CHAMPION ET AL.
Gilbert, M., M. F. Karwaski, S . Bernatchez, N. M. Young, E. Taboada, J. Michniewicz, A. M. Cunningham and w. w. Wakarchuk. 2002. The genetic bases for the variation in the lipooligosaccharide of the mucosal pathogen, Campylobacter jejuni. Biosynthesis of sialylated ganglioside mimics in the core oligosaccharide. J. Biol. Chem. 277:327-337. Goon, S., C. P. Ewing, M. Lorenzo, D. Pattarini, G. Majam, and P. Guerry. 2006. A sigma28-regulated nonflagella gene contributes to virulence of Campylobacter jejuni 81-176. Infect. lmmun. 74:769-772. Gundogdu, O., S . D. Bentley, M. T. Holden, J. Parkhill, N. Dorrell, and B. W. Wren. 2007. Re-annotation and re-analysis of the Campylobacter jejuni NCTCl1168 genome sequence. BMC Genomics 8~162. Hendrix, R. W. 2002. Bacteriophages: evolution of the majority. Theoretical Population Biol. 61:471-480. Hinchliffe, S . J., K. E. Isherwood, R. A. Stabler, M. B. Prentice, A. Rakin, R. A. Nichols, P. C. Oyston, J. Hinds, R. W. Titball, and B. W. Wren. 2003. Application of DNA microarrays to study the evolutionary genomics of Yersinia pestis and Yersinia pseudotuberculosis. Genome Res. 13:2018-2029. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. lmmun. 74:4694-4707. Howard, S . L., W., Gaunt, J. Hinds, A. A. Witney, R Stabler, and B. W. Wren. 2006. Application of comparative phylogenomics to study the evolution of Yersinia enterocolitica and to identify genetic differences relating to pathogenicity. J. Bacteriol. 188: 3645-3653. Hu, L., and D. J. Kopecko. 1999. Campylobacter jejuni 81-176 associates with microtubules and dynein during invasion of human intestinal cells. Infect. lmmun. 67:4171-4182. Karlyshev, A. V., 0. L. Champion, C. Churcher, J. R. Brisson, H. C. Jarrell, M. Gilbert, D. Brochu, F. St Michael, J. Li, W. W. Wakarchuk, I. Goodhead, M. Sanders, K. Stevens, B. Stevens, J. Parkhill, B. W. Wren, and C. M. Szymanski. 2005. Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol. Microbiol. 55:90-103. Karlyshev, A. V., and B. W. Wren. 2001. Detection and initial characterization of novel capsular polysaccharide among diverse Carnpylobacterjejuni strains using Alcian blue dye. J. Clin. Microbiol. 39:279-284. Leonard, E. E., 11, T. Takata, M. J. Blaser, S. Falkow, L. S. Tompkins, and E. C. Gaynor. 2003. Use of an open-reading framespecific Campylobacter jejuni DNA microarray as a new genotyping tool for studying epidemiologically related isolates. J. Infect. Dis. 187:691-694. .eonard, E. E., 11, L. S . Tompkins, S . Falkow, and I. Nachamkin. 2004. Comparison of Campylobacter jejuni isolates implicated in Guillain-Barre syndrome and strains that cause enteritis by a DNA microarray. Infect. lmmun. 72:1199-1203. .ouwen, R. P., A. van Belkum, J. A. Wagenaar, Y. Doorduyn, R. Achterberg, and H. P. Endtz. 2006. Lack of association between the presence of the pVir plasmid and bloody diarrhea in Campylobacter jejuni enteritis. J. Clin. Microbiol. 44:18671868. McGovern, K. J., T. G. Blanchard, J. A. Gutierrez, S . J. Czinn, S . Krakowka, and P. Youngman. 2001. Gamma-glutamyltransferase is a Helicobacter pylori virulence factor but is not essential for colonization. Infect. Immun. 69:4168-4173. Miller, W. G., B. M. Pearson, J. M. Wells, C. T. Parker, V. V. Kapitonov, and R. E. Mandrell. 2005. Diversity within the Campylobacter jejuni type I restriction-modification loci. MicrobiolOD 151:337-351.
Morgan, G. J., G. F. Hatfull, S . Casjens, and R. W. Hendrix. 2002. Bacteriophage Mu genome sequence: analysis and comparison with Mu-like prophages in Haemophilus, Neisseria and Deinococcus. J. Mol. Biol. 317:337-359. Oelschlaeger, T. A., P. Guerry, and D. J. Kopecko. 1993. Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proc. Natl. Acad. Sci. USA 90:6884-6888. On, S. L., N. Dorrell, L. Petersen, D. D. Bang, S . Morris, S . J. Forsythe, and B. W. Wren. 2006. Numerical analysis of DNA microarray data of Campylobacter jejuni strains correlated with survival, cytolethal distending toxin and haemolysin analyses. lnt. J. Med. Microbiol. 296:353-363. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S . Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S . Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pearson, B. M., C. Pin, J. Wright, K. I'Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS. Lett. 554~224-230. Poly, F., T. Read, D. R. Tribble, S . Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. lmmun. 75:34253433. Prendergast, M. M., D. R Tribble, S . Baqar, D. A. Scott, J. A. Ferris, R. I. Walker, and A. P. Moran. 2004. In vivo phase variation and serologic response to lipooligosaccharide of Campylobacter jejuni in experimental human infection. Infect. lmmun. 72:916-922. Russell, R G., M. J. Blaser, J. I. Sarmiento, and J. Fox. 1989. Experimental Campylobacter jejuni infection in Macaca nemestrina. Infect. lmmun. 57:1438-1444. Stabler, R A., D. N. Gerding, J. G. Songer, D. Drudy, J. S . Brazier, H. T. Trinh, A. A. Witney, J. Hinds, and B. W. Wren. 2006. Comparative phylogenomics of Clostridium dificile reveals clade specificity and microevolution of hypervirulent strains. J. Bacteriol. 188:7297-7305. Suerbaum, S., C. Josenhans, T. Sterzenbach, B. Drescher, P. Brandt, M. Bell, M. Droge, B. Fartmann, H. P. Fischer, Z. Ge, A. Horster, R. Holland, K. Klein, J. Konig, L. Macko, G. L. Mendz, G. Nyakatura, D. B. Schauer, Z. Shen, J. Weber, M. Frosch, and J. G. Fox. 2003. The complete genome sequence of the carcinogenic bacterium Helicobacter hepaticus. Proc. Natl. Acad. Sci. USA 100:7901-7906. Taboada, E. N., R. R. Acedillo, C. D. Carrillo, W. A. Findlay, D. T. Medeiros, 0. L. Mykytczuk, M. J. Roberts, C. A. Valencia, J. M. Farber, and J. H. Nash. 2004. Large-scale comparative genomics meta-analysis of Campylobacter jejuni isolates reveals low level of genome plasticity. J. Clin. Microbiol. 42:45664576. Tholema, N., M. Vor der Bruggen, P. Maser, T. Nakamura, J. I. Schroeder, H. Kobayashi, N. Uozumi, and E. P. Bakker. 2005. All four putative selectivity filter glycine residues in KtrB are essential for high affinity and selective K+ uptake by the KtrAB system from Vibrio alginolyticus. J. Biol. Chem. 280:4114641154. Tracz, D. M., M. Keelan, J. Ahmed-Bentley, A. Gibreel, K. Kowalewska-Grochowska, and D. E. Taylor. 2005. pVir and bloody diarrhea in Campylobacter jejuni enteritis. Emerg. Infect. Dis. 11:838-843.
CHAPTER 4
van Belkum, A., S., Scherer, L. van Alphen, and H. Verbrugh. 1998. Short-sequence DNA repeats in prokaryotic genomes. Microbiol. Mol. Biol. Rev. 62:275-293. Wagner, D., J. Maser, B. Lai, Z. Cai, C. E. Barry 111, K. Honer Zu Bentrup, D. G. Russell, and L. E. Bermuda. 2005. Elemental analysis of Mycobacterium avium-, Mycobacterium tuberculosis-, and Mycobacterium smegmatis-containing phagosomes in-
COMPARATIVE GENOMICS OF c. rmuw
71
dicates pathogen-induced microenvironments within the host cell's endosomal system. J. Immunol. 174:1491-1500. Wagner, P. L., and M. K. Waldor. 2002. Bacteriophage control of bacterial virulence. Infect. Immun. 70:3985-3993. Yao, R., D. H. Burr, and P. Guerry. 1997. CheY-mediated rnodulation of Campylobacter jejuni virulence. Mol. Microbiol. 23: 1021-1031.
74
MILLER
NJC genomes were limited; for some species, e.g., C. cuwus and C. hominis, the total genomic data available consisted of the 16s rDNA sequence and perhaps the sequence of a few genes. Thus, for the NJC species, the genetic basis of virulence, host preference and colonization could not be identified. The availability of eight new NJC genomes (for the purposes of this chapter, C. jejuni subsp. doylei is considered an NJC) permits genomic comparisons between the NJC group and between the NJC genomes and the C. jejuni subsp. jejuni genomes. Such comparisons will identify common virulence determinants, surface structures, metabolic pathways, signal transduction and chemotaxis proteins, and other features contained within all Campylobacter strains. Ideally, in addition, Campylobacter genome comparisons might identify genes and pathways specific to particular Campylobacter species, providing the genetic basis for phenotypes observed for that species and identifying also the genetic basis underlying differences in host range, colonization, and pathogenicity. Campylobacter genomic comparisons permit also analyses regarding evolution and the origin of particular Campylobacter species and/or subspecies. Additionally, genomic analysis of the genus Campylobacter is enhanced by the presence of multiple related Campylobacterales genomes. These include the ge)~ nome of Arcobacter butzleri (Miller et al., 2 0 0 7 ~the three Helicobacter pylori genomes (strains 26695 [Tomb et al., 19971, J99 [Alm, 19991, and HPAGl [GenBank accession number NC0008086.1]), and the genomes of Helicobacter hepaticus (Suerbaum et al., 2003), Helicobacter acinonychis (Eppinger et al., 2006), Wolinella succinogenes (Baar et al., 2003), and Sulfurimonas denitrificans (formerly Thiomicrospira denitrificans [Takai et al., 20061; GenBank accession number CP000153.1). Additionally, the genomes of more distantly related Epsilonproteobacteria, such as Sulfurovum and Nitratiruptor (Nakagawa et al., 2007), have been completed. This chapter will compare and contrast the genomes of Campylobacter species other than C. jejuni subsp. jejuni. Where relevant, the NJC genomes will be compared with related genomes, especially those of C. jejuni subsp. jejuni and A. butzleri. Elements important in overall genome structure, such as genomic islands, prophage, and insertion sequences (ISs) will be discussed. Individual topics, including restriction modification (R-M), metabolism, and virulence will be presented. Discussion of surface structures (e.g., lipooligosaccharide, capsule, and C. fetus subsp. fetus S-layer proteins) and signal transductiod chemotaxis will be presented elsewhere in this book (chapters 27, 28, 23, and 20, respectively). Finally, genomic data from the organisms C. jejuni subsp.
doylei 269.97, C. fetus subsp. fetus 82-40, C. curvus 525.92, C. concisus 13826, and C. hominis ATCC BAA-381 are from Fouts et al. (unpublished data).
GENERAL FEATURES OF THE NJC GENOMES The first Campylobacter genome to be sequenced, C. jejuni subsp. jejuni NCTC 11168 (Parkhill et al., 2000), was approximately 1.6 Mb in size with a G+C content of 30.6%. Analysis of the NJC genomes (Table 1) indicates that these genomes are similar in size (approximately 1.53 to 1.97 Mb). The G+C content of most of the NJC genomes is 29.7 to 34.5%, with the exception of C. curvus 525.92, which has the much higher G+C content of 44.5%. These numbers contrast with the genome sequence of the related organism A. butzleri RM4018, which has both a larger size (2.3 Mb) and lower G+C content (27%). The total number of coding sequences (CDSs) per Campylobacter genome is proportional to the size of the genome, with an average gene size range of 988 to 1,051 bp. Finally, the CDSs with assigned function are fairly consistent across the sequenced Campylobacter genomes, with an assigned function/ total CDSs average of 46%. Pairwise BLASTP comparisons identified 8 80 proteins conserved across all of the NJC proteomes. Of these 880 proteins, 753 were conserved further across the family Campylobacteraceae (i.e., Campylobacter + Arcobacter), and 569 were conserved still further across the order Campylobacterales (Campylobacteraceae + Helicobacteraceae). The identities of the majority of proteins within the core protein set are, predictably, proteins with housekeeping functions, e.g., those involved in protein synthesis, transcription, DNA replication, flagellar biosynthesis, and biosynthesis of purines and pyrimidines. However, 294 proteins in the core protein set have been classified as “conserved hypothetical” or have been assigned only a general function. The majority of proteins within the core protein set have homologs in other bacterial taxa: only 27 of the 880 proteins would be considered unique, within Campylobacterales, to Campylobacter. Only a few of these 27 proteins, e.g., SerC, have assigned function; the remainder are conserved hypothetical. The proteomes of C. concisus 13826 and C. hominis ATCC BAA-381 were not included in the BLASTP analysis. Because the exclusion of C. lari from the pairwise BLASTP analysis would lead to a larger core protein set as a result of the inclusion of multiple biosynthetic genes (see below), it is likely that the size of the core protein set will diminish as new Campylobacter proteomes are factored into the analysis.
Table 1. General features of the sequenced Cumpylobucteruceae genomes" Parameter Genome size (bp) Yo G+C CDS numbers" Pseudogenesd Assigned function Generallunknown function No. of prophagelgenomic islands1CRISPRs No. of rRNA operons Plasmids (kb) No. of IS elements No. of poly G:C tracts (>8 bp) Average tract size No. of restrictionlmodification systems Type I (hsd) Type I1 Type 111 No. of two component systems Response regulator Sensor histidine kinase
Cii
Cii
3 1;3;4;180 6 18 9.4
-1,660,000 34.5 1,670 ND 79 1 879 0/1/0 3 3;110 0 75 13.6
1,525,460 29.7 1,500 13 736 764 11010 3 46 0 15 9.7
29 2 1
1 2 0
49 69 29
10 6"
9 7
7 4
1,616,554 30.6 1,622 32 670 952 O/l/O 3 37;45 0 19 9.3
1,845,106 30.6 1,728 125 811 917
1,684,122 31.4 1,639 18 785 853
11.511
o/o/o
3 0 3 74 10.6
19 3 1
2 1 0
9 6
9 7
3 0 0 29 9.4
1,777,831 30.3 1,800 47 758 1042 21211 3 0 0 24 9.9
1 3 0
10 6
01011
RM2100
RM2228
81-176
1,641,481 30.6 1,597 27 833 764
cu RM319.5'
Cjd 269.97
Cii RM1221
11168
cc
Ab
Cur 525.92
RM4018
1,773,615 33.3 1,710 27 795 915 011/1 3 0 0 30 9
1,971,264 44.5 1,875 21 743 1,132 01312 3 0 0 9 10
2,341,251 27.1 2,259 5 1,011 1,248 1/3/0 5 0 0 0 NIAf
19 2 0
1 1 0
1 19 0
0 0 0
8 6
12 12
15 14
42 37
Cf 82-40
Cl
Cjj, Campylobacter jejuni subsp. jejuni: Cjd, C. jejuni subsp. doylei: Cc, C. coli: Cu, C. upsaliensis; Cl, C. lari; Ck C. fetus subsp. fetus; Cur, C. curvus; Ab, Arcobacter butzleri; CRISPRs, clustered regularly interspaced short palindromic repeats. bNumbers based on preliminary analysis of the RM3195 draft genome. "Total does not include pseudogenes, where applicable. dTotals include potential contingency genes. 'ND, not determined. fNA, not applicable. g Contains putative pseudogenes or contingency genes. Two sensor histidine kinases are putative pseudogenes. a
76
MILLER
By use of the core protein set, pairwise average amino acid identities between any two Campylobacter proteomes were calculated (Fig. 1).For example, the 880 core proteins of C. jejuni subsp. doylei 269.97 had an average amino acid identity per protein of 87.2% when compared by BLASTP to their homologs within the 880 core proteins of C. coli RM2228. Average amino acid identity values ranged from 97.6% for C. jejuni subsp. doylei 269.97 and the related C. jejuni subsp. jejuni genomes to 58.4 to 59.6% for C. curvus 525.92 and C. fetus subsp. fetus 82-40 when compared with any of the thermophilic Campylobacter genomes. The amino acid identity values between C. fetus subsp. fetus and C. curvus were 64.9%, indicating that although they demonstrate similar amino acid identity values when compared with the thermophilic campylobacters, they are similarly unrelated to each other. The most related genomes to C. jejuni subsp. jejuni were, in order: C. jejuni subsp. doylei, C. coli, C. upsaliensis, C. lari, and C. fetus subsp. fetuslC. curvus. Amino acid identities varied by 1.1% between the C. jejuni subsp. jejuni genomes; however, the level of strain-to-strain variation in the NJC genomes is presently unknown. Multilocus sequence typing (MLST) of the NJC genomes suggests that similar levels of strain-to-strain variation exist within these species (Miller et al., 2005), especially within C. fetus subsp. fetus, where interstrain variation is remarkably low (van Bergen et al., 2005). However, the seven genes utilized by MLST represent a very small fraction of the total genome, and MLST genes are generally housekeeping genes with intrinsically lower levels of variation. Strain-to-strain variation would also be expected to be influenced by lateral gene transfer. For example, the sequenced genomes of C. jejuni subsp. jejuni RM1221 and C. jejuni subsp. doylei 269.97 contain regions that originated in C. coli and C. upsaliensis, respectively. The C. coli region within strain
I
cii 98.9
I
I
RM1221 encodes core proteins and thus formed part of the 1.1% variation within C. jejuni subsp. jejuni. MLST analysis of other Campylobacter species has identified other instances of putative lateral gene transfer (data not shown). Unfortunately, a more thorough analysis of strain-to-strain variation at the sequence level within the NJC species will require sequencing additional genomes. However, this solution is not presently feasible because of the high cost of genomic sequencing. A more reasonable middle ground would entail sequencing of a much larger gene set than the seven sequenced by MLST. This gene set would have to be small enough to enable the analysis of multiple strains but large enough to present an accurate representation of intraspecific variation.
PLASMIDS Plasmids of a wide range of sizes have been observed in many NJC species, including C. coli (Taylor and Tracz, 2005), C. upsaliensis (Owen and Hernandez, 1990; Stanley et al., 1994) and other, nonthermophilic campylobacters (Boosinger et al., 1990; Waterman et al., 1993; Willoughby et al., 2005; data not shown). The primary phenotypic marker associated with Campylobacter plasmids is antibiotic resistance. Genes encoding resistance to tetracycline, kanamycin, and chloramphenicol have been identified on Cumpylobacter plasmids (Trieber and Taylor, 2000). Cumpylobacter plasmids fall into two basic groups: cryptic plasmids and megaplasmids. Although cryptic plasmids are small, generally <5 kb in size, and megaplasmids are much larger molecules, in Campylobacter as large as 180 kb, the primary distinction between the two types of plasmids is related to
Cid
cc
cu
c/
Cf
Cur
97.6 100
87.7 87.2 I00
77.6 77.2 77.4 100
71.9 71.6 71.7 69.8 I00
59.4 59.3 59.5 58.4 59.3
59.6 59.5 59.6 58.7 59.6
I
i
i I
i
I
i
I
Cjj cid Cc CU
c/
Figure 1. Average amino acid identities of the Cumpylobucter proteomes. Pairwise BLASTP comparisons of the Cumpylobacter proteomes were performed. For each protein from each proteome, the protein with the most significant homology from the other proteomes was identified. Significant homology was defined as >30% amino acid sequence identity with at least 75% sequence overlap. From these matches, a set of 880 core proteins, conserved across the Cumpylobacter proteomes, was identified. Values represent average amino acid identities between the core protein sets of any two given Cumpylobucter proteomes. See Table 1, footnote a, for abbreviations.
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. 7ETUNr CAMPYLOBACTER SPP.
the genes encoded by the plasmids. Cryptic plasmids are defined as plasmids that encode only one or two replication proteins and, usually but not always, one or two mobilization proteins. Megaplasmids are distinguished by the presence of several genes that encode proteins involved in plasmid conjugation. Cryptic Plasmids Multiple NJC cryptic plasmids have been identified and characterized (Kienesberger et al., 2007; Miller et al., 2007a; Waterman et al., 1993). In some strains, multiple plasmids are present (Kienesberger et al., 2007; Miller et al., 2007a). One such strain is C. coli RM2228, which contains three cryptic plasmids, pCC2228-1 (1.3 kb), pCC2228-2 (3.3 kb) and pCC2228-3 (4.3 kb), and the 180-kb megaplasmid pCC2228-4 (Miller et al., 2007a). The presence of three cryptic plasmids in strain RM2228 suggests that these three plasmids represent multiple incompatibility (Inc) groups (Miller et al., 2007a). The existence of multiple Inc groups within this strain is evidenced further by predicted differences in the replication origin and replication apparatus of each cryptic plasmid. pCC2228-1 is unusual in both its small size (1.3 kb) and the presence of only one CDS on the plasmid. Comparison of this CDS to the GenBank nr database suggests that this CDS encodes a replication protein of the Firmicutes RepL family. RepL-family plasmids have been shown to replicate via a rollingcircle mechanism (Hefford et al., 1997), although no data are available to indicate that pCC2228-1 replicates by a similar mechanism. Rolling circle plasmids are not unique in Campylobacterales. Such plasmids have been identified in Helicobacter (Kleanthous et al., 1991); however, the replication proteins from these plasmids belong to the unrelated Rep-2 family. pCC2228-2 is a typical theta-replicating, iteroncontaining plasmid related to the majority of characterized Campylobacter plasmids, including most of the Campylobacter shuttle plasmids. Iterons are binding sites for the cognate replication (Rep) protein, and binding of Rep to the iterons is necessary for replication and copy number control (Chattoraj, 2000). pCC2228-3 is also an iteron-containing plasmid. However, it differs from pCC2228-2 in that the replication protein, RepC, is unrelated to the RepAB proteins of pCC2228-2, and that the plasmid organization is repCliteronlmob instead of iteronlrepAB1 mob. Most of the characterized NJC cryptic plasmids resemble pCC2228-2 in both gene content and iteronlrepABlmob organization. One exception is the C. fetus subsp. venerealis plasmid pCFVlO8 (Kienes-
77
berger et al., 2007), which is similar in organization to pCC2228-3 and pCJ1170. Miller et al. (2007a) proposed that as many as nine Inc groups exist within the Campylobacter cryptic plasmids. This estimate was based on the presence of both unique iteron repeat sequences and phylogenetically distinct replication proteins. Plasmids from the same putative Inc group have been isolated from different species. For example, plasmids pCC2228-3 and pCJ1170, isolated from C. coli and C. jejuni subsp. jejuni, respectively, contain identical iteron repeats and Rep proteins (Miller et al., 2007a). The existence of similar plasmids in different Campylobacter species indicates that some cryptic plasmids have an expanded host range. For instance, the pIP1455-derived shuttle vectors have been used to transform C. jejuni subsp. jejuni, C. coli, C. lari, and C. fetus subsp. fetus (Kienesberger et al., 2007; Miller et al., 2000; data not shown). pIP1455-derived plasmids do not replicate in C. hyointestinalis, although plasmids derived from C. hyointesintalis replicate also in C. fetus (Waterman et al., 1993).Additionally, pIP1455-derived plasmids have not been used successfully to transform C. upsaliensis (data not shown); however, this may reflect enhanced restriction-modification in this species rather than host range. Other plasmids, such as pCFV108, have been demonstrated also to possess a narrower host range (Kienesberger et al., 2007). Megaplasmids Potentially, the most interesting of the Campylobacter plasmid types are the megaplasmids. This is due to their much greater size, with a concomitant larger number of CDSs, and their similarity to characterized Campylobacter genomic islands (Fouts et al., 2005). Unfortunately, sequence information exists for only four NJC megaplasmids: the 44.7-kb C. coli megaplasmid pCC3 1 (Batchelor et al., 2004) and the three megaplasmids described by Fouts et al. (2005): the 180-kb C. coli megaplasmid pCC22284, the 110-kb C. upsaliensis megaplasmid pCU3 1952, and the 46-kb C. lari megaplasmid pCL2100. All four megaplasmids encode homologs of the Tra type IV secretionlplasmid conjugation proteins, termed Cmg (Campylobacter mating genes) by Batchelor et al. (2004), or other proteins- involved in plasmid conjugation and/or DNA transfer. In addition, pCC2228-4 and pCC3 1 encode antibiotic resistance genes: both plasmids encode the tetracycline resistance gene tetO, and pCC2228-4 encodes putative kanamycin (kan) and hygromycin-B (hyg) resistance genes. The tetO gene in both plasmids resides in an area of high G+C content (39 to 40%), suggesting
78
MILLER
lateral gene transfer (Batchelor et al., 2004). However, the TetO protein of pCC2228-4 is only 93% identical to the TetO protein of pCC3 1 and is almost identical (99% identity) to the TetO protein of C. jejuni subsp. jejuni 81-176 megaplasmid pTet. Intriguingly, the pCC2228-4 kan and hyg genes are contained within a putative mobile element similar to clostridial transposons (see below). Additionally, pCC2228-4 contains three identical copies of an IS, ISCcol. The remainder of the genes contained on the four Campylobacter megaplasmids encodes mainly proteins of unknown function. Nevertheless, several of these genes have homologs in Epsilonproteobacterial or other eubacterial taxa or, significantly, homologs within Campylobacter genomic islands (Fouts et al., 2005). pCC31 contains several genes homologous to genes from the Actinobacillus actinomycetemcomitans plasmid pVT745 (Batchelor et al., 2004), and pCC2228-4 contains genes homologous to genes from Vibrio spp. It is likely that the four NJC megaplasmids described here represent only a small subset of the total NJC megaplasmids. It is also likely that the gene content of NJC megaplasmids is modulated by lateral gene transfer and the presence of repeated sequences and mobile elements. Recombination between repeat sequences and mobile elements would result necessarily in a more modular megaplasmid structure, leading to new combinations of genes. This is especially true with pCC2228-4, which contains four mobile elements and where approximately 25% of the megaplasmid is composed of repeated sequences.
HOMOPOLYMERIC G:C TRACTS, CONTINGENCY GENES, PSEUDOGENES, AND MISMATCH REPAIR Homopolymeric G:C tracts have been identified in all of the NJC genomes (Fouts et al., 2005) (Table 1). The number of G:C tracts 8 bp or longer ranges from nine in C. cuwus 525.92 to at least 75 in C. upsaliensis RM3195. The average length of these tracts is similar to those reported in C. jejuni subsp. jejuni (approximately 9 to 10 bp) with the exception of C. upsaliensis; homopolymeric G:C tracts in C. upsaliensis RM3195 are as long as 19 bp and average nearly 14 bp in length. Neither the length nor number of homopolymeric G:C tracts in the NJC genomes is proportional to the G+C content of the genome; C. cuwus 525.92 has the fewest number of G:C tracts but has the highest G+C content (44.5%; Table 1).Similar to homopolymeric G:C tracts in C. jejuni subsp. jejuni (Gilbert et al., 2005; Miller et al., 2005b; Parkhill et al., 2000), homopolymeric G:C
tracts within the NJC genomes are located often within genes involved in synthesis of surface structures and restriction/modification genes. For example, 46 of 74 of the C. jejuni subsp. doylei 269.97 homopolymeric G:C tracts are located within genes related to surface structure/restriction-modification. Hypervariability at these loci has been demonstrated to lead to changes in surface structures (Guerry et al., 2002; Karlyshev et al., 2005; Linton et al., 2000). It is likely that hypervariable G:C tracts in the NJC genomes would lead to similar changes in phenotype. Contingency (phase-variable) genes, characterized by hypervariable G:C tracts, have been identified in the C. coli, C. lari, and C. upsaliensis genomes (Fouts et al., 2005); no data are available for the remaining NJC genomes. Although putative contingency genes within the thermophilic NJC genomes have been identified, the proportion of these genes with respect to total G:C tract-containing genes is small. Identification of contingency genes within the NJC genomes has relied on a sufficient level of sequencing coverage. Tracts with a low level of coverage (e.g., two to three reads) may appear stable simply because not enough data are available to detect hypervariability. Thus, additional experiments will be necessary to determine whether each homopolymeric tract is stable or hypervariable. Additionally, it is possible that some tracts may be hypervariable only under certain environmental conditions. Therefore, tracts that appear stable after growth on laboratory media may become hypervariable when the strain is inoculated into a host. Finally, stability of homopolymeric tracts can vary between strains. This was demonstrated in C. jejuni subsp. jejuni 30 and RM1516, where the homopolymeric G:C tracts within identical hsdS genes were stable in strain 30 and hypervariable in strain RM1516 (Miller et al., 2005b). Several genes within the NJC genomes are frameshifted or have other defects and are considered putative pseudogenes (Table 1).Many of these frameshifts are centered on a homopolymeric G:C tract; thus, it is possible that these G: C-tract-containing pseudogenes are phase variable. As before, additional investigations will be necessary to determine the nature of these putative contingency genes. Despite the presence of putative contingency genes within the population of pseudogenes, contained within each NJC genome are gene fragments or genes with a combination of authentic (nonhypervariable) frameshifts and point mutations. The C. jejuni subsp. doylei 269.97 genome is predicted to contain 125 pseudogenes. Twenty-nine of these pseudogenes in strain 269.97 contain homopolymeric G:C tracts and are potentially phase variable. The remainder are related
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. TETUNI CAMPYLOBACTER SPP.
to genes encoding proteins involved in signal transduction (e.g., methyl-accepting chemotaxis proteins), transport (e.g., feoB), restriction/modification (e.g., hsd), and metabolism (e.g., pro& sdhABC, and nap). In two instances, the origin of pseudogenes in strain 269.97 can be ascribed to insertional inactivation, either via a genomic island or an IS element. However, there are no definite data on why C. jejuni subsp. doylei 269.97 contains so many putative pseudogenes relative to the other NJC genomes. The genome of strain 269.97 does contain 74 homopolymeric G:C tracts. Indeed, there is a rough correlation between the number of G:C tracts and the number of pseudogenes. However, these tracts account only for 23% of the total pseudogenes in C. jejuni subsp. doylei 269.97. It is more likely that a root cause exists that accounts for the presence of both pseudogenes and contingency genes; this cause is possibly related to DNA repair. An unexpected feature of the A. butzleri RM4018 genome was the absence of homopolymeric G:C tracts (Table 1; Miller et al., 2 0 0 7 ~ a) ~feature shared also by the genome of the related S. denitrificans. Therefore, comparison of the A. butzleri and S. denitrificans genomes to other genomes of the epsilon subdivision would perhaps permit construction of a model describing the formation of homopolymeric G:C tracts. Analysis of these two genomes revealed the presence of two to four members of the DnaQ superfamily (Miller et al., 2 0 0 7 ~ genes ) ~ that were absent in the other Epsilonproteobacterial genomes. The DnaQ superfamily includes the 3 ’ 4 5 ’ exonuclease ExoX (Viswanathan and Lovett, 1999). The 3 ’ 4 5 ’ exonucleases are key components of the methyl-directed mismatch repair (MMR) system. MMR depends on three functions: the MutSLH endonuclease complex, chromosomal methylation (to identify parental/daughter strands), and multiple 5’-3 ’ and 3 ‘ 4 5 ’ single-strand DNA exonucleases (Iyer et al., 2006). Although genes encoding 5 ’ 4 3 ’ single-strand DNA exonucleases (e.g., recJ) have been identified in Campylobacter and Helicobacter, no protein with predicted 3 ’ 4 5 ’ exonuclease activity has been identified. Multiple mutations in the Escherichia coli 3 ‘ 4 5 ’ exonucleases resulted in a 32fold increase in +1 frameshifts and an 11-fold increase in -1 frameshifts (Viswanathan and Lovett, 1998). Therefore, it is possible that the presence of homopolymeric G:C tracts can be explained in part by the state of MMR in an organism: organisms such as A. butzleri RM4018, which are predicted to encode both 5 ’ 4 3 ’ and 3 ’ 4 . 5 ’ single-strand DNA exonucleases, would be predicted to contain a fully functional MMR system.
79
GENOME STRUCTURE Genomic Colinearity between the Campylobacter Genomes Total genomic colinearity between C. jejuni subsp. jejuni and the NJC genomes is quite low. The exceptions are the C. coli RM2228 and C. jejuni subsp. doylei 269.97 genomes. The RM2228 and C. jejuni subsp. jejuni NCTC 11168 genomes are remarkably colinear (Fig. 2A), with only a small number of inversions in the RM2228 genome relative to the NCTC 11168 genome. Only two of these are of any substantial size and are clustered near the predicted origin of replication (approximately bp 1).The C. jejuni subsp. doylei 269.97 genome contains a large genomic inversion, relative to the NCTC 11168 genome, bounded by IS elements (Fig. 2B); otherwise, the C. jejuni subsp. doylei and C. jejuni subsp. jejuni genomes would be largely colinear. For the remainder of the NJC genomes, pairwise comparisons of colinearity resemble Fig. 2C, in which the gene orders of C. lari RM2100 and C. jejuni subsp. jejuni NCTC 11168 were compared, where no obvious large syntenic regions exist. Although overall colinearity between the NJC genomes is generally low, pockets of localized synteny can be identified within the genomes. Such syntenic pockets represent clusters of genes that remain linked within all members of the genus. Analysis of the Campylobacter genomes identified 44 clusters of colinear genes with a minimum size of five CDSs (Table 2). These clusters range in size from 5 to 21 CDSs, with an average size of 8 CDSs. In some cases, these clusters contain genes that encode similar functions or components of a single pathway. For example, cluster 18 encodes tricarboxylic acid (TCA) cycle enzymes, and cluster 44 encodes SecY and 20 ribosomal proteins. However, for most of the clusters, no common biological function can be defined, and it is unclear why the genes in these clusters remain linked throughout the genus. The total number of genes present- in these clusters is 357, indicating that the majority of Campylobacter genes are not located in syntenic clusters and are not tightly linked. As with the core protein set, the C. concisus and C. hominis genomes were not analyzed for localized synteny; therefore, it is likely that some of these 44 clusters will be absent from future lists that include additional Campylobacter genomes. Bacteriophage Mu (CMLP1 and AMLP1) The bacteriophage Mu-like element CMLPl (Campylobacter Mu-like - _ phage) was identified first in jejuni subsp. jejuni R&i1221(Fouts et al., 2005).
80
MILLER
C. jejuni NCTC 11168
1800 b
1600
Q)
6
1400
(0 (v
1200
.-
2
3 1000
D
.-c 2 .%
'
800 600 400 200
0
C.jejuni NCTC 11168
C
C. jejuniNCTC 11168
Figure 2. Pairwise dot plot comparisons of selected Cumpylobucter genomes. Each protein in the C. jejuni subsp. jejuni NCTC 11168 proteome was compared with the proteomes of C. coli RM2228 (A), C. jejuni subsp. doylei 269.97 ( B ) , and C. luri RM2100 (C) by BLASTP analysis. Each dot represents a protein with significant homology, using the parameters from Fig. 1, to a C. jejuni subsp. jejuni NCTC 11168 protein. The x and y axis values represent gene numbers.
Amplification of other NJC strains with a MuB primer set has detected CMLP-related elements in other Campylobacter species, such as C. coli, C. lari, C. upsaliensis, C. helveticus, C. fetus subsp. fetus, C. mucosalis, C. hyointestinalis, C. concisus, and C. curvus
(data not shown), although none of the sequenced NJC genomes is predicted to contain a CMLP-related element. CMLP-related elements have also been identified in non-Campylobacter Epsilonproteobacterial species, e.g., Helicobacter bilis, A. butzleri, A. skirro-
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. TETUNI CAMPYLOBACTER SPP.
81
Table 2. Localized regions of synteny within Campylobacter Avg. aa identity (vs Cjj)" Region no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
Locus CjOO59c-CjOO68 CjOO98-CjOll8 CjOl23c-CjO12 7c CjOl29c-CjO138 CjO147c-CjO155c CjO156c-CjO166 CjOl86c-CjO193c CjO232c-CjO236c Cj0273-CjO2 77 Cj0281c-CjO288c CjO317-CjO322 CjO328c-CjO333c CjO365c-CjO369c CjO394c-CjO405 CjO472-CjO4 79 CjO495-CjO499 Cj0509c-CjO5 16 Cj0531-el0538 CjO539-CjO545 CjO580c-CjO590 Cj0595c-CjO600 CjO63Oc-CjO635 CjO639c-CjO652 CjO699c-CjO714 CjO795c-CjO8OOc Cj080l-Cj0807 Cj089 1c-CjO897c CjO925-CjO930 CjO954c-CjO96 1c CjO992c-CjO997 CjlO56c-CjlO62 CjlO9Oc-CjlO95 CjlO99-CjllO5 Cjl123c-Cjl127c Cjll75c-Cjll82c Cj1226c-Cj123 1 Cjl27Oc-Cjl282 Cjl366c-Cjl3 70 Cjl487c-Cjl495c Cjl592-Cj1596 Cjl606c-Cjl610 Cjl634c-Cjl638 Cjl639-Cjl645 Cjl688c-Cjl708c
No. of genes 10 21 5 10 9 10 8 5 5 8 6
6 5 12 8 5 8 8 7 11 6 6 14 16 6 7 7 6 8 6 7 6 7 5 8 6 12 5 9 5 5 5 7 21
Cjd
Cc
Cl
Cu
Cff
Cur
97 98 98 97 98 98 98 98 99 98 97 98 97 98 99 96 97 99 97 96 98 97 98 98 97 98 97 98 98 96 99 97 98 96 98 98 99 99 98 99 99 98 99 99
88 87 86 83 82 81 87 88 91 89 87 90 85 80 97 80 83 96 83 78 86 84 82 85 80 87 86 85 84 84 88 89 79 87 86 92 85 88 85 98 86 85 87 98
71 70 63 72 69 63 77 82 80 76 73 81 59 65 88 56 69 83 67 66 72 66 63 76 58 74 68 75 68 70 74 71 68 65 70 73 71 68 77 93 68 66 70 91
83 80 73 74 75 67 77 82 84 77 79 85 69 72 93 68 77 91 72 71 76 68 72 81 71 77 76 78 74 71 76 83 71 72 75 80 79 75 77 94 75 71 75 93
60 57 52 58 60 50 62 60 67 61 61 70 47 55 81 50 59 63 51 56 59 54 52 66 52 63 58 68 57 58 58 59 55 53 58 62 60 59 63 86 57 55 54 82
58 58 53 59 65 48 62 66 69 65 60 70 46 56 82 51 60 60 50 56 54 57 52 66 50 63 61 69 59 59 64 57 56 53 58 61 61 60 61 85 56 54 55 83
amino acid; Cjj, Campylobacter jejuni subsp. jejuni; Cjd, C. jejuni subsp. doylei; Cc, C. coli; Cl, C. lari; Cu, C. upsaliensis; Cfk C. fetus subsp. fetus; Cur, C. cuwus.
aaa,
wii, and A. cryaerophilus (data not shown). The genome of the A. butzleri RM4018 contains a Mu-like phage, termed AMLP1 (Arcobacter -Mu-like phage) (Miller et al., 2 0 0 7 ~ ) Al&ough . several of th; proteins encoded by AMLP1 have homologs within CMLP1, there are differences between the two phages, such as the absence of the CMLP1 extracellular DNase in AMLP1. DNA microarray analysis and preliminary sequence analysis of other CMLP elements also indicate differences in gene content be-
tween the CMLP elements (Parker et al., 2006; data not shown), suggesting that CMLPl and AMLP1 are just two members of a much larger family of Mu-like bacteriophage prevalent within the Epsilonproteobacteria. Other Prophage Bacteriophage Mu-like elements are identified in bacterial genomes as a result of the presence of sev-
82
MILLER
era1 well-conserved components, such as the genes that encode the Mu transposase. However, members of other less well-characterized prophage families can be difficult to identify, and confirmation of such prophage relies often on the presence of a few phagerelated genes and one or two putative capsid genes. Moreover, a protein with low similarity to a phagerelated protein does not always signify a bacteriophage origin. Thus, for several of the NJC integrated elements, additional investigations will be necessary ( e g , bacteriophage induction) before an element can be identified definitively as a prophage. Nevertheless, the integrated element CLIEl (Campylobacter -lari -integrated element) in C. lari m 2 l O O is almost certainly a limbdoid-type bacteriophage, similar to the element CJIE4 (Campylobacter jejuni integrated element) in C. j e j u s subsp. jejuni-RM1221. Additionally, one of the integrated elements in C. jejuni subsp. doylei 269.97 may be a prophage, although this label is more tenuous.
ulence functions were identified within these genome islands. In fact, no biological role has been ascribed to any of the nonprophage Campylobacter genomic islands. However, two NJC genomic islands are particularly noteworthy: the C. curuus 525.92 element CURIE2 (Curvus integrated element) and the C. jejuni subspxoylei 269.97 eiement DIE6. CURIE2 contains the sole Type I R-M locus in C. curvus 525.92, suggesting that this R-M locus was acquired through lateral transfer. This is similar to A. butzleri RM4018, in which the only DNA methyltransferase present in the genome is contained within a genomic island (Miller et al., 2 0 0 7 ~ )DIE6 . is a clear representation of lateral gene transfer between C. upsaliensis and C. jejuni subsp. doylei. DIE6 contains 26 CDSs, all encoding proteins with homologs of high similarity (? = 97.5% identity) to proteins of C. upsaliensis RM3195. The C. upsaliensis CDSs are colinear with the C. jejuni subsp. doylei CDSs and are predicted to encode mainly glycosyltransferases and other proteins involved in capsular biosynthesis.
Genomic Islands/Integrated Plasmids
CFUSPR Elements
As with identification of prophage, the distinction between genomic island and integrated plasmid can be subtle and not discerned readily. Certainly, the presence of plasmid-related genes, such as the plasmid-conjugation-related tra genes, implies a plasmid origin. One such integrated plasmid is the element DIE4 (Doylei integrated element), which contains at least Tl tra genes. Integrated plasmids can be identified also through homology to characterized megaplasmids, as was the case with the C. jejuni subsp. jejuni RM1221 element CJIE3 (Fouts et al., 2005). However, the remaining NJC genomic islands do not have significant homology to existing characterized plasmids, indicating either that these elements have a nonplasmid origin or that the ancestor plasmids of these islands have not yet been identified. In the NJC genomes, genomic islands have been identified in C. jejuni subsp. doylei, C. upsaliensis, C. fetus subsp. fetus, and C. curvus. Consistent with the integrated elements of C. jejuni subsp. jejuni RM1221 (Fouts et al., 2005), many NJC genomic islands contain a phage-related integrase/ recombinase-encoding gene on one end, insertion points in or in close proximity to tRNAs (or the tmRNA in DIE1) and terminal repeats of 15 to 32 bp. The size of the NJC genomic islands ranges from 7 CDSs in C. fetus subsp. fetus 82-40 to DIE4, which contains 52 CDSs. Unfortunately, the NJC genomic islands are predicted to mostly encode conserved hypothetical proteins of unknown or at best general function. No drug-resistance genes or genes encoding putative vir-
CRISPRs (clustered regularly interspaced short palindromic repeats) have been identified in -40% of the bacterial and archaeal genomes (Kunin et al., 2007). These repeats are generally 25 to 50 bp in length and are separated by nonrepetitive spacer sequences of similar length. The spacer sequences are often unique, and spacers in some CRISPRs have been demonstrated to be of extrachromosomal origin, i.e., phage and plasmids (Bolotin et al., 2005). CRISPR-associated (Cas) protein families are present only in CRISPR-containing strains and are always found adjacent to a CRISPR (Haft et al., 2005). CRISPR elements have been organized into at least eight different subtypes (Haft et al., 2005) on the basis of factors such as gene composition, repeat length, and Cas sequence similarity. In addition to the Cas proteins (Casl-7), other CRISPR-associated protein families unique to particular CRISPR subtypes, such as Csn, Csm, Cmr, and Csh, have also been identified (Haft et al., 2005). CRISPR elements of the Nmeni subtype (Haft et al., 2005) have been identified in C. jejuni subsp. jejuni (Fouts et al., 2005; Schouls et al., 2003). Similarly, C. jejuni subsp. doylei 269.97 contains also an Nmeni-subtype CRISPR element. These elements have repeat lengths of 36 bp and contain only Casl, Cas2, and the CRISPR-associated protein Csnl. Although the C. jejuni subsp. doylei CRISPR repeat sequences are identical to those of C. jejuni subsp. jejuni NCTC 11168 and RM1221, the spacer sequences are unique, and the C. jejuni subsp. doylei CRISPR contains six repeats instead of the five pres-
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. 7ElUNl CAMPYLOBACTER SPP.
ent in strain NCTC 11168 and the four present in strain RM1221. No CRISPR elements were identified in the genomes of C. coli RM2228, C. lari RM2100, and C. upsaliensis RM3195 (Fouts et al., 2005). However, CRISPRs are present in the genomes of C. curvus 525.92 and C. fetus subsp. fetus 82-40. In fact, C. cuwus 525.92 contains two CRISPRs, and the repeat sequences of the two CRISPRs are different. The subtype of the first cannot be identified because of the absence of associated Cas proteins; this CRISPR contains nine repeats of 30 bp with an average spacer length of 36.4 bp. The second CRISPR element is of the Hmari subtype and contains four repeats of 25 bp and an average spacer length of 40.7 bp. This subtype is defined by the presence of Cas proteins Casl6 and two additional CRISPR-associated proteins, Cshl and Csh2. The C. fetus subsp. fetus CRISPR element is highly unusual. This element is bounded by two CRISPR arrays with an additional repeat sequence two genes downstream. The left CRISPR array contains 22 repeats, and the right CRISPR array contains 27 repeats. All of the repeats, including the downstream singlet, have identical 3 0-bp sequences, and the spacers within the arrays are 34 to 37 bp, with an average length of 35.9 bp. Another unusual feature is the size (22 CDSs) of the element and the organization and identities of the CRISPR-associated proteins: Cas proteins Casl-6 are linked to the right CRISPR array; however, several phage-related proteins, including a putative capsid protein, are linked to the left CRISPR array. The C. fetus subsp. fetus CRISPR element subtype is undefined and perhaps novel; however, BLASTP analysis indicates that the Cas proteins are similar to clostridial Cas proteins (e.g., C. kluyveri, C. novyi, and C. tetani) and the Cas proteins of Syntrophomonas wolfei. The presence of phage-related genes within the C. fetus subsp. fetus 82-40 CRISPR element is intriguing, considering that CRISPR elements have been demonstrated to provide resistance to bacteriophage in Streptococcus therrnophilus (Barrangou et al., 2007). In S. therrnophilus, phage resistance was demonstrated to be conferred by the addition of one to four novel spacers (with concomitant addition of repeats) to the internal (Cas-linked) end of the CRISPR array (Barrangou et al., 2007). These novel spacers are identical to phage sequences, and the presence of a spacer sequence identical to a phage sequence provides resistance against bacteriophage that contain that spacer sequence (Barrangou et al., 2007). It is possible that the Cas proteins serve to degrade bacteriophage DNA and insert small fragments as novel CRISPR spacers. In this case, the phage proteins in the C. fetus subsp. fetus CRISPR
83
element may have resulted from an error in which a larger than normal bacteriophage fragment was inserted. Another possibility is that the phage sequences originated via recombination between the terminal spacer sequence and a bacteriophage, although the small size of the spacer makes this unlikely. BLASTN analysis of the C. fetus subsp. fetus spacer sequences did not identify any homologous sequences within the NCBI nr database. It should prove interesting to characterize both additional C. fetus subsp. fetus CRISPR elements and C. fetus subsp. fetus bacteriophage and investigate phage resistance in this organism. ISs and Transposons The genomes of H. acinonychis Sheeba, W. succinogenes DSM 1740, and S. denitrificans ATCC 33889, as well as all three sequenced H. pylori genomes, are predicted to contain several ISs. However, the genomes of C. jejuni subsp. jejuni NCTC 11168 and RM1221 do not contain any IS elements, outside of a degenerate IS606-like element adjacent to the 5 s rRNA gene of rrnC. The genomes of C. lari RM2100, C. upsaliensis RM3195, C. fetus subsp. fetus 82-40, and C. curvus 525.92 also do not contain IS elements, suggesting that IS elements, although common within the Epsilonproteobacteria, are rare in Campylobacter. Nevertheless, IS elements have been identified in two Carnpylobacter genomes: C. jejuni subsp. doylei 269.97 and C. coli RM2228. Three copies of an IS60.5-like element (Kersulyte et al., 1998), termed here ISCjdl, are present in the C. jejuni subsp. doylei 269.97 genome. All three copies of ISCjdl are identical at the nucleotide level and are predicted to encode two transposases: the 12.3kDa IS200-like TnpA and the 47.7-kDa IS1341-like TnpB. ISCjdl is 1,851 bp, and its right end is similar (12/14 bp) to the right end of IS60.5; also, the target sequence (5’-TTTAAA) is similar to that of IS605 (Kersulyte et al., 1998). The orientation of ISCjd1-1 is reversed relative to the orientation of the other two elements. C. jejuni subsp. doylei 269.97 contains a large chromosomal inversion whose breakpoints are bounded by elements ISCjdl-1 and ISCjd1-3 (Fig. 2). Because these two elements have opposite orientations, it is likely that this inversion arose via homologous recombination between these IS elements. Similar IS60.5-mediated genomic rearrangements have been observed in H. pylori (Kersulyte et al., 1998). The C. coli RM2228 contains six copies of a novel 1,870-bpY IS606-related IS termed ISCcol (Fouts et al., 2005). Three copies are located on the chromosome; the other three copies are located on the 180-kb megaplasmid pCC2228-4. All six copies
Next Page 84
MILLER
are nearly identical at the nucleotide level; however, two of the genomic copies contain identical 344-bp deletions within tnpA. Although IS606 and ISCcol share a common target sequence (5’-TTTAT) and have nearly identical (12/13 nt) right ends, the two ISs differ in the orientation of the transposase genes: tnpA and tnpB are transcribed divergently in IS606 and transcribed in the same direction in ISCcol. In addition to the IS elements present in the C. coli RM2228 genome and the megaplasmid pCC2228-4, pCC2228-4 is predicted to contain a mobilizable transposon similar to the clostridial transposons Tn44.52 and Tn44.53 (Adams, 2002). Mobilizable transposons encode mobilization functions but are not conjugative, relying on conjugative machinery provided in trans to transfer into recipient strains. Mobilizable transposons contain a resolvase rather than a transposase and contain a transfer origin, oriT, within the transposon; transposition proceeds through a nonreplicative circular intermediate (Adams, 2002). The putative mobilizable transposon of pCC2228-4 contains homologs of the clostridial proteins TnpX (resolvase), TnpV (excisionase), and TnpY in addition to a MobA/MobL family mobilization protein. The mobilizable transposons of Clostridium often contain antibiotic resistance genes, e.g., Tn44.51 and Tn4453 confer chloramphenicol resistance (Adams, 2002). Similarly, the C. coli mobilizable transposon contains two antibiotic resistance genes: aphA-3, which confers resistance to kanamycin, and an aminoglycoside phosphotransferase gene similar to an E. coli hygromycin-B resistance gene. The transposon encodes also a putative pyroglutamyl peptidase. The presence of identical copies of ISCcol on the strain RM2228 chromosome and megaplasmid suggests that Hfr (high frequency of recombination) strain formation may be possible wGhin Campylobacter. In E. coli, the conjugative F episome cointegrates into the chromosome via homologous recombination between IS elements common to both the chromosome and the F plasmid, resulting in an Hfr strain. Because the F episome contains an oriT transfer origin, conjugation between the Hfr strain and a suitable recipient generally leads to transfer of chromosomal sequences in addition to plasmid sequences. Thus, analogous Hfr formation in Campylobacter would result necessarily in significant lateral gene transfer between Campylobacter strains and perhaps different Campylobacter species. Therefore, Campylobacter IS elements would be predicted to affect greatly genome structure through both IS-mediated inversions/deletions and Hfr-mediated large-scale transfer of genetic material.
EVOLUTION AND PHYLOGENY Comparisons of multiple related genomes within a taxon can lead to insights about the evolution of that taxon and the formation of specific lineages or insights about the origin of particular species or subspecies. An example of metabolic lineages identified from comparing the Campylobacter genomes are the three dihydroorotase (DHOase) lineages identified within the Epsilonproteobacteria. DHOase is encoded by pyrC and catalyzes the second step in pyrimidine biosynthesis. In mammalian cells, DHOase is one domain of the CAD multifunctional protein, along with carbamoyl phosphate synthetase and aspartate transcarbamoylase (ATCase) (McPhail and Shepherdson, 2006). In some bacterial taxa, DHOase is a homodimer, not associated with ATCase (Brichta et al., 2004). Bacteria often contain two pyrC genes, encoding the type I1 homodimeric DHOase and a defective type I DHOase related to the multifunctional complex enzyme and demonstrated to associate with ATCase (Schurr et al., 1995). The defective DHOases lack a histidine zinc-binding domain critical for enzymatic activity (Schurr et al., 1995). The Campylobacter genomes are predicted to encode both the type I1 and nonfunctional type I DHOases, with the exception of C. fetus, C. cuwus, and C. concisus, which encode only a type I DHOase. Unlike the DHOases of the other campylobacters, these type I DHOases contain both zinc-binding domains and are presumably functional. Also, consistent with the identification of a functional ATCaseassociated DHOase, pyrC is linked in these three genomes to pyrB, which encodes the ATCase. Thus, two separate DHOase lineages are present within Campylobacter; all of the sequenced Helicobacteraceae genomes are predicted to encode two pyrC genes and would be members of the lineage that includes the thermophilic campylobacters. A third DHOase lineage is exemplified by the A. butzleri RM4018 genome that encodes only the homodimeric type I1 DHOase. Members of this lineage include the deepsea-vent Epsilonproteobacteria Sulfurovum and Nitratiruptor. The biological significance of these three lineages is unclear. It has been proposed that functional type I DHOases provide thermal protection for unstable intermediates and permit growth at higher temperatures (McPhail and Shepherdson, 2006). However, this appears unlikely in Campylobacter because members of the C. fetus DHOase lineage generally grow at lower temperatures than the thermophilic campylobacters. Genomic comparisons and analysis can sometimes lead to new perspectives regarding the origin of species or subspecies. One such example is the or-
Previous Page CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. 7E7UNI CAMPYLOBACTER SPP.
igin of C. jejuni subsp. doylei. The defining characteristic of C. jejuni subsp. doylei is the absence of nitrate reductase activity (Steele and Owen, 1988). DNA microarray analysis and genomic sequencing revealed defects in the nitrate reductase locus nap, specifically in the two genes encoding the large and small subunits of nitrate reductase, NapA and NapB, respectively (Miller et al., 2007b). In eight C. jejuni subsp. doylei strains tested, all napA loci contained a deletion, and the napB loci of seven strains contained either deletions or point mutations leading to truncated proteins. The napA loci of an additional 19 C. jejuni subsp. doylei isolates also contained deletions. Unexpectedly, the size of the napA deletions in all 27 C. jejuni subsp. doylei isolates was identical (2,761 bp) with identical endpoints. The presence of these identical napA deletions, identified in strains isolated across four continents, suggests that C. jejuni subsp. doylei arose from C. jejuni subsp. jejuni as a result of a single evolutionary event. Divergence at napB occurred subsequently as the constraint maintaining an active nitrate reductase was removed with the loss of NapA. The putative C. jejuni subsp. jejuni ancestor of C. jejuni subsp. doylei has not been identified.
METABOLISM AND RESPIRATION Embden-Meyerhof-Parnas and EntnerDoudoroff Pathways Members of the genus Campylobacter cannot utilize sugars as a sole carbon source. The inability to metabolize sugars can be traced to the absence of two key components of the Embden-Meyerhof-Parnas (EMP or glycolytic) pathway. The first component is the phosphoenolpyruvate-dependent phosphotransferase system, which simultaneously transports and phosphorylates sugars. The second component is phosphofructokinase, which catalyzes the conversion of fructose-6-phosphate to fructose-1,6 bisphosphate. Some members of the Helicobacteraceae, e.g., H. acinonychis and H. pylori (but not H. hepaticus), contain enzymes of the Entner-Doudoroff (ED) pathway. Like the EMP pathway, the ED pathway converts glucose-6-phosphate to glyceraldehyde-3-phosphate but utilizes a gluconate-6-phosphate intermediate rather than a fructose-6-phosphate intermediate. Thus, the ED pathway obviates the need for phosphofructokinase. Helicobacters are also missing the terminal step of the EMP pathway catalyzed by pyruvate kinase. However, the ED pathway produces also pyruvate in addition to glyceraldehyde-3-phosphate. Thus, the presence of the ED pathway in H. pylori, in addition to a predicted hexose transporter and glucokinase, would indicate that H. pylori can
85
utilize glucose as a sole carbon source; glucose utilization in H. pylori was demonstrated by Mendz et al. (1993). In addition to the EMP enzymes conserved throughout Campylobacter, C. jejuni subsp. doylei 269.97 is predicted to encode all of the ED pathway enzymes. In addition, this strain is predicted to encode a putative glucose/hexose transporter and a glucokinase homologous to proteins from Helicobacter. Therefore, the presence of these latter two proteins plus the combined enzymes of the EMP and ED pathways would suggest that C. jejuni subsp. doylei 269.97, like H. pylori, can utilize at least glucose as a sole carbon source. However, no data are available to indicate that C. jejuni subsp. doylei can utilize glucose or other carbohydrates. A surprising feature of the ED pathway in C. jejuni subsp. doylei is that the pathway-encoding genes are organized in a cluster and located between the isoleucinyl-tRNA and 23s rRNA genes in the rrnB rRNA locus. The presence of these Helicobacter-related genes in C. jejuni subsp. doylei suggests that this cluster was acquired from a member of this genus. Helicobacter strains that possess the ED pathway also contain a phosphoglucose isomerase (PGI) distinct from those found in the other Epsilonproteobacteria. These PGI proteins are similar to those from multiple bacterial taxa, including the Gammaproteobacteria, whereas the PGI proteins from Campylobacter, Arcobacter, and H. hepaticus are similar to Thermotoga and archaeal taxa. c. jejuni subsp. doylei 269.97 is predicted to encode both PGI types. Another example of dichotomy in the EMP enzymes is represented by fructose bisphosphate aldolase (FBA). Curiously, however, the Helicobacter FBAs are members of the type IIB group, which includes FBAs from Thermotoga and archaeal taxa, whereas Campylobacter FBAs (including C. jejuni subsp. doylei FBA) are members of the type IIA group, which includes FBAs from the Gammaproteobacteria. TCA Cycle
A complete oxidative tricarboxylic acid (TCA) cycle is present in C. jejuni subsp. jejuni, as evidenced by the presence of all three subunits of succinate dehydrogenase (encoded by sdhABC). The genome of C. jejuni subsp. doylei 269.97 contains the sdhABC genes. However, these genes are heavily fragmented and are unlikely to be functional. Genes encoding succinate dehydrogenase are absent from the C. coli, C. lari, C. upsaliensis, and C. fetus subsp. fetus genomes; however, succinate dehydrogenase is predicted to be encoded by both C. cuwus and C. concisus. Thus, complete oxidative TCA cycles would be
86
MILLER
predicted to be present in these latter two species and C. jejuni subsp. jejuni, whereas C. coli, C. upsaliensis, and C. fetus subsp. fetus would be predicted to contain a branched noncyclic pathway, similar to that present within H. pylori (Marais et al., 1999). Campylobacter lari RM2100 does not encode citrate synthase, aconitase, isocitrate dehydrogenase, succinate dehydrogenase, or succinyl-CoA synthetase (W. G. Miller et al., unpublished data). Therefore, it would appear that the TCA cycle is missing from this organism. The absence of &A, which encodes citrate synthase, was noted first during the development of an MLST method for C. lari (Miller et al., 2005a) because gltA is a member of the C. jejuni subsp. jejuni MLST gene set. Also, the absence of gltA was noted during development of an MLST method for the related species C. insulaenigrae (Stoddard et al., 2007), suggesting that the TCA cycle may be absent in this species as well. Campylobacter lari RM2100 is predicted to encode pyruvate carboxylase, malate dehydrogenase, fumarase, and fumarate reductase. Therefore, in strain RM2100, pyruvate can be interconverted into oxaloacetate, malate, fumarate, and succinate, presumably to provide metabolic precursors. Formation of acetyl-CoA via oxidative decarboxylation of pyruvate is catalyzed in Campylobacter by a pyruvate: flavodoxin oxidoreductase (Daucher and Kreig, 1995). Pyruvate:flavodoxin oxidoreductases are usually oxygen labile, and it has been proposed that this enzyme contributes in part to the microaerophily of Campylobacter (Kelly, 2005). Instead of pyruvate:flavodoxin oxidoreductase, the genome of the aerobic microorganism A. butzleri RM4018 is predicted also to encode all three subunits of pyruvate dehydrogenase, an enzyme found often in aerobic bacteria. Additionally, strain RM4018 is predicted to encode two fumarases: fumarase C, encoded also by Campylobacter and Helicobacter, and fumarase A. Fumarase A is an iron-sulfur-clustercontaining enzyme that is unstable under aerobic conditions, whereas fumarase C is stable in the presence of oxygen. The presence of TCA cycle enzymes with different oxygen labilities suggests that A. butzleri RM4018 is able to grow at a much wider range of oxygen concentrations than either Campylobacter or Helicobacter. Indeed, A. butzleri RM4018 has been demonstrated to grow anaerobically, microaerobically, and aerobically (Miller et al., 2 0 0 7 ~ ) . Respiration On the basis of the different environmental niches inhabited by the NJC genomes, it might be predicted that the respiratory pathways present in
these strains would be substantially diverse. Indeed, comparative analysis of the NJC genomes reveals a large degree of variation in the electron transport pathways. Most of the respiratory proteins described in C. jejuni subsp. jejuni NCTC 11168 (Kelly, 2005) are present also in the NJC genomes. The respiratory pathways of C. jejuni subsp. doylei 269.97 are the most limited among the characterized NJC genomes, and the pathways of C. curvus 525.92 are the most extensive. Besides the aforementioned deletions at the nitrate reductase (nap) locus, C. jejuni subsp. doylei 269.97 is also not predicted to encode the trimethyl amine-N-oxide (TMAO) reductase TorA (homologous to CjO264c). Unlike the nap locus, loss of torA is not due to genomic deletions but rather due to multiple frameshifts, resulting in a presumably nonfunctional protein. Intriguingly, both cytochrome c peroxidases, homologs of Cj0020 and Cj0358, are also present in strain 269.97 as pseudogenes. Although C. jejuni subsp. doylei 269.97 does not reduce nitrate, it is predicted to encode the NrfAH nitrite reductase. The genome of C. curvus 525.92 is predicted to encode several electron acceptors novel to Campylobacter. Two proteins are predicted members of the hybrid-cluster protein/prismane family; members of this family have two Fe/S clusters: a [4Fe-4S] cubane cluster and a hybrid [4Fe-2S-20] cluster. One protein has similarity to carbon monoxide dehydrogenases, and the other, also encoded by C. fetus subsp. fetus 82-40, has similarity to hydroxylamine reductases. However, both types are similar and can be interconverted via a single amino acid change (He0 et al., 2002; Wolfe et al., 2002). Hybrid-cluster proteins are expressed optimally under anaerobic conditions in the presence of nitrate or nitrite (Wolfe et al., 2002). C. fetus subsp. fetus and C. curvus 525.92 are predicted also to encode a putative hydroxylamine oxidoreductase that catalyzes the conversion of hydroxylarnine to nitrite. C. fetus subsp. fetus 82-40 contains homologs of the C. jejuni subsp. jejuni nitrite reductase NrfAH; however, homologs of these proteins were not found in C. curvus. Campylobacter curvus 525.92, C. fetus subsp. fetus 82-40, and A. butzleri RM4018 contain the NorB nitric oxide reductase large subunit. The cognate NorC nitric oxide reductase small subunit is not predicted to be encoded by any of these three strains, and no data are available as to whether these NorB homologs retain full activity in the absence of NorC. Campylobacter curvus and C. fetus subsp. fetus contain also the NosZ nitrous oxide reductase that reduces N 2 0 to N2. The presence of the prismane family proteins along with the putative hydroxylamine oxidoreductase and the NorB and NosZ homologs suggests that
CHAPTER 5
*
COMPARATIVE GENOMICS OF NON-C. TETUNI CAMPYLOBACTER SPP.
nitrogen metabolism in C. fetus subsp. fetus and C. curvus may be more complex than in the other NJC species. An additional electron acceptor present in C. curvus 525.92 is the polysulfide reductase PsrABC. Polysulfide reductase was characterized first in the Helicobacteraceae species W. succinogenes (Krafft et al., 1995). In W. succinogenes, Psr uses electrons obtained from hydrogenases or formate dehydrogenase to reduce polysulfides (e.g., tetrasulfide S,2- and pentasulfide SS2-) via a methyl-menaquinone-6 electron carrier (Dietrich and Klimmek, 2002). Reduction of polysulfides generates hydrogen sulfide; thus, presence and activity of this enzyme in strain 525.92 is consistent with the H,S production phenotype observed in C. cuwus strains. Finally, in addition to five putative TMAO reductases contained in C. curvus 525.92, this strain is also predicted to contain a novel DmsABC anaerobic DMSO reductase. In addition to the HyaABCD Ni/Fe hydrogenase, C. curvus 525.92 and C. fetus subsp. fetus 8240 contain a second hydrogenase, the hydrogenase-4 encoded by hyfABCEFGHI. As in E. coli, the hyf cluster of C. curvus is linked to focB, which encodes a putative formate/nitrate transporter (Skibinski et al., 2002); focB is not linked to the hyf cluster in C. fetus subsp. fetus. Linked also to both clusters are putative hydrogenase maturation genes hycH and h y d . Two additional variable respiration-related loci are present within the NJC genomes, cydAB and the NADH:quinone oxidoreductase complex (NDH-1) encoded by the 14-gene nu0 cluster. CydAB is one of two terminal oxidases in C. jejuni subsp. jejuni, the other being the cb-type cytochrome c oxidase encoded by ccoNOPQ. All NJC genomes contain both the CcoNOPQ and CydAB terminal oxidases, with the exception of C. lari and C. fetus subsp. fetus, which do not contain the CydAB terminal oxidase. The CydAB terminal oxidase is found also only in certain Helicobacteraceae species: H. pylori and H. acinonychis do not contain this oxidase, but W. succinogenes and H. hepaticus are predicted to encode CydAB. The NJC genomes all contain similar NDH1 complexes with the exception of C. fetus subsp. fetus 82-40. The NDH-1 complex of this strain is similar to the NDH-1 complex of A. butzleri RM4018 and differs from the other complexes in one important aspect. The NDH-1 complex is composed of 14 subunits: two subunits, NuoE and NuoF, are critical for binding and oxidizing NADH (Smith et al., 2000). In C. jejuni subsp. jejuni and H. pylori, the nuoE and nuoF genes have been replaced by genes encoding proteins of unknown function. Thus, it is not certain whether the NDH-1 complexes in these two species are functional (Kelly, 2005). However, the A. butzleri and C. fetus subsp. fetus NDH-
87
1 complexes contain NuoE and NuoF homologs, distinguishing these two complexes from the NDH-1 complexes of the other NJC genomes. The A. butzleri and C. fetus subsp. fetus Nu0 complexes are also distinct as a result of the fusion of the nuoC and nuoD CDSs, supporting further their similarity. Curiously, the A. butzleri RM4018 nuo complex has a second copy of gltA, which encodes citrate synthase, located between the nuoF and nuoG genes (Miller et al., 2 0 0 7 ~ ) .It is unknown at present whether this gltA encodes a functional citrate synthase, although the high similarity between the two enzymes would suggest that it is functional. It is also not known why this second gltA copy is located in the nuo cluster. The presence of anaerobic electron acceptors, e.g., fumarate reductase, would indicate that campylobacters such as C. jejuni subsp. jejuni should grow anaerobically. However, C. jejuni subsp. jejuni is incapable of anaerobic growth (Kelly, 2005). Sellars et al. (2002) proposed that the anaerobic growth of C. jejuni subsp. jejuni was restricted as a result of the presence of oxygen-requiring metabolic reactions in this organism, specifically reduction of ribonucleotides by the class I ribonucleoside-diphosphate reductase (RNR) encoded by the genes nrdA and nrdB. Class I enzymes contain a tyrosyl radical that is generated only in the presence of oxygen. However, the genomes of C. fetus subsp. fetus 82-40 and C. curvus 525.92 (as well as A. butzleri Rh44018) are predicted to encode the NrdD anaerobic-type class I11 RNR and anaerobic reductase activator protein NrdG, in addition to the NrdAB class I RNR. All three class I11 RNR-containing strains can grow anaerobically, supporting this hypothesis. Genomic Reduction In addition to the TCA cycle and respiratory genes absent from the C. lari RM2100 genome, several other genes common to Campylobacter are absent also from this genome (W. G. Miller et al., unpublished data). Many of these genes encode proteins involved in amino acid biosynthesis, including argCOBD, gltB, gltD, proA, proB, trpAFDE, trpC, metEF, metAB, and leuABCD. In addition to these amino acid biosynthetic genes, genes related to the biosynthesis of pantothenate, thiamine, tetrahydrofolate, and biotin are absent. Thus, strain RM2100 would be predicted to be multiply auxotrophic. In some cases, only particular branches or sections of biosynthetic pathways are missing. For example, the aromatic amino acid pathway from erythrose4-phopshate to chorismate is present in strain RM2100, as are the branches from chorismate to both phenylalanine and tyrosine; however, the path-
88
MILLER
way leading to synthesis of tryptophan is missing. Also, strain RM2100 is predicted to encode ArgG and ArgH; thus, theoretically, RM2100 could synthesize arginine if provided with an exogenous source of citrulline. Finally, genes encoding the hemin and enterochelin transporters ChuABCD and CeuBCDE are missing from strain RM2100, as is the gene encoding the autoinducer-2 production protein, LuxS. The mechanism leading to this genomic reduction in C. lari RM2100 is unknown. C. lari genes described in this and previous sections have been deleted and are not present within the RM2100 genome as fragmented or frameshifted pseudogenes. In fact, loss of these genes would require 21 separate deletions. C. jejuni subsp. doylei 269.97 is also missing the ceuBCDE gene cluster. However, in this strain, loss of the ceu cluster is linked to insertion of the DIE2 genomic island. With the exception of the prophage CLIE1, C. lari RM2100 contains no genomic islands or mobile elements. Thus, genomic rearrangements modulated by these elements are not the cause of the C. Zari genomic reduction. About half of the deletions do occur at syntenic breakpoints, suggesting that gene loss in strain RM2100 may be due to genome rearrangement. However, the other half of the deletions occur within localized syntenic regions. Notably, one such region in C. jejuni subsp. jejuni contains (in order) the oor gene cluster, sucCD, mdh, icd, and Cj0530. In C. Zari RM2100, sucCD and icd have been deleted as a result of two separate and presumably independent events. These multiple deletions within the C. lari RM2100 genome would be predicted to have a profound effect on metabolism, colonization, and the overall lifestyle of this strain. However, strain RM2100 is a human clinical isolate; therefore, it is unlikely that these deletions have a serious detrimental effect on pathogenicity in humans. The absence of gZtA from both C. lari and C. insulaenigrae might suggest that many of the genes absent from strain RM2100 are absent also from C. insulaenigrae. Absence of this gene from both species may imply a common ancestor and/or may reflect the similar environmental niches that both species occupy: C. lari in shorebirds and shellfish, and C. insulaenigrae in marine mammals (Stoddard et al., 2007). It is possible that genomic reduction in C. Zari is confined to strain RM2100. Additional investigations will be necessary to determine the extent of genome reduction in C. Zari and perhaps in the related species C. insuZaenigrae. An intriguing insight into genomic reduction in Campylobacter relates to motility in C. hominis. Preliminary analysis of the Campylobacter hominis ATCC BAA-381 genome indicates that this strain
does not encode any genes involved in motility: flagellar genes, flagellar-related genes (e.g., pflA), chemotaxis genes, or genes encoding methyl-accepting chemotaxis (MCP domain) proteins. Similar to the genes absent in C. Zari, these motility-related loci are present in the C. jejuni subsp. jejuni NCTC 11168 genome in at least 30 unlinked chromosomal locations. Therefore, from the C. jejuni subsp. jejuni perspective, multiple independent deletion events would have had to occur to result in the loss of motility genes in C. hominis. However, another explanation is likely. The motility genes in the deep-sea-vent Epsilonproteobacterial taxon Nitratiruptor are heavily clustered, with one cluster containing over 40 chemotaxis and flagellar genes. Additionally, in A. butxZeri RM4018, a flagellar cluster of 23 genes is present in the genome. It is interesting to note that the more evolutionarily “distant” the Epsilonproteobacterial taxon, the more the motility genes are clustered. Thus, it would be much more probable for loss of the multiple motility genes to result from deletion of one or a few clusters rather than 30 independent deletion events. A similar mechanism of clustered loci may explain the genomic reduction in C. Zari. Clustering of such related loci may also be indicative of an origin via lateral gene transfer (Nakagawa et al., 2007). Finally, an obvious consequence of loss of motility in C. horninis would be reduced virulence, perhaps contributing to the putative commensal nature of this organism.
RESTRICTION/MODIFICATION SYSTEMS Compared with H. pylori, Campylobacter species contain relatively few R-M systems. With the exception of C. upsaliensis RM3195, the NJC strains contain at most five R-M systems: generally one Type I system encoded by the hsd locus and a few Type I1 or I11 R-M systems, and a small number of DNA methyltransferases. In addition, like C. jejuni subsp. jejuni and H. pylori, several NJC R-M genes contain homopolymeric G: C tracts and are therefore putative contingency genes. Type I (hsd) R-M Systems The Type I R-M enzyme is a multisubunit complex composed of HsdR (endonuclease), HsdM (methyitransferase), and HsdS (specificity subunit) subunits in the stoichiometry R,S,M,. On the basis of several factors, including DNA sequence similarity and immunological cross-reactivity, Type I R-M systems have been divided into four families: IA, IB, IC, and ID (Titheradge et al., 2001; Fig. 3). HsdM pro-
CHAPTER 5
A
COMPARATIVE GENOMICS OF NON-C. TETUNI CAMPYLOBACTER SPP.
89
W. succinogenesHsdM S. denitrificans HsdM H. pylon’HP1403 ” ‘ [ H . pyloriJHP1423
H.acinonychis HsdM H. py/oriJHP0415 H.pylori HP0463 H.pvlori HPAGI 0438
I
looL
C, upsaliensis HsdMl Ciistrain 81-176 HsdMl
Figure 3. Phylogenetic tree depicting the similarities among HsdM polypeptides from Cumpylobucterules and from selected bacterial taxa. Bootstrap values (>75%) are based on 500 replicates and are indicated as percentages. Cumpylobucter HsdM proteins are bold. Representatives of the IA (IA),IB (IB), IC (IC) and ID (ID) families are included for comparison. Type I families, identified previously (A, B, AB, C, D, and F), are shaded in gray. See Table 1, footnote a, for abbreviations.
teins within a family are well conserved, with HsdM proteins from different families having low amino acid sequence identity. An analysis of 73 C. jejuni subsp. jejuni strains demonstrated that the C. jejuni subsp. jejuni Type I systems are members of at least three families: IC and the novel “IAB” and “IF” families (Miller et al., 2005b). Extension of this analysis to the NJC strains (Fig. 3) emphasizes the familial organization of the Campylobacter Type I R-M systems. C. fetus subsp. fetus 82-40 HsdM is 80% identical to S. denitrificans ATCC 33889 HsdM but only 21 and 33% identical to the HsdM proteins of C. curvus 525.92 and C. concisus 13826, respectively. Similarly, C. curvus HsdM (ccIB’y)demonstrates greater homology to Streptococcus pneumoniae R6 HsdM (“IB”; 61.2%) than to the HsdM protein from the related C. concisus (“IC”; 20.6%). On the basis of the analysis presented in Fig. 3, at least five additional Type I families would be predicted to exist within the Epsilonproteobacteria.
Like H. pylori, C. upsaliensis RM3195 is predicted to contain three hsd loci: locus 1 is a member of a novel family that includes the C. jejuni subsp. jejuni 81-176 locus cjj812 76-0776-0780, and loci 2 and 3 are members of the IB and IC families, respectively. The hsdM gene of locus 3 contains a 15-bp G:C tract; therefore, hsdM3 may be a contingency gene. Homopolymeric G:C tracts have been identified within IC hsd genes. The IC hsdS genes of C. jejuni RM1516 and 30 also contain G:C tracts, and the homopolymeric tract of RM1516 hsdS was demonstrated to be hypervariable (Miller et al., 2005b). Similar to the C. upsaliensis RM3195 hsdM3 gene, other NJC hsd loci contain putative contingency geneslpseudogenes. The hsd locus of C. lari RM2100 contains a frameshift within hsdM. Campylobacter lari hsdM does not contain a homopolymeric G:C tract but does contain a homopolymeric A tract adjacent to the frameshift, suggesting that C. lari hsdM may be a contingency gene. Additionally,
90
MILLER
both hsd loci of C. jejuni subsp. doylei 269.97 contain potential defects: the Type IC locus hsdS gene contains a homopolymeric G:C tract, whereas the Type IAB locus hsdR and hsdM genes are degenerate.
mosomal methylation patterns, making the chromosome remarkably resistant to restriction. This is probably the explanation for the observation that many C. upsaliensis strains are refractory to analysis by pulsed-field gel electrophoresis (data not shown).
Type II/III R-M Systems Excluding C. upsaliensis RM3195, the NJC genomes encode few Type I1 or Type I11 restriction enzymes, ranging from one putative Type I1 pseudogene in C. curuus to three restriction enzymes in C. jejuni subsp. doylei 269.97. One Type I1 restriction enzyme, the C. jejuni subsp. jejuni enzyme CjeI, is encoded by C. coli, C. upsaliensis, C. fetus subsp. fetus, and C. curvus. However, the C. upsaliensis and C. curuus homologs are fragmented and presumably nonfunctional, and it is not known whether the other two homologs recognize the same target sequence as the C. jejuni subsp. jejuni enzyme. The Type 11s enzyme encoded by Cj0031/32 is also found in C. jejuni subsp. doylei, C. coli, and C. lari. As in C. jejuni subsp. jejuni NCTC 11168, the C. coli Cj0032/32 homolog is a contingency gene; however, the C. jejuni subsp. doylei and C. lari homologs do not contain poly G:C tracts. Eight Type I1 and Type I11 restriction enzymes are predicted to be encoded by the C. upsaliensis RM3195 genome (Fouts et al., 2005; data not shown). However, the genes encoding two Type I1 restriction enzymes (including CjeI) are fragmented and may be nonfunctional. Three of the eight R-M systems are homologous to R-M systems from H. pylori, and one Type I1 system is similar to the Haemophilus influenzae Hin4II system (Azarinskas et al., 2006), although in strain RM3195, the two Hin4II genes encoding the m5C and m6A DNA methyltransferases have been fused into one open reading frame.
C . upsaliensis Methyltransferases In addition to the Type I, 11, and I11 methyltransferases, the C. upsaliensis RM3 195 genome also contains at least 15 additional genes that encode adenineand cytosine-specific DNA methyltransferases (Fouts et al., 2005). The majority (9 of 15) of these methyltransferases have homologs within H. pylori, whereas 5 of the remaining methylases are similar to proteins from other taxa: e.g., Mod-7 is 40% identical to a methyltransferase from Thermoplasma volcanium, and Mod-17 is 51% identical to a methyltransferase from Bacillus stearothermophilus. It is unclear what biological role these multiple methylases play in C. upsaliensis. Undoubtedly, the combined action of the suite of methyltransferases within C. upsaliensis would lead to dense, overlapping chro-
VIRULENCE/PATHOGENICITY LOCI Cytolethal Distending Toxin One of the major proposed virulence factors in Campylobacter is the cytolethal distending toxin (CDT), an exotoxin encoded by the cdtA, cdtB, and cdtC genes. CDT has been demonstrated to irreversibly block eukaryotic cells in the G2 phase of the cell cycle (Whitehouse et al., 1998). The genomes of the NJC strains C. coli RM2228, C. lari RM2100, C. upsaliensis RM3195, and C. fetus subsp. fetus 82-40 all contain the complete cdt operon, consistent with previous studies investigating CDT production in these species (Asakura et al., 2007; Bang et al., 2001; Mooney et al., 2001; Ohya et al., 1993). On the basis of genomic analyses, C. concisus 13826, C. curvus 525.92, and C. jejuni subsp. doylei 269.97 are not predicted to encode CDT. A CDT-like toxin was observed in some strains of C. concisus (Engberg et al., 2005); therefore, either CDT is encoded only by a subset of C. concisus strains, or the CDT-like toxin is encoded by an as yet unidentified genomic determinant. The cdt operon in C. jejuni subsp. doylei 269.97 is represented only by a remnant of the cdtA gene. Deletions within the cdt operon were observed in all eight C. jejuni subsp. doylei strains characterized by Parker et al. (2007), suggesting that absence of CDT is characteristic of the C. jejuni subsp. doylei subspecies. Although the C. coli, C. lari, C. upsaliensis, and C. fetus subsp. fetus genomes encode CDT, the CDT subunits demonstrate a substantial amount of diversity when compared with the average amino acid identities between the species. For example, the average amino acid identity between C. jejuni subsp. jejuni and C. coli proteins is 87.7% (Fig. l),yet the amino acid identities between the C. jejuni subsp. jejuni and C. coli CdtA, CdtB, and CdtC subunits are 47.4%, 67.2%, and 49.4%, respectively. Similar levels of diversity were reported by Asakura et al. (2007), who reported CdtABC identities of 47.6Y0, 67.3%, and 47.9%. Differences between these two sets of numbers probably reflect strain variation in cdt sequence. In some instances, variation between the subunits can lead to changes in activity. For example, Asakura et al. (2007) demonstrated that CDTs from different strains within a given species possessed vari-
COMPARATIVE GENOMICS OF NON-C. TETUNI CAMPYLOBACTER SPP.
CHAPTER 5
able levels of activity. In addition, CDTs from C. coli were generally inactive. Asakura et al. (2007) cloned and sequenced the cdt loci of C. coli and C. fetus; the positions of the cdt operons reported correspond to the positions of cdt operons in the C. coli RM2228 and C. fetus subsp. fetus 82-40 genomes. However, a second cdt cluster is present in the C. fetus subsp. fetus 82-40 genome. This locus (CFF8240-0951-0957) is unusual in that it contains five additional cdt genes in the order cdtC2-cdtA2-cdtC3-cdtB2-cdtA3;a putative frameshifted cdtB gene is present downstream of the first cdtC gene in the cluster. Thus, C. fetus subsp. fetus 82-40 contains three cdtA genes, at least two cdtB genes, and three cdtC genes. Even more surprising is the fact that very little similarity exists between the various C. fetus subsp. fetus CDT subunits; the greatest amount of sequence identity is between CdtB1 and CdtB2 (55.9%; Fig. 4). It is not known if the C. fetus subsp. fetus CDT subunits independently assort, i.e., whether CdtB1 can complex with any of the CdtA or CdtC subunits. If so, then C. fetus subsp.
fetus can potentially encode 18 unique exotoxins. It is likely that differences in CDT activity reflect changes in critical amino acids within the CDT subunits (Asakura et al., 2007). Therefore, it is possible that the C. fetus subsp. fetus CDTs, as a whole, possess a range of activities. It is not known whether other C. fetus subsp. fetus strains also possess multiple cdt operons. Other Virulence Factors The remainder of the virulence factors, identified initially in C. jejuni subsp. jejuni, are found in different subsets of the NJC genomes. For example, the fibronectin adhesin CadF is predicted to be encoded by all of the NJC genomes, as is the Campylobacter invasion antigen, CiaB, and the putative virulence determinant MviN. However, the JlpA lipoprotein is encoded only by C. coli RM2228 and C. upsaliensis RM3195 and not by the C. lari, C. fetus subsp. fetus, or C. cuwus genomes. Similarly, the PEBla adhesin subunit is not encoded by C. lari RM2100 or by C. curvus 525.92. Curiously, CiaB
A Cjlll68-CdtA Cjl221-CdtA Cc2228-CdtA Cl2100-CdtA Cu3195-CdtA Cff-CdtAl
C ff-CdtA2 Cff-CdtA3 Hhep-CdtA
Cjlll68-Cdtl Cjl221-CdtB Cc2228-CdtB C12100-CdtB Cu3195-CdtB Cff-CdtBl Cff-CdtB2 Hhep-CdtB
~
~
I
67.2 66.5 71.7 59.3
59.3
60.6
I
~
1
66.5
91
t
58.5
Figure 4. Pairwise comparisons of the Cumpylobucterules cytolethal distending toxin subunits. Pairwise comparisons of the CdtA (A; lower triangle), CdtB (B) and CdtC (A; upper triangle) subunits were performed by BLASTP. Values represent amino acid identities.
92
MILLER
and MviN are also encoded by the Nautiliales taxon Caminibacter and by the deep-sea-vent Epsilonproteobacterial taxa Sulfurovum and Nitratiruptor (Nakagawa et al., 2007). The presence of putative virulence determinants, like CiaB, in nonpathogenic taxa leads to several conclusions: (i) the deep-rooted Epsilonproteobacterial taxa (e.g., Nitratiruptor) are pathogenic, (ii) the function of CiaB in these taxa is not the same as the function in C. jejuni subsp. jejuni, (iii) critical amino acid differences between the CiaB proteins result in decreased or nonexistent activity in the nonpathogenic taxa, and (iv) CiaB may not be a true virulence determinant. It is highly unlikely that organisms such as Nitratiruptor are pathogenic; thus, it is probably safe to exclude the first option. Also, there is evidence indicating that CiaB is a virulence determinant (Raphael et al., 2005). This leaves options (ii) and (iii). It is entirely possible that the biological role of CiaB has evolved over time. Thus, the function that this protein had in an ancestral Epsilonproteobacterial organism may not resemble the function that it currently possesses in Campylobacter. The scenario in which the function of CiaB has evolved into its current state is more likely than the scenario in which the CiaB virulence determinant in taxa such as Caminibacter became nonfunctional as a result of the accumulation of point mutations (option iii). A fifth option is that the biological function of CiaB is identical in all taxa, but that the virulence phenotype only manifests in Campylobacter inside the human host. Additional experimental data will be necessary to define the role of these virulence determinants in the NJC genomes. Very few additional virulence determinants were observed after analysis of the NJC genomes. Two such determinants were identified in the genomes of C. fetus subsp. fetus 82-40 and C. jejuni subsp. doylei 269.97. C. fetus subsp. fetus 82-40 contains a putative filamentous hemagglutinin of the HecA family. The putative hecA gene in this strain is linked to the associated hecB gene. A hecAlhecB homolog is also in the genome of A. butzleri RM4018 (Miller et al., 2 0 0 7 ~ )Campylobacter . coli RM2228 also contains a hecA CDS, but this sequence is heavily fragmented and is not predicted to encode a presumably functional hemagglutinin; however, strain RM2228 does encode a presumably functional HecB protein. The genome of C. jejuni subsp. doylei 269.97 is predicted to encode an extracellular serine protease of the IgAl family. IgAl protease family proteins have been also demonstrated to cleave human coagulation factor V (Brunder et al., 1997), hemoglobin (Otto et al., 1998), human chorionic gonadotrophin hormone, granulocyte-macrophage colony stimulating factor,
the CD8 surface antigen of cytotoxic T lymphocytes, and lysosomal glycoproteins (Mistry and Stockley, 2006). It may not be coincidental that the two strains that contain hemagglutinid blood-factor-related virulence determinants are members of species isolated often from human blood cultures. Additional experiments will be necessary to determine whether these virulence factors are functional in vivo.
CONCLUSIONS Genomic data for the NJC species are incomplete: at the time of this writing, the annotation for many of these genomes ( e g , C. upsaliensis RM3195) has not been finalized. Therefore, many novel and interesting CDSs may still be hidden within the NJC genomes. Also, it is likely that additional Campylobacter genomes will soon be sequenced, further adding to the genomic data available. Analysis of the Campylobacter genomes is also enhanced by the existence of other Epsilonproteobacterial genomes; several NJC CDSs have homologs not within Campylobacter but in related taxa, such as A. butzleri, W. succinogenes, and H. pylori. Thus, knowledge of genomes such as those of Arcobacter sulfidicus, Helicobacter mustelae, and Caminibacter-all three nearing completion-will also prove beneficial. Nevertheless, despite the incomplete nature of the NJC genomes, these genomes have greatly expanded our knowledge of Campylobacter biology. Many of the features described in this chapter were identified after in silico analysis of the NJC genomes: thus, experimental evidence for many of these features is absent. Therefore, many of the pathways and systems described and hypotheses proposed will need to be supported by future investigations. However, one benefit of multiple, related sequenced genomes is the fact that CDSs are often conserved across multiple Campylobacter taxa. For example, 33% of the core Campylobacter proteins are conserved hypothetical proteins, or proteins assigned only a general function. Thus, experimental data related to one of these proteins are likely to apply to its homologs within the core protein set, potentially increasing the experimental database for multiple CDSs en bloc. Comparative analysis of the NJC genomes has also identified novel mobile elements, genomic islands, and prophage. Additionally, it has provided evidence of lateral gene transfer between Cumpylobacter species. Comparative genome analysis can therefore be utilized in investigations related to genome structure and evolution of the genus. Major contributors to strain variation in Campylobacter are the genomic islands and integrated elements. Al-
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. lElUNI CAMPYLOBACTER SPP.
though the current biological role of these genomic islands appears to be the synthesis of a plethora of conserved hypothetical proteins, as new experimental data are obtained, it is likely that the functions of the proteins encoded by these elements will be ascertained. Analysis of the genomic islands should also include a more thorough characterization of the family of plasmids and prophage that exist within Campylobacter; this is because the proposed origin of genomic islands, and for that matter CRISPR elements, is extrachromosomal. Finally, genomic data for the NJC genomes will lead to new typing methods and perhaps culture methods. Indeed, completion of the NJC C. coZi, C. Zuri, and C. upsuliensis genomes resulted in novel MLST methods for these three species (Miller et al., 2005a) and other sequence-based typing methods (Klena et al., 2004). Similar sequence data also led to novel MLST methods for Arcobacter spp. (W. G. Miller et al., unpublished data). Many of the NJC genomes are difficult to culture and are often overlooked in clinical samples, presumably as a result of differences in atmospheric, growth temperature, and media requirements or as a result of variable resistance to commonly used antibiotics. Comparative genomic analysis may improve culturing and detection techniques and provide a more accurate picture of the role that the NJC genomes play in human illnesses. REFERENCES Adams, V., D. Lyras, K. A. Farrow, and J. I. Rood. 2002. The clostridial mobilisable transposons. Cell. Mol. Life Sci. 59:20332043. A h , R. A., L. S. Ling, D. T. Moir, B. L. King, E. D. Brown, P. C. Doig, D. R. Smith, B. Noonan, B. C. Guild, B. L. deJonge, G. Carmel, P. J. Tummino, A. Caruso, M. Uria-Nickelsen, D. M. Mills, C. Ives, R. Gibson, D. Merberg, S. D. Mills, Q. Jiang, D. E. Taylor, G. F. Vovis, and T. J. Trust. 1999. Genomicsequence comparison of two unrelated isolates of the human gastric pathogen Helicobacter pylori. Nature 397:176-180. Asakura, M., W. Samosornsuk, M. Taguchi, K. Kobayashi, N. Misawa, M. Kusumoto, K. Nishimura, A. Matsuhisa, and S. Yamasaki. 2007. Comparative analysis of cytolethal distending toxin (cdt) genes among Campylobacter jejuni, C. coli and C. fetus strains. Microb. Pathog. 42:174-183. Azarinskas, A., 2. Maneliene, and A. Jakubauskas. 2006. Hin411, a new prototype restriction endonuclease from Haemophilus influenzae RFL4: discovery, cloning and expression in Escherichia coli. 1.Biotechnol. 123:288-296. Baar, C., M. Eppinger, G. Raddatz, J. Simon, C. Lanz, 0. Klimmek, R. Nandakumar, R. Gross, A. Rosinus, H. Keller, P. Jagtap, B. Linke, F. Meyer, H. Lederer, and s. C. Schuster. 2003. Complete genome sequence and analysis of Wolinella succinogenes. Proc. Natl. Acad. Sci. USA 100:11690-11695. Bang, D. D., F. Scheutz, P. Ahrens, K. Pedersen, J. Blom, and M. Madsen. 2001. Prevalence of cytolethal distending toxin (cdt) genes and CDT production in Campylobacter spp. isolated from Danish broilers. J. Med. Microbiol. 50:1087-1094.
93
Barrangou, R., C. Fremaux, H. Deveau, M. Richards, P. Boyaval, S. Moineau, D. A. Romero, and P. Horvath. 2007. CRISPR provides acquired resistance against viruses in prokaryotes. Science 315:1709-1712. Batchelor, R. A., B. M. Pearson, L. M. Friis, P. Guerry, and J. M. Wells. 2004. Nucleotide sequences and comparison of two large conjugative plasmids from different Campylobacter species. Microbiology 150:3507-3517. Bolotin, A., B. Quinquis, A. Sorokin, and S. D. Ehrlich. 2005. Clustered regularly interspaced short palindrome repeats (CRISPRs) have spacers of extrachromosomal origin. Microbiology 151:2551-2561. Boosinger, T. R., W. T. Blevins, J. V. Heron, and J. L. Sunter. 1990. Plasmid profiles of six species of Campylobacter from human beings, swine, and sheep. Am. J. Vet. Res. 51:718-722. Brichta, D. M., K. N. Azad, P. Ralli, and G. A. ODonovan. 2004. Pseudomonas aemginosa dihydroorotases: a tale of three pyrCs. Arch. Microbiol. 182:7-17. Brunder, W., H. Schmidt, and H. Karch. 1997. EspP, a novel extracellular serine protease of enterohaemorrhagic Escherichia coli 0157:H7 cleaves human coagulation factor V. Mol. Microbiol. 24:767-778. Chattoraj, D. K. 2000. Control of plasmid DNA replication by iterons: no longer paradoxical. Mol. Microbiol. 37:467-476. Daucher, J. A., and N. R. Kreig. 1995. Pyruvate-ferredoxin oxidoreductase in Campylobacter species. Can. J. Microbiol. 41: 198-201. Dietrich, W., and 0. Klimmek. 2002. The function of methylmenaquinone-6 and polysulfide reductase membrane anchor (PsrC) in polysulfide respiration of Wolinella succinogenes. Eur. J. Biochem. 269:1086-1095. Engberg, J., D. D. Bang, R. Aabenhus, F. M. Aarestrup, V. Fussing, and P. Gerner-Smidt. 2005. Campylobacter concisus: an evaluation of certain phenotypic and genotypic characteristics. Clin. Microbiol. Infect. 11:288-295. Eppinger, M., C. Baar, B. Linz, G. Raddatz, C. Lanz, H. Keller, G. Morelli, H. Gressmann, M. Achtman, and S. C. Schuster. 2006. Who ate whom? Adaptive Helicobacter genomic changes that accompanied a host jump from early humans to large felines. PLoS Genet. 2:e120. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. Deboy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Friedman, C. R., J. Neimann, H. C. Wegener, and R. V. Tauxe. 2000. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations., p. 121-138. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter. ASM Press, Washington, DC. Gilbert, M., P. Godschalk, C. T. Parker, H. Endtz, and W. Wakarchuk. 2005. Genetic basis for the variation in the lipooligosaccharide outer core of Campylobacter jejuni and possible association of glycosyltransferase genes with post-infectious neuropathies, p. 219-248. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Guerry, P., C. M. Szymanski, M. M. Prendergast, T. E. Hickey, C. P. Ewing, D. L. Pattarini, and A. P. Moran. 2002. Phase variation of Campylobacter jejuni 8 1-176 lipooligosaccharide affects ganglioside mimicry and invasiveness in vitro. Infect. Immun. 70:787-793.
94
MILLER
Haft, D. H., J. Selengut, E. F. Mongodin, and K. E. Nelson. 2005. A guild of 45 CRISPR-associated (Cas) protein families and multiple CRISPR/ Cas subtypes exist in prokaryotic genomes. PLoS Comput. Biol. 1:e60. Hefford, M. A., Y. Kobayashi, S . E. Allard, R. J. Forster, and R M. Teather. 1997. Sequence analysis and characterization of pOM1, a small cryptic plasmid from Butyrivibrio fibrisolvens, and its use in construction of a new family of cloning vectors for Butyrivibrios. Appl. Environ. Microbiol. 63:1701-1711. Heo, J., M. T. Wolfe, C. R. Staples, and P. W. Ludden. 2002. Converting the NiFeS carbon monoxide dehydrogenase to a hydrogenase and a hydroxylamine reductase. 1. Bacteriol. 184: 5894-5897. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Iyer, R. R., A. Pluciennik, V. Burdett, and P. L. Modrich. 2006. DNA mismatch repair: functions and mechanisms. Chem. Rev. 106:302-323. Karlyshev, A. V., 0. L. Champion, C. Churcher, J. R. Brisson, H. C. Jarrell, M. Gilbert, D. Brochu, F. St Michael, J. Li, W. W. Wakarchuk, I. Goodhead, M. Sanders, K. Stevens, B. White, J. Parkhill, B. W. Wren, and C. M. Szymanski. 2005. Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol. Microbiol. 55:90-103. Kelly, D. J. 2005. Metabolism, electron transport and bioenergetics of Campylobacter jejuni: implications for understanding life in the gut and survival in the environment, p. 275-292. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Press, Norfolk, United Kingdom. Kersulyte, D., N. S . Akopyants, S . W. Clifton, B. A. Roe, and D. E. Berg. 1998. Novel sequence organization and insertion specificity of IS605 and IS606: chimaeric transposable elements of Helicobacter pylori. Gene 223:175-186. Kienesberger, S., G. Gorkiewicz, M. M. Joainig, S . R. Scheicher, E. Leitner, and E. L. Zechner. 2007. Development of experimental genetic tools for Campylobacter fetus. Appl. Environ. MiCrobiOl. 73:4619-4630. Kleanthous, H., C. L. Clayton, and S . Tabaqchali. 1991. Characterization of a plasmid from Helicobacter pylori encoding a replication protein common to plasmids in gram-positive bacteria. Mol. Microbiol. 5:2377-2389. Klena, J. D., C. T. Parker, K. Knibb, J. C. Ibbitt, P. M. Devane, S. T. Horn, W. G. Miller, and M. E. Konkel. 2004. Differentiation of Campylobacter coli, Campylobacter jejuni, Campylobacter lari, and Campylobacter upsaliensis by a multiplex PCR developed from the nucleotide sequence of the lipid A gene IpxA. 1.Clin. Microbiol. 425549-5557. Krafft, T., R. Gross, and A. Kroger. 1995. The function of Wolinella succinogenes psr genes in electron transport with polysulphide as the terminal electron acceptor. Eur. J. Biochem. 230: 601-606. Kunin, V., R. Sorek, and P. Hugenholtz. 2007. Evolutionary conservation of sequence and secondary structures in CRISPR repeats. Genome Biol. 8:R61. Linton, D., M. Gilbert, P. G. Hitchen, A. Dell, H. R Morris, W. W. Wakarchuk, N. A. Gregson, and B. W. Wren. 2000. Phase variation of a beta-1,3 galactosyltransferase involved in generation of the ganglioside GM1-like lipo-oligosaccharide of Campylobacter jejuni. Mol. Microbiol. 37501-5 14. Mandrell, R. E., and W. G. Miller. 2006. Campylobacter, p. 476521. In Y. Motarjemi and M. Adams (ed.), Emerging Foodborne Pathogens. Woodhead Publishing Ltd., Cambridge, England.
Marais, A., G. L. Mendz, S. L. Hazel], and M. G. F. 1999. Metabolism and genetics of Helicobacter pylori: the genome era. Microbiol. Mol. Biol. Rev. 63 :642-674. McPhail, D., and M. Shepherdson. 2006. The aspartate transcarbamoylase-dihydroorotase complex in Deinococcus radiophiltrs has an active dihydroorotase. Arch. Microbiol. 185:78-8 1. Mendz, G. L., S . L. H a d , and B. P. Burns. 1993. Glucose utilization and lactate production by Helicobacter pylori.]. Gen. Microbiol. 139:3023-3028. Miller, W. G., A. H. Bates, S . T. Horn, M. T. Brandl, M. R. Wachtel, and R. E. Mandrell. 2000. Detection on surfaces and in Caco-2 cells of Campylobacter jejuni cells transformed with new gfp, yfp, and cfp marker plasmids. Appl. Environ. Microbiol. 665426-5436. Miller, W. G., S. Heath, and R. E. Mandrell. 2007a. Cryptic plasmids isolated from Campylobacter strains represent multiple, novel incompatibility groups. Plasmid 57:108-117. Miller, W. G., and R. E. Mandrell. 2005. Prevalence of Campylobacter in the food and water supply: incidence, outbreaks, isolation and detection., p. 101-163. In M. E. Konkel and J. M. Ketley (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Scientific Press, Norfolk, United Kingdom. Miller, W. G., S . L. On, G. Wang, S. Fontanoz, A. J. Lastovica, and R. E. Mandrell. 2005a. Extended multilocus sequence typing system for Campylobacter coli, C. lari, C. upsaliensis, and C. helveticus.]. Clin. Microbiol. 43:2315-2329. Miller, W. G., C. T. Parker, S. Heath, and A. J. Lastovica. 2007b. Identification of genomic differences between Campylobacter jejuni subsp. jejuni and C. jejuni subsp. doylei at the nap locus leads to the development of a C. jejuni subspeciation multiplex PCR method. BMC Microbiol. 7:11. Miller, W. G., C. T. Parker, M. Rubenfield, G. L. Mendz, M. M. S. M. Wosten, D. W. Ussery, J. F. Stolz, T. T. Binnewies, P. F. Hallin, G. Wang, J. A. Malek, A. Rogosin, L. H. Stamler, and R E. Mandrell. 2007c. The complete sequence and analysis of the Epsilonproteobacterium. Arcobacter butzleri. PLoS One 2: e1358. Miller, W. G., B. M. Pearson, J. M. Wells, C. T. Parker, V. V. Kapitonov, and R. E. Mandrell. 2005b. Diversity within the Campylobacter jejuni type I restriction-modification loci. Microbiology 151:337-351. Mistry, D., and R. A. Stockley. 2006. IgAl protease. Int. 1. Biochem. Cell Biol. 38: 1244-1248. Mooney, A., M. Clyne, T. Curran, D. Doherty, B. Kilmartin, and B. Bourke. 2001. Campylobacter upsaliensis exerts a cytolethal distending toxin effect on HeLa cells and T lymphocytes. Microbiology 147:735-743. Nakagawa, S., Y. Takaki, S . Shimamura, A. L. Reysenbach, K. Takai, and K. Horikoshi. 2007. Deep-sea vent E-proteobacterial genomes provide insights into emergence of pathogens. Proc. Natl. Acad. Sci. USA 104:12146-12150. Ohya, T., K. Tominaga, and M. Nakazawa. 1993. Production of cytolethal distending toxin (CLDT) by Campylobacter fetus subsp. fetus isolated from calves. J. Vet. Med. Sci. 55507-509. Otto, B. R., S . J. van Dooren, J. H. Nuijens, J. Luirink, and B. Oudega. 1998. Characterization of a hemoglobin protease secreted by the pathogenic Escherichia coli strain EB1.J. Exp. Med. 188:1091-1103. Owen, R. J., and J. Hernandez. 1990. Occurrence of plasmids in “Campylobacter upsaliensis” (catalase negative or weak group) from geographically diverse patients with gastroenteritis or bacteraemia. Eur. J. Epidemiol. 6:111-117. Parker, C. T., W. G. Miller, S . T. Horn, and A. J. Lastovica. 2007. Common genomic features of Campylobacter jejuni subsp. doylei strains distinguish them from C. jejuni subsp. jejuni. BMC Microbiol. 7 5 0 .
CHAPTER 5
COMPARATIVE GENOMICS OF NON-C. 7E7UNI CAMPYLOBACTER SPP.
Parker, C. T., B. Quinones, W. G. Miller, S. T. Horn, and R. E. Mandrell. 2006. Comparative genomic analysis of Campylobacter jejuni strains reveals diversity due to genomic elements similar to those present in C. jejuni strain RM1221. J. Clin. Microbiol. 44:4125-4135. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Raphael, B. H., M. R. Monteville, J. D. Klena, L. A. Joens, and M. E. Konkel. 2005. Interactions of Campylobacter jejuni with non-professional phagocytic cells., p. 397-413. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Press, Norfolk, United Kingdom. Schouls, L. M., S . Reulen, B. Duim, J. A. Wagenaar, R J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41:15-26. Schurr, M. J., J. F. Vickrey, A. P. Kumar, A. L. Campbell, R. Cunin, R. C. Benjamin, M. S . Shanley, and G. A. O'Donovan. 1995. Aspartate transcarbamoylase genes of Pseudomonas putida: requirement for an inactive dihydroorotase for assembly into the dodecameric holoenzyme. J. Bacteriol. 177:1751-1759. Sellars, M. J., S. J. Hall, and D. J. Kelly. 2002. Growth of Campylobacter jejuni supported by respiration of fumarate, nitrate, nitrite, trimethylamine-N-oxide, or dimethyl sulfoxide requires oxygen. J. Bacteriol. 184:4 187-4 196. Skibinski, D. A., P. Golby, Y. S. Chang, F. Sargent, R. Hoffman, R. Harper, J. R. Guest, M. M. Attwood, B. C. Berks, and S. C. Andrews. 2002. Regulation of the hydrogenase-4 operon of Escherichia coli by the sigma(S'I)-dependent transcriptional activators FhlA and HyfR. J. Bacteriol. 184:6642-6653. Smith, M. A., M. Finel, V. Korolik, and G. L. Mendz. 2000. Characteristics of the aerobic respiratory chains of the microaerophiles Campylobacter jejuni and Helicobacter pylori. Arch. Microbiol. 174: 1-10. Stanley, J., C. Jones, A. Burnens, and R. J. Owen. 1994. Distinct genotypes of human and canine isolates of Campylobacter upsaliensis determined by 16s rRNA gene typing and plasmid profiling. J. Clin. Microbiol. 32:1788-1794. Steele, T. W., and R. J. Owen. 1988. Campylobacter jejuni subsp. doylei subsp. nov., a subspecies of nitrate-negative campylobacters isolated from human clinical specimens. Int. J. Syst. Bacteriol. 38:3 16-3 18. Stoddard, R. A., W. G. Miller, J. E. Foley, J. Lawrence, F. M. Gulland, P. A. Conrad, and B. A. Byrne. 2007. Campylobacter insulaenigrae Isolates from northern elephant seals (Mirounga angustirostris) in California. Appl. Environ. Microbiol. 73:17291735. Suerbaum, S., C. Josenhans, T. Sterzenhach, B. Drescher, P. Brandt, M. Bell, M. Droge, B. Fartmann, H. P. Fischer, Z. Ge, A. Horster, R. Holland, K. Klein, J. Konig, L. Macko, G. L. Mendz, G. Nyakatura, D. B. Schauer, 2. Shen, J. Weber, M.
95
Frosch, and J. G. Fox. 2003. The complete genome sequence of the carcinogenic bacterium Helicobacter hepaticus. Proc. Natl. Acad. Sci. USA 100:7901-7906. Takai, K., M. Suzuki, S. Nakagawa, M. Miyazaki, Y. Suzuki, F. Inagaki, and K. Horikoshi. 2006. Sulfurimonas paralvinellae sp. nov., a novel mesophilic, hydrogen- and sulfur-oxidizing chemolithoautotroph within the Epsilonproteobacteria isolated from a deep-sea hydrothermal vent polychaete nest, reclassification of Thiomicrospira denitrificans as Sulfurimonas denitrificans comb. nov. and emended description of the genus Sulfurimonas. Int. J. Syst. Evol. Microbiol. 56: 1725-1733. Taylor, D. E., and D. M. Tracz. 2005. Mechanisms of antimicrobial resistance in Campylobacter, p. 193-204. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Titheradge, A. J., J. King, J. Ryu, and N. E. Murray. 2001. Families of restriction enzymes: an analysis prompted by molecular and genetic data for type ID restriction and modification systems. Nucleic Acids Res. 29:4195-4205. Tomb, J. F., 0. White, A. R. Kerlavage, R. A. Clayton, G. G. Sutton, R. D. Fleischmann, K. A. Ketchum, H. P. Klenk, S. Gill, B. A. Dougherty, K. Nelson, J. Quackenbush, L. Zhou, E. F. Kirkness, S . Peterson, B. Loftus, D. Richardson, R. Dodson, H. G. Khalak, A. Glodek, K. McKenney, L. M. Fitzegerald, N. Lee, M. D. Adams, J. C. Venter, et al. 1997. The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388539-547. Trieber, C. A., and D. E. Taylor. 2000. Mechanisms of antibiotic resistance in Campylobacter, p. 441-454. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. American Society for Microbiology, Washington, DC. van Bergen, M. A., K. E. Dingle, M. C. Maiden, D. G. Newell, L. van der Graaf-Van Bloois, J. P. van Putten, and J. A. Wagenaar. 2005. Clonal nature of Campylobacter fetus as defined by multilocus sequence typing. J. Clin. Microbiol. 435888-5898. Viswanathan, M., and S. T. Lovett. 1998. Single-strand DNAspecific exonucleases in Escherichia coli. Roles in repair and mutation avoidance. Genetics 149:7-16. Viswanathan, M., and S. T. Lovett. 1999. Exonuclease X of Escherichia coli. A novel 3'-5' DNase and DnaQ superfamily member involved in DNA repair. J. Biol. Chem. 274:30094-30100. Waterman, S. R., J. Hackett, and P. A. Manning. 1993. Isolation of a restriction-less mutant and development of a shuttle vector for the genetic analysis of Campylobacter hyointestinalis. Gene 125:19-24. Whitehouse, C. A., P. B. Balbo, E. C. Pesci, D. L. Cottle, P. M. Mirabito, and C. L. Pickett. 1998. Campylobacter jejuni cytolethal distending toxin causes a G2-phase cell cycle block. Infect. Immun. 66:1934-1940. Willoughby, K., P. F. Nettleton, M. Quirie, M. A. Maley, G. Foster, M. Toszeghy, and D. G. Newell. 2005. A multiplex polymerase chain reaction to detect and differentiate Campylobacter fetus subspecies fetus and Campylobacter fetus subspecies venerealis: use on UK isolates of C. fetus and other Campylobacter spp. J. Appl. Microbiol. 99:758-766. Wolfe, M. T., J. Heo, J. S. Garavelli, and P. W. Ludden. 2002. Hydroxylamine reductase activity of the hybrid cluster protein from Escherichia coli. J. Bacteriol. 1845898-5902.
11. CLINICAL AND EPIDEMIOLOGIC ASPECTS OF CAMPYLOBACTER INFECTIONS
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 6
Clinical Aspects of Campylobacter jejuni and Campylobacter coli Infections MARTIN J. BLASERAND JORGENENGBERG
Clinical Pathology
In this chapter, we describe the clinical aspects of infection with Campylobacter jejuni (C. jejuni subsp. jejuni) and C. coli, which are the main causes of Campylobacter enteritis in humans. Our account includes extraintestinal infections that may arise either as a complication of enteritis or as an apparently spontaneous event, but not infections due to species other than C. jejuni and C. coli, which are described in chapter 7 (which includes infection due to C. jejuni subsp. doylei). Campylobacter-associated GuillainBarr6 syndrome has taken on such importance that it also merits its own chapter (chapter 13). An overview of Campylobacter-associated illness is shown diagrammatically in Fig. l.
The essential lesion in Campylobacter enteritis is an acute inflammatory enteritis, which commonly extends down the intestine to affect the colon and rectum. Terminal ileitis and cecitis with mesenteric adenitis are prominent features. Because autopsy and surgical material are rare, nearly all our knowledge of the histological changes in the gut are derived from biopsy specimens obtained at proctosigmoidoscopy. These are described below (see “Colitis”). For infection to become established, campylobacters must first survive gastric acidity to colonize the jejunum and ileum. Lowering of gastric acidity, therefore, facilitates infection, which is well established in relation to Salmonella infections. A casecontrol study indicated that therapy with the proton pump inhibitor omeprazole roughly doubles the risk of acquiring Campylobacter enteritis; this effect was not seen with H, receptor antagonists, perhaps reflecting their lesser potency (Neal et al., 1996). Colonization of the gut mucosa depends on unimpaired bacterial motility and the ability of bacteria to attach to the surface of mucosal cells. The rapid motility and spiral shape of campylobacters enable them to move easily through the viscous mucus overlying the mucosa. These and other aspects of pathogenesis are discussed in section I11 of this book.
DESCRIPTION OF DISEASE
Campylobacter enteritis is an acute diarrheal disease with clinical manifestations like those of other acute bacterial gut infections of the intestinal tract, such as salmonellosis or shigellosis (Table 1).Clinically, Campylobacter enteritis cannot be distinguished from these infections, although the presence of a prodromal period of fever without diarrhea, intense abdominal pain, or prostration would favor a diagnosis of Campylobacter enteritis. A definitive diagnosis can only be made by detecting campylobacters in the feces. There does not seem to be any clear difference between infections caused by C. jejuni and C. coli when assessed by frequency of diarrhea, blood in stool, abdominal pain, fever, vomiting, mean duration of illness, or admission to hospitals.
Immune response Antibodies to Campylobacter antigens appear in the serum from about the fifth day of illness, peak within 2 to 4 weeks, and then decline over several
Martin J. Blaser Department of Medicine, New York University School of Medicine, 550 First Ave., OBV A-606, New York, NY 10016. Jsrgen Engberg Department of Clinical Microbiology, 445 Hvidovre Hospital, Ketteglrd All6 30, DK-2650 Hvidovre, Denmark.
99
100
BLASER AND ENGBERG
Immunodeficient
Healthy
Intestinal Complications Extra-Intestinal Infections Bacteremia Metastatic seeding infections
T
fl
I
Late Onset Complications * Reactive arthritis * Guillain-Bsrrk syndrome Postinfectious irritable bowel syndrome Figure 1. Diagram illustrating an overview of illnesses due to C. jejuni and C. coli.
months (Black et al., 1988; Blaser and Duncan, 1984; Strid et al., 2001). Intestinal antibodies are also produced (Black et al., 1992). Volunteer studies show that Campylobacter infection confers short-term immunity to the homologous strain (Black et al., 1988, 1992), but it is not known for how long it lasts, nor how broad the immunity is from a single infection. Sequential infections arising 6 to 18 months apart have been recorded, invariably with a strain that is different from the original one (Kell and Ellis, 1985; M. B. Skirrow, unpublished data). However, under
Table 1. Clinical features of Campylobacter enteritis derived from surveys of community outbreaks in which 50 or more people were affected and analyzed" Symptom
No. of Outbreaksb
Fever Diarrhea Headache Abdominal pain Myalgia Vomiting" Blood in feces
25 26 21 26 5
20 7
Frequency of symptom (%) Mean
Median
Range
50 84 41 79 42 15 15
52 85 47 80 37 11 13
6-75 52-100 6-69 56-99 28-59 1-42 0.5-32
"Compiled from the following references: Fahey et al. (1995); Hoskins and Davies (1983); Itoh et al. (1983); Jones et al. (1981); Matsuaki and Katayama (1984); McNaughton et al. (1982); Melby et al. (1990); Mentzing (1981); Porter and Reid (1980); Potter et al. (1983); Rogol et al. (1983); Sacks et al. (1986); Stalder et al. (1983); Stehr-Green et al. (1991); Taylor et al. (1982); Tettmar and Thornton (1981); Vogt et al. (1982); D. A. Robinson, unpublished data. bNumbers of outbreaks from which data are obtained. "Excluding waterborne outbreaks because the exceptionally high frequency of vomiting in several waterborne outbreaks suggested the additional presence of norovirus.
natural conditions, immunity is gained from relatively few infections, despite the heterogeneity of C. jejuni strains. In developing countries where repeated infection is common in early childhood, infection rates decline with age, fewer infections are associated with diarrhea, and the duration and magnitude of convalescent excretion of campylobacters are reduced. This is paralleled by a progressive increase in specific serum immunoglobulin (Ig) A antibody (Blaser et al., 1986). In addition, in developing countries, Campylobacter enteritis is virtually absent in older children and adults. However, this evidence does not prove conclusively that immunity depends on repeated reexposure to the organism; other factors have been observed to influence immunity as well. In a seroepidemiologic study among rural children in Wisconsin, increasing age and farm residence were associated with C. jejuni seropositivity, and the clinic visit rate for diarrhea was 46% lower among farm-resident children compared with those who did not reside on farms. The findings are consistent with repeated antigenic stimulation in a farm environment with subsequent development of clinical immunity (Belongia et al., 2003). The importance of the humoral response is indicated by the problems experienced by subjects with hypogammaglobulinemia, who are likely to develop prolonged and sometimes severe infections (see below). Campylobacter-specific IgA antibody is secreted in the breast milk of immune mothers and partially protects against infection in their infants (Nachamkin et al., 1994). Considerations of the immune response are clinically relevant
CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
101
in that illness gets increasingly milder with age among children in developing countries.
to have more severe illness than those whose illness starts with diarrhea.
Infective dose
Diarrheal stage
There are many influencing factors, but in general, the infective dose of C. jejuni is often low. Infection has been induced with doses as low as 500 bacteria in experimental human infection (Black et al., 1988; Robinson, 1981). This low figure is consistent with the large size of milk-borne and waterborne outbreaks of Campylobacter enteritis, which may further lower the infective dose by buffering action or rapid washthrough of stomach contents. Higher doses lead to higher attack rates.
The onset of diarrhea betrays the intestinal nature of the infection. It is commonly profuse, watery, and bile stained, and it is sometimes prostrating. Surveys have shown that at least 50% of patients attending emergency rooms have 10 or more bowel movements per day. After 1 or 2 days of diarrhea, frank blood appears in the stools in about 15% of patients (30% of hospitalized patients), indicating a progression of infection in the tissues of the colon and rectum (see section “Colitis” below). Large variations in invasion frequency have been observed among strains of C. jejuni, and the occurrence of the plasmid pVir has been associated with bloody stool, suggesting that clinically severe (bloody) diarrhea may in part be strain dependent (Tracz et al., 2005). However, a recent study could not support an association between the presence of the pVir plasmid and bloody diarrhea and suggests a role for other virulence determinants (Louwen et al., 2006). In a large waterborne outbreak, the frequency of bloody diarrhea was approximately one-third of culture-confirmed infections, compared with less than 3% of non-culture-confirmed infections, and it reflects the age and severity bias in surveillance activity that is exclusively based on passive case detection from samples received by the health care system (Engberg et al., 1998). In two studies, gross or occult blood in the stools was shown to be more frequent than in Salmonella or Shigella enteritis (Blaser et al., 1983; Rao and Fuller, 1992). Nausea is a frequent symptom, but only about 15% of patients vomit. A particular feature of Campylobacter enteritis is abdominal pain, which may become continuous and sufficiently intense so as to mimic acute appendicitis; this is the most frequent reason for admission of patients with Campylobacter enteritis to the hospital. Although the pain is usually central, it can radiate to the right iliac fossa and make diagnosis difficult (see below). Outbreaks of Campylobacter enteritis in schools have been reported in which 30% of children had abdominal pain without diarrhea (Matsuaki and Katayama, 1984; Wilson et al., 1983).
Incubation period The mean incubation period of Campylobacter enteritis, calculated from 17 point-source outbreaks, is 3.2 days, with a range of 18 h to 8 days. The extremes of the range are suspect because the outliers might not represent infection from the defined source. A range of 1to 7 days is a reasonable working estimate. We emphasize that the mean incubation period of 3 days is longer than that of most other intestinal infections, a point of significance when questioning patients with suspected food-borne infection. Onset and prodrome The clinical consequences depend in part on the virulence of the infecting strain, the challenge dose, and the susceptibility of the patient. As described above, people who have had previous Campylobacter infection, such as those who habitually drink raw milk, or children brought up in developing countries will probably experience no symptoms. Others are likely to become ill. The disease described below is that shown by patients sufficiently ill to seek medical attention, but unrecorded milder degrees of illness are undoubtedly common. The onset is often abrupt with cramping pains in the abdomen, quickly followed by diarrhea, but about 30% of patients experience a nonspecific influenza-like prodrome with one or more symptoms of fever, headache, dizziness, and myalgia. Rigors have been recorded in up to 22% of patients, and fever may be sufficiently high to cause convulsions in children (Jones et al., 1981) or delirium in adults. Meningismus also has been observed. The prodrome can be highly misleading in the absence of abdominal symptoms, which may not appear for two or even three days. Patients with prodromal symptoms tend
Recovery stage After a variable period, usually about 3 to 4 days into the illness, the diarrhea begins to ease and the patient’s condition improves, although abdominal pain may persist for several more days. The mistake many patients make at this stage is to eat too much
102
BLASER AND ENGBERG
of erythromycin (Paulet and Coffernils, 1990). The proportion of patients admitted to the hospital shows even greater variation between surveys (0.5 to 32%). Most surveys indicate a hospitalization range of 5 to 10%. A fatal outcome is rare and usually confined to elderly patients or those with another serious disease (Skirrow et al., 1993; Smith and Blaser, 1985). The burden to society of Campylobacter-associated illness is discussed in more detail in chapter 8.
too soon, which invites a prompt return of symptoms, especially abdominal pain. Minor relapses have been reported in 15 to 25% of patients who seek care from physicians (Blaser et al., 1979; Drake et al., 1981; Pitkanen et al., 1983). Weight loss of up to 5 kg is a common outcome. Patients continue to excrete campylobacters in their feces for several weeks after they have clinically recovered, unless the infection has been treated with antibiotics. Estimation of excretion period depends on the sensitivity of detection methods. With direct plating methods, 50% of samples are culture negative after 3 weeks, but in a study that used more sensitive methods, a mean excretion period of 37.6 days (maximum 69 days) was found (Kapperud et al., 1992). Long-term carriage has only been observed in patients with immune deficiency, notably hypogammaglobulinemia and AIDS (see below). The question of convalescent Campylobacter excretion is academic: it has never been proved that such individuals transmit infection. A diagram of the typical course of Campylobacter enteritis is provided in Fig. 2.
SPECIAL ASPECTS OF INFECTION
Campylobacter Enteritis in Children The pattern of infection in children, especially newborn infants, differs sufficiently from the above description to warrant special consideration. Fortunately, illness is seldom severe. Fever is often absent, but vomiting and the passage of blood in the stools are more frequent. In a hospital-based study, 92% of children under 1 year of age had frank blood in their stools (Karmali and Fleming, 1979). In the newborn, the passage of blood, often without obvious diarrhea, can mimic intussusception, with the risk of an unnecessary laparotomy being performed. In cases in which the infant shows more severe illness, the clinical picture can resemble necrotizing enterocolitis (Torphy and Bond, 1979). The high frequency of blood in the stools of infected children reflects the frequency of colitis, although few have performed sigmoidoscopic examinations in the young (Guandalini et al., 1983). Chronic colitis also has been described (Heyman et al., 1982). Pancolitis was recorded in a 14-year-old boy (Lambert et al., 1979) and a 4-year-old girl who apparently died from Campylobacter infection superimposed on Crohn’s disease (Coffin et al., 1982). In another case, a 7-year-old boy with suspected bowel
Morbidity The average duration of illness is difficult to measure because of a multiplicity of variables, such as the immune status of the host, the virulence of the infecting strain, and the criteria used to define illness. A survey of nine outbreaks affecting about 1,500 people showed a mean duration of illness of 4.6 days, but in one of the outbreaks, a third of the patients were ill for more than 7 days (Millson et al., 1991). A survey of sporadic cases in Norway recorded a mean figure of 3.8 days lost from work or school and 14.6 days for the presence of symptoms (Kapperud et al., 1992). There is a remarkable case of a young man who allegedly had Campylobacter enteritis for 17 years and who was ultimately cured by a course
-- a/
I *
Days
,
Incubation period
Ingest ion
1 I
1
2
3
4
5
6
7
Fever
Abdominal pain
-/
Diarrhoea
1-111
Blood in faeces
Campylobacters
in faeces
50%
Figure 2. Diagram illustrating the typical course of Cumpylobucter enteritis. Reprinted from D. Greenwood, R. Slack, and J. Peutherer (ed.), Medical Microbiology, 15th ed. (Churchill Livingstone, Edinburgh, 1997).
CHAPTER 6
CLINICAL ASPECTS OF c. r E r m I AND
obstruction was found by radiography to have colonic distension, fluid levels, and aphthous ulcers of his colon, all apparently due to Campylobacter infection (Bentley et al., 1985). Neonatal Infection Neonatal infection usually is acquired at birth from infected mothers. The mother may not be having diarrhea at the time, nor is there always a history of diarrhea (Anders et al., 1981; Karmali et al., 1984). At least five outbreaks in neonatal nurseries have been reported in which there has been nosocomial spread of infection from some communally used items, such as inadequately disinfected thermometers or baby baths (Goossens et al., 1986; Hershkowici et al., 1987; Rusu and Lucinescu, 1988; Terrier et al., 1985; van Dijk and van der Straaten, 1988). In one such case, C. jejuni was isolated from a rubber bath plug in the affected nursery (van Dijk and van der Straaten, 1988). In another outbreak, 11 infants developed C. jejuni meningitis (Goossens et al., 1986). There is a single report of a 3*/2-week-old boy with hemolytic anemia associated with C. jejuni gastroenteritis (Damani et al., 1993). Rashes Urticaria appearing late in Campylobacter enteritis has been described on several occasions (Bradshaw et al., 1980; Bretag et al., 1984; Hoskins and Davies, 1983; Lopez-Brea et al., 1984). Most remarkable was when 12 boys sought care for urticaria near the end of an outbreak of Campylobacter enteritis in a United Kingdom boarding school, in which some 400 boys had symptoms; not all of the 12 boys with urticaria had had diarrheal symptoms (Hoskins and Davies, 1983). Erythema nodosum appearing 1 to 2 weeks after Campylobacter enteritis has been reported in four patients, all women (Ashworth and Engblish, 1984; Eastmond et al., 1983; Ezpeleta et al., 1992; Lambert et al., 1979). A patient whose forearms, elbows, and shins were affected had terminal ileitis (Lambert et al., 1979). Immune complex vasculitis has been reported in association with Campylobacter enteritis, but the patients had underlying diseases that could have contributed to the skin pathology (Autenrieth et al., 1996; Nagaratnam et al., 1990; Schsnheyder et al., 1995). Infection in Immunodeficient Patients Hypogammaglobulinemia and AIDS are the conditions most often associated with more severe
c. COLI INFECTIONS
103
Campylobacter infection. Chronic carriage of campylobacters with recurrent enteritis, often with bacteremia, are typical problems. A study of Campylobacter infection in patients infected with human immunodeficiency virus (HIV), 76% of whom had AIDS, showed that 10% had bacteremia (Molina et al., 1995), a rate much higher than in immunocompetent hosts. Another study showed that bacteremia in HIV-infected patients was often a severe debilitating febrile illness requiring multiple and prolonged courses of antimicrobial therapy (Tee and Mijch, 1998). The incidence of Campylobacter infection in patients with AIDS has been calculated to be 40-fold higher compared with immunocompetent hosts (Sorvillo et al., 1991). In a survey of 41 hypogammaglobulinemic patients, 5 had experienced at least one episode of C. jejuni septicemia, 3 with an erysipelas-like cellulitis (Kerstens et al., 1992). Repeated courses of antimicrobial treatment are often needed in these patients, and this carries with it the risk that the infecting strain will become resistant, often compounding an intractable problem. Limited experience suggests that the best approach is to combine an antimicrobial agent with an immunoglobulin preparation. A commercial IgM (Pentaglobin) was used successfully with antibiotics to cure two hypogammaglobulinemic patients of their C. jejuni infections (Borleffs et al., 1993), and the bactericidal properties of the preparation were demonstrated in vitro against several serotypes of C. jejuni (Autenrieth et al., 1995). The anti-Campylobacter IgM antibodies present in this preparation appear essential because eight other immunoglobulins containing only IgG lacked a bactericidal effect (Autenrieth et al., 1995). Another report describes the successful combination of maternal plasma with ciprofloxacin in treating an intractable 2-year-long C. jejuni infection in a 7-year-old boy with X-linked agammaglobulinemia (Autenrieth et al., 1996).
INTESTINAL COMPLICATIONS Appendicitis Abdominal pain in Campylobacter enteritis may be intense and continuous, and it may radiate to the right iliac fossa. In most patients, the appendix is not affected (“pseudo-appendicitis”), and the pain is caused by terminal ileitis and mesenteric adenitis. Yet occasionally there is genuine appendicitis, and campylobacters can be isolated from the inflamed appendixes (Chan et al., 1983; Mtgraud et al., 1982; Morlet and Glancy, 1986; Pearson et al., 1982). True
104
BLASER AND ENGBERG
guarding and rebound tenderness demand surgical attention whether or not Campylobacter infection is thought to be present. Although appendicitis associated with Campylobacter enteritis seems to occur infrequently, in actuality, it is most likely often overlooked as a result of appropriate specimens not being routinely cultured. A survey of 25 l children admitted to a surgical unit with suspected appendicitis showed that 6 (2.4%) had Campylobacter infection, 3 of whom (1.2%) had genuine appendicitis, and 3 had mesenteric adenitis with normal appendixes (Pearson et al., 1982). In a larger survey of 533 older children and adults, the Campylobacter infection rate was almost the same (2.8%), but only one patient showed minor histological evidence of acute appendicitis (Puylaert et al., 1989). The other infected patients were shown by graded-compression ultrasound to have a characteristic picture of mural thickening of the terminal ileum and cecum with enlarged mesenteric lymph nodes, but no image of the appendix. As a result of these sonograms, 26 planned appendectomies were canceled (Puylaert et al., 1989). The proportion of patients with Campylobacter enteritis referred for suspected appendicitis is small. In a large milk-borne outbreak in the United Kingdom (Jones et al., 1981), the fraction was 0.1%, and among sporadic infections reported to the Communicable Disease Surveillance Centre, London, the figure was 0.27% (unpublished data). Nevertheless, even these small proportions (1to 400 to 1 to 1,000 infections) could account for hundreds of cases a year (or more) because Campylobacter infections in the population are so common.
Colitis Sigmoidoscopy and analysis of rectal biopsy specimens show that elements of colitis and proctitis are present in most patients with Campylobacter enteritis sufficiently severe to have these tests done (Blaser et al., 1980; Colgan et al., 1980; Lambert et al., 1979; McKendrick et al., 1982; Price et al., 1979). In some patients, colitis is a dominant feature, and then the problem is to distinguish the condition from acute nonspecific inflammatory bowel disease (IBD). Stool cultures should be obtained as soon as possible, and it is worth determining whether the local laboratory can undertake rapid diagnosis by direct microscopic examination of feces for campylobacters. Cultures that use filtration techniques to identify other Carnpylobacter and related species should also be done. There are conflicting views on how reliably the distinction between Campylobacter infection and
IBD can be made by sigmoidoscopy and analysis of rectal biopsy samples, with some workers finding little difficulty (Mee et al., 1985; Price et al., 1979) but others encountering problems in some patients (Lambert et al., 1979; McKendrick et al., 1982). First, the changes seen in Campylobacter enteritis are indistinguishable from those of other acute bacterial infections of the gut, such as those caused by Salmonella and Shigella infections. Endoscopic appearances range from mildly hyperemic intact epithelium to mucosal edema, granularity, friability, spontaneous bleeding, and patchy aphthous type ulceration (Colgan et al., 1980; Lambert et al., 1979; McKendrick et al., 1982; Mee et al., 1985). Active colitis may extend at least as far as the splenic flexure, and one patient has been described as having ulcers up to 3 X 5 cm and a “cobblestone” mucosa, but with “skip areas” of normal mucosa (Loss et al., 1980). Histology shows acute inflammation of the mucosa with edema, infiltration by polymorphonuclear leukocytes, and crypt abscess formation (Colgan et al., 1980; Lambert et al., 1979; McKendrick et al., 1982; Price et al., 1979; van Spreeuwel et al., 1985). In one study, the degree of histological abnormality was found to correlate with a history of blood in the stools, but not with fever, abdominal pain, or diarrhea (Colgan et al., 1980). In contrast to IBD, the lesions are focal, there is little or no distortion of the crypts or mucosa, and there is no striking depletion of mucus cells. However, the appearances overlap, especially in patients with Campylobacter enteritis who have had the infection for more than a week, in whom chronic inflammatory cells may be numerous (McKendrick et al., 1982). The main distinguishing histological features are listed in Table 2. It is important to remember that an apparent flare-up of IBD may be due to an intercurrent intestinal infection, such as Campylobacter enteritis (Goodman et al., 1980; Newman and Lambert, 1980). Pathogens must be looked for in such patients so that appropriate antimicrobial treatment can be started at the earliest opportunity. It is unlikely that Campylobacter infection plays any part in the cause of IBD; no excess of patients with antibodies to Campylobacter antigens were found in serological surveys of patients with IBD (Blaser et al., 1984a, 1984b; Melby and Kildebo, 1988).
Toxic Megacolon Toxic megacolon complicating Campylobacter enteritis has been reported in three patients, all adult women (Gould, 1985; McKinley et al., 1980; Stephenson and Cotton, 1985). Unless perforation of the bowel is suspected, it is best to avoid surgical
CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
105
Table 2 . Comparison of histological features that help to differentiate Cumpylobucter and other infective causes of proctocolitis from acute inflammatory bowel disease (IBD)" Feature
Infective colitis
Distribution of lesions Crypt architecture Mucus cell depletion Inflammatory cells Distribution of cells Epithelioid granulomata Basal lymphoid aggregates Isolated giant cells
IBD
Focal and segmental Preserved Absent or mild Mainly polymorphs in focal collections; crypt abscess formation Mainly mucosal Absent Absent In upper part of glandular epithelium
Diffuse and general Distorted, atrophic Chronic inflammatory cells (especially plasma cells) Submucosa may be involved Often present Present Basally located
aBased on data from Surawicz and Belic (1984).
treatment in such patients and to rely instead on vigorous antimicrobial treatment. Intestinal Hemorrhage Severe intestinal hemorrhage has been reported. One such patient was a previously healthy 24-yearold nurse who had to undergo emergency hemicolectomy for a massively bleeding ulcer in the terminal ileum (Michalak et al., 1980). Ileostomy Stoma Ulceration Patients with an ileostomy may develop partial strangulation of their stoma if they get Cumpylobucter enteritis, presumably through congestion and edema. In two patients with long-established ileostomies resulting from total colectomy for IBD, the stoma became extensively ulcerated but eventually healed without lasting damage (Meuwissen et al., 1981; Skirrow, 1981). Perirectal Abscess
A perirectal abscess in a 64-year-old woman yielded C. jejuni and Citrobucter freundii by culture 3 weeks after she had experienced diarrhea presumed to have been caused by the C. jejuni strain (Krajden et al., 1986). EXTRAINTESTINAL INFECTION Bacteremia Bacteremia in Cumpylobucter enteritis is seldom reported, but it probably occurs commonly as a transient event in the early stages of infection, especially in patients with rigors and high fever. There are three reasons why C. jejuni bacteremia is not detected more frequently: (i) blood cultures are rarely taken
early in the disease; (ii) C. jejuni is generally sensitive to the bactericidal properties of normal serum, unlike C. fetus, which is serum resistant; and (iii) not all methods of detecting bacteremia are equally sensitive for Cumpylobucter species (Wang and Blaser, 1986). A survey of laboratory reports made to the Communicable Disease Surveillance Centre, London, showed an average bacteremia rate of 1.5 per 1,000 intestinal infections, but there was wide variation with age (Skirrow et al., 1993). The highest rate (5.9 per 1,000) was in patients aged 65 years or more, and the lowest (0.3 per 1,000) was in children aged 1 to 4 years (Fig. 3 ) . Rates in male subjects were nearly twice as high as those in female subjects. Although some patients had immunodeficiency or other underlying disease, 71% were apparently normal (in contrast to patients with C. fetus infection, most of whom had some predisposing disorder). C. jejuni strains belonging to Penner serogroups HS:4 and HS: 18 were more frequent among blood than fecal isolates, but subsequent genotypic analysis of these and additional strains did not suggest that blood isolates were especially invasive or serum resistant (Jackson et al., 1997a). In a smaller, more recent study performed at two centers in Denmark, the bacteremia rate was found to be 8 per 1,000 intestinal infections (Schsnheyder et al., 1995). C. jejuni can cause serious septicemic infection in patients whose immune system is severely compromised (see above). A fatal outcome was reported in two patients who had undergone splenectomy for thalassemia, both of whom had iron overload (Jackson et al., 1997b; Meyrieux et al., 1996). Infection of the Hepatobiliary System Hepatitis
A mild degree of hepatic inflammation may be a regular feature of Cumpylobucter enteritis. Several re-
106
BLASER AND ENGBERG
4
I I
5
I
<1
1-4
5-14
15-24 25-34 35-44 45-54 55-64 265
Age in Years
Figure 3. Distribution of Cumpylobacter bacteremia cases in England and Wales, 1981 to 1991. Solid line indicates number of cases (n = 374); dashed line, cases per 1,000 intestinal infections. Reprinted from Skirrow et al. (1993).
ports have shown slightly increased serum transaminases in hospitalized patients with Campylobacter enteritis (Jackson et al., 1997b; McKendrick et al., 1982; Pitkanen et al., 1983). Some strains of C. jejuni have been shown to produce hepatotoxic factors in mice (Kita et al., 1992), and in experimentally infected rhesus monkeys, the liver and gallbladder were found to be the most consistently colonized organs after the intestines (Fitzgeorge et al., 1981). However, obvious clinical hepatitis is rare. Two young men were found to have enlarged livers and increased liver enzymes 1 week after the onset of Campylobacter enteritis (Humphrey, 1993). In three other patients with clinical hepatitis, analysis of liver biopsy samples showed evidence of hepatitis: a 52year-old man had mild mononuclear cell infiltration in portal areas, with variation in hepatocyte and nuclear size (Reddy and Thomas, 1982); a 48-year-old alcoholic man had cholestatic hepatitis with lymphohistiocytic infiltration of portal tracts and C. jejuni bacteremia (Nahum et al., 1982); and a previously healthy 50-year-old woman, also with bacteremia (with C. coli), had neutrophil infiltration of the portal tracts with adjacent areas of necrosis (Ampelas et al., 1982). More recently, an 82-year-old woman admitted to the hospital with rigors, fever, and watery diarrhea due to C. jejuni was found to have acute hepatitis (Korman et al., 1997). Significantly, the hepatitis resolved in all six of these patients with treatment directed against the infecting Campylobacter strain. Cholecystitis
C. jejuni is a bile-tolerant organism, so it is not surprising that it is sometimes found in the gallblad-
der in association with acute or acute-complicatingchronic cholecystitis (Darling et al., 1979; Drion et al., 1988; Gerritsen van der and Veringa, 1993; Mertens and De, 1979; Pereira et al., 1981). Patients may or may not give a history of recent diarrhea. Prospective searches for campylobacters in bile from several hundred cholecystectomy samples yielded none, so biliary infection appears to be uncommon (Darling et al., 1979).
Pancreatitis Acute pancreatitis has been reported in a few patients, all adults, during the acute stages of Campylobacter enteritis. Abdominal pain was a dominant symptom in each (Castilla et al., 1989; De Bois et al., 1989; Dutronc et al., 1995; Gallagher et al., 1981; Ponka and Kosunen, 1981). An 88-year-old woman had acute pancreatitis and C. jejuni bacteremia without diarrhea (Ezpeleta et al., 1992). The pancreatitis resolved satisfactorily in all of these patients after they had received appropriate antimicrobial treatment. The mechanism by which campylobacters cause pancreatitis is unknown. The possibility that a mild degree of pancreatitis is a regular feature of the disease was suggested by a Finnish study of 188 hospitalized patients with Campylobacter enteritis. In that study, 22% of patients were considered to have pancreatitis on the basis of increased serum amylase or lipase levels (Pitkanen et al., 1983). However, in a study of 22 other patients, none was found to have increased enzyme levels (Murphy et al., 1991).
Next Page CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
107
Campylobacter Abortion and Perinatal Infection
Renal and Urinary Tract Disease
C. fetus has been traditionally regarded as the cause of epizootic abortion of sheep and cattle, but it was only after the implication of C. jejuni and C. coli in human infections that the role of these organisms in veterinary medicine became apparent. A survey in the United Kingdom showed that C. jejuni and C. coli account for about 40 and 20% of ovine abortion cases, respectively; only 40% are caused by C. fetus. This predilection of campylobacters for uterine and fetal tissues in ruminants is fortunately uncommon in humans. Nevertheless, there have been nearly 30 reports of septic abortion or stillbirth due to Campylobacter infection. About half were caused by C. fetus, one by C. hyointestinalis, and the remaining half by C. jejuni or C. coli. The average stages of gestation for women infected with C. fetus or with C. jejuni are 28 and 19 weeks, respectively. However, five infants infected in utero with C. fetus were born sufficiently late in gestation to survive premature birth, whereas none of the C. jejuni-infected infants survived; only two were born late enough to have a chance of survival. The pathology in humans resembles that of Campylobacter abortion in sheep, namely an acute placentitis that is severe enough to cause fetal death by placental insufficiency, although infection eventually reaches the fetus via amniotic fluid. The campylobacters probably reach the placenta by hematogenous spread from the intestinal tract, although this theory is unproven. There is no evidence that infection ascends from the genital tract. Although most infected mothers have fever and about half have diarrhea shortly before or at the time of abortion, maternal illness is seldom pronounced, and there may be little to indicate the nature of the problem. Twenty patients with Campylobacter abortion were reviewed by Simor et al. (1986), and several more patients have been described since then (Denton and Clarke, 1992; Moscuna et al., 1989; Simor and Ferro, 1990). A 28-year-old woman with acquired agammaglobulinemia had recurrent abortions after a persistent attack of C. jejuni diarrhea (Pines et al., 1983). The factors that allow campylobacters to infect the human placenta are unknown. In a study in the United Kingdom, C. jejuni and C. coli were isolated from the rectal swabs of 0.2% of women in labor, most of whom had no symptoms relating to the organism (Youngs and Roberts, 1985). Thus, many thousands of pregnant women must develop the infection without any untoward effect. Some infections are almost certainly overlooked, but even allowing for this, Campylobacter-induced abortion remains a rare event.
There are reports of eight patients, five of them children under the age of 5 years, who developed hemolytic-uremic syndrome within a few days of developing Campylobacter enterocolitis (Chamovitz et al., 1983; Denneberg et al., 1982; Dickgiesser, 1983; May et al., 1986; Shulman and Moel, 1983). An exception is the case of an adult patient in whom the interval was 1 month (Delans et al., 1984). Most of the patients had blood in their stools, and all recovered except one, a 4-month-old Bangladeshi infant who died (Haq et al., 1985). A further report describes a 72-year-old woman who developed the closely related syndrome of thrombotic thrombocytopenic purpura 5 days after the onset of nonbloody C. jejuni diarrhea (Morton et al., 1985). Because only one of these reports states that enterohemorrhagic Escherichia coli were not isolated, it is possible that the isolation of campylobacters was coincidental to undiscovered enterohemorrhagic E. coli infection. Several of the incidents arose before the role of enterohemorrhagic E. coli in hemolytic-uremic syndrome was known or before routine measures for their isolation were implemented. Nephritis Rare instances of nephritis, mainly glomerulonephritis, complicating Campylobacter enteritis have been reported. An 18-year-old man developed mild acute glomerulonephritis during an attack of C. jejuni enteritis (Menck, 1981). A 33-year-old man was shown to have moderately severe diffuse proliferative endocapillary glomerulonephritis by analysis of renal biopsy specimens a few days into an attack of C. jejuni diarrhea; he required hemodialysis but recovered completely (Maidment et al., 1985). A 5-year-old girl developed glomerulonephritis with pulmonary hemorrhage and anemia (Goodpasture’s syndrome) 3 to 4 weeks after the onset of C. jejuni diarrhea. Analysis of renal biopsy samples showed immune complexmediated crescentic glomerulonephritis, and C. jejuni antigen was identified in the glomeruli. The patient developed progressive renal failure (Andrews et al., 1989). Mesangial IgA glomerulonephritis has been observed in two men 15 and 26 years old, both of whom had hematuria within 2 days of the onset of diarrhea (Carter and Cimolai, 1991; M. B. Skirrow, unpublished data). There is a single report of selflimited tubulointerstitial nephritis diagnosed in a 20year-old man 8 days after the onset of C. jejuni enteritis (Rautelin et al., 1987).
Previous Page 108
BLASER AND ENGBERG
Prostatitis and urinary tract infection The urinary tract seems an unlikely place to find campylobacters because they do not tolerate acid conditions well, but there are two reports of apparent C. jejuni urinary infection. In the first, the main site of infection was thought to be the prostate (Davies and Penfold, 1979), but in the other, there appeared to be cystitis in a 6-year-old girl (Feder et al., 1986). The authors of that report saw Campylobacter-like organisms in the urine and obtained appropriate cultures for them. As they indicate, infections will be missed unless cultures specific for campylobacters are set up when morphologically suspect bacteria are seen in urine samples. Miscellaneous Extraintestinal Infections Focal infection Focal infection with C. jejuni or C. coli is rare. Most such infections are in subjects with immunodeficiency or other predisposing conditions such as cirrhosis, and presumably they arise from bacteremia. Reports of focal infection, other than infection of the female reproductive tract (see above), are listed in Table 3. Peritonitis There are four reports of spontaneous peritonitis, two due to C. jejuni, one to C. coli, and one in which the differentiation was not made. Three patients had long-standing alcoholic cirrhosis, and the other had cirrhosis due to congestive heart failure (Doming0 et al., 1985; Ho et al., 1987; McNeil et al., 1984; Schmidt et al., 1980). Less rare is Campylobacter peritonitis complicating continuous am-
bulatory peritoneal dialysis, which usually arises during an attack of Campylobacter enteritis. It is unknown whether the bacteria seed the peritoneal cavity from the bloodstream, by transluminal migration from the intestinal, or by contamination of the catheter fittings. A report of eight patients and a review of other case reports are given by Wood et al. (1992), and there has been one other subsequent report (Webster and Farrell, 1993). Myocarditis Myocarditis has been reported within a week of C. jejuni enteritis in two young men. The cardinal feature in both patients was severe chest pain radiating to the right arm. The first patient, aged 27 years, was believed to have perimyocarditis (Ponka et al., 1980). The second patient, aged 23 years, had more convincing evidence of cardiac damage in that his cardiac enzymes were quite high (Florkowski et al., 1984). Electrocardiogram changes were observed in both cases: T-wave inversion, slight increase of ST segment, and lengthening of the PR interval on serial tracings. No signs of lasting damage were found in either patient 2 months after the episode. There is a brief report of transient atrial fibrillation associated with C. jejuni enteritis in three patients, all over the age of 50 years (Kell and Ellis, 1985). Splenic rupture Spontaneous rupture of the spleen was reported in a 71-year-old man 1 week after the onset of explosive diarrhea due to C. jejuni enteritis (Frizellaand Rietveld, 1993). Recovery was uneventful after partial splenectomy. There was no other factor to account for the rupture; the authors indicated that
Table 3. Focal infections due to C. jejuni and C. coli Reference
Focus of infection
Age ( yr) I sex of patient
Predisposing condition
Organism
Muytjens and Hoogenhout (1982) Pedler and Bint (1984)
Chest wall abscess
72/F
Mastectomy scar, postirradiation
C. jejuni
Osteitis of foot
57/M
C. jejuni
Peterson et al. (1993) Pasticci et al. (1992) Unpublished data" Schieven et al. (1991) Norrby et al. (1980) Ritchie et al. (1987) Unpublished data" Unpublished data"
Prosthetic hip sepsis Septic arthritis of knee Septic arthritis of shoulder Acute bursitis Meningitis, nonneonatal Subdural sepsis Empyema Empyema
Site of previously removed histiocytoma AIDS Rheumatoid arthritis Not known Chronic bursitis Long-standing ventricular shunt Previous hemispherectomy Not known Not known
a
Public Health Laboratory Service, London.
60/M 51/F 90lM 81IM 34/M 21/2/F 70/M 52/M
C. jejuni C. jejuni Campylobacter spp. C. jejuni C. jejuni C. jejuni C. jejuni C. coli
CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
spontaneous splenic rupture has been associated with other enteric pathogens, including Salmonella Dublin. Because Campylobacter infection is so common, it is not certain which of these complications are secondary events and which are unrelated adventitious occurrences. Late-Onset Complications Reactive arthritis and Reiter’s syndrome The reactive arthritis (ReA) that sometimes follows Campylobacter enteritis is no different from that associated with Salmonella or other intestinal bacterial infections. In a review of 29 patients, the mean interval between the onset of bowel symptoms and the appearance of pain and swelling of the joints was 14 days, with a range of 3 days to 6 weeks (Peterson, 1994). Ankles, knees, wrists, and the small joints of the hands and feet are most commonly affected, often in migratory fashion (Peterson, 1994; Schaad, 1982). Little is known about the mechanism by which Campylobacter causes ReA. A recent study examined the lipooligosaccharide biosynthesis gene cluster of infecting C. jejuni strains causing joint disease and uncomplicated gastroenteritis, and the authors identified lipooligosaccharide biosynthesis gene cluster A as a potential bacterial marker for an increased risk of developing postinfectious joint inflammation (Bergman et al., 2007). Campylobacter-associated ReA incidence and prevalence varies widely among reports because of differences in exposure and in case ascertainment, lack of diagnostic criteria for ReA, and perhaps the genetic makeup and the age of exposed individuals. For instance, clinical examination of a subset of patients with self-reported joint symptoms in a recent prospective study revealed joint symptoms from other causes unrelated to the preceding infection in 37% of patients (Schiellerup, 2006). A review of the literature available to date suggests that the incidence of Campylobacter ReA may occur in 1 to 5% of those infected (Pope et al., 2007). The duration of arthritis ranges from several weeks to several months, or occasionally a year. Although the arthritis can be incapacitating, full recovery is the rule. Acute erosions of the hamate bone of the wrist and distal ends of the clavicles were found in a 46-year-old man (Ebright and Ryan, 1984). The full Reiter’s syndrome has been observed in some patients with Cumpylobacter-associated arthritis (Leung et al., 1980; Peterson, 1994; Saari and Kauranen, 1980; Schaad, 1982), and uveitis in the absence of arthritis has been reported in two middle-aged women, one of whom had hypogammaglobulinemia (Lever et al., 1984) and the other of whom had conjunctivitis several weeks before the onset of uveitis (Howard et al., 1987).
109
In a recent prospective case-case comparison study of 1,003 sporadic Campylobacter infections, no significant correlation between HLA B27 and occurrence of ReA was found, whereas in 812 patients with Salmonella, Yersinia, or Shigella infection, there was a significant correlation between ReA and HLA B27 (Schiellerup, 2006). However, possession of the HLA B27 tissue antigen carries a significantly increased risk of a more severe course of ReA, and this also accounts for Campylobacter-associated ReA. Two recent studies also show that a severe acute course of gastrointestinal symptoms is correlated to a higher incidence of ReA (Locht and Krogfelt, 2002; Schiellerup, 2006). The long-term prognosis for postCampylobacter ReA is not defined in the current literature. However, in another study, HLA B27 was associated with a strong predisposition to ReA and patients who possess it had, on average, higher erythrocyte sedimentation rate values (84.6 versus 44.6) than B27-negative patients (Peterson, 1994). At least 60% (and probably closer to 80%) of ReA patients possess the B27 antigen. However, it is difficult to generalize about the frequency of ReA in Campylobucter enteritis patients because much depends on the prevalence of HLA B27 in the population. Estimates ranging from 0 to 1.7% have been given in community outbreaks of the disease (Eastmond et al., 1983; Jones et al., 1981; Melby et al., 1990; Mentzing, 1981; Millson et al., 1991). A calculation of 5% in hospital patients has been reported in Scandinavia, where 14% of the population is HLA B27 positive (Pitkanen et al., 1983). A long-term follow up of 66 patients affected in a food-borne outbreak of C. jejuni enteritis in Sweden suggested that nonspecific rheumatic disorders are a common sequela of infection (Bremell et al., 1991). Four patients experienced chronic or relapsing rheumatic symptoms beginning 3 to 8 months after infection and persisting for 5 years, but in the absence of a control group, the significance of this finding is questionable. Guillain-BarrC syndrome The association of Campylobacter enteritis with Guillain-BarrC syndrome, which emerged during the mid-1980s, greatly improved our understanding of the morbidity of the disease. In view of the importance of Guillain-Barrt syndrome and the complexity of its pathogenesis, it is given special consideration in chapter 13. Postinfectious irritable bowel syndrome Late diarrheal complications are not infrequent after acute episodes of infectious disease (Spiller,
110
BLASER AND ENGBERG
2007). A major question is whether long-term diarrhea is postinfectious irritable bowel syndrome (PIIBS) or represents closer surveillance of events that are occurring at similar rates in individuals without infectious diarrhea. Among patients with unselected IBS, 6 to 17% believe that their symptoms began after an infection; conversely, prospective studies following patients with acute bacterial gastroenteritis show a 4 to 31% incidence of PI-IBS (Spiller, 2007). An increasing body of evidence Suggests that after Campylobacter enteritis, PI-IBS is occurring at rates greater than expected from controls (Dunlop et al., 2003; Marshall et al., 2006; Thornley et al., 2001). Studies of patients after infection indicate PIIBS rates that are about 5 to 20% after subtracting the rate from unaffected controls (Dunlop et al., 2003; Marshall et al., 2006; Thornley et al., 2001). In a study of individuals 3 months after Campylobacter infection, PI-IBS, chiefly of the diarrheapredominant type, occurred in 103 (13.8%) of the 747 subjects (Thornley et al., 2001). Rectal biopsy samples were examined in patients with new-onset PI-IBS, and they matched with those of asymptomatic patient controls and healthy volunteers. Enterochromaffin cell counts were significantly higher in the PI-IBS patients than in the two control groups. Multivariant analysis showed that both increased enterochromaffin cell counts and clinical depression were important predictors of developing PI-IBS (Thornley et al., 2001). The overlay of physical and psychosomatic syndromes is illustrated by these associations. One other interesting observation is that among C. jejuni isolates from patients, those that were toxigenic to HEp-2 cells were more often isolated from patients who later developed PI-IBS (Thornley et al., 2001). Altogether, a body of evidence is emerging that PI-IBS develops in a proportion of patients after Campylobacter enteritis, in which some level of diarrheal symptoms can persist for months, at least-.
DIAGNOSIS AND CLINICAL INVESTIGATIONS Bacteriology As mentioned previously, a definitive diagnosis of Campylobacter enteritis can only be made bacteriologically. Because the laboratory diagnosis of the infection is described in chapter 12, all that is necessary here is to stress the importance of delivering fecal samples to the laboratory with a minimum of delay. Samples should be refrigerated while awaiting transport. If delays of more than a day are likely, transport medium, such as a standard Cary-Blair type medium, should be used. Rectal swabs as a substitute
for feces should only be used as a last resort, in which case the use of transport medium is essential. Rapid diagnosis may be helpful in patients with suspected appendicitis or with acute colitis masquerading as acute IBD, or for definitive early antibacterial therapy. The microscopic examination of wet feces or fecal smears for campylobacters can provide a rapid diagnosis, but the sensitivity of microscopy is variable (36 to 90%), and it may not be available in all laboratories. Hematology and Biochemistry It is only in patients with complications that hematologic and biochemical analyses are likely to be needed, but it is useful to be aware of the changes to be found in patients with Campylobacter enteritis. These values relate mainly to series of patients sufficiently ill to be treated in the hospital. Leucocyte counts are usually normal or only slightly high (Pitkanen et al., 1983). Mean leucocyte counts (109/liter) in three series of patients were 7.9 (Svedhem and Kaijser, 1980), 8.9 (Pentland, 1979), and 10.0 (DeWitt et al., 1985). A particular feature of Campylobacter enteritis not seen in other forms of acute diarrhea, except shigellosis, was a high percentage of band forms (mean 15.1%), especially in the presence of only a moderate total leucocyte count (DeWitt et al., 1985). The erythrocyte sedimentation rate is usually raised, and high C-reactive proteins were recorded in all acutely ill patients in one series (Pitkanen et al., 1983). Biochemical values in Campylobacter enteritis are usually normal, although metabolic acidosis was detected in 32% of hospital patients in one study (Pitkanen et al., 1983). Mildly increased serum transaminases have been reported in 14 to 25% of patients (Drake et al., 1981; McKendrick et al., 1982; Pitkanen et al., 1983) and slightly increased alkaline phosphatase values in 10% (Drake et al., 1981). Endoscopy and Radiography The role of endoscopy and rectal biopsy in the management of patients has already been described above (see “Colitis”). Only patients with severe or complicated infection are likely to require radiographic investigation. Dilated loops of intestine with fluid levels on plain abdominal radiographs were observed in 5 of 20 patients in one study of hospital patients (Bradshaw et al., 1980), but not in 14 patients in another study (Pentland, 1979). Contrast radiography has shown the presence of pancolitis (Blaser et al., 1980; Lambert et al., 1979), nodular thickening of the terminal ileum (Lambert et al.,
CHAPTER 6
-
CLINICAL ASPECTS OF C. lElUNI AND C. COLI INFECTIONS
1982), right-sided colonic ulceration and dilation (McKinley et al., 1980), and multiple aphthous type ulceration, which in one patient mimicked Crohn’s disease (Bentley et al., 1985). Barium enema examination strongly suggested colonic carcinoma in two patients, so much so that one patient was subjected to hemicolectomy for what was found to be nothing more than inflammatory edema, congestion, and lymphoid hyperplasia (Doberneck, 1983; Noble et al., 1982). The increasing availability and usage of abdominal computed tomographic scanning should facilitate differentiation between appendicitis and mesenteric adenitis.
TREATMENT
No specific treatment is required for most patients with Campylobacter enteritis, other than the oral replacement of fluid and electrolytes lost through diarrhea and vomiting. Antimicrobial therapy plays a limited role because the patient will likely be showing signs of improvement by the time a bacteriological diagnosis has been made. This does not necessarily reflect a slow laboratory service; it has repeatedly been shown that patients seek medical advice only after they have been ill for an average of 4 to 6 days (Anders et al., 1982; Mandal et al., 1984; Pai et al., 1983). This late presentation of patients affected early trials of antibiotic therapy, but anecdotal evidence suggested that antimicrobial agents were effective when given early in the disease. This prompted trials of erythromycin and fluoroquinolones given empirically when patients first sought care for diarrhea. Such studies confirmed that illness was shortened and that campylobacters disappeared from the feces, provided that the infecting strain was susceptible to the antimicrobial used (Dryden et al., 1996; Goodman et al., 1990; Mattila et al., 1993; Pichler et al., 1987; Salazar Lindo et al., 1986). Trials in developing countries showed less effect, but the results were confounded by the simultaneous presence of other pathogens or the high frequency of antibiotic-resistant strains (Taylor et al., 1987). Clear instances of improvement with antimicrobial therapy have been observed in patients with severe or chronic infection, often in cases where other measures have failed (Bentley et al., 1985; Blaser et al., 1980; Noble et al., 1982). Erythromycin Erythromycin was the first antimicrobial agent to be used, and in general, macrolides remain the agents
111
of choice. Table 4 shows data on macrolide resistance among C. jejuni, C. coli, and C. jejuni and C. coli combined, isolated from human sources around the world since 1997 (Engberg et al., 2005). There are notable differences in resistance rates between countries and species. Almost all studies report a higher frequency of erythromycin resistance in C. coli than in C. jejuni, with rates reported in proportions ranging from 0 to 20% in C. jejuni and 0 to 29% in C. coli. In a number of industrialized countries, a higher proportion of C. coli, including macrolide-resistant C. coli, have been reported among travel-related infections than among those domestically acquired. The trend over time for macrolide resistance shows stable, low rates in most countries, which is comforting because erythromycin-or alternatively one of the newer macrolides, such as clarithromycin or azithromycin-is the drug of choice for treating C. jejuni and C. coli enteritis. Development of resistance to macrolides in Cumpylobucter during therapy has not been documented in humans. The origin of resistant strains has been linked to the veterinary use of antibiotics of the macrolide-lincosamide group and is discussed in more detail in chapter 36. Fluoroquinolones and Other Antimicrobials Ciprofloxacin and other fluoroquinolones were hailed with great enthusiasm when they first appeared, but their use has been severely compromised by increasing resistance rates in many countries (Fig. 4). In some countries, the increase in resistance has been remarkably rapid and considerable; in others, resistance rates have increased more steadily. For example, 86% of human C. jejuni isolates were quinolone resistant in Hong Kong (Chu et al., 2004). Thus, in the high-endemic quinolone resistance areas, fluoroquinolones cannot be recommended for community-acquired bacterial diarrhea because the predominant causes are often Campylobacter spp. Although lower frequencies are reported from other regions, recent trends show a clear and worrying tendency of emerging quinolone resistance in many countries. As discussed in chapter 36, the use of fluoroquinolones (mainly enrofloxacin) in veterinary medicine is correlated with an increase in quinolone resistance in food animals; in retail food of animal origin, especially poultry products; and most importantly, in human Cumpylobacter infections. Veterinary use of fluoroquinolones is not the only selection pressure that acts on Campylobacter to
112
BLASER AND ENGBERG
Table 4. Macrolide resistance among C. jejuni, C. coli, and C. jejuni and C. coli combined, isolated from human sources around the world since 1997" Country
C. jejuni
C. coli
Argentina Australia Austria Belgium Bosnia and Herzegovina Canada Chile Denmark
3b 3
6b
4 20 0-12 6 0-7'1 0-7d
6 25
4-21
Ob
Ob
3 0-4
11 0-29
Egypt Finland France Germany India Indonesia Ireland Italy The Netherlands New Zealand Norway Mexico Spain Sweden Thailand United Kingdom
C. jejuni and C. coli
<1-2 22
0'13'
6
0 2 1 4'13' 3
0-2 '1 <1-3 14 2-5 3 lb-2 1-3
United States Vietnam
24 6"/11'
35 0'15d 17'-26 25
1-5
4-9
0
0
2-1 1
Reference(s) Fernandez (2001) Sharma et al. (2003) Feierl et al. (2001, 2003); Feierl (2004) Vandenberg et al. (2003) Uzunovic-Kamberovic (2003) Gaudreau and Gilbert (2003); Gibreel et al. (2004) Fernandez (2001) Danish Zoonosis Centre (2004a; Engberg et al. (2004); Engberg (2004) Putnam et al. (2003) Rautelin et al. (2003) Mtgraud and Prouzet-Maulion (2004) Luber et al. (2003); Steinbruckner et al. (2001); Wagner et al. (2003) Jain et al. (2005) Tjaniadi et al. (2003) Lucey et al. (2002) Pezzotti et al. (2003) VANTURES (2003b) Goodchild et al. (2001) N O M I N O M - V E T (2004); Afset and Maeland (2001) Tuz-Dzib et al. (1999) Campos et al. (2001); Saenz et al. (2000) Osterlund et al. (2003); Ronner et al. (2004) Bodhidatta et al. (2002); Isenbarger et al. (2002) Campylobacter Sentinel Surveillance Scheme Collaborators (2002); Moore et al. (2001); Rao et al. (2005); Wickins et al. (2001) Centers for Disease Control and Prevention (2004); Gupta et al. (2004); Nachamkin et al. (2002) Isenbarger et al. (2002)
"Reprinted from Engberg et al., 2005. bIsolates exclusively from children. Isolates acquired domestically. dIsolates acquired abroad.
select for quinolone resistance. Resistance occurs naturally, but the selection and dissemination of resistance are inevitable results of antibiotic use. Fluoroquinolone use in humans can in itself lead to the emergence of quinolone-resistant Campylobacter in treated infections. At least five case-control studies have specifically addressed risk factors for quinolone Campylobacter infections in the United States, the United Kingdom, Denmark, and Canada (Campylobacter Sentinel Surveillance Scheme Collaborators, 2002; Engberg et al., 2004; Johnson et al., 2007; Kassenborg et al., 2004; Smith et al., 1999). Four of the five studies evaluated current or recent treatment with antimicrobials. An association between treatment with a fluoroquinolone before stoolspecimen collection and having a quinolone-resistant Campylobacter infection was only observed in the study by Smith and colleagues (1999). They showed that treatment with a fluoroquinolone before stool
culture accounted for a maximum of 15% of resistant isolates in Minnesota during 1996 and 1998. Quinolone use in humans is not the major selective force for quinolone resistance among Campylobacter spp. causing human infection. Foreign travel was identified as a risk factor in all five studies, and this is in agreement with recent surveillance data from a number of countries. The studies also show a significant difference in quinolone resistance rates between travel-related infections and domestically acquired infections, and they document the importance of stratifying susceptibility data by travel status (Engberg, 2006; Engberg et al., 2005). Travel-related infections from destinations with recognized high quinolone resistance in Campylobacter in poultry, and high risk of acquiring Campylobacter infections are unsurprisingly associated with significantly higher prevalence of quinolone resistance compared with infections acquired domesti-
CHAPTER 6
CLINICAL ASPECTS OF C. lETUNI AND C. COLI INFECTIONS
113
100
80 60 40
20
0 r W #
w
Figure 4. Trends in quinolone resistance rates among C. coli and C. jejuni isolates from humans from 11 countries, 1989 to 2006. The bars represent both nalidixic acid and fluoroquinolone resistance and are based on mean values of resistance from numerous reports. Year in parentheses is year of licensure for use in veterinary medicine in each country. Canada and the United States banned veterinary use of fluoroquinolones in 1997 and 2005, respectively. Updated and modified from Engberg et al. (2001). References therein plus the following: Bodhidatta et al. (2002); Boonmar et al. (2005); Centers for Disease Control and Prevention (2003, 2004, 2007); Danish Zoonosis Centre (2004a, 2004b, 2005, 2006, 2007); Engberg et al. (2004); Feierl (2004); Feierl et al. (2001, 2003, 2007); Gallay et al. (2007); Migraud and Prouzet-Maulion (2004); Pezzotti et al. (2003); Rautelin et al. (2003); Sanders et al. (2002); Tribble et al. (2007); VANTURES (2003, 2004, 2005); Wickins et al. (2001); Prouzet-Maulion and Megraud (personal communication); J. Engberg (unpublished data).
cally. The significantly lower prevalence of quinolone resistance among domestically acquired Campylobacter probably reflects less veterinary use of fluoroquinolones in these countries. For example, in Australia, where fluoroquinolones have not been licensed for use in food production animals and only cooked chicken products may be imported, little fluoroquinolone resistance has been found in domestically acquired human infections (Unicomb et al., 2006). In contrast, although foreign travel is associated with quinolone-resistant infections in the United States, the majority of quinolone-resistant infections are domestically acquired (Gupta et al., 2004; Kassenborg et al., 2004). Table 5 summarizes information from casecomparison studies evaluating the duration of illness in patients infected with quinolone-resistant Campylobacter strains versus quinolone-sensitive Campylo-
bacter strains (Campylobacter Sentinel Surveillance Scheme Collaborators, 2002; Engberg et al., 2004; Kassenborg et al., 2004; Smith et al., 1999). The recent study by Nelson et al. (2004) evaluated the duration of illness across a variety of analytical models, including a multivariable analysis of variance model, and identified a consistent correlation between quinolone resistance and prolonged duration of diarrhea. Although the results from these studies are not all statistically significant, the estimates all point in the same direction, and taken together, they suggest that patients infected with quinolone-resistant strains have a longer duration of illness. Whether patients with resistant infections may experience a longer duration of illness because the antibiotic provided to them simply does not work against resistant Campylobacter, whether it may be due to a possible coselection of virulence traits in re-
114
BLASER AND ENGBERG
Table 5 . Studies evaluating the duration of illness in patients infected with quinolone-resistant or quinolone-susceptible Campylobacter strains a Resistant Reference No. of patients
Smith et al. (1999) Neimann et al. (2001)b The Campylobacter Sentinel Surveillance Scheme Collaborators (2002) Domestically acquired infection Travel-related infection Engberg et al. (2004)” Nelson et al. (2004)d Model A Model B Model C
Sensitive
Duration of diarrhea (days)
No. of patients
Duration of diarrhea (days)
69 5
10 14
115 31
7 9
0.03 0.13
86
12.7 11.8 13.2
381
13.5 11.2 10.3
0.56 0.66 0.001
26 7 9
9 12 8
264 56 76
7 6 6
0.04 0.04 0.2
-
“Reprinted from Engberg et al., 2005. *Stratified by treatment, but not on antimicrobial agent used for treatment. ‘Analysis not stratified by treatment. dModel A, analysis of 290 persons who did not receive antidiarrheal medications; model B, analysis of 63 persons who did not receive antimicrobial agents or antidiarrheal medications; model C, analysis of 85 persons who received only fluoroquinolone antimicrobial agents.
sistant strains, or both remains to be fully determined. Tetracyclines and chloramphenicol are alternative antibiotics when resistance or some other reason excludes the former drugs, but up to 60% of strains may be resistant to tetracyclines. Amoxicillin plus clavulanic acid (but not sulbactam or tazobactam) appears to be effective (Hakanen et al., 2003; Lachance et al., 1993). Serious systemic infection should be treated with an arninoglycoside, such as gentamicin, or imipenem. A macrolide may still be needed to clear infection from the gut.
INDICATIONS FOR ANTIMICROBIAL THERAPY The categories of patients likely to benefit from antimicrobial therapy include those who remain acutely ill and show no signs of improvement at the time a bacteriological diagnosis is made, those who have unusually severe or complicated infection, those who have systemic infection, those who are immunosuppressed or have another predisposition to infection, and those who live in an institution or closed group where the risks of the spread of disease are high. In this last case, treatment is provided to elirninate the organism from the stools of infected patients. Erythromycin was used successfully in this way to control a persistent outbreak of Cumpylobacter infection in a day care nursery in Israel (Ashkenazi et al., 1987).
Acknowledgments. We thank Martin Skirrow for his contributions to this chapter in the second edition and for his many contributions over more than three decades of study to our understanding of the clinical nature of Campylobacter enteritis.
REFERENCES Afset, J. E., and J. A. Maeland. 2001. [Erythromycin and ciprofloxacin resistant Campylobacter jejuni]. Tidsskr. Nor Laegeforen. 121:2152-2154. Ampelas, M., C . Perez, J. Jourdan, B. Nalet, A. Raynaud, J. M. Emberger, and H. Michel. 1982. [Hepatitis caused by Campylobacter coli (author’s transl.)]. Now. Presse Med. 11593-595. Anders, B. J., B. A. Lauer, and J. W. Paisley. 1981. Carnpylobacter gastroenteritis in neonates. Am. J. Dis. Child 135:900-902. Anders, B. J., B. A. Lauer, J. W. Paisley, and L. B. Reller. 1982. Double-blind placebo controlled trial of erythromycin for treatment of Campylobacter enteritis. Lancet i:13 1-132. Andrews, P. I., G. Kainer, L. C. Yong, V. H. Tobias, and A. R. Rosenberg. 1989. Glomerulonephritis, pulmonary hemorrhage and anemia associated with Campylobacter jejuni infection. Aust. N.Z. J. Med. 19:721-723. Ashkenazi, S., Y. Danziger, Y. Varsano, J. Peilan, and M. Mimouni. 1987. Treatment of Campylobacter gastroenteritis. Arch. Dis. Child. 62:84-85. Ashworth, J., and J. S. C. Engblish. 1984. Recurrent erythema nodosom and prolonged Campylobacter jejuni excretion. BMJ 288:830. Autenrieth, I. B., V. Schuster, J. Ewald, D. Harmsen, and H. W. Kreth. 1996. An unusual case of refractory Campylobacter jejuni infection in a patient with X-linked agammaglobulinemia: successful combined therapy with maternal plasma and ciprofloxacin. Clin. Infect. Dis. 23526-531. Autenrieth, I. B., A. Schwarzkopf, J. H. Ewald, H. Karch, and R. Lissner. 1995. Bactericidal properties of Campylobacter jejunispecific immunoglobulin M antibodies in commercial immunoglobulin preparations. Antimicrob. Agents Chemother. 39: 19651969.
CHAPTER 6
CLINICAL ASPECTS OF C. 7E7u1vI AND C. COLI INFECTIONS
Belongia, E. A., P. H. Chyou, R. T. Greenlee, G. Perez-Perez, W. F. Bibb, and E. 0. DeVries. 2003. Diarrhea incidence and farm-related risk factors for Escherichia coli 0157:H7 and Campylobacter jejuni antibodies among rural children. J. Infect. Dis. 1879460-1468. Bentley, D., J. Lynn, and J. W. Laws. 1985. Campylobactercolitis with intestinal aphthous ulceration mimicking obstruction. Br. Med. J. (Clin. Res. Ed.) 291:634. Bergman, M., A. Heikema, M. Kuijf, et al. 2007. Development of joint disease is associated with Campylobacter jejuni strains harboring lipo-oligosaccharide biosynthese gene cluster A. Zoonoses Public Health 54(Suppl. 1):19-155. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Black, R. E., D. Perlman, M. L. Clements, M. M. Levine, and M. J. Blaser. 1992. Human volunteer studies with Campylobacter jejuni, p. 207-215. In Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Blaser, M. J., I. D. Berkowitz, F. M. LaForce, J. Cravens, L. B. Reller, and W. L. Wang. 1979. Campylobacter enteritis: clinical and epidemiologic features. Ann. Intern. Med. 91:179-185. Blaser, M. J., and D. J. Duncan. 1984. Human serum antibody response to Campylobacter jejuni infection as measured in an enzyme-linked immunosorbent assay. Infect. Immun. 44:292298. Blaser, M. J., D. Hoverson, I. G. Ely, D. J. Duncan, W. L. Wang, and W. R. Brown. 1984a. Studies of Campylobacter jejuni in patients with inflammatory bowel disease. Gastroenterology 86: 33-38. Blaser, M. J., R. A. Miller, J. Lacher, and J. W. Singleton. 1984b. Patients with active Crohn's disease have elevated serum antibodies to antigens of seven enteric bacterial pathogens. Gastroenterology 87:88 8-894. Blaser, M. J., R. B. Parsons, and W. L. Wang. 1980. Acute colitis caused by Campylobacter fetus ss. jejuni. Gastroenterology 78: 448-453. Blaser, M. J., D. N. Taylor, and P. Echeverria. 1986. Immune response to Campylobacter jejuni in a rural community in Thailand. J. Infect. Dis. 153:249-254. Blaser, M. J., J. G. Wells, R. A. Feldman, R. A. Pollard, and J. R. Allen. 1983. Campylobacter enteritis in the United States. A multicenter study. Ann. Intern. Med. 98:360-365. Bodhidatta, L., N. Vithayasai, B. Eimpokalarp, C. Pitarangsi, 0. Serichantalergs, and D. W. Isenbarger. 2002. Bacterial enteric pathogens in children with acute dysentery in Thailand: increasing importance of quinolone-resistant Campylobacter. Southeast Asian J. Trop. Med. Public Health 33:752-757. Boonmar, S., L. Sangsuk, K. Suthivarakom, P. Padungtod, and Y. Morita. 2005. Serotypes and antimicrobial resistance of Campylobacter jejuni isolated from humans and animals in Thailand. Southeast Asian J. Trop. Med. Public Health 36:130-134. Borleffs, J. C., J. F. Schellekens, E. Brouwer, and M. RozenbergArska. 1993. Use of an immunoglobulin M containing preparation for treatment of two hypogammaglobulinemic patients with persistent Campylobacter jejuni infection. Eur. J. Clin. Microbiol. Infect. Dis. 12:772-775. Bradshaw, M. J., R. Brown, J. H. Swallow, and J. A. Rycroft. 1980. Campylobacter enteritis in Chelmsford. Postgrad. Med. J. 56:80-84. Bremell, T., A. Bjelle, and A. Svedhem. 1991. Rheumatic symptoms following an outbreak of Campylobacter enteritis: a five year follow up. Ann. Rheum. Dis. 50:934-938. Bretag, A. H., R. S. Archer, H. M. Atkinson, and W. H. Woods. 1984. Circadian urticaria: another Campylobacter association. Lancet i:954.
115
Campos, J. A. G., T. Alarcbn, M. MenCndez, and M. L6pez-Brea. 2001. Antibiotic resistance in thermophilic enteropathogenic Campylobacter spp. clinical isolates from children, p. 6. Abstr. 1l t h Int. Workshop Campylobacter Helicobacter Relat. Organisms, Freiburg, Germany, 1 to 5 September 2001. Campylobacter Sentinel Surveillance Scheme Collaborators. 2002. Ciprofloxacin resistance in Campylobacter jejuni: case-case analysis as a tool for elucidating risks at home and abroad. J. Antimicrob. Chemother. 50561-5 68. Carter, J. E., and N. Cimolai. 1991. IgA nephropathy associated with Campylobacter jejuni enteritis. Nephron 58:lOl-102. Castilla, L. M., M. Castro, and P. Guerrero. 1989. Acute pancreatitis associated with Campylobacter enteritis. Dig. Dis. Sci. 34: 961-962. Centers for Disease Control and Prevention. 2003. National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS): 2001 Annual Report. Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention. 2004. National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS): 2002 Annual Report. Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention. 2007. National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS): Human Isolates Final Report, 2004. Centers for Disease Control and Prevention, Atlanta, GA. http: //www.cdc.gov
/narms/NARMSAnnualReport2004.pdf. Chamovitz, B. N., A. I. Hartstein, S. R Alexander, A. B. Terry, P. Short, and R. Katon. 1983. Campylobacter jejuni-associated hemolytic-uremic syndrome in a mother and daughter. Pediatrics 71:253-256. Chan, F. T., G. Stringel, and A. M. Mackenzie. 1983. Isolation of Campylobacter jejuni from an appendix. J. Clin. Microbiol. 18: 422-424. Chu, Y. W., M. Y. Chu, K. Y. Luey, Y. W. Ngan, K. L. Tsang, and K. M. Kam. 2004. Genetic relatedness and quinolone resistance of Campylobacter jejuni strains isolated in 2002 in Hong Kong. J. Clin. Microbiol. 42:3321-3323. Coffin, C. M., P. L'Heureaux, and L. P. Dehner. 1982. Campylobacter-associated enterocolitis in childhood. Report of a fatal case. Am. J. Clin. Pathol. 78:117-123. Colgan, T., J. R. Lambert, A. Newman, and S. C. Luk. 1980. Campylobacter jejuni enterocolitis. A clinicopathologic study. Arch. Pathol. Lab. Med. 104571-574. Damani, N. N., C. A. Humphrey, and B. Bell. 1993. Haemolytic anaemia in Campylobacter enteritis. J. Infect. 26: 109-1 10. Danish Zoonosis Centre. 2004a. Annual Report on Zoonoses in Denmark 2003. Danish Zoonosis Centre, Ministry of Food, Agriculture and Fisheries, Saborg, Denmark. Danish Zoonosis Centre. 2004b. DANMAP 2003. Use of Antimicrobial Agents and Occurrance of Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in Denmark. Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), Copenhagen, Denmark. http: // www.food.dtu.dk / Admin / Public / DWSDownload.aspx?File = Files%2fFiler %2fZoonosecentret%2fPublikationer%2fDanmap %2fDanmap-2004.pdf. Danish Zoonosis Centre. 2005. DANMAP 2004. Use ofdntimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in Denmark. Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), Copenhagen, Denmark. http: //www.food.dtu.dk/Admin/Public/DWSDownload.aspx? File= Files%2fFiler%2fZoonosecentret%2fPublikationer%2fDanmap %2fDanmap-2005.pdf.
116
BLASER AND ENGBERG
Danish Zoonosis Centre. 2006. D A N I " 2005. Use ofdntimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in Denmark. Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), Copenhagen, Denmark. http: //www.food.dtu.dk/Admin/Public/DWSDownload.aspx?File= Files%2fFiler%2fZoonosecentret%2fPublikationer%2~anmap %2fDanmap-2005.pdf. Danish Zoonosis Centre. 2007. DANMAP 2006. Consumption of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria from Food Animals, Food and Humans in Denmark. Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), Copenhagen, Denmark. http: // www.food.dtu.dk/ Admin / Public / DWSDownload.aspx ? File = Files%2fFiler%2fZoonosecentret%2fPublikationer%2fDanmap %2fDanmap-2006.pdf. Darling, W. M., R. N. Peel, M. B. Skirrow, and A. E. Mulira. 1979. Campylobacter cholecystitis. Lancet i:1302. Davies, J. S., and J. B. Penfold. 1979. Campylobacter urinary infection. Lancet i:1091-1092. De Bois, M. H. W., M. C. Shoemaker, S. D. J. vander Werf, and J. B. C. M. Polyeart. 1989. Pancreatitis associated with Campylobacter jejuni infection: diagnostics by ultrasonography. BM] 298: 1004. Delans, R. J., J. D. Biuso, S. R. Saba, and G. Ramirez. 1984. Hemolytic uremic syndrome after Campylobacter-induced diarrhea in an adult. Arch. Intern. Med. 144:1074-1076. Denneberg, T., M. Friedberg, L. Holmberg, C. Mathiasen, K. 0. Nilsson, R. Takolander, and M. Walder. 1982. Combined plasmapheresis and hemodialysis treatment for severe hemolyticuremic syndrome following Campylobacter colitis.Acta Paediatr. Scand. 71:243-245. Denton, K. J., and T. Clarke. 1992. Role of Campylobacter jejuni as a placental pathogen.]. Clin. Pathol. 45:171-172. DeWitt, T. G., K. F. Humphrey, and G. V. Doern. 1985. White blood cell counts in patients with Campylobacter-induced diarrhea and in controls. J. Infect. Dis. 152:427-428. Dickgiesser, A. 1983. [Campylobacter infection and the hemolyticuremic syndrome]. Immun. Infect. 11:71-74. Doberneck, R. C. 1983. Campylobacter colitis mimicking colonic cancer during barium enema examination. Surgery 93508-509. Domingo, P., B. Mirelis, A. Gimeno, and R. Cabezas. 1985. [Spontaneous peritonitis caused by Campylobacter jejuni in a cirrhotic patient]. Med. Clin. (Barc.) 84:416. Drake, A. A., M. J. Gilchrist, J. A. Washington, K. A. Huizenga, and R. E. Van Scoy. 1981. Diarrhea due to Carnpylobacterfetus subspecies jejuni. A clinical review of 63 cases. Mayo Clin. Proc. 56:414-423. Drion, S., C. Wahlen, and P. Taziaux. 1988. Isolation of Campylobacter jejuni from the bile of a cholecystic patient. ]. Clin. Microbiol. 26:2 193-21 94. Dryden, M. S., R. J. Gabb, and S. K. Wright. 1996. Empirical treatment of severe acute community-acquired gastroenteritis with ciprofloxacin. Clin. Infect. Dis. 22: 1019-1025. Dunlop, S. P., D. Jenkins, K. R. Neal, and R. C. Spiller. 2003. Relative importance of enterochromaffin cell hyperplasia, anxiety, and depression in postinfectious IBS. Gastroenterology 125: 165 1-1 659. Dutronc, Y., M. Duong, M. Buisson, P. Chavanet, and H. Porter. 1995. Acute pancratitis and Campylobacter jejuni infection. Med. Mal. Infect. 25:1168-1169. Eastmond, C. J., J. A. Rennie, and T. M. Reid. 1983. An outbreak of Campylobacter enteritis-a rheumatological follow-up survey. 1.Rheumatol. 10:107-108. Ebright, J. R., and L. M. Ryan. 1984. Acute erosive reactive arthritis associated with Campylobacter jejuni-induced colitis.Am. J. Med. 76:321-323.
Engberg, J. 2006. Contributions to the epidemiology of Campylobacter infections. A review of clinical and microbiological studies. Dan. Med. Bull. 53:361-389. Engberg, J., F. M. Aarestrup, D. E. Taylor, P. Gerner-Smidt, and I. Nachamkin. 2001. Quinolone and macrolide resistance in Campylobacter jejuni and C. coli: resistance mechanisms and trends in human isolates. Emerg. Infect. Dis. 7:24-34. Engberg, J., P. Gerner-Smidt, F. Scheutz, E. M. Nielsen, S. L. W. On, and K. Mcrlbak. 1998. Water-borne Campylobacter jejuni infection in a Danish town-a 6-week continuous source outbreak. Clin. Microbiol. Infect. 4:648-656. Engberg, J., M. Keelan, P. Gerner-Smidt, and D. E. Taylor. 2005. Antimicrobial resistance in Campylobacter. In F. M. Aarestrup (ed.), Antimicrobial Resistance in Bacteria of Animal Origin. ASM Press, Washington, DC. Engberg, J., J. Neimann, E. M. Nielsen, F. M. Aarestrup, and V. Fussing. 2004. Quinolone-resistant Campylobacter infections: risk factors and clinical consequences. Emerg. Infect. Dis. 10: 1056-1 063. Ezpeleta, C., P. R. de Ursua, F. Obregon, F. Goni, and R. Cisterna. 1992. Acute pancreatitis associated with Campylobacter jejuni bacteremia. Clin. Infect. Dis. 15:1050. Fahey, T., D. Morgan, C. Gunneburg, G. K. Adak, F. Majid, and E. Kaczmarski. 1995. An outbreak of Campylobacter jejuni enteritis associated with failed milk pasteurisation. ]. Infect. 3 1: 137-143. Feder, H. M., Jr., M. Rasoulpour, and A. J. Rodriguez. 1986. Campylobacter urinary tract infection. Value of the urine Gram's stain. JAMA 256:2389. Feierl, G. 2004. Jahresbericht 2003 der Nationalen Referenzzentrale fur Campylobacter, Mitteilungen der Sanitatsverwaltung,p. 3-7. Institute of Hygiene, Graz, Austria. Feierl, G., C. Berghold, J. Posch, G. Gorkiewicz, E. Daghofer, and E. Marth. 2003. Epidemiology, clinical features and therapy of campylobacteriosis in Styria, Austria. Int. 1. Med. Microbiol. 293(Suppl. 35):134. Feierl, G., A. Grisold, U. Wagner, A. Baduara, G. Gorkiewicz, E. Leitner, and E. Marth. 2007. Epidemiology of campylobacteriosis and trends of resistance in Austria. Zoonoses Public Health 54(Suppl. 1):34. Feierl, G., U. Wagner, B. Sixl, A. Grisold, E. Daghofer, and E. Marth. 2001. Epidemiology of campylobacteriosis and development of resistance in Styria, Austria, p. 9. Abstr. 11th Int. Workshop Campylobacter Helicobacter Related Organisms, Freiburg, Germany, 1 to 5 September 2001. Fernandez, H. 2001. Emergence of antimicrobial resistance in Campylobacter: the consequences for incidence, clinical course, epidemiology and control, p. 67-72. In The Increasing Incidence of Human Campylobacteriosis. Report and Proceedings of a WHO Consultation of Experts. Copenhagen, Denmark, 21-25 November 2000. Department of Communicable Disease Surveillance and Response, World Health Organization, Geneva. Fitzgeorge, R. B., A. Baskerville, and K. P. Lander. 1981. Experimental infection of rhesus monkeys with a human strain of Campylobacter jejuni. J. Hyg. (Lond.) 86:343-351. Florkowski, C. M., R. B. Ikram, I. M. Crozier, H. Ikram, and M. E. Berry. 1984. Campylobacter jejuni myocarditis. Clin. Cardiol. 7558-559. Frizella, F. A., and J. A. Rietveld. 1993. Spontaneous splenic rupture associated with Campylobacter jejuni infection. Br. 1.Surg. 81~716-718. Gallagher, P., P. Chadwick, D. M. Jones, and L. Turner. 1981. Acute pancreatitis associated with Campylobacter infection. Br. J. Surg. 68:383. Gallay, A., V. Prouzet-Maulion, I. kempf, P. Lehours, L. Labadi, H. Gewenaelle, C. Camou, E. EspiC, H. De Valk, J. C. Desen-
CHAPTER 6
CLINICAL ASPECTS OF C. lETUNI AND C. COLI INFECTIONS
clos, and F. MCgraud. 2007. Recent trends of Campylobacter resistance to antibiotic among humans, broilers and pigs in France, 1999-2005. Zoonoses Publ. Health 54(Suppl. 1):45. Gaudreau, C., and H. Gilbert. 2003. Antimicrobial resistance of Campylobacter jejuni subsp. jejuni strains isolated from humans in 1998 to 2001 in Montreal, Canada. Antimicrob. Agents Chemotber. 47:2027-2029. Gerritsen van der, H. A., and E. M. Veringa. 1993. Cholecystitis caused by Campylobacter jejuni. Clin. Infect. Dis. 17:133. Gibreel, A., D. M. Tracz, L. Nonaka, T. M. Ngo, S. R. Connell, and D. E. Taylor. 2004. Incidence of antibiotic resistance in Campylobacter jejuni isolated in Alberta, Canada, from 1999 to 2002, with special reference to tet(0)-mediated tetracycline resistance. Antimicrob. Agents Chemother. 48:3442-3450. Goodchild, C., B. Dove, D. Riley, and A. J. Morris. 2001. Antimicrobial susceptibility of Campylobacter species. N.Z. Med. J. 114560-561. Goodman, L. J., G. M. Trenholme, R L. Kaplan, et al. 1990. Empiric antimicrobial therapy of domestically acquired acute diarrhea in urban adults. Arch. Intern. Med. 150541-546. Goodman, M. J., K. W. Pearson, D. McGhie, S. Dutt, and S. G. Deodhar. 1980. Campylobacter and Giardia lamblia causing exacerbation of inflammatory bowel disease. Lancet ii: 1247. Goossens, H., G. Henocque, L. Kremp, J. Rocque, R. Boury, G. Alaino, L. Vlaes, W. Hemelhof, C. Van den Borre, M. Macart, and J. P. Butzler. 1986. Nosocomial outbreak of Campylobacter jejuni meningitis in newborn infants. Lancet ii:146-149. Gould, S. R. 1985. Toxic megacolon complicating Campylobacter colitis. BM] 291:1580. Guandalini, S., S. Cucchiara, G. De Ritis, G. Capano, A. Caprioli, V. Falbo, V. Giraldi, A. Guarino, P. Vairano, and A. Rubino. 1983. Campylobacter colitis in infants. J. Pediatr. 102:72-74. Gupta, A., J. M. Nelson, T. J. Barrett, R. V. Tauxe, S. P. Rossiter, C. R. Friedman, K. W. Joyce, K. E. Smith, T. F. Jones, M. A. Hawkins, B. Shiferaw, J. L. Beebe, D. J. Vugia, T. RabatskyEhr, J. A. Benson, T. P. Root, and F. J. Angulo for the NARMS Working Group. 2004. Antimicrobial resistance among Campylobacter strains, United States, 1997-2001. Emerg. Infect. Dis. 10:1102-1 109. Hakanen, A. J., M. Lehtopolku, A. Siitonen, P. Huovinen, and P. Kotilainen. 2003. Multidrug resistance in Campylobacter jejuni strains collected from Finnish patients during 1995-2000.1. Antimicrob. Chemother. 52: 1035-1039. Haq, J. A., K. M. Rahman, and M. S. Akbar. 1985. Haemolyticuraemic syndrome and Campylobacter. Med. J. Aust. 142:662663. Hershkowici, S., M. Barak, A. Cohen, and J. Montag. 1987. An outbreak of Campylobacter jejuni infection in a neonatal intensive care unit. J. Hosp. Infect. 954-59. Heyman, M. B., V. I. Paterno, and M. E. Ament. 1982. Campylobacter colitis-a cause of chronic diarrhea in children. West. 1.Med. 137:243-245. Ho, H., M. J. Zuckerman, and S. M. Polly. 1987. Spontaneous bacterial peritonitis due to Campylobacter coli. Gastroenterology 92~2024-2025. Hoskins, T., and J. R. Davies. 1983. Clinical features of a large outbreak of milk-borne Campylobacter enteritis in a boys’ boarding school, p, 14. In A. D. Pearson, M. B. Skirrow, B. Rowe, J. R. Davies, and D. M. Jones (ed.), Campylobacter II. Proceedings of the Second International Workshop on Campylobacter Infections. Public Health Laboratory Service, London, United Kindgom. Howard, R. S., N. J. Sarkies, and M. D. Sanders. 1987. Anterior uveitis associated with Campylobacter jejuni infection. 1.Infect. 14~186-187.
117
Humphrey, K. S . 1993. Campylobacter infection and hepatocelMar injury. Lancet 341:49. Isenbarger, D. W., C. W. Hoge, A. Srijan, C. Pitarangsi, N. Vithayasai, L. Bodhidatta, K. W. Hickey, and P. D. Cam. 2002. Comparative antibiotic resistance of diarrheal pathogens from Vietnam and Thailand, 1996-1999. Emerg. Infect. Dis. 8:175180. Itoh, T., Y. Yanagawa, A. Kai, M. Takahashi, M. Inaba, I. Takano, S. Sakai, and M. Ohashi. 1983. Epidemiological and bacteriological studies on fifteen outbreaks of Campylobacter enteritis in Tokyo.]. Jpn. Assn. Infect. Dis. 57576-586. Jackson, C. J., A. J. Fox, D. R. Wareing, E. M. Sutcliffe, and D. M. Jones. 1997a. Genotype analysis of human blood isolates of Campylobacter jejuni in England and Wales. Epidemiol. Infect. 118:8 1-89. Jackson, N., M. Zaki, A. R. Rahman, M. Nazim, M. N. Win, and S. Osman. 1997b. Fatal Campylobacter jejuni infection in a patient splenectomised for thalassaemia. ]. Clin. Patbol. 50:436437. Jain, D., S. Sinha, K. N. Prasad, and C. M. Pandey. 2005. Campylobacter species and drug resistance in a north Indian rural community. Trans. R. SOC. Trop. Med. Hyg. 99:207-214. Johnson, J. Y. M., L. M. McMullen, P. Hasselback, M. Louie, G . Jhangri, and L. D. Saunders. 2007. Risk factors for ciprofloxacin resistance in reported Campylobacter infections in southern Alberta. Epidemiol. Infect. 2007(3 August): 1-10. Jones, P. H., A. T. Willis, D. A. Robinson, M. B. Skirrow, and D. S. Josephs. 1981. Campylobacter enteritis associated with the consumption of free school milk. 1.Hyg. (Lond.) 87:155-162. Kapperud, G., J. Lassen, S. M. Ostroff, and S. Aasen. 1992. Clinical features of sporadic Campylobacter infections in Norway. Scand. J. Infect. Dis. 24:741-749. Karmali, M. A., and P. C. Fleming. 1979. Campylobacter enteritis in children. J. Pediatr. 945277533. Karmali, M. A., B. Norrish, H. Lior, B. Heyes, A. Monteath, and H. Montgomery. 1984. Campylobacter enterocolitis in a neonatal nursery. 1.Infect. Dis. 149:874-877. Kassenborg, H. D., K. E. Smith, D. J. Vugia T. Rabatsky-Ehr, M. R. Bates, M. A. Carter, N. B. Dumas, M. P. Cassidy, N. Marano, R. V. Tauxe, and F. J. Angulo for the Emerging Infections Program FoodNet Working Group. 2004. Fluoroquinolone-resistant Campylobacter infections: eating poultry outside of the home and foreign travel are risk factors. Clin. Infect. Dis. 38(Suppl. 3):S279-S284. Kell, R. J. A., and M. E. Ellis. 1985. Transient atrial fibrillation in Campylobacter jejuni infection. BMJ 291:1542. Kerstens, P. J., H. P. Endtz, J. F. Meis, W. J. Oyen, R. J. Koopman, P. J. van den Broek, and J. W. van der Meer. 1992. Erysipelaslike skin lesions associated with Campylobacter jejuni septicemia in patients with hypogammaglobulinemia. Eur. J. Clin. Microbiol. Infect. Dis. 11:842-847. Kita, E., F. Nishikawa, N. Kamikaidou, A. Nakano, N. Katsui, and S . Kashiba. 1992. Mononuclear cell response in the liver of mice infected with hepatotoxigenic Campylobacter jejuni. J. Med. Microbiol. 37:326-331. Korman, T. M., C. C. Varley, and D. W. Spelman. 1997. Acute hepatitis associated with Campylobacter jejuni bacteraemia. Eur. 1. Clin. Microbiol. Infect. Dis. 16:678-681. Krajden, S., C. J. Burul, and M. Fuksa. 1986. Campylobacter jejuni associated with a perirectal abscess. Can. J. Surg. 29:228. Lachance, N., C. Gaudreau, F. Lamothe, and F. Turgeon. 1993. Susceptibilitiesof beta-lactamase-positive and -negative strains of Campylobacter coli to beta-lactam agents. Antimicrob. Agents Cbemotber. 37:1174-1176. Lambert, M., E. Marion, E. Coche, and J. P. Butzler. 1982. Campylobacter enteritis and erythema nodosum. Lancet i:1409.
118
BLASER AND ENGBERG
Lambert, M. E., P. F. Schofield, A. G. Ironside, and B. K. Mandal. 1979. Campylobacter colitis. Br. Med. J. 1:857-859. Leung, F. Y., G. 0. Littlejohn, and C. Bombardier. 1980. Reiter’s syndrome after Campylobacter jejuni enteritis. Arthritis Rheum. 23:948-950. Lever, A. M., J. M. Dolby, A. D. Webster, and A. B. Price. 1984. Chronic Campylobacter colitis and uveitis in patient with hypogammaglobnlinaemia. Br. Med. J. (Clin. Res. Ed.) 288:53 1. Locht, H., and K. A. Krogfelt. 2002. Comparison of rheumatological and gastrointestinal symptoms after infection with Campylobacter jejunitcoli and enterotoxigenic Escherichia coli. Ann. Rheum. Dis. 61:448-452. Lopez-Brea, M., P. M. Fontelos, M. Baquero, and L. Aragon. 1984. Urticaria associated with Campylobacter enteritis. Lancet i:1354. Loss, R. W., Jr., J. C. Mangla, and M. Pereira. 1980. Campylobacter colitis presenting as inflammatory bowel disease with segmental colonic ulcerations. Gastroenterology 79: 138-140. Louwen, R. P. L., A. Van Belkum, J. A. Wagenaar, Y. Doorduyn, R. Achterberg, and H. P. Endtz. 2006. Lack of association between the presence of the pVir plasmid and bloody diarrhea in Campylobacter jejuni enteritis. J. Clin. Microbiol. 44:18671868. mber, P., J. Wagner, H. Hahn, and E. Bartelt. 2003. Antimicrobial resistance in Campylobacter jejuni and Campylobacter coli strains isolated in 1991 and 2001-2002 from poultry and humans in Berlin, Germany. Antimicrob. Agents Chemother. 47: 3 825-3 830. .ucey, B., B. Cryan, F. O’Halloran, P. G. Wall, T. Buckley, and S. Fanning. 2002. Trends in antimicrobial susceptibility among isolates of Campylobacter species in Ireland and the emergence of resistance to ciprofloxacin. Vet. Rec. 151:317-320. Maidment, C. G., D. B. Evans, R. A. Coulden, and S. Thiru. 1985. Campylobacter enteritis complicated by glomerulonephritis. J. Infect. 10:177-178. Mandal, B. K., M. E. Ellis, E. M. Dunbar, and K. Whale. 1984. Double-blind placebo-controlled trial of erythromycin in the treatment of clinical Campylobacter infection. J. Antimicrob. Chemother. 13:619-623. Marshall, J. K., M. Thabane, A. X. Garg, W. F. Clark, M. Salvadori, and S. M. Collins. 2006. Incidence and epidemiology of irritable bowel syndrome after a large waterborne outbreak of bacterial dysentery. Gastroenterology 131:445-450. Matsuaki, S., and A. Katayama. 1984. Studies of outbreaks of food poisoning due to Campylobacter jejuni between 1980 and 1982 in Yamaguchi prefecture, Japan. Yamaguchi J. Vet. Med. 115356. Mattila, L., H. Peltola, A. Siitonen, H. Kyronseppa, I. Simula, and M. Kataja. 1993. Short-term treatment of traveler’s diarrhea with norfloxacin: a double-blind, placebo-controlled study during two seasons. Clin. Infect. Dis. 17:779-782. May, T., A. Gerard, P. Voiriot, J. L. Schmit, C. Lion, and P. Canton. 1986. [Campylobacter jejuni enteritis associated with hemolytic-uremic syndrome]. Presse Med. 15:803-804. McKendrick, M. W., A. M. Geddes, and J. Gearty. 1982. Campylobacter enteritis: a study of clinical features and rectal mucosal changes. Scand. J. Infect. Dis. 14:35-38. McKinley, M. J., M. Taylor, and M. H. Sangree. 1980. Toxic megacolon with Campylobacter colitis. Conn. Med. 44:496-497. McNaughton, R. D., R. Leyland, and L. Mueller. 1982. Outbreak of Campylobacter enteritis due to consumption of raw milk. Can. Med. Assoc. J. 126:657-658. McNeil, N. I., S. Buttoo, and G. L. Ridgway. 1984. Spontaneous bacterial peritonitis due to Campylobacter jejuni. Postgrad. Med. J. 60:487-488.
Mee, A. S., M. Shield, and M. Burke. 1985. Campylobactercolitis: differentiation from acute inflammatory bowel disease. J. R. Soc. Med. 78:2 17-223. Mtgraud, F., and V. Prouzet-Maulton. 2004. kvolution de la rCsistance des campylobacters aux antibiotiques en France (19862002), p. 156-158. Institut de Veille Sanitaire, Saint-Maurice, France. Mtgraud, F., C. Tachoire, J. Latrille, and J. M. Bondonny. 1982. Appendicitis due to Campylobacter jejuni. Br. Med. J. (Clin. Res. Ed.) 285~1165-1166. Melby, K., 0. P. Dahl, L. Crisp, and J. L. Penner. 1990. Clinical and serological manifestations in patients during a waterborne epidemic due to Campylobacter jejuni. J. Infect. 21:309-3 16. Melby, K., and S. Kildebo. 1988. Antibodies against Campylobacter jejunilcoli in patients suffering from campylobacteriosis or inflammatory bowel disease. NIPH Ann. 11:47-52. Menck, H. 1981. [Campylobacter jejuni enteritis complicated by glomerulonephritis]. Ugeskr. Laeger 143: 1020-1021. Mentzing, L. 0. 198 1. Waterborne outbreaks of Campylobacter enteritis in central Sweden. Lancet ii:352-354. Mertens, A., and S. M. De. 1979. Campylobacter cholecystitis. Lancet i:1092-1093. Meuwissen, S. G. M., P. J. M. Bakkar, and P. J. G. M. Rietra. 198 1. Acute ulceration of ileal stoma due to Campylobacter fetus subspecies jejuni. BMJ 282:1362. Meyrieux, V., G. Monneret, A. Lepape, M. Chomarat, and V. Banssillon. 1996. Fatal septic shock with multiple organ failure due to Campylobacter jejuni. Clin. Infect. Dis. 22:183-184. Michalak, D. M., J. Perrault, M. J. Gilchrist, R. R. Dozois, J. A. Carney, and P. F. Sheedy. 1980. Campylobacter fetus ss. jejuni: a cause of massive lower gastrointestinal hemorrhage. Gastroenterology 79:742-745. Millson, M., M. Bokhout, J. Carlson, L. Spielberg, R. Aldis, A. Borczyk, and H. Lior. 1991. An outbreak of Campylobacterjejuni gastroenteritis linked to meltwater contamination of a municipal well. Can. J. Public Health 82:27-31. Molina, J., I. Casin, P. Hausfater, E. Giretti, Y. Welker, J. Decazes, V. Garrait, P. Lagrange, and J. Modai. 1995. Campylobacter infections in HIV-infected patients: clinical and bacteriological features. AIDS 9:881-885. Moore, J. E., M. Crowe, N. Heaney, and E. Crothers. 2001. Antibiotic resistance in Campylobacter spp. isolated from human faeces (1980-2000) and foods (1997-2000) in Northern Ireland: an update. J. Antimicrob. Chemother. 48:455-457. Morlet, N., and R. Glancy. 1986. Campylobacter jejuni appendicitis. Med. J. Aust. 14556-57. Morton, A. R., R. Yu, S. Waldek, A. M. Holmes, A. Craig, and K. Mundy. 1985. Campylobacter induced thrombotic thrombocytopenic purpura. Lancet ii:1133-1134. Moscuna, M., Z. Gross, R. Korenblum, M. Volfson, and M. Oettinger. 1989. Septic abortion due to Campylobacter jejuni. Eur. J. Clin. Microbiol. Infect. Dis. 8:800-801. Murphy, S., N. J. Beeching, S. J. Rogerson, and A. D. Harries. 1991. Pancreatitis associated with Salmonella enteritis. Lancet 338571. Muytjens, H. L., and J. Hoogenhout. 1982. Campylobacter jejuni isolated from a chest wall abscess. Clin. Microbiol. Newslett. 4: 166. Nachamkin, I., S. H. Fischer, X. H. Yang, 0. Benitez, and A. Cravioto. 1994. Immunoglobulin A antibodies directed against Campylobacter jejuni flagellin present in breast-milk. Epidemiol. Infect. 112:359-3 65. Nachamkin, I., H. Ung, and M. Li. 2002. Increasing fluoroquinolone resistance in Campylobacter jejuni, Pennsylvania, USA, 1982-2001. Emerg. Infect. Dis. 8:1501-1503.
CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
Nagaratnam, N., T. K. Goh, and D. Ghoughassian. 1990. Campylobacter jejuni-induced vasculitis. Br. J. Clin. Pract. 44:636637. Nahum, H. D., E. Kaloustian, P. Baumer, P. A. Felten, and J. Lubetski. 1982. [Biliary retention syndrome in Campylobacter jejuni septicemia]. N o w . Presse Med. 11:1805-1806. Neal, K. R., H. M. Scott, R. C. Slack, and R. F. Logan. 1996. Omeprazole as a risk factor for Campylobacter gastroenteritis: case-control study. BMJ 312:414-415. Neimann, J., K. Mdbak, J. Engberg, F. M. Aarestrup, and H. C. Wegener. 2001. Longer duration of illness among Campylobacter patients treated with fluoroquinolones. CHR02001. Abstr. 1lth Int. Workshop Campylobacter Helicobacter Relat. Organisms, Freiburg, Germany, 1 to 5 September 2001. Nelson, J. M., K. E. Smith, D. J. Vugia, T. Rabatsky-Ehr, S. D. Segler, H. D. Kassenborg, S. M. Zansky, K. Joyce, N. Marano, R. M. Hoekstra, and F. J. Angulo. 2004. Prolonged diarrhea due to ciprofloxacin-resistant Campylobacter infection. J. Infect. Dis. 190:1150-1157. Newman, A., and J. R. Lambert. 1980. Cumpylobacter jejuni causing flare-up in inflammatory bowel disease. Lancet ii:919. Noble, C. J., D. J. Hibbert, and G. J. Patel. 1982. Campylobacter colitis: a case with unusual radiological features.J. Infect. 5:199200. NORM/NORM-VET. 2004. NORM/NORM-VET 2003. Usage of antimicrobial agents and occurrence of antimicrobial resistance in Norway. Tromsn, Norway. Norrby, R., R. V. McCloskey, G. Zackrisson, and E. Falsen. 1980. Meningitis caused by Campylobacter fetus ssp jejuni. Br. Med. J. 280:1164. Osterlund, A., M. Hermann, and G. Kahlmeter. 2003. Antibiotic resistance among Campylobacter jejunilcoli strains acquired in Sweden and abroad: a longitudinal study. Scand. J. Infect. Dis. 3.5~478-481. Pai, C. H., F. Gillis, E. Tuomanen, and M. I. Marks. 1983. Erythromycin in treatment of Campylobacter enteritis in children. Am. J. Dis. Child 137:286-288. Pasticci, M. B., E. Baratta, F. A. Del, A. Gillio, F. Baldelli, and S. Pauluzzi. 1992. Campylobacter jejuni-an unusual cause of infectious arthritis. Postgrad. Med. J. 68:151-152. Panlet, P., and M. Coffernils. 1990. Very long term diarrhoea due to Campylobacter jejuni. Postgrad. Med. J. 66:410-411. Pearson, A. D., D. P. Drake, D. Brookfield, S. O’Connel, W. G. Suckling, M. J. Knill, E. Ware, and J. R. Knott. 1982. Cumpylobacter infections in patients presenting with diarrhoea, mesenteric adenitis and appendicitis. In D. G. Newell (ed.), Campylobacter: Epidemiology, Pathogenesis and Biochemistry, p. 147-151. MTP Press, Lancaster, United Kingdom. Pedler, S. J., and A. J. Bint. 1984. Osteitis of the foot due to Campylobacter jejuni. J. Infect. 8:84-85. Pentland, B. 1979. Campylobacter enteritis-an in-patient study. Scott. Med. J. 24:299-301. Pereira, M. S., S. D. Lipton, and J. K. Kim. 1981. Acute cholecystitis and Campylobacter fetus. Ann. Intern. Med. 94:821. Peterson, M. C. 1994. Rheumatic manifestations of Campylobacter jejuni and C. fetus infections in adults. Scand. J. Rheurnatol. 23: 167-170. Peterson, M. C., R. W. Farr, and M. Castiglia. 1993. Prosthetic hip infection and bacteremia due to Campylobacter jejuni in a patient with AIDS. Clin. Infect. Dis. 16:439-440. Pezzotti, G., A. Serafin, I. Lnzzi, R Mioni, M. Milan, and R. Perin. 2003. Occurrence and resistance to antibiotics of Campylobacter jejuni and Campylobacter coli in animals and meat in northeastern Italy. lnt. J. Food Microbiol. 82:281-287. Pichler, H. E. T., G. Diridl, K. Stickler, and D. Wolf. 1987. Clinical efficacy of ciprofloxacin compared with placebo in bacterial diarrhea. Am. J. Med. 82:329-332.
119
Pines, A., E. Goldhammer, J. Bregman, N. Kaplinsky, and 0. Frankl. 1983. Campylobacter enteritis associated with recurrent abortions in agammaglobulinemia.Acta Obstet. Gynecol. Scand. 62~279-280. Pitkanen, T., A. Ponka, T. Pettersson, and T. U. Kosunen. 1983. Campylobacter enteritis in 188 hospitalized patients. Arch. lntern. Med. 143:215-219. Ponka, A., and T. U. Kosunen. 1981. Pancreas affection in association with enteritis due to Campylobacter fetus ssp. jejuni, Acta Med. Scand. 209:239-240. Ponka, A., T. Pitkanen, T. Pettersson, S. Aittoniemi, and T. U. Kosunen. 1980. Carditis and arthritis associated with Campylobacter jejuni infection. Acta Med. Scand. 208:495-496. Pope, J. E., A. Krizova, A. X. Garg, H. Thiessen-Philbrook, and J. M. Ouimet. 2007. Campylobacter reactive arthritis: a systematic review. Semin. Arthritis Rheum. 117:237-257. Porter, I. A., and T. M. S. Reid. 1980. A milk-borne outbreak of Campylobacter infection. J. Hyg. (Lond.) 84:415-419. Potter, M. E., M. J. Blaser, R. K. Sikes, A. F. Kaufmann, and J. G. Wells. 1983. Human Campylobacter infection associated with certified raw milk. Am. J. Epidemiol. 117:475-483. Price, A. B., J. Jewkes, and P. J. Sanderson. 1979. Acute diarrhoea: Campylobacter colitis and the role of rectal biopsy. J. Clin. Pathol. 32:990-997. Putnam, S . D., R. W. Frenck, M. S. Riddle, A. El-Gendy, N. N. Taha, B. T. Pittner, R. Abu-Elyazeed, T. F. Wierzba, M. R. Rao, S. J. Savarino, and J. D. Clemens. 2003. Antimicrobial susceptibility trends in Campylobacter jejuni and Campylobacter coli isolated from a rural Egyptian pediatric population with diarrhea. Diag. Microbiol. Infect. Dis. 47:601-608. Puylaert, J. B., R. J. Vermeijden, S. D. van der Werf, L. Doornbos, and R. K. Koumans. 1989. Incidence and sonographic diagnosis of bacterial ileocaecitis masquerading as appendicitis. Lancet ii: 84-86. Rao, D., J. R. Rao, E. Crothers, R. McMullan, D. McDowell, A. McMahon, P. J. Rooney, B. C. Millar, and J. E. Moore. 2005. Increased erythromycin resistance in clinical Campylobacter in Northern Ireland-an update. J. Antimicrob. Chemother. 552. Rao, G. G., and M. Fuller. 1992. A review of hospitalized patients with bacterial gastroenteritis. J. Hosp. Infect. 20:105-111. Rautelin, H., A. Vierikko, M. L. Hanninen, and M. Vaara. 2003. Antimicrobial susceptibilities of Campylobacter strains isolated from Finnish subjects infected domestically or from those infected abroad. Antimicrob. Agents Chemother. 47:102-105. Rautelin, H. I., A. V. Outinen, and T. U. Kosunen. 1987. Tubulointerstitial nephritis as a complication of Cumpylobacter jejuni enteritis. A case report. Scand. J. Urol. Nephrol. 21:151-152. Reddy, K. R, and E. Thomas. 1982. Campylobacter jejuni enterocolitis and hepatitis. Gastroenterology 82: 1156. Ritchie, P. M., J. C. Forbes, and P. Steinbok. 1987. Subdural space Campylobacter infection in a child. CMAJ 137:45-46. Robinson, D. A. 1981. Infective dose of Cumpylobacter jejuni in milk. Br. Med. J. 282:1584. Rogol, M., I. Sechter, H. Falk, Y. Shtark, S. Alfi, Z. Greenberg, and R. Mizrachi. 1983. Waterborne outbreak of Carnpylobacter enteritis. Eur. J. Clin. Microbiol. 2588-590. Ronner, A. C., E. 0. Engvall, L. Andersson, and B. Kaijser. 2004. Species identification by genotyping and determination of antibiotic resistance in Campylobacter jejuni and Campylobacter coli from humans and chickens in Sweden. Int. J. Food Microbiol96: 173-179. Rusu, V., and S. Lucinescu. 1988. The value of Campylobacter jejunilcoli serotyping for the evaluation of an enteritis outbreak in newborn infants. p. 138-139. In B. Kaijser and E. Falsen (ed.), Campylobacter N. Proceedings of the Fourth International Work-
120
BLASER AND ENGBERG
shop on Campylobacter Infections. University of Goteborg, GOteborg, Sweden. Saari, K. M., and 0. Kauranen. 1980. Ocular inflammation in Reiter’s syndrome associated with Campylobacter jejuni enteritis. Am. J. Ophthalmol. 90572-573. Sacks, J. J., S. Lieb, L. M. Baldy, S. Berta, C. M. Patton, M. C. White, W. J. Bigler, and J. J. Witte. 1986. Epidemic campylobacteriosis associated with a community water supply. Am. J. Public Health 76:424-428. Saenz, Y., M. Zarazaga, M. Lantero, M. J. Gastanares, F. Baquero, and C. Torres. 2000. Antibiotic resistance in Campylobacter strains isolated from animals, foods, and humans in Spain in 1997-1998. Antimicrob. Agents Chemother. 44:267-271. Salazar Lindo, E., R. B. Sack, E. Chea Woo, B. A. Kay, Z. A. Piscoya, R. Leon Barua, and A. Yi. 1986. Early treatment with erythromycin of Campylobacter jejuni-associated dysentery in children. J. Pediatr. 109:355-360. Sanders, J. W., D. W. Isenbarger, S. E. Walz, L. W. Pang, D. A. Scott, C. Tamminga, B. A. Oyofo, W. C. Hewitson, J. L. Sanchez, C. Pitarangsi, P. Echeverria, and D. R. Tribble. 2002. An observational clinic-based study of diarrheal illness in deployed United States military personnel in Thailand: presentation and outcome of Campylobacter infection. Am. J. Trop. Med. Hyg. 67~533-538. Schaad, U. B. 1982. Reactive arthritis associated with Campylobacter enteritis. Pediatr. Infect. Dis. 1:328-332. Schiellerup, P. 2006. Reactive arthritis complication bacterial enteritis-a case-case comparison study. Thesis. Statens Serum Institut, Copenhagen, Denmark. Schieven, B. C., D. Baird, C. L. Leatherdale, and Z. Hussain. 1991. Campylobacter jejuni infected bursitis. Diagn. Microbiol. Infect. Dis. 14507-508. Schmidt, U., H. Chmel, Z. Kaminski, and P. Sen. 1980. The clinical spectrum of Campylobacter fetus infections: report of five cases and review of the literature. Q. J. Med. 49:431-442. Schmheyder, H. C., P. Scrgaard, and W. Frederiksen. 1995. A survey of Campylobacter bacteremia in three Danish counties, 1989 to 1994. Scand. J. Infect. Dis. 27:145-148. Sharma, H., L. Unicomb, W. Forbes, S. Djordjevic, M. Valcanis, C. Dalton, and J. Ferguson. 2003. Antibiotic resistance in Campylobacter jejuni isolated from humans in the Hunter Region, New South Wales. Commun. Dis. Intell. 27(Suppl.):80-88. Shulman, S. T., and D. Moel. 1983. Campylobacter infection. Pediatrics 72:437. Simor, A. E., and S. Ferro. 1990. Campylobacter jejuni infection occurring during pregnancy. Eur. J. Clin. Microbiol. Infect. Dis. 9:142-144. Simor, A. E., M. A. Karmali, T. Jadavji, and M. Roscoe. 1986. Abortion and perinatal sepsis associated with Campylobacter infection. Rev. Infect. Dis. 8:397-402. Skirrow, M. B. 1981. Acute ulceration of ileal stoma due to Campylobacter fetus subspecies jejuni. BMJ 110567-573. Skirrow, M. B., D. M. Jones, E. Sutcliffe, and J. Benjamin. 1993. Campylobacter bacteraemia in England and Wales, 198 1-91. Epidemiol. Infect. 110567-573. Smith, G. S., and M. J. Blaser. 1985. Fatalities associated with Campylobacter jejuni infections. JAMA 253:2873-2875. Smith, K. E., J. M. Besser, C. W. Hedberg, F. T. Leano, J. B. Bender, J. H. Wicklund, B. P. Johnson, K. A. Moore, and M. T. Osterholm. 1999. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992-1998. N.Engl. J. Med. 340:15251532. Sorvillo, F. J., L. E. Lieb, and S. H. Waterman. 1991. Incidence of campylobacteriosis among patients with AIDS in Los Angeles County. J. Acquir. Immune. Defic. Syndr. 4598-602.
Spiller, R. C. 2007. Role of infection in irritable bowel syndrome. J. Gastroenterol. 42(Suppl.):41-47. Stalder, H., R. Isler, W. Stutz, M. Salfinger, S. Lauwers, and W. Vischer. 1983. Outbreak of Campylobacter enteritis involving over 500 participants in a jogging rally, p. 143. In A. D. Pearson, M. B. Skirrow, B. Rowe, J. R. Davies, and D. M. Jones (ed.), Campylobacter 11, Proceedings of the Second International Workshop on Campylobacter Infections. Public Health Laboratory Service, London, United Kingdom. Stehr-Green, J., P. Mitchell, C. Nicholls, S. McEwan, and A. Payne. 1991. Campylobacter enteritis-New Zealand. Morb. Mortal Wkly. Rep. 40:116-117. Steinbruckner, B., F. Ruberg, M. Vetter-Knoll, and M. Kist. 2001. Antimicrobial susceptibility of Campylobacter jejuni and Campylobacter coli isolated in Freiburg from 1992-2000, p. 8. Abstr. 11th Int. Workshop Campylobacter Helicobacter Related Organisms,Freiburg, Germany, 1 to 5 September 2001. Stephenson, T. J., and D. W. K. Cotton. 1985. Toxic megacolon complicating Campylobacter colitis. BMJ 291:1242. Strid, M. A., J. Engberg, L. B. Larsen, K. Begtrup, K. Mdbak, and K. A. Krogfelt. 2001. Antibody responses to Campylobacter infections determined by an enzyme-linked immunosorbent assay: 2-year follow-up study of 210 patients. Clin. Diagn. Lab. Immunol. 8:314-319. Surawicz, C. M., and L. Belic. 1984. Rectal biopsy helps to distinguish acute self-limited colitis from idiopathic inflammatory bowel disease. Gastroenterology 86:104-113. Svedhem, A., and B. Kaijser. 1980. Campylobacter fetus subspecies jejuni: a common cause of diarrhea in Sweden. J. Infect. Dis. 142~353-359. Taylor, D. N., M. J. Blaser, P. Echeverria, C. Pitarangsi, L. Bodhidatta, and W. L. Wang. 1987. Erythromycin-resistant Campylobacter infections in Thailand. Antimicrob. Agents Chemother. 31:43 8-442. Taylor, D. N., B. W. Porter, C. A. Williams, H. G. Miller, C. A. Bopp, and P. A. Blake. 1982. Campylobacter enteritis: a large outbreak traced to commercial raw milk. West.J. Med. 137:365369. Tee, W., and A. Mijch. 1998. Campylobacter jejuni bacteremia in human immunodeficiency virus (H1V)-infected and non-HIVinfected patients: comparison of clinical features and review. Clin. Infect. Dis. 26:91-96. Terrier, A., M. Altwegg, P. Bader, and A. voncraevenitz. 1985. Hospital epidemic of neonatal Campylobacter jejuni infection. Lancet ii:1182. Tettmar, R. E., and E. J. Thornton. 1981. An outbreak of Campylobacter enteritis affecting an operational Royal Air Force unit. Public Health 95:69-73. Thornley, J. P., D. Jenkins, K. Neal, T. Wright, J. Brough, and R. C. Spiller. 2001. Relationship of Campylobacter toxigenicity in vitro to the development of postinfectious irritable bowel syndrome. J. Infect. Dis. 184:606-609. Tjaniadi, P., M. Lesmana, D. Subekti, N. Machpud, S. Komalarini, W. Santoso, C. H. Simanjuntak, N. Punjabi, J. R. Campbell, W. K. Alexander, H. J. Beecham 111, A. L. Corwin, and B. A. Oyofo. 2003. Antimicrobial resistance of bacterial pathogens associated with diarrheal patients in Indonesia. Am. J. Trop. Med. Hyg. 68~666-670. Torphy, D. E., and W. W. Bond. 1979. Campylobacter fetus infections in children. Pediatrics 64:898-903. Tracz, D. M., M. Keelan, J. Ahmed-Bentley, A. Gibreel, K. Kowalewska-Grochowska, and D. E. Taylor. 2005. pVir and bloody diarrhea in Campylobacter jejuni enteritis. Emerg. Infect. Dis. 11:838-843. Trihble, D. R., J. W. Sanders, L. W. Pang, C. Mason, C. Pitarangsi, S. Baqar, A. Armstrong, P. Hshieh, A. Fox, E. A. Maley, C.
CHAPTER 6
CLINICAL ASPECTS OF C. TETUNI AND C. COLI INFECTIONS
Lebron, D. J. Faix, J. V. Lawler, G. Nayak, M. Lewis, L. Bodhidatta, and D. A. Scott. 2007. Traveler's diarrhea in Thailand: randomized, double-blind trial comparing single-dose and 3-day azithromycin-based regimens with a 3-day levofloxacin regimen. Clin. Infect. Dis. 44:338-346. Tuz-Dzib, F., M. L. Guerrero, L. E. Cervantes, L. K. Pickering, and G. M. Ruiz-Palacios. 1999. Increased incidence of quinolone resistance among clinical isolates of Campylobacter jejuni in Mexico. Proceedings of the 10th International Workshop on Campylobacter, Helicobacter, and Related Organisms, Baltimore, MD. Unicomb, L. E., J. Ferguson, R. J. Stafford, R. Ashbolt, M. D. Kirk, N. G. Becker, M. S. Patel, G. L. Gilbert, M. Valcanis, and L. Mickan for the Australian Campylobacter Subtyping Study Group. 2006. Low-level fluoroquinolone resistance among Campylobacter jejuni isolates in Australia. Clin. Infect. Dis. 42:13681374. Uzunovic-Kamberovic, S. 2003. Antibiotic susceptibility of Campylobacter jejuni and Campylobacter coli human isolates from Bosnia and Herzegovina. J. Antimicrob. Chemother. 5 1:10491051. van Dijk, W. C., and P. J. van der Straaten. 1988. An outbreak of Campylobacter jejuni infection in a neonatal intensive care unit. J. Hosp. Infect. 11:91-92. van Spreeuwel, J. P., G. C. Duursma, C. J. Meijer, R. Bax, P. C. Rosekrans, and J. Lindeman. 1985. Campylobacter colitis: histological immunohistochemical and ultrastructural findings. Gut 26~945-951. Vandenberg, O., Y. Glupczynski, S. Ibekwem, K. Houf, A. Dediste, N. Douat, P. Retore, G. Zissis, and J. P. Butzler. 2003. Trends in antimicrobial susceptibility among isolates of Campylobacter species isolated from humans in 1996 to 2002 in Belgium. lnt. J. Med. Microbiol. 293(Suppl. 35):l-148. VANTURES, the Veterinary Antibiotic Usage and Resistance Surveillance Working Group. 2003. MARAN 2002. Monitoring of Antimicrobial Resistance and Antibiotic Usage in Animals in The Netherlands in 2002. http://www.cidc-lelystad.wur.nl/ NR/ rdonlyres/ 7F79ACE6-OFD2-41AB-8lB2-BB17FA89603Cl 11378 IMARAN2002web4.pdf. VANTURES, the Veterinary Antibiotic Usage and Resistance Surveillance Working Group. 2004. MARAN 2003. Monitoring of
121
Antimicrobial Resistance and Antibiotic Usage in Animals in The Netherlands in 2003. http://www.cidc-lelystad.wur.nl/ NR/ rdonlyres/ 7F79ACE6-OFD2-41AB-8lB2-BB17FA89603C/ 11381/MARAN2003webl.pdf. VANTURES, the Veterinary Antibiotic Usage and Resistance Sur"2004. Monitoring of veillance Working Group. 2005. i Antimicrobial Resistance and Antibiotic Usage in Animals in The Netherlands in 2004. http: //www.cidc-lelystad.wur.nl/ NR / rdonlyres / 7F79ACE6-OFD2-41AB-81B2-BB17FA89603C/ 11382/MARAN2004webl.pdf. Vogt, R. L., H. E. Sours, T. Barrett, R. A. Feldman, R. J. Dickinson, and L. Witherell. 1982. Campylobacter enteritis associated with contaminated water. Ann. Intern. Med. 96:292-296. Wagner, J., M. Jabbusch, M. Eisenblatter, H. Hahn, C. Wendt, and R. Ignatius. 2003. Susceptibilities of Campylobacter jejuni isolates from Germany to ciprofloxacin, moxifloxacin, erythromycin, clindamycin, and tetracycline. Antimicrob. Agents Chemother. 47:2358-236 1. Wang, W. L., and M. J. Blaser. 1986. Detection of pathogenic Campylobacter species in blood culture systems. J. Clin. Microbiol. 23:709-714. Webster, P. B., and D. J. Farrell. 1993. Campylobacter peritonitis as a complication of continuous ambulatory peritoneal dialysis. Aust. J. Med. Sci. 14:68-70. Wickins, H. V., R. Thwaites, and J. A. Frost. 2001. Drug resistance in Campylobacter jejuni and Campylobacter coli in England, and Wales 1993-2001. Abstr. 11th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Freiburg, Germany, 1 to 5 September 2001. Wilson, P. G., J. R Davies, T. W. Hoskins, K. P. Lander, H. Lior, D. M. Jones, and A. D. Pearson. 1983. Epidemiology of an outbreak of milk-borne enteritis in a residential school, p. 143. In A. D. Pearson, M. B. Skirrow, B. Rowe, J. R. Davies, and D. M. Jones (ed.), Campylobacter 11. Proceedings of the Second lnternational Workshop on Campylobacter Infections. Public Health Laboratory Service, London, United Kingdom. Wood, C. J., V. Fleming, J. Turnidge, N. Thomson, and R. C. Atkins. 1992. Campylobacter peritonitis in continuous ambulatory peritoneal dialysis: report of eight cases and a review of the literature. Am. J. Kidney Dis. 19:257-263. Youngs, E. R., and C. Roberts. 1985. Campylobacter carriage and pregnancy. Br. J. Obstet. Gynaecol. 9 2 5 4 1 4 4 2 .
Campylobuctcr, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 7
Clinical Significance of Campylobacter and Related Species Other Than Campylobacter jejuni and Campylobacter coli ALBERT J. LASTOVICA AND BAN MISHU ALLOS
subsp. fetus is now recognized as a major cause of septic abortion in domestic animals (Garcia et al., 1983). Since 1947, C. fetus has also been implicated as a causative agent of a range of human intestinal and extraintestinal illnesses. C. fetus is the type species of the genus Campylobacter (VCron and Chatelain, 1973), and it is separated into two subspecies, C. fetus subsp. fetus and C. fetus subsp. venerealis. This division stems from the realization that two distinct disease entities could be attributed to two varieties of strains (Florent, 1959), which were subsequently accorded subspecies status on the basis of DNA-DNA hybridization studies (Harvey and Greenwood, 1983). C. fetus subsp. fetus colonizes the intestine and causes sporadic abortion in sheep and cattle, usually late in gestation. C. fetus subsp. venerealis is adapted to the bovine genital tract and causes infertility by destroying the embryo early in gestation. This disease, known as bovine vibriosis, is of major concern to the cattle industry (Garcia et al., 1983). The principal habitat of C. fetus subsp. venerealis is the prepuce of asymptomatic bulls, which form a reservoir of infection for cows. The status of C. fetus subsp. venerealis as a human pathogen is apparently minimal, but because of the difficulty of separating subsp. venerealis from subsp. fetus by phenotypic methods (mainly growth in the presence of 1%glycine), uncertainty remains. Hum et al. (1997) developed a PCR assay for the identification and differentiation of the two subspecies, which, when compared with traditional phenotyping, indicated that strains are often misidentified. Strains confirmed as subsp. venerealis by genotyping methods have been isolated from the vagina of a woman with bacterial vaginosis in Sweden, and from two homosexual men in Australia (Holst et al., 1987;
Campylobacters are the most frequently identified bacteria causing diarrhea in humans, particularly in very young children, in both developed and developing countries (Allos, 2001). Campylobacters have also been associated with other clinical conditions such as bacteremia Guillain-BarrC syndrome, hemolytic-uremic syndrome, pancreatitis, and reactive arthritis. The genus Campylobacter currently comprises 17 species, of which 14 have been isolated from humans. Historically, more than 95% of Campylobacter strains isolated and identified in cases of human disease have been c. jejuni subsp. jejuni or coli. However, the isolation techniques currently used in many diagnostic laboratories may not support the growth of other, potentially pathogenic non-jejuni, non-coli Campylobacter species. These organisms may be fastidious, requiring special atmospheric and temperature conditions as well as prolonged incubation, or they may be unable to tolerate the antibiotics commonly included in selective media plates. The disease potential of these non-jejuni, non-coli Campylobacter species is beginning to be appreciated, particularly in areas where they are being looked for. This chapter will describe the microbiology, epidemiology, and clinical features of infection with Campylobacter species other than C. jejuni subsp. jejuni and C. coli that are associated with human disease.
c.
CAMPYLOBACTER FETUS McFaydean and Stockman (1913) first recognized campylobacters as a causative agent of fetal infection and abortion in sheep. In 1919, these organisms, then called Vibrio fetus, were reported to cause abortion in cattle (Smith, 1919). Campylobacter fetus
Albert J. Lastovica Department of Biotechnology, University of the Western Cape, Private Bag X17, Bellville 7535, South AfBan Mishu Allos Vanderbilt University School of Medicine, Division of Infectious Diseases, A 2200 MCN, Nashville, TN rica. 37232.
123
124
LASTOVICA AND ALLOS
Salama et al., 1992a). Sexual transmission of C. fetus subsp. venerealis in humans has not been demonstrated. Our concern in this chapter is with C. fetus subsp. fetus, which will hereafter be referred to simply as C. fetus. Microbiology
C. fetus, like other campylobacters, grows in a microaerobic atmosphere (6 to 12% CO,). These organisms do not ferment carbohydrates and are catalase, oxidase, and nitrate reductase positive (Table 1). C. fetus strains grow well at 25 and 37”C, but unlike C. jejuni, will not usually grow at 42°C (Dediste et al., 1998). C. fetus is resistant to nalidixic acid, susceptible to cephalothin, and lacks pyrazimidase activity (Burnens and Nicolet, 1993), which are useful characteristics for distinguishing C. fetus from C. jejuni. Although early investigators (Butzler and Skirrow, 1979; Edmonds et al., 1985) reported that the isolation of C. fetus from stools was rare, its natural habitat is the intestine. Many clinical laboratories incubate samples at 42°C and may use a medium that contains cephalothin, conditions favoring the isolation of C. jejuni but not of C. fetus. C. fetus grows slowly, and isolation from blood culture can take several weeks (Wang and Blaser, 1986). Because other enteric flora grow more rapidly than campylobacteria, antibiotic-containing media traditionally have been used for the isolation of campylobacters from stool specimens. However, isolation of C. fetus and other non-jejuni Campylobacter species from stools may be unsuccessful because of the use of antibioticcontaining media to which these organisms are susceptible. Filtration of stools through 0.65-pm filters onto the surface of antibiotic-free blood agar plates permits the passage of the smaller C. fetus bacteria while retaining larger contaminating microorganisms (Fig. 1) (Lastovica, 2006). PCR assays for C. fetus subsp. fetus and C. fetus subsp. venerealis have been published (Oyarzabal et al., 1997; Schultz et al., 2006). Epidemiology The source of many C. fetus infections in humans is likely zoonotic. The major habitat of C. fetus is the intestine, and it is commonly isolated from healthy sheep and cattle (Table 2). The organism is found in abundance in the genital tracts of aborting sheep and cattle and their products of conception (Garcia et al., 1983). It has also been found in poultry, swine, and reptiles (Smibert, 1984). Feces from infected animals may contaminate soil, freshwater,
and carcasses during abattoir processing (Blaser et al., 1983), and human infection most probably results from consumption of contaminated food or water. The only exposure common to 10 Californian patients with malignancy or other serious disease who developed C. fetus sepsis was “nutritional therapy” that entailed consuming raw calf‘s liver during the week before they became ill (CDC, 1981). Exposure to raw beef, raw beef liver, or improperly cooked pork shortly before illness has been reported in several patients with C. fetus bacteremia (Miki et al., 2005; Rao et al., 1990). Klein et al. (1986) reported an outbreak in Wisconsin of C. jejuni and C. fetus infection caused by drinking raw milk. The presence of C. fetus was detected solely because the C. fetus strains grew (atypically) at 42”C, the standard temperature for isolating C. jejuni. In an outbreak of C. fetus diarrhea that affected 18 of 90 members of a Hutterite colony in Alberta, Canada, persons who worked in the colony’s abattoir were 2.03 times more likely to have diarrhea than others (Rennie et al., 1994). The modes of transmission of C. fetus infection in humans are not well understood. Despite evidence which indicates zoonotic transmission, more than two-thirds of patients with C. fetus bacteremia lived in an urban environment and had no known exposure to farm animals (Bokkenheuser, 1970; Tremblay et al., 2003). Nosocomial spread of C. fetus infection in a neonatal intensive care unit, in which four infants developed meningitis, has been recorded (Morooka et al., 1996). Clinical Features The clinical features of diarrheal disease due to C. fetus infection in healthy individuals is similar to infection by C. jejuni (Rennie et al., 1994). Sequelae are uncommon, and most patients do not require antibiotic treatment. Although C. fetus infection may occur at any age, most infections occur in elderly persons. In an analysis of 111 C. fetus strains isolated in Quebec, Canada, from 1983 to 2000 (Tremblay et al., 2003), 53% occurred in persons older than 70 years. The annual incidence of infection was 0.1 per 100,000 population; however, this is likely an underestimate because many infections may be neither detected nor reported (Tremblay et al., 2003). C. fetus infection usually occurs in compromised patients, more than 75% of whom are men with serious medical conditions (e.g., diabetes mellitus, atherosclerosis, liver cirrhosis, chronic alcoholism, asplenia, and AIDS) and patients being treated with immunosuppressive agents (Devlin and McIntyre, 1983; Guerrant et al., 1976; Monno et al., 2004; Sakran et al., 1999). In a study of almost 100 patients
Table 1. Phenotypic and biochemical characteristics of Campylobacter, Arcobacter, and intestinal Helicobacter species of clinical significance"
Species or subspecies
C. jejuni subsp. jejuni biotype I f C. jejuni subsp. jejuni biotype 2f C. jejuni subsp. doylei C. coli C. fetus subsp. fetus C. upsaliensis C. lari C. hyointestinalis C. sputorum C. concisus C. mucosalis
c. cuwusi
C. rectus' C. showae C. gracilisi.' A. butzleri' A. cryaerophilus' H. cinaedi" H. fennelliae m.n H. pullorum
Catalase
Nitrate reduction
Arylsulfatase
Pyrazinamidase
Hippurate hydrolisis
Nalidixic acidb
Cephalothin'
HIS production Rapid TSIe Lead H,Sd acetate
Growth at:
250c
420c
Indoxyl acetate
+
++ ++
+ +
-
-
+
+ + + + + + + + + + + + + + +
++ +
+
-
+
(+I
+
5+ 5+ 3+ 5+ 5+ 3+ ND ND -
-
(-) -
+ + + ND
+ +
(+)
-
+
+
ND
ND
+, positive; (+), most strains positive; -, negative; (-), most strains negative; ND, not determined; R, resistant; (R), most strains resistant; S, susceptible; ( S ) , most strains susceptible. bNalidixic acid (30-pg disk). Cephalothin (30-pg disk). dMethod of Skirrow and Benjamin (1980). 'Triple sugar iron. "Biotypes 1 and 2 of Skirrow's scheme (1980). 8 Some strains grow better in HI-enhanced rnicroaerophilic conditions. hUPTC variant positive. Biovar paraureolytic positive. iPitting colonies on blood agar. kNonmotile and oxidase negative. 'Aerobic growth at 30°C. Spreading, noncolonial growth. "Hypochlorite odor. a
+ +
-
H~
required
Urease
126
LASTOVICA AND ALLOS
Figure 1. A 4-day-old pure growth of C. fetus after filtration of a stool specimen onto an antibiotic-free blood-agar plate.
with C. fetus infection, the male-to-female ratio was 1.1 to 1.0, and 69% of isolates were from blood, 20% from stools, and 11% from other typically sterile sites (Tremblay et al., 2003). Intestinal infections may have serious complications such as toxic megacolon (Kalkay et al., 1983). C. fetus accounts for more bacteremia in adults than any other Campylobacter species except C. jejuni (Lastovica, 1996; Skirrow et al., 1993). Infections due to C. fetus were previously considered primarily to cause bacteremia in elderly men with chronic underlying illness, but AIDS patients may now represent the most typical population (Font et al., 1997; Rao et al., 1990). Infections also may occur in patients who have undergone solid or hematopoietic organ transplantation (Heng et al., 2002; Monno et al., 2004). C. fetus may cause a prolonged, relapsing illness that is characterized by chills, fever, and myalgia, and that lacks an identified focus of infection (Blaser et al., 1983; Guerrant et al., 1976;). Sec-
Table 2. Sources and disease associations of non-jejuni, non-coli Campylobacter and related bacteria" Species or subspecies
Recognized sources
Human disease association
Animal disease association
Cattle Cats, dogs, ducks, monkeys Pigs, cattle, hamsters
Septicemia, enteritis, abortion, meningitis Septicemia (rarely) Enteritis, septicemia Enteritis, septicemia6
Bovine, and ovine spontaneous abortion Bovine infectious infertility Canine, feline gastroenteritis Porcine and bovine enteritis
Pigs, birds (including poultry)
None
?c
Enteritis, septicemia
?
C. concisusd
Cats, dogs, chickens, monkeys, seals, mussels, oysters (river and seawater) Humans
?
C. mucosalis C. sputorum biovar sputorum C. jeiuni subsp. doylei H. cinaedi
Pigs Humans, cattle, pigs, sheep Humans Humans, hamsters
H. fennelliae
Humans
H. pullorurn A. butzleri
A. skirrowii
Poultry Pigs, bulls, horses, cattle, chickens, primates, ostriches, ducks, water, sewage Pigs, bulls, poultry, sheep, sewage, horses Sheep, bulls, pigs, chickens, ducks
Periodontal disease, enteritis, septicemia None Abscesses Enteritis, septicemiab Enteritis, septicemia: proctocolitis Enteritis, septicemia: proctocolitis Enteritis Enteritis, septicemia
H. rappini
Humans, mice, sheep
C. fetus subsp. fetus
Cattle, sheep
C. fetus subsp. venerealis C. upsaliensis C. hyointestinalis subsp. hyointestinalis C. hyointestinalis subsp. lawsonii C. lari
A. cryaerophilus
Enteritis, septicemia None Enteritis, septicemia6
Porcine necrotic enteritis ? ?
Hamster enteritis ?
Avian hepatitis Porcine, bovine, primate gastroenteritis, porcine abortion Bovine, porcine, ovine, equine abortion Porcine, equine abortion; ovine, bovine gastroenteritis Ovine abortion
"Data from Bolton et al. (1987); Flores et al. (1990); Foster et al. (2004); Fox et al. (1989); Garcia et al. (1983); Holst et al. (1987); Lammerding et al. (1996); Lastovica (2006); Patton et al. (1989); Sandstedt and Ursing (1991); Tanner et al. (1984); Vandamme et al. (1992b). 'Children and HIV patients. Includes C. curvus and C. rectus. ?,unknown.
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
ondary seeding to an organ may occur, which can lead to complications and occasionally death (Collins et al., 1964; Viejo et al., 2001). Bacteremia due to C. fetus may be primary, arising from the gastrointestinal tract, or secondary, arising from infection of another site. Infection may occur as a direct result of invasion of the bloodstream via venipuncture or by an intravenous line (Guerrant et al., 1976). Although the organism is rarely isolated from feces, diarrhea precedes or accompanies bacteremia in nearly half the cases (Guerrant et al., 1976; Lastovica, 1996). C. fetus is occasionally isolated from pediatric patients (Table 3). High fevers frequently occur, but usually they are well tolerated, and the mortality is approximately 20% (Dickgiesser et al., 1983; Rao et al., 1990). C. fetus exhibits a distinct affinity for vascular tissue, and infections have been associated with thrombophlebitis, cellulitis, and mycotic aneurysms (Carbone et al., 1985; Cone et al., 2003; Montero et al., 1997). Reviews describe 26 cases of C. fetusassociated endocarditis (Bar et al., 1998; Farrugia et al., 1994); all the patients had some kind of predisposing disorder such as liver or hepatic cirrhosis, rheumatic or ischemic heart disease, or alcoholism. A 52-year-old man developed C. fetus endocarditis when he ate raw meat shortly after having a tooth extracted (Miki et al., 2005). Kanj et al. (2001) described a 14-year-old girl with P-thalassemia and C. fetus pericarditis who improved with pericardiectomy and ampicillin treatment. C. fetus is recognized as a causative agent of septic abortions in animals, and it is now known to cause perinatal sepsis and fetal loss
Table 3. Distribution of Campylobacter and related species isolated from the diarrhetic stools of children at the Red Cross Children’s Hospital, Cape Town, South Africa, from 1 October 1990 to 30 June 2007” Species or subspecies
No.
YO
C. jejuni subsp. jejuni C. concisus C. upsaliensis C. jejuni subsp. doylei H. fennelliae C. coli C. hyointestinalis H. cinaedi CLO / HLO A. butzleri C. fetus subsp. fetus C. curvtrs, C.rectus, and H. rappini C. sputorum biovar sputorum and C. lari Total
1,956 1,503 1,414 43 1 337 181 57 51 31 20 9 9 7 6,006
32.57 25.02 23.54 7.18 5.61 3.01 0.95 0.85 0.52 0.33 0.15 0.15 0.12 100.00
“Data from Lastovica (2006) and unpublished results. bCLOIHLO, Cumpylobucter or Helicobacter organisms that could not be fully characterized.
127
in humans. In pregnant animals, ingestion of C. fetus leads to intestinal infection, followed by bacteremia, and because the organism has a high affinity for placental tissues, there is a risk of infection of the placenta and fetus (Miller et al., 1959). Similarly, gastrointestinal colonization, with or without enteritis, followed by a hematogenous spread and then a placental infection, probably occurs in the rare cases of infection in women. C. fetus infection in pregnant women is usually recognized during the third trimester, although abortions at 9.5 and 14 weeks have been reported (Steinkraus and Wright, 1994). The illness is usually mild in the mothers, but outcome is often poor in the infants. Neonates may be infected transplacentally or during delivery. Even with appropriate antimicrobial therapy, the average mortality of fetuses and neonates is about 70% (Sauerwein et al., 1993; Steinkraus and Wright, 1994). Fetal infection and abortion may only become apparent several weeks after maternal infection. Salpingitis and turboovarian abscess due to C. fetus have been described in nonpregnant women (Brown and Sautter, 1977; McGechie et al., 1982). The central nervous system may be infected by C. fetus, with meningoencephalitis being the most common manifestation in adults (Dronda et al., 1998; Gorkiewicz et al., 2002). Subarachnoid hemorrhages, brain abscesses (La Scola et al., 1998), cerebral infarctions, and subdural abscesses (Mendelson et al., 1986) have been reported; although two-thirds of the patients survive, neurological sequelae are frequent. In neonates, the prognosis is usually worse, but not invariably so (La Scolea, 1985). The cerebrospinal fluid typically shows polymorphonuclear pleocytosis. Subdural effusion may also be present. C. fetus may remain latent in an immunocompromised host, after bacteremic seeding in a bony focus, only to be reactivated years later (Neuzil et al., 1994). Eradication of such an infection probably requires surgical excision of the infected area. C. fetus can be an unobtrusive inhabitant of the gut, but with decreased host immunity, the organism may invade the mucosa and cause generalized infection. Postoperative prosthetic hip joint (Chambers et al., 2005) and knee (David et al., 2005) infections due to C. fetus have been reported, as has chronic osteomyelitis of the ankle (Bracikowski et al., 1984) and spine (Yamashita et al., 1999). Other forms of C. fetus infection are septic arthritis (Fick et al., 1979), lung abscess (Lawrence et al., 1971), empyema (Targan et al., 1977), gluteal abscess (Otero et al., 1994), chorioamnionitis (Viejo et al., 2001), cholangitis (Alvarez et al., 2000), and cholecystitis (Takatsu et al., 1997). Infection of the peritoneum by C. fetus has been reported in peritoneal dialysis patients; it was postu-
128
LASTOVICA AND ALLOS
lated that direct contamination of the catheter had occurred (Wens et al., 1985). Spontaneous C. fetus peritonitis has occurred in alcoholic patients with cirrhosis (Targan et al., 1976), possibly as a result of impaired reticuloendothelial clearance of portal bacteremia. The pathogenesis of C. fetus infection is covered in detail in chapter 23. Treatment
C. fetus infections are often prolonged and result in relapse, but most patients will recover with appropriate antibiotic treatment and medical procedures. The prognosis depends on the severity of the underlying illness and the rapidity of application of antibiotic treatment. Infection by c. fetus can be lethal in some debilitated persons and could hasten the death of others. Healthy immunocompetent patients usually have self-limiting enteritis or bacteremia with no sequelae and rarely need antibiotics. However, patients with systemic C. fetus infection often require extended parenteral therapy. Erythromycin may not be sufficient to treat infections, but ampicillin, carbapenems, or third-generation cephalosporins are usually effective against established C. fetus infections (Fujihara et al., 2006; Herve et al., 2004). Although tigecycline has been shown to be effective against C. jejuni and C. coli, its effectiveness in treating infections with C. fetus has not been studied (RodriguezAvial et al., 2006). Patients with hypogammaglobulinemia and persistent C. fetus bacteremia may require lifelong antibiotic therapy. Intravenous immunoglobulins are not effective in treating immunodeficient persons with C. fetus infection because the serum from normal persons usually does not contain opsonizing antibodies to C. fetus (Neuzil et al., 1994).
CAMPYLOBACTER UPSALIENSIS C. upsaliensis is a recognized human pathogen in both healthy and immunocompromised patients. It causes bacteremia and acute or chronic diarrhea (Lastovica, 1996). This organism, originally described as a catalase-negative or weak Campylobacter, has been associated with the hemolytic uremic syndrome (Carter and Cimolai, 1996) and with spontaneous human abortion (Gurgan and Diker, 1994). C. upsaliensis was first isolated from the stools of healthy and diarrhetic dogs in 1983. DNA-DNA hybridization studies indicated that this was a new species (Sandstedt and Ursing, 1991), and the name upsaliensis was adopted in honor of the Swedish town, Uppsala, where the organism was first isolated (International
Union of Microbiological Societies, 199 1). Analysis of 96 C. upsaliensis strains from three continents showed that 84% belonged to one of five serotypes (Lentzsch et al., 2004). These data point to clonal expansion of C. upsaliensis and indicate a high degree of antigen conservation during evolution. A review by Bourke et al. (1998) has summarized available data on C. upsaliensis. C. helveticus is a closely related species found in dogs and cats, but to date, it has not been isolated from humans (Stanley et al., 1992). Microbiology
C. upsaliensis is a thermotolerant Campylobacter species that usually grows well at 42"C, but not at 25°C. Some strains will grow better in an H,enhanced microaerophilic atmosphere (Table 1).This organism is catalase negative or only weakly positive (Sandstedt and Ursing, 1991); it is hippurate and aryl sulfatase negative but nitrate reductase and indoxyl acetate positive. A striking and distinguishing characteristic of C. upsaliensis is its intense susceptibility to both nalidixic acid and cephalothin, with inhibitory zones of up to 80 mm for cephalothin (Lastovica et al., 1989). Because cephalothin is often used in media for the isolation of Campylobacter, cephalothin-susceptible C. upsaliensis cannot be isolated with such media. Detection and the Cape Town Protocol The so-called Cape Town protocol for the isolation of C. upsaliensis and other campylobacters from stool specimens without the use of selective media was developed at the Red Cross Children's hospital in Cape Town. The method involves filtration of stools through a membrane filter onto antibioticfree blood agar plates (Fig. 1) and subsequent incubation in an H,-enhanced microaerobic atmosphere (Lastovica, 2006). With the use of this protocol since 1990, stool cultures positive for Campylobacter or related microorganisms rose to 21.8% from the 7.1% obtained with antibiotic-containing selective plates and conventional microaerobic incubation previously used (Lastovica, 2006). This South African diagnostic microbiology laboratory could only begin to isolate C. upsaliensis from stool specimens once the isolation protocol was changed from antibiotic-containing media to the Cape Town protocol (Lastovica, 2006). The use of membrane filtration in this protocol has the added advantage of permitting the isolation of antibiotic-susceptible campylobacters other than C. upsaliensis. On the basis of the difference in colony
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
morphology on primary isolation and subsequent biochemical and serological confirmation, 16.2% of the stools of South African children with gastroenteritis had multiple isolates of two to five species, or of several different serotypes of C. jejuni (Lastovica, 2006). C. upsaliensis was frequently coisolated with C. jejuni subsp. jejuni, C. jejuni subsp. doylei, and particularly with Helicobacter fennelliae (Lastovica, 2006). If infection by more than one species is suspected, considerable care must be taken to separate the domed colonies of C. upsaliensis from the spreading, noncolonial growth of Helicobacter fennelliae or H. cinaedi before positive identification can be undertaken (Fig. 2). In an Australian study of 676 hospitalized patients with gastroenteritis, 75 Campylobacter strains were isolated on blood-free medium with a selective supplement, but concurrent isolation onto antibioticfree blood agar overlaid with a membrane filter yielded 213 Campylobacter strains. Some strains could only be isolated by the membrane filter technique (Albert et al., 1992). Loades et al. (2005) showed that every commercial Campylobacter selective media tested was inhibitory to some stains in a panel of 59 isolates of a variety of Campylobacter, Helicobacter, and Arcobacter spp. In a direct comparison, the Cape Town protocol proved superior to selective media for the isolation of C. upsaliensis from clinical specimens (Lastovica and Le ROUX,2001). A process of membrane filtration onto antibiotic-free media and incubation in an H,-enhanced microaerophilic atmo-
Figure 2. Culture plate showing the spreading, noncolonial growth of H. fennelliae (top) contrasted with the domed colonies of C. upsaliensis (bottom).
129
sphere is a simple, efficient, and cost-effective alternative to antibiotic-containing selection media, and this method permits the isolation of a whole spectrum of clinically relevant Campylobacter, Helicobacter, and Arcobacter species (Lastovica, 2006). Epidemiology C. upsaliensis is widespread on all continents and is commonly isolated from the intestinal contents of dogs. C. upsaliensis comprised 64% of the 98 Campylobacter strains isolated from dogs in Sweden over a 2-year period (Sandstedt et al., 1983). The organism has been isolated from dogs with chronic diarrhea (Davies et al., 1984; Modolo and Giuffrida, 2004), healthy puppies and kittens (Hald and Madsen, 1997), asymptomatic cats (Fox et al., 1989), and asymptomatic vervet monkeys (Lastovica et al., 1991). Although the exact source of C. upsaliensis infection in humans in unknown, zoonotic transmission is a distinct possibility. Four of seven patients with C. upsaliensis infection in one study reported animal contact (Patton et al., 1989). Stool isolates of C. upsaliensis that appeared to be the same strain were isolated from a 53-year-old man with bloody diarrhea and his healthy 3-year-old dog (Goossens et al., 1991). In another study, identical plasmid profiles were detected in human and dog isolates (Owen et al., 1985). Gurgan and Diker (1994) documented the isolation of C. upsaliensis in blood and fetoplacental specimens of a woman who experienced a spontaneous abortion at 18 weeks’ gestation; a C. upsaliensis strain was isolated from her asymptomatic cat, and protein profile analysis confirmed a strong similarity between the human and feline isolates. Although these observations are suggestive, animal-to-human transmission of C. upsaliensis still remains to be unequivocally proved. Indirect evidence suggests that person-to-person transmission of C. upsaliensis is possible. Goossens et al. (1995) documented C. upsaliensis infection in 34 children in four day care centers in Brussels, Belgium. On the basis of several molecular typing methods, it was demonstrated that the outbreaks of C. upsaliensis infection in three centers were due to the same strain. This C. upsaliensis strain was closely related to the strain isolated from the fourth day care center. The proportion of all campylobacters that are C. upsaliensis is relatively high when fecal samples are prospectively cultured for C. upsaliensis. From 1990 to 2007, at a pediatric hospital in South Africa, 1,414 strains of C. upsaliensis were isolated, which accounted for 23.5% of all the campylobacters and related organisms isolated and identified (Table 3). This
130
LASTOVICA AND ALLOS
high isolation rate of C. upsaliensis is attributed to application of the Cape Town protocol (Lastovica, 2006). Only a single C. upsaliensis strain was obtained in a study of 631 stools from Thai children (Taylor et al., 1991). A Canadian study of 915 Campylobacter stool isolates indicated that only 7 (0.1%) were C. upsaliensis (Taylor et al., 1989). In a survey of 3 94 Campylobacter clinical blood culture strains isolated in England and Wales, two (0.8%) were C. upsaliensis (Skirrow et al., 1993). By contrast, a South African study of 221 Campylobacter strains obtained from children's blood cultures indicated that 39 (18%) were C. upsaliensis (Lastovica, 1996). Differences in isolation and culture protocols used in various laboratories could account in part for the wide variation in C. upsaliensis prevalence. Alternatively, data from South Africa compared with other studies may reflect differences in prevalence and exposure to C. upsaliensis, or possibly differences in colonization and the nature of C. upsaliensis infection in different geographical areas. Characterization of Isolates
C. upsaliensis has a plasmid carriage rate of about 90%, much higher than other species of Campylobacter (Da Silva Tatley et al., 1992; Owen and Hernandez, 1990). Digestion of C. upsaliensis chromosomal DNA with HaeIII has indicated that less than 20% of the strains examined were related (Owen and Hernandez, 1990). Although serotyping C. upsaliensis strains has been of limited value (Lastovica et al., 1989), sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) protein profile analysis (Owen et al., 1989) has proved useful for the differentiation of individual isolates of C. upsaliensis. Pulsed-field gel electrophoresis (PFGE)
indicated molecular heterogeneity in a study of 20 C. upsaliensis strains (Bourke et al., 1996). Identification of C. upsaliensis can be done by PCR assays based on the 16s rRNA gene (Linton et al., 1996), the GTPase gene (Van Doorn et al., 1997), or the GlyA gene (Steele et al., 1985b). Miller et al. (2005) have developed a multilocus sequence typing scheme for C. coli, C. lari, C. upsaliensis, and C. helveticus. Realtime PCR may ultimately provide an efficient method for differentiating Campylobacter species (Jensen et al., 2005). Clinical Features The usual symptoms associated with C. upsaliensis infection are gastrointestinal and include watery diarrhea, abdominal cramps, vomiting, and low-grade fever (MCgraud and Bonnet, 1986; Patton et al., 1989; Taylor et al., 1989). Although most patients recover quickly, others may be ill for several weeks (MCgraud and Bonnet, 1986; Patton et al., 1989; Taylor et al., 1989). In a study of 99 patients with C. upsaliensis in their stools, in which the onset of the symptoms was abrupt, 92% had diarrhea, 14% had vomiting, and 6% had fever (Goossens et al., 1990). Symptoms persisted for more than 1 week in 16% of these patients, while 25% had bloody stools and 10% had fecal leukocytes. In a study of 882 C. upsaliensis-infected children (Table 4), the average age was 19.4 months (range, 1 month to 10 years). Loose stools were present in 89% and watery stools in lo%, and they formed in less than 1% of the children. Gastroenteritis symptoms were present in 88%, vomiting in 9%, and fever (temperature >38"C) in 6% of these patients. Less than 4% were coinfected with Salmonella or Shigella. However, 8% of these patients had Ascaris, Tricuris, Cryptosporidium, or
Table 4. Clinical features of infection with non-jejuni, non-coli Cumpylobucter species and related organisms obtained from children in Cape Town, South Africa, 1990 to 2007 % of patients with:
Organism
Coexisting enteric
No. of patients Diarrhea
C. jejuni subsp. jejuni C. concisus C. upsaliensis C. jejuni subsp. doylei H. fennelliae C. hyointestinalis H. cinaedi A. butzleri C. fetus subsp. fetus
1,432 911 882 358 253 51 38 15 6
85 80 88 79 79 74 79 100 100
Vomiting
6 3 9 6 8 12 8 60 17
Fever
5 9 6 4 5 4
8 20 17
pathogen:
Preexisting conditions
Bacterial
Parasitic
6
7 6 6 8 10 12 8 0 0
4
3 4 5 6 8 20 17
9 18 1 10 33 39 26 40 100
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
Giardia detected in their stools. Underlying illnesses such as kwashiorkor, marasmus, convulsions, hepatitis, anemia, or tuberculosis were present in 13% of these children (Lastovica et al., 1989). Two of 8 C. upsaliensis bacteremic patients had recent abdominal surgery (Patton et al., 1989), and 8 of 16 children with C. upsaliensis bacteremia had gastrointestinal symptoms. These observations suggest that the bacteremias may have been the result of intestinal infections (Lastovica et al., 1989). Most patients with C. upsaliensis bacteremia have other serious underlying medical conditions (Lastovica et al., 1989; Patton et al., 1989). C. upsaliensis has been isolated from the breast abscess of a patient who reported no animal contact or gastrointestinal symptoms (Gaudreau and Lamonthe, 1992), and C. upsaliensis has been linked to the hemolytic-uremic syndrome (Carter and Cimolai, 1996). Pathogenesis Because C. upsaliensis may produce both inflammatory and noninflammatory diarrhea in humans, the organism’s virulence may depend in part on hostcell invasion. C. upsaliensis is capable of invading cell lines of gastrointestinal origin more efficiently than nonintestinal cell lines (Mooney et al., 2003), suggesting that the organism has a tropism for the gastrointestinal tract. Specific intestinal cell factors may influence the organism’s ability to be internalized, which likely affects its disease-causing abilities. Sylvester et al. (1996) demonstrated that C. upsaliensis is capable of binding to CHO and HEp-2 cells in tissue culture. Surface proteins in the range 50 to 90 kDa on C. upsaliensis isolates could bind to phosphatidylethanolamine, a putative cell membrane receptor. Biotin-labeled C. upsaliensis strains also bound in a concentration-dependent fashion to human small-intestine mucin, implying that that C. upsaliensis express an adhesin or adhesins capable of recognizing a specific mucin epitope or epitopes (Sylvester et al., 1996). Binding to mucins may possibly influence bacterial access to cell membrane receptors and thus influence host resistance to infection. Mooney et al. (2001) observed that C. upsaliensis releases a cytolethal distending toxin that affected HeLa and T lymphocyte cells. They also described invasion of cultured epithelial and primary human small intestinal cells, and electron microscopy revealed C. upsaliensis within Caco-2 cell cytoplamic vacuoles (Mooney et al., 2003). Fouts et al. (2005) described a novel putative virulence locus, licABCD for C. upsaliensis, similar to genes present in Neisseria (Serino and Virji, 2002). These licABCD genes encode proteins involved in the acquisition of cho-
13 1
line, the synthesis of phosphorylcholine, and the transfer of phosphorylcholine to lipooligosaccharide or teichoic/lipoteichoic acids to facilitate attachment to host cells (Serino and Virji, 2002). Although these observations are suggestive but not conclusive of virulence, proof of C. upsaliensis as a human enteropathogen will require additional studies. Treatment The most active antimicrobial agents available for treatment of C. upsaliensis infection are fluoroquinolones (Preston et al., 1990). Erythromycin was once considered the preferred treatment in Campylobacter infections, but 4 to 18% of C. upsaliensis isolates are resistant to erythromycin (CDC, 1981; Moore et al., 2006; Patton et al., 1989). Infections with C. upsaliensis have been treated successfully with amoxicillin-clavulanate, cefotaxime, and doxycycline (Goossens et al., 1991, 1995; Russell et al., 1992).
CAMPYLOBACTER HYOINTESTINALIS In 1983, C. hyointestinalis was identified and suggested as a possible cause of proliferative enteritis in pigs (Gebhart et al., 1983). This organism has subsequently been isolated from human stools, particularly those from children with watery, nonbloody diarrhea. The name hyointestinalis is derived from the Latin hyo, “hog,” and intestinalis, “pertaining to the intestine.” On the basis of phenotypic and genomic methods, On et al. (1995a) described Campylobacter hyointestinalis subsp. lawsonii, subsp. nov., and Campylobacter hyointestinalis subsp. hyointestinalis subsp. nov. The pathogenic role of C. hyointestinalis subsp. lawsonii in animals and humans is unknown at present (On, 1996). Microbiology and Diagnosis
C. hyointestinalis is closely related to C. fetus, and like C. fetus, it is catalase and nitrate reductase positive, indoxyl acetate negative, susceptible to cephalothin, and resistant to nalidixic acid. C. hyointestinalis differs from C. fetus in its copious production of H,S in triple sugar iron agar in the presence of H,, often entirely blackening lead acetate strips. C. hyointestinalis grows under microaerophilic conditions, but some strains require additional hydrogen (Edmonds et al., 1987; Lastovica, 2006). These isolates can be differentiated from other H,-requiring isolates by the catalase and other tests (Table 1)(Vandamme et al., 1992b). All strains of C. hyointestinalis
132
LASTOVICA AND ALLOS
will grow at 37"C, and some strains will also grow at 42°C. Lack of aryl sulfatase activity and intolerance to 3.5% NaCl are useful phenotypic tests for the diagnosis of C. hyointestinalis (Burnens and Nicolet, 1993). Because cephalothin is a constituent of many Campylobacter-selective media, C. hyointestinalis, similar to other cephalothin-susceptible campylobacters, is underdetected. Application of the Cape Town protocol has proved to be efficient for the isolation of C. hyointestinalis (Lastovica, 2006). PFGE and SDS-PAGE have been used to differentiate strains of C. hyointestinalis from each other as well as from other Campylobacter species (Costas et al., 1987; Salama et al., 1992b). DNA probes have been useful for detection of C. hyointestinalis in swine with proliferative enteritis (Gebhart et al., 1990). Oligodeoxynucleotide probes have been constructed for both C. hyointestinalis and C. fetus based on 16s rRNA sequence data (Wesley et al., 1991). A PCR assay based on the 16s rRNA gene has been developed for the detection of C. hyointestinalis (Linton et al., 1996). Epidemiology C. hyointestinalis has been consistently isolated from the intestines of pigs with proliferative enteritis, but not from asymptomatic pigs or from pigs with other enteric diseases (Gebhart et al., 1985). Its role in this disease remains uncertain. C. hyointestinalis has been isolated from hamsters, cattle, reindeer, and nonhuman primates (Table 2) (Hanninen et al., 2002; Russell et al., 1992; Walder et al., 1983). In one study, 17% of fecal samples from 300 steers contained C. hyointestinalis (Inglis et al., 2005). Transmission of C. hyointestinalis from pig to a human with chronic diarrhea and vomiting was verified by whole-cell protein electrophoresis and 16s rRNA gene sequencing data (Gorkiewicz et al., 2002). Ohya and Nakazawa (1992) reported that 21 (77%) of 29 C. hyointestinalis strains from swine with proliferative enteropathy can produce a cytotoxin, although at present, the role of this cytotoxin in human disease is uncertain. Clinical Features and Treatment In 1986, the first reported human illness was obtained with the isolation of a C. hyointestinalis strain from the stool of a homosexual man with proctitis. The patient's symptoms resolved after appropriate antibiotic treatment (Fennel1et al., 1986). Since then, C. hyointestinalis strains have been isolated from stool specimens of four patients, all of whom had
nonbloody, watery diarrhea (Edmonds et al., 1987). An 8-month-old girl, the youngest patient, had drunk unpasteurized milk, and the oldest, a 78-year-old woman, exhibited fever and vomiting. There were no leukocytes present in their stools, and both patients recovered with appropriate antibiotic therapy. The other two patients, both homosexual men, had abdominal cramps, and one was febrile. One patient recovered with trimethoprim-sulfamethoxazoletreatment. No antibiotic treatment was administered to the other patient, who continued to have intermittent diarrhea and cramps for several months (Edmonds et al., 1987). A case of C. hyointestinalis-associated diarrhea was reported in a 52-year-old woman who was immunodeficient because of an evolutive chronic myeloid leukemia (Minet et al., 1988). This patient was febrile and had a nonbloody, watery diarrhea. In another study, five strains of C. hyointestinalis were isolated from five members of the same family who had previously drunk unpasteurized milk (Salama et al., 1992b). In this family outbreak, only in the index case, a 5-month-old girl, was diarrhea present; she was also infected with C. jejuni. These five C. hyointestinalis strains were examined by PFGE. Three of the strains had identical genome patterns; the other two had completely different patterns and appeared to be unrelated (Salama et al., 1992b). In a review of patients infected with C. hyointestinalis, Breynaert et al. (1998) examined the clinical features of nine patients. Seven of the patients (four men and three women with a mean age of 63 years) were adults, and two were young children. Six patients experienced diarrhea, and five had abdominal pain. The youngest asymptomatic patient was a l-year-old girl with constipation, and the oldest was an 89-year-old woman with a myocardial infarction. None of these patients was seriously immunocompromised, but most of the adult patients had a history of neurological or vascular disease (Breynaert et al., 1998). The clinical features of 51 South African pediatric strains of C. hyointestinalis were examined (Table 4). The average age of the C. hyointestinalis patients was 16 months (range, l month to 7 years). Stools from these children were loose in 77%, watery in 21%, and formed in <2% of the cases. Blood was present in 13% of the stools and fecal leukocytes in 18%. Thirty-nine percent of these patients had underlying illness such as anemia, hepatitis, tuberculosis, convulsions, kwashiorkor, and marasmus. Other clinical features are listed in Table 4. Two extraintestinal isolates of C. hyointestinalis were obtained from the blood cultures of a 22-year-old man after bone marrow transplantation and from a 7-monthold girl with chronic diarrhea (Lastovica, 1996).
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
C. hyointestinalis is susceptible to a range of antimicrobial agents including ampicillin, amoxicillin, amoxicillin-clavulanic acid, doxycycline, and ciprofloxacin (Inglis et al., 2005; Laatu et al., 2005). Resistance has been observed to metronidazole, lincomycin, tetracycline, and erythromycin (Inglis et al., 2005; Laatu et al., 2005).
CAMPYLOBACTER LARl C. lari is a nalidixic acid-resistant, thermophilic Campylobacter first isolated from seagulls in 1980 by Skirrow and Benjamin (1980). Although the first human isolate was from an asymptomatic 6-year-old boy, C. Zari can produce acute diarrheal illness in normal hosts and can cause bacteremia in immunocompromised patients. A subgroup of C. Zuri has the unusual capability (for a Cumpylobucter) of hydrolyzing urea; strains are usually sensitive to nalidixic acid. Known by the acronym UPTC (urease-positive thermophilic Cumpylobacter), they are plentiful in natural water and shellfish. C. Zari are microaerophilic (some isolates may require H,-enhanced microaerobic growth conditions) campylobacters that grow at 42°C but not usually at 25°C. They are resistant to nalidixic acid, cephalosporins, vancomycin, and trimethoprim. Most are oxidase and nitrate reductase positive and do not hydrolyze hippurate (Table 1) (Tauxe et al., 1985). Epidemiology
C. lari has been isolated from a variety of environmental and animal sources (Table 2). Of 312 riverine samples collected in a United Kingdom study, 134 yielded campylobacters, and 7 (5%) of these were C. Zuri (MCgraud et al., 1988). In another survey of surface waters in Norway, two (2%) of 96 Campylobucter samples cultured were C. Zari (Brennhoud et al., 1992). These isolations may be significant because water is an established vehicle for the transmission of campylobacters to humans (Taylor et al., 1983). C. Zuri isolation rates were 8% from herring gulls, 29% from kittiwakes (Glunder and Petermann, 1989), and 7% from crows (Maruyama et al., 1990). In a study of the prevalence of Campylobacter spp. in oysters and mussels in The Netherlands, Van Doorn et al. (1998) found that of 44 Campylobacter isolates, 38 were C. luri. C. luri has been isolated from pigs in Sweden (Lindblom et al., 1990), poultry in Peru (Tresierra-Ayala et al., 1994) and Tanzania (Kazawala et al., 1993), fresh vegetables in Canada
133
(Park and Sanders, 1992), and occasionally from dogs (Benjamin et al., 1983). Diagnosis Because the spectrum of clinical disease described in association with C. lari is similar to that seen with other campylobacters, the diagnosis of C. luri depends on the isolation and identification of the organisms from cultured specimens. C. luri is distinct from C. jejuni in several ways, most notably in its resistance to nalidixic acid, to which most C. jejuni isolates are sensitive. C. Zari may not be correctly identified because many clinical microbiology laboratories may not routinely test isolates for nalidixic acid resistance. Nalidixic acid-susceptible isolates of C. Zari have been reported (Bolton et al., 1987; MCgraud et al., 1988), as have nalidixic acid-resistant C. jejuni strains (Walder et al., 1983), but C. lari is negative in the hippurate hydrolysis test, whereas C. jejuni is positive (Table l),and C. jejuni strains hydrolyze indoxyl acetate, whereas C. Zari does not (Popovic-Uroic et al., 1990). PCR assays specific for C. Zari based on the 16s rRNA gene (Linton et al., 1996) and the flaA gene (Sekizuka et al., 2005) are available. Van Doorn and associates (Stamp et al., 1993) developed a PCR assay based on a novel GTPase gene that utilizes speciesspecific probes for C. jejuni, C. coli, C. lari, and C. upsaliensis. This PCR hybridization assay offers rapid and specific identification of thermophilic CampyZobacter spp. Of 38 C. lari strains isolated from mussels and oysters, Van Doorn et al. (1998) found that the C. Zuri isolates were a more heterogeneous group of isolates when compared with C. jejuni than previously thought. On the basis of sequence information, a novel PCR reverse hybridization line probe assay was developed by these researchers to permit the specific and rapid detection of different c. lari variants (Van Doorn et al., 1998). A multilocus sequence typing system is available for the identification and differentiation of C. coZi, C. Zari, C. upsuliensis, and C. helventicus in mixed cultures (McGechie et al., 1982). A multiplex of PCR assays have been devised that distinguish C. coli, C. jejuni, C. Zari, and C. upsuliensis (Klena et al., 2004).
Clinical Features
C. luri is an enteric pathogen for both immunocompetent and immunocompromised hosts. Although the first human isolates of C. lari were from asymptomatic persons (Benjamin et al., 1983; Skirrow and Benjamin, 1980), the pathogenicity of C. Zari
Next Page 134
LASTOVICA AND ALLOS
was only realized in 1984 with the description of a fatal case of C. luri bacteremia in a severely immunocompromised patient (Lentzsch et al., 2004). Persistent C. luri bacteremia has been reported in a human immunodeficiency virus (H1V)-infected patient (Vargas et al., 1992) and in a patient with X-linked agammaglobulinemia Uirapongsananuruk et al., 2006). Bacteremia was associated with a purulent pleurisy in an 80-year-old debilitated patient (Bruneau et al., 1998). C. luri accounted for 2 (0.8%) of 394 Cumpylobucter strains isolated from the blood cultures of patients in England and Wales (Skirrow et al., 1993). C. luri is associated with diarrheal disease, as it induced colitis in a 32-year-old HIV-positive woman who required extensive antimicrobial therapy before symptoms improved (Dronda et al., 1998). C. luri was isolated from the diarrhetic stools of five immunocompetent persons, two of whom were hospitalized. Illness lasted from 1 week to 4 months (median, 2 weeks), and all patients recovered completely (Tauxe et al., 1985). None of these patients was febrile, but four had abdominal cramps, four had contact with pets, and four had eaten chicken in the week before symptoms became apparent. A commonsource waterborne outbreak of C. luri infection occurred in Ontario in 1985 (Borczyk et al., 1987b). Gastroenteritis occurred among construction workers who drank water that had become contaminated with surface water from Lake Ontario, which had a large population of seagulls. Of 162 ill persons, 87% had diarrhea, 70% abdominal pain, and 20% fever, and vomiting, nausea, malaise, and headaches were also noted (Borczyk et al., 1987b). Only one patient reported bloody stools, and the mean duration of illness was 4 days (range, 1 to 10 days). Of the 125 stool samples cultured, 7 yielded C. luri, which is probably an underestimate because collection was delayed and specimens were transported in dry containers, reducing the bacterial viability. UPTC variants of C. luri were isolated in France from the stools of two compromised adult patients (an ovarian cancer patient and an alcoholic) with diarrhea, and from the inflamed appendix of an immunocompetent 10-year-old boy (Mtgraud et al., 1988). C. luri was isolated from the blood of a 79year-old woman with leukemia (Schrader et al., 1991). A urinary tract infection due to a UPTC variant was reported in an alcoholic man with cirrhosis (Btzian et al., 1990). C. luri-associated reactive arthritis has been reported (Goudswaard et al., 1995). C. luri strains are capable of producing both cytotoxic and cytotonic factors Uohnson and Lior, 1986), but their role in the disease process is still unknown.
Treatment
C. luri infections involved with uncomplicated diarrhea are usually self-limiting and generally do not require antibiotic therapy. When indicated, erythromycin, aminoglycosides, clindamycin, and chloramphenicol have been successfully used (Evans and Riley, 1992; Nachamkin et al., 1984). C. luri is resistant to third-generation cephalosporins, vancomycin, penicillin, and trimethoprim-sulfamethoxazole (Evans and Riley, 1992;). Fluoroquinolone-resistant strains have been reported in HIV-infected persons (Evans and Riley, 1992;).
CAMPYLOBACTER JEJUNl SUBSP. DOYLE1 On the basis of DNA hybridization studies, C. jejuni has been divided into two subspecies: C. jejuni subsp. jejuni, and C. jejuni subsp. doylei (Smibert, 1984). C. jejuni subsp. doylei was named after L. P. Doyle, who first isolated these organisms (Steele and Owen, 1988). The clinical relevance of C. jejuni subsp. doylei is just beginning to be appreciated. Clinical and other aspects of C. jejuni subsp. jejuni are covered in detail in chapter 6. Microbiology and Diagnosis The inability to reduce nitrate to nitrite (Steele and Owen, 1988) is the determining phenotypic characteristic that distinguishes C. jejuni subsp. doylei from C. jejuni subsp. jejuni and all other campylobacters (Table 1). Although H. fennelliue is also nitrate reductase negative, it has a spreading colony morphology on blood agar plates, has a strong hypochlorite smell (Fennel1et al., 1984), and is resistant to polymyxin B (Burnens and Nicolet, 1993). These characteristics readily differentiate C. jejuni subsp. doylei from H. fennelliue. C. jejuni subsp. doylei grows poorly at 42°C and is usually hippuricase and catalase positive (Table 1).These organisms are sensitive to nalidixic acid, but unlike C. jejuni subsp. jejuni, they are also susceptible to cephalothin (Table 1).Application of the Cape Town protocol is an efficient method for obtaining these microorganisms (Lastovica, 2006). Differences in colony morphology at primary isolation and subsequent characterization by biochemical and serological tests indicate that C. jejuni subsp. doylei can be coisolated with C. jejuni subsp. jejuni, C. upsuliensis, and H. fennelliue (Lastovica, 2006). Care must taken to separate the discrete, domed colonies of jejuni subsp. doylei from the noncolonial spreading growth of H. fennelliue in suspected cases of mixed infection (Fig. 2). Miller et
c.
Previous Page CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
135
al. (2007) have developed a nap multiplex PCR assay based on the nitrate reductase locus that can be used to unambiguously subspeciate C. jejuni subsp. jejuni and C. jejuni subsp. doylei. A microarray study by Parker et al. (2007) demonstrated that C. jejuni subsp. doylei lacks various metabolic, transport, and virulence functions that were present in C. jejuni subsp. jejuni.
gest a pathogenic and possibly invasive role for C. jejuni subsp. doylei. Little is known of the virulence mechanisms in C. jejuni subsp. doylei, and in a comparative genomic indexing study (Parker et al., 2007), virulence factors such as cytolethal distending toxin were unable to be detected in the C. jejuni subsp. doylei strains examined.
Epidemiology and Clinical Features
Hz-REQUIRING CAMPYLOBACTERS
Urease-negative, gastric Campylobacter-like organisms, which were originally called GCL02 isolates and subsequently C. jejuni subsp. doylei, were identified in the gastric antral biopsy specimens of six patients (Owen et al., 1985). Similar microorganisms were found in the feces of young Australian children hospitalized with gastritis (Steele et al., 1985a, 1985b). Other studies have shown that this microorganism may be associated with diarrhea in children (FCrnandez et al., 1997; Morey, 1996). Of 631 Thai children with diarrhea, 93 (15%) had campylobacters in their stool, and 1 (1.1%) of these was a C. jejuni subsp. doylei strain (Taylor et al., 1991). In a Belgian study of 15,185 stools, 802 Campylobacter isolates were cultured, and C. jejuni subsp. doylei accounted for 4 (0.5%) of 787 strains. C. jejuni subsp. doylei was detected in 2 (0.7%) of 288 Italian children with enteritis (Musmanno et al., 1998). In a South African study of 6,006 Campylobacter and Helicobacter isolates obtained from the diarrheal stools of children over a 17-year period, 431 (7.2%) of the strains were C. jejuni subsp. doylei (Table 3). These isolates were obtained from gastroenteritis patients whose average age was 20 months (range, 1 month to 11 years). Eighty-two percent of the children had loose stools, 18% had watery stools, 15% had bloody stools, and 24% had fecal leukocytes. Other clinical features are indicated in Table 4. In a separate study of South African pediatric Campylobacter blood cultures, 53 (24%) of 221 isolates were C. jejuni subsp. doylei (Lastovica, 1996). The average age of these children was 12 months (range, 2 to 30 months) Twenty-six of 53 children had diarrhea, often chronic, suggesting that intestinal infection preceded systemic infection. Thirty of the 53 patients had severe protein deficiency diseases, such as marasmus and kwashiorkor. C. jejuni subsp. doylei comprised less than 8% of the campylobacters found in stool (Table 3) but formed 24% of the Campylobacter blood cultures seen at the same pediatric hospital (Lastovica, 1996). Morey (1996) reported that in aboriginal children in central Australia, C. jejuni subsp. doylei accounted for a staggering 85% of Campylobacter bacteremias. These observations sug-
Six Campylobacter species have an essential growth requirement for hydrogen or formate. Five of them are found in the gingival flora of the human mouth, notably in periodontal pockets of diseased gums. C. rectus (Latin rectus, “straight”) and C. curvus (Latin cuwus, “curved”) were originally described as Wolinella species (Tanner et al., 1984, 1987). C. concisus (Latin concisus, “concise”) and C. showae (Latin showae, referring to Showa University, where the organism was first isolated) are biochemically similar organisms (Etoh et al., 1993; Tanner et al., 1981). C. gracilis was formerly regarded as a Bacteroides species but was reclassified in 1995 (Vandamme et al., 1995). A similar organism, also wrongly classified as a Bacteroides sp., “B.” ureolyticus, is now considered to belong to the family Camp y lobacteraceae and is awaiting suitable generic assignation (Vandamme et al., 1995). C. mucosalis is an H,-requiring species that has been associated with proliferative enteritis in pigs (Lawson and Rowland, 1984) (Table 2). Microbiology Apart from the H, or formate requirement of these species, another feature they have in common (except for C. showae) is that they are all catalase negative, Many isolates are susceptible to cephalothin and nalidixic acid (30-pg disks), often with inhibitory zones up to 50 mm in diameter (Lastovica et al., 1993; Vandamme et al., 1991). C. cuwus may be difficult to differentiate from C. concisus because the indoxyl acetate assay is not always infallible (Lauwers et al., 1991), and serological (Tanner et al., 1984) or other methods may be required. Two suspected cases of C. mucosalis enteritis in Italian school children were reported (Figura et al., 1993). However, these isolates were only characterized by variable phenotypic criteria, such as colony color. A subsequent investigation of these presumptive C. mucosalis strains by another researcher who used DNA-DNA hybridization techniques indicated these strains were misidentified C. concisus strains (Lastovica et al., 1993). Occasionally, some isolates of C. hyointestinalis, C.
136
LASTOVICA AND ALLOS
jejuni subsp. doylei, C. upsaliensis or C. lari may require an H,-enriched microaerophilic atmosphere for growth. However, these strains can easily be differentiated from the above hydrogen-requiring species by differences in the available phenotypic tests (Lastovica, 2006) (Table 1).C. rectus, C. curvus, C. gracilis and “B.” ureolyticus can produce agar-corroding (pitting) colonies. C. gracilis is unique (for a Campylobacter) because it is nonmotile (aflagellate) and oxidase negative. C. rectus and C. cuwus may appear oxidase negative in traditional tests, but they are positive in the modified test of Tarrand and Groschel (1982). A PFGE study obtained 51 patterns from 53 C. concisus isolates, indicating considerable genetic heterogeneity (Matsheka et al., 2002). A random amplification of polymorphic DNA assay is available for the differentiation of C. concisus strains (Matsheka et al., 2006). Epidemiology and Clinical Features
C. rectus was isolated from 80% of 1,654 adults and children with periodontitis in a 2-year study (Rams et al., 1993) and from periodontitis patients with inflammatory bowel disease (Van Dyke et al., 1986). A recent intriguing study showed C. rectus and C. concisus were isolated from endoscopic biopsy samples of seven patients with Barrett’s esophagitis but not from controls, which is suggestive of a link between these bacteria and the initiation, maintenance, or exacerbation of this disease (Macfarlane et al., 2007). Chronic oral infection with C. rectus also has been postulated as a possible contributor to the cause of the severe vasculopathy known as Buerger’s disease (Iwai et al., 2005). Reports of additional C. rectus isolates associated with gastroenteritis have not been published, possibly because less than optimal conditions were used for the isolation of this fastidious microorganism. C. curzlus has been recovered from periodontal and septicemia patients (Tanner et al., 1987). It has been isolated from diarrheal stools of children in Belgium (Lauwers et al., 1991) and South Africa (Table 3), and from the blood culture of a man with liver abscesses (Wetsch et al., 2006). Oral bacteria are rarely reported to cause extraorbital infection. Han et al. (2005) reported isolating under anaerobic conditions and verified by 16s rRNA assays of C. rectus in a breast abscess and C. cuwus in a liver abscess and a bronchial abscess. All three patients recovered with antibiotic treatment. The association of C. concisus with human periodontal disease is well known (Tanner et al., 1981, 1987), but a direct causal role has not been established. C. concisus may have been responsible for osteomyelitis of the sacrum in a patient with diabetes
and a sacral decubitus ulcer (Engberg et al., 2005; Johnson et al., 1985). Evidence that C. concisus is an enteric pathogen is equivocal. Of 14 C. concisus isolates studied in Belgium, 7 were from patients with diarrhea and 1 was from the blood of a man with carcinoma of the bronchus (Vandamme et al., 1989). One study found no significant difference in C. concisus isolation rates between patients with diarrhea and controls (Totten et al., 1985). Studies in South Africa (Samie et al., 2007) and Belgium (Engberg et al., 2000) of adults and children with diarrhea have shown the prevalence of C. curzlus in patients’ stools was not different from controls. In another South African study of 1,519 Campylobacter isolates, 187 (12.3%) required hydrogen for growth (Lastovica et al., 1993). One isolate was from an adult diarrheal stool, one was from the blood culture of an 18-dayold infant, one was from the duodenal tissue sample of an adult, and 184 were from pediatric diarrheal stools. All 187 H,-requiring Campylobacter isolates were characterized as C. concisus on the basis of phenotypic testing. Ninety-two isolates were tested and confirmed as C. concisus by DNA-DNA hybridization (Lastovica et al., 1993). The mean age of the pediatric patients was 24 months (range, 4 days to 11 years). Twenty-one percent of the stools were watery; the rest were loose. Blood was present in 14% of the stools and fecal leukocytes in 27%. Preexisting clinical problems in 18% of the patients included anemia, metabolic disturbances, tuberculosis, and kwashiorkor. Additional data are provided in Table 4. Istavan et al. (2004) detected and characterized a membrane-bound hemolytic phospholipase in clinical strains of C. concisus isolated from Australian children with gastroenteritis. The presence of this potential virulence factor suggests C. concisus is an opportunistic pathogen. Danish investigators who used molecular techniques have proposed subpopulations of C. concisus (“genomospecies”) that have more pathogenic potential than others (Aabenhus et al., 2002). They found some genomospecies were more likely to be found in stools of patients with diarrhea, whereas others were more likely to be detected only in healthy controls (Aabenhus et al., 2005). Thirty C. concisus isolates from pediatric stools were tested against eight antimicrobial agents, and ciprofloxacin was the most effective agent examined (Greig et al., 1993). All strains were susceptible to tetracycline, ampicillin, and gentamicin. Activity of the cephalosporins was variable. In a study of 457 pediatric C. concisus strains isolated over the period 1998 to 2005, resistance to ciprofloxacin increased from 7 to 18%, and erythromycin resistance increased from 5 to 22% (Moore et al., 2006). Twenty Belgian C. concisus strains were susceptible to tetracycline and am-
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
picillin; and resistance to gentamicin, ciprofloxacin, and erythromycin was 5% (Vandenberg et al., 2006). C. gracilis and “B.” ureolyticus are both associated with periodontal disease but have been isolated from other sites. C. gracilis is found in deep-seated abscesses, particularly in the head and neck region, and “B.” ureolyticus from more superficial lesions such as ulcers, soft tissue infections, and urethritis (Akhtar and Eley, 1992; Johnson and Finegold, 1987; Johnson et al., 1985; Vandamme et al., 1995). C. mucosalis can be recovered at high concentrations (up to 10* CFU/g) from the diseased intestinal mucosa of porcine intestinal adenomatosis cases, but experimental infection does not always cause the disease (Lawson and Rowland, 1984). There is no known association of C. mucosalis and human disease.
CAMPYLOBACTER SPUTORUM Although not strictly hydrogen requiring, C. sputorum shares many of the features with the foregoing species, notably in being catalase negative, and it also forms part of the human gingival flora. C. sputorum was considered to have three biovars: biovar sputorum, living in the mouth and intestinal tract; biovar bubulus, living in the genital tract of healthy cattle and sheep; and biovar fecalis, found mainly in the feces of sheep and cattle. The biochemical differences between biovar bubulus and biovar sputorum are unreliable, and biovar bubulus is no longer recognized (On et al., 1998). However, a group of catalasenegative, urease-positive Campylobacter strains isolated from cattle feces and from a patient with enteritis in Canada were found to be C. sputorum and now form a new biovar, paraureolyticus (On et al., 1998). C. sputorum biovar sputorum has been isolated from the human lung and from abscesses of the groin and axillary areas (Borczyk et al., 1987a; On et al., 1992), from gastroenteritis patients (Table 3 ) , and from the blood of a knee abscess patient (Tee et al., 1998b).
ARCOBACTER SPECIES The genus Arcobacter was originally described as aerotolerant campylobacters, but after DNA-rDNA and DNA-DNA hybridization studies, these organisms were moved from the genus Campylobacter to the genus Arcobacter (Vandamme and De Ley, 1991). Arcobacters are morphologically similar to campylobacters, but they differ sharply in that they grow over a wide temperature range (15 to 42°C) (Lehner
137
et al., 2005) and under aerobic atmospheric conditions, which Campylobacter and Helicobacter spp. cannot. There are six recognized Arcobacter species: A. butzleri, A. cryaerophilus, A. nitrofigilis, A. skirrowii, A. cibarius, and A. halophilus (Snelling et al., 2006). The type species, A. nitrofigilis, originally isolated from plant roots, has not been isolated from humans or animals. A. skirrowii has been isolated from aborted cattle, sheep and pig fetuses, and humans. A. butzleri was first described by Kiehlbauch et al. (1991a), who recognized two distinct genetic groups among 78 isolates of aerotolerant “campylobacters” referred to the CDC in the United States. Group 1 contained the type strain of C. cyaerophila (now A. cryaerophilus). Group 2 was named C. butzleri (now A. butzleri). Biochemical differences of these two species are outlined in Table 1. A. cyaerophilus and A. butzleri grow poorly or not at all at 42°C on blood agar plates, but they will grow on MacConkey plates and are generally resistant to cephalothin (Table 1).A. cibarius (Houf et al., 2005) was isolated from chicken carcasses, and A. halophilus (Donachie et al., 2005) was isolated from a hypersaline lagoon. Epidemiology Livestock animals and surface and groundwaters represent a significant source of Arcobacter (Lehner et al., 2005). One hundred strains of A. butzleri and 41 strains of A. butzleri-like or Arcobacter spp. were isolated from six drinking water plants in Germany (Jacob et al., 1998). Two strains of A. butzleri were isolated from a contaminated well in Idaho (Moore et al., 2006). Survival studies, conducted at 5”C, a temperature typical of groundwater, indicated that A. butzleri can remain viable for up to 16 days in groundwater (Rice et al., 1999). A. cryaerophilus was first isolated from porcine, equine, and bovine feces and from aborted porcine and bovine fetuses (Boudreau et al., 1991; Ellis et al., 1977), and from urban sewage (Stamp et al., 1993). In a study of 15 broiler chickens obtained from a poultry abattoir and 10 broiler chickens from a supermarket, all 25 carcasses yielded A. butzleri. Three supermarket and 10 abattoir carcasses also carried A. cyaerophilus, and 2 abattoir carcasses carried A. skirrowii (Atabay et al., 1998). A Canadian study indicated that poultry appears to be a major reservoir for A. butzleri: 121 of 125 broiler chicken carcasses examined were positive for A. butzleri after abattoir processing. A. butzleri was recovered from whole chicken and ground chicken and turkey samples from retail stores (Lammerding et al., 1996). Arcobacters also may contaminate the poultry-processing environment (Gude et
138
LASTOVICA AND ALLOS
al., 2005). In a study of 130 raw meat samples, a strain of A. butzleri was identified in a pork sample (Zanetti et al., 1996). The organisms may be transmitted to piglets both vertically (i.e., in utero) and horizontally (Ho et al., 2006b). Characterization of Isolates
A multiplex PCR assay has been developed to identify Arcobacter isolates and to distinguish A. butzleri from other arcobacters (Harmon and Wesley, 1997). On the basis of 23s rRNA gene polymorphisms, Hurtado and Owen (1997) have proposed a PCR scheme for the rapid identification and speciation of Campylobacter and Arcobacter spp. Rhotyping and restriction fragment-length polymorphism analysis have been reported to be able to distinguish between A. cryaerophilus and A. butzleri (Kiehlbauch et al., 1991b). Clinical Features
Arcobacter butzleri can produce diarrhea and associated gastrointestinal symptoms in humans. A. butzleri was the fourth most common “atypical” campylobacter isolated and was associated with persistent, watery diarrhea in a study of more than 65,000 stool samples (Vandenberg et al., 2004). A. butzleri has been isolated from human diarrhetic stools, blood cultures, and the peritoneal fluid of appendicitis patients (Kiehlbauch et al., 1991). A. butzleri was isolated from the feces of diarrheic children in Thailand (Taylor et al., 1991) and South Africa (Samie et al., 2007) (Table 3). Underlying illness such as anemia, tuberculosis, and pneumonia were present in 40% of these children, whose average age was 19 months. Additional clinical details are provided in Table 4. In a survey of fecal samples in Belgium, Desiste and colleagues (1998) found that A. butzleri made up 7.4% of Carnpylobacter and Arcobacter isolations. The absence of fever (temperature >38”C) was noted in most patients, but watery diarrhea occurred in 13 of the 16 patients, and vomiting and abdominal pain occurred in about half of them. Neither blood nor inflammatory exudates was detected by microscopy of stools. Half the patients were treated with amoxicillin-clavulanate and rapidly recovered; the remainder improved spontaneously with conservative management. An outbreak of A. butzleri infection affected 10 Italian children, aged 2 to 5 years, none of whom had diarrhea or fever, but all had episodes of recurrent abdominal cramps. Each episode lasted up to 2 h and occurred several times a day for up to 10 days
(Vandamme et al., 1992a). These children felt well between attacks, and the illness was self-limiting. All isolates belonged to a single serogroup and had identical protein profiles. Most of the children seroconverted to this strain. The sequential timing of these infections suggested person-to-person spread. A. butzleri was isolated from the blood of a preterm neonate, probably infected in utero (On et al., 1995b). Treatment with penicillin and cefotaxime was successful in resolving the infection. A. butzleri appears to be pathogenic for nonhuman primates. It was found in 14 (6%) of 222 nonhuman primates with diarrhea. A. butzleri was isolated from colonic specimens obtained at necroscopy from 3 (4%) of 76 macaques. All three animals had active colitis (Anderson et al., 1993). The stools of 7 (39%) of 18 infant Macaca nemestrina monkeys (apparently asymptomatic), when cultured weekly from birth to 1year of age, also yielded A. butzleri (Russell et al., 1992). A. cryaerophilus has been isolated from human diarrhetic stools and blood cultures (Dediste et al., 1998; Kiehlbauch et al., 1991). An A. cyaerophilus strain was isolated from the blood of a 72-year-old uremic woman with hematogenous pneumonia that resulted from an infected atrioventricular fistula (Hsueh et al., 1997). She had experienced a period of diarrhea 2 months before this episode, and she was successfully treated with ceftizoxime and tobramycin. A 73-year-old man developed persistent diarrhea attributed to A. skimowii infection (Wybo et al., 2004), an organism rarely isolated from humans. A. butzleri is generally susceptible to fluoroquinolones, aminoglycosides, and tetracyclines; >20% of isolates are resistant to ampicillin and erythromycin (Kabeya et al., 2004; Vandenberg et al., 2006). Pathogenicity Neonatal piglets have been used as models to determine relative pathogenicities based on fecal shedding and tissue colonization of Arcobacter (Wesley et al., 1996). One-day-old cesarean-derived colostrum-deprived piglets were infected by mouth with three field strains and type strains of A. butzleri, A. cryaerophilus, and A. skirrowii. Arcobacter spp. were detected at least once in rectal swab samples of all but one of the experimentally infected piglets. At necroscopy, Arcobacter spp. were cultured from the liver, kidney, ileum, or brain tissue of two infected piglets, but no severe gross pathology was noted. These data suggest that Arcobacter spp., particularly A. butzleri, can colonize neonatal pigs (Wesley et al., 1996). Two recent reviews summarize clinical presentation, pathogenicity, and other aspects of Arco-
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
bacter infection (Ho et al., 2006a; Houf and Stephan, 2007).
HELICOBACTER CINAEDI AND HELICOBACTER FENNELLIAE Initially, H. cinaedi and H. fennelliae were called CLOs (Campylobacter-like organisms). Helicobacter cinaedi derives from the Latin word cinaedi, “of a homosexual.” C. fennelliae was named after Cynthia Fennell, the microbiologist who first isolated these organisms from the rectal swabs of homosexual men. In a taxonomic revision, the names of these organisms were subsequently amended to Helicobacter cinaedi and Helicobacter fennelliae (Vandamme et al., 1991). Microbiology
H. cinaedi and H. fennelliae grow poorly if at all under microaerobic conditions and require H,-enhanced microaerobic conditions for optimum growth. These bacteria grow at 37°C but grow poorly or not at all at 42°C (Table 1).H. fennelliae may be differentiated from H. cinaedi by SDS-PAGE profiles (On et al., 1991), serology (Flores et al., 1989), and differences in aryl sulfatase activity (Burnens and Nicolet, 1993) (Table 1).A useful diagnostic test is the smell of a mature growth of H. fennelliae, which has an odor similar to bleach and which is absent in H. cinaedi and campylobacters (Fennell et al., 1984). Although Campylobacter and Arcobacter species form domed colonies on freshly prepared agar plates, both H. cinaedi and H. fennelliae produce a flat, spreading growth, without discrete colonies (Fennell et al., 1984). This growth may be missed on primary isolation plates, particularly if the domed colonies of a Campylobacter species are present as well. H. fennelliae is often coisolated with C. jejuni subsp. jejuni, C. jejuni subsp. doylei, or C. upsaliensis (Albert et al., 1992; Lastovica, 2006). In cases of suspected mixed infection, extreme care must be taken to separate the noncolonial spreading growth of H. fennelliae or H. cinaedi from the domed colonies of Campylobacter (Fig. 2). Strain Characterization Thirty-four human and animal strains of H. cinaedi and four animal and human strains of H. fennelliae were characterized by phenotypic and molecular tests (Kiehlbauch et al., 1995). Results indicated that most isolates of H. cinaedi formed a single group, both phenotypically and by DNA-DNA hy-
139
bridization studies. Subgroups were distinguishable by ribotyping. Two human H. fennelliae strains had similar but different ribotyping patterns from those of the type strains and from each other, and had no bands in common with either of the H. fennelliae animal strains. Molecular phylogenic analysis of the 16s rRNA gene of 3 1 South African H. fennelliae clinical isolates indicated that they have a closer affinity to a described Australian strain than to the type strain, NCTC 11612 (Lastovica and Smuts, 2007). Epidemiology and Clinical Features Enteric helicobacters were first recognized as human pathogens when they were isolated from 26 of 158 stools of homosexual men with diarrhea or proctitis examined at a sexually transmitted disease clinic (Lentzsch et al., 2004). These organisms were also isolated from 6 of 75 asymptomatic homosexual men but were not isolated from 150 heterosexual men and women (Lentzsch et al., 2004). In a study of the diarrheal stools of homosexual or bisexual men in Baltimore and Washington, DC, 9 (27%) of 33 patients had Campylobacter or Helicobacter in their stools, 2 patients had H. cinaedi infections, and 1 had a CL03 strain (Laughton et al., 1988). Two CL03 strains were isolated from the stools of homosexual men with proctitis (Flores et al., 1989; Totten et al., 1985). CLOs have been isolated from up to 8% of the stools of homosexual men with diarrhea or proctitis (Laughton et al., 1988). In a California study, CLOs were not identified in the stools or colonic brushings of 27 homosexual men with diarrhea (Wilcox et al., 1990). The clinical features of CLO infections, including diarrhea, abdominal cramps, and hematochezia, were similar to those in C. jejuniinfected patients (Quinn et al., 1984). CLO infections also produced fever, anal discharge, and pain (Quinn et al., 1984). Sigmoidoscopic examinations of infected patients indicated mucosal bleeding and ulcers. Fecal leukocytes were present in most patients, and histological examination revealed crypt abscesses and polymorphonuclear leukocytes scattered through the lamina propria. A subsequent report described HIVinfected patients with H. cinaedi in their stools who experienced chronic but mild symptoms lasting several weeks (Grayson et al., 1989), but without blood or polymorphonuclear leukocyte cells in the stools. Diarrhea consisted of three to four loose stools per day and was not associated with systemic illness. In addition to gastroenteritis, H. cinaedi and H. fennelliae can cause bacteremia, particularly in HIVpositive patients (Burman et al., 1995). Chills, lowgrade fever, lethargy, and malaise are common in patients with H. cinaedi bacteremia, and not all have
140
LASTOVICA AND ALLOS
preceding gastrointestinal symptoms. A homosexual man with AIDS developed recurrent H. cinaedi bacteremia during 2 months of intermittent diarrhea, and fecal incontinence. His blood cultures yielded a second organism that was identified as H. fennelliae, and cultures were consistently positive, despite treatment with ciprofloxacin and trimethoprim-sulfamethoxazole. The patient died 4 months later, although the role of H. cinaedi and H. fennelliae was not clearly defined (Ng et al., 1987). Another AIDS patient developed bacteremia due to H. fennelliae (Kemper et al., 1993). In a survey of 394 blood culture Campylobacter isolates from patients in England and Wales, 2 (0.50%) isolates were H. fennelliae and 1 (0.25%) was H. cinaedi (Skirrow et al., 1993). A study of 221 pediatric blood culture isolates in a pediatric hospital in South Africa indicated that 7 (3.2%) strains were H. fennelliae and 5 (2.3%) were H. cinaedi (Lastovica, 1996). At this same hospital, 6,006 Carnpylobacter or Helicobacter isolates were obtained from 26,933 diarrheal stools (Table 3). There were 337 strains (5.61%) of H. fennelliae and 51 (0.95%) of H. cinaedi detected (Table 3). The average age of the infected children was 18 months (range, 2 weeks to 11 years). Seventy-three percent of the stools contained blood, and 8% had fecal leukocytes. More than 2.5% of these patients had underlying illnesses such as anemia, pneumonia, metabolic disturbances, and nutritional problems. Additional clinical details are listed in Table 4. These reports suggest that H. cinaedi and H. fennelliae infections may occur more frequently than previously thought in the immunocompetent and heterosexual populations. Gastroenteritis due to H. cinaedi or H. fennelliae have been reported in heterosexual men, women, and children in various studies (Burnens et al., 1993; Grayson et al., 1989; Lastovica, 2006; Uckay et al., 2006; Wilcox et al., 1990). The source of human infection of these organisms is not known, although apparent transmission from animals has been recorded. H. cinaedi was isolated from the cerebrospinal fluid and blood of a 5day-old neonate whose mother had a mild diarrheal illness during the third trimester of her pregnancy. The mother had cared for pet hamsters during the first two trimesters of her pregnancy (Orlicek et al., 1993). H. cinaedi has been isolated from 72% of commercially available hamsters (Gebhart et al., 1989) and from the stools of diarrheic dogs (Burnens et al., 1992). Pathogenesis Evidence for the pathogenicity of H. cinaedi and H. fennelliae comes from several sources. The asso-
ciation of these organisms with homosexual men who had proctitis and enteritis, but not with asymptomatic men, suggests a causal relationship. Bacteremia in patients who are immunocompromised and the presence of fecal leukocytes indicate that these organisms play a pathogenic role. Flores et al. (1990) studied the effects of experimental H. cinaedi and H. fennelliae infection in infant macaque monkeys. When four monkeys were challenged with H. cinaedi, two animals subsequently developed diarrhea, and H. cinaedi was isolated from the stools and blood cultures of all four. When challenged with H. fennelliae, all four monkeys became bacteremic, and two animals developed diarrhea. H. fennelliae was isolated from their stools, and prolonged rectal colonization was observed in all animals examined (Flores et al., 1990). A putative virulence factor in H. cinaedi has been described by Taylor et al. (2003), who demonstrated the production of cytolethal distending toxin, which causes an irreversible cell-cycle block at the G2/M transition. Treatment Although H. cinaedi and H. fennelliae infections have not been fatal, to date, some patients have displayed a slow clinical response to antimicrobial therapy. Antimicrobial agents that have demonstrated in vitro activity against H. cinaedi and H. fennelliae include ampicillin, tetracycline, rifampin, nalidixic acid, chloramphenicol, and gentamicin (Flores et al., 1985). However, 28% of strains were resistant to erythromycin and clindamycin, and 17% were resistant to sulfamethoxazole in one study (Flores et al., 1985). Like campylobacters, these organisms are resistant to trimethoprim, and most of them are resistant to metronidazole. In South Africa, 13% of H. cinaedi and H. fennelliae pediatric stool and blood culture isolates were resistant to erythromycin (Moore et al., 2006). Fluoroquinolones may be the best treatment for severe or persistent H. cinaedi and H. fennelliae infection. Two patients with persistent H. cinaedi bacteremia were successfully treated with oral ciprofloxacin (Burman et al., 1995).
OTHER HELICOBACTER A N D CAMPYLOBACTER SPECIES
H. canis is a species that resembles H. fennelliae. It was originally found in dogs (with and without diarrhea), but it has also been isolated from a 5%year-old boy with diarrhea (Stanley et al., 1993). H. pullorum was described as a species isolated from poultry liver, duodenum, and cecum, and from hu-
CHAPTER 7
-
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
mans with diarrhea (Ceelen et al., 2005; Stanley et al., 1994). H. westmeadii is another newly recognized species isolated from the blood of two AIDS patients with febrile diarrhea in Australia (Trivett-Moore et al., 1997). H. rappini (formerly Flexispira rappini) is a fusiform bacterium with spiral periplasmic fibrils and bipolar tufts of sheathed flagella. These organisms were first isolated from aborted lambs, then subsequently from patients with diarrhea (Romero et al., 1988), from the blood cultures of a 7-year-old girl with pneumonia (Tee et al., 1998a), and from a 65year-old febrile man undergoing hemodialysis (Sorlin et al., 1999). Solnick (2003) has provided a comprehensive review article on Helicobacter species other than H. pylori. In 2000, Logan et al. reported the isolation and characterization of C. lanienae sp. nov. from healthy abattoir workers exposed to cattle or pig carcasses. In 2001, Lawson et al. recovered C. hominis sp. nov. from the feces of healthy humans. In 2004, Foster at al. isolated C. insulaenigrae sp. nov. from three seals and a porpoise. The clinical relevance of these newly recognized Campylobacter species has yet to be determined. PCR assays for the simultaneous detection and differentiation of the genera and individual species of Campylobacter, Helicobacter, and Arcobacter may make this task easier (Hill et al., 2006; Marshall et al., 1999).
CONCLUSION New species of Campylobacter and related genera are being identified on a regular basis. Many of these “atypical” campylobacters may play a greater role in causing human and animal disease than previously recognized. Because methods originally formulated for the isolation of C. jejuni will often fail to support the growth of non-jejuni, non-coli Campylobacter species, these fastidious organisms are most likely underdetected in clinical specimens. Appreciation and application of an efficient protocol is essential for the isolation of non-jejuni, non-coli Campylobacter species for surveillance, epidemiological, and other studies. Reservoirs of newly described non-jejuni, non-coli Campylobacter species have been found in animals such as pigs, cattle, dogs, foxes, and rodents. Nonmammalian species such as birds and shellfish have recently been implicated as reservoirs for these organisms, and surface and groundwater are known to harbor non-jejuni, non-coli Campylobacter species and related organisms. At present, the role that these newly described Campylobacter species play in the disease process is not understood. Additional research and surveillance, such as that provided by FoodNet (Allos et al., 2004), is required to
141
better define the prevalence, persistence, and pathogenicity of emerging campylobacters and related genera.
REFERENCES Aabenhus, R., S. L. On, B. L. Siemer, H. Permin, and L. P. Andersen. 2005. Delineation of Campylobacter concisus genomospecies by amplified fragment length polymorphism analysis and correlation of results with clinical data. J. Clin. Microbiol. 43: 5091-5096. Aabenhus, R., H. Permin, S. L. On, and L. P. Andersen. 2002. Prevalence of Campylobacter concisus in diarrhoea of immunocompromised patients. Scand. J. Infect. Dis. 34:248-252. Akhtar, N., and A. Eley. 1992. Restriction endonuclease analysis and ribotyping differentiate genital and nongenital strains of Bacteroides ureolyticus. J. Clin. Microbiol. 30:2408-2414. Albert, M. J., W. Tee, A. Leach, V. Asche, and J. L. Penner. 1992. Comparison of a blood free medium and a filtration technique for the isolation of a Campylobacter spp. from diarrhoea1 stools of hospitalized patients in central Australia. J. Med. Microbiol. 37:176-179. Allos, B. M. 2001. Campylobacter jejuni infections: update on emerging issues and trends. Clin. Infect. Dis. 32:1201-1206. Allos, B. M., R. J. Moore, P. M. Griffin, and R. V. Tauxe. 2004. Surveillance for sporadic foodbourne disease in the 21st century: the FoodNet perspective. Clin. Infect. Dis. 38(Suppl 3):S115s120. Alvarez, 0. A., F. Vanegas, G . L. Maze, G. W. Gross, and M. Lee. 2000. Polymicrobial cholangitis and liver abscess in a patient with the acquired immunodeficiency syndrome. South. Med. J. 93:232-234. Anderson, K. F., J. A. Kielbauch, D. C. Anderson, H. M. McClure, and I. K. Wachmuth. 1993. Arcobacter (Campylobacter) butzleri-associated diarrheal illness in a non-human primate population. Infect. Immun. 61:2220-2223. Atabay, H. I., J. E. Corry, and S. L. On. 1998. Diversity and prevalence of Arcobacter spp. in broiler chickens. J. Appl. Microbiol. 84:1007-1016. Bar, W., G. Mirquez de Bar, H.-M. Nitschke, A. Schiessler, G. Mauff, A. Goldmann, B. Steinbrueckner, G. Harter, and M. Kist. 1998. Endocarditis associated with Campylobacter fetus, p. 162-165. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms.Institute of Child Health, Cape Town, South Africa. Benjamin, J., S. Leaper, R. J. Owen, and M. B. Skirrow. 1983. Description of Campylobacter laridis, a new species comprising the nalidixic acid resistant thermophilic Campylobacter (NARTC) group. Cur. Microbiol. 8:231-238. BCzian, M . C., G. Ribou, C. Barberis-Giletti, and F. Mtgraud. 1990. Isolation of a urease positive thermophilic variant of Campylobacter lari from a patient with urinary tract infection. Eur. J. Clin. Microbiol. Infect. Dis. 9:895-897. Blaser, M. J., D. N. Taylor, and R. A. Feldman. 1983. Epidemiology of Campylobacter jejuni infections. Epidemiol. Rev. 5: 157-176. Bokkenheuser, V. 1970. Vibrio fetus infection in man. Ten new cases and some epidemiological observations. Am. J. Epidemiol. 91:400-409. Bolton, F. J., D. Coates, D. N. Hutchinson, and A. F. Godfree. 1987. A study of thermophilic Campylobacter in a river system. J. Appl. Bacteriol. 62:167-176. Borczyk, A., H. Lior, A. McKeown, and H. Svendsen. 1987a. Isolations of Campylobacter sputorum associated with human in-
142
LASTOVICA AND ALLOS
fection, p. 166-167. In B. Kaijser, and E. Falsen (ed.), Campylobacter N . University of Goteborg, Goteborg, Sweden. Borczyk, A., S. Thompson, D. Smith, and H. Lior. 1987b. Waterborne outbreak of Campylobacter laridis-associated gastroenteritis. Lancet i:164-165. Boudreau, M., R. Higgins, and K. R. Mittal. 1991. Biochemical and serological characterization of Campylobacter cryaerophila. J. Clin. Microbiol. 2954-58. Bourke, B., V. L. Chan, and P. Sherman. 1998. Campylobacter upsaliensis: waiting in the wings. Clin. Microbiol. Rev. 11:440449. Bourke, B., P. M., Sherman, D. Woodward, H. Lior, and V. L. Chan. 1996. Pulsed-field gel electrophoresis indicates genotypic heterogeneity among Campylobacter upsaliensis strains. FEMS Microbiol. Lett. 14357-61. Bracikowski, J. P., I. E. Hess, and M. F. Rein. 1984. Campylobacter osteomyelitis. Southern Med. J. 77:1611-1613. Brennhoud, O., G. Kapperud, and G. Langeland. 1992. Survey of thermotolerant Campylobacter spp. and Yersinia spp. in three surface water sources in Norway. Int. J. Food Microbiol. 15: 327-328. Breynaert, J., P. Vandamme, and S. Lauwers. 1998. Review of 9 cases of human infection with Campylobacter hyointestinalis, p. 428-431. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. Institute of Child Health, Cape Town, South Africa. Brown, W. J., and R Sautter. 1977. Campylobacter fetus septicemia with concurrent salpingitis. J. Clin. Microbiol. 6:72-75. Bruneau, B., L. Burc, C. Bizet, N. Lambert-Zechovsky, and C. Branger. 1998. Purulent pleurisy caused by Campylobacter lari. Eur. J. Clin. Microbiol. Infect. Dis. 17:185-188. Burman, W. J., D. L. Cohn, R. R. Reeves, and M. L. Wilson. 1995. Multifocal cellulitis and monoarticular arthritis as manifestations of Helicobacter cinaedi bacteremia. Clin. Infect. Dis. 20564570. Burnens, A. P., B. Angeloy-Wick, and J. Nicolet. 1992. Comparison of Campylobacter carriage rates in diarrheic and healthy pet animals.]. Vet. Med. Ser. B 39:175-180. Burnens, A. P., and J. Nicolet. 1993. Three supplementary diagnostic tests for Campylobacter species and related organisms. J. Clin. Microbiol. 31:708-710. Burnens, A. P., J. Stanley, V. B. Schaad, and J. Nicolet. 1993. Novel Campylobacter-like organism resembling Helicobacter fennelliae isolated from a boy with gastroenteritis and from dogs. J. Clin. Microbiol. 31:1916-1917. Butzler, J. P., and M. B. Skirrow. 1979. Campylobacter enteritis. Clin. Gastroenterology 8:737-765. Carbone, K. M., M. C. Heinrich, and T. C. Quinn. 1985. Thrombophlebitis and cellulitis due to Campylobacter fetus ssp. fetus. Medicine 64:244-250. Carter, J. E., and N. Cimolai. 1996. Hemolytic-uremic syndrome associated with acute Campylobacter upsaliensis gastroenteritis. Nephron 74:489. (Letter.) Ceelen, L., A. Decostere, G. Verschraegen, R. Ducatelle, and F. Haesebrouck. 2005. Prevalence of Helicobacter pullorum among patients with gastrointestinal disease and clinically healthy persons. J. Clin. Microbiol. 43:2984-2986. Centers for Disease Control. 198 1. Campylobacter sepsis associated with “nutritional therapy”-California. Morb. Mortal. Wkly. Rep. 30:294-295. Chambers, S . T., S. C. Morpeth, and H. M. Laird. 2005. Campylobacter fetus prosthetic hip joint infection: successful management with device retention and review. J. Infect. 50:258261. Collins, H. S., A. Blevins, and E. Baxter. 1964. Protracted bacteremia and meningitis due to Vibrio fetus. Arch. Intern. Med. 113~361-364.
Cone, L. A., P. B. Dreisbach, J. Hirschberg, C. Shekar, L. P. Dreisbach, and W. Salatich. 2003. Cellulitis and septic arthritis caused by Campylobacter fetus and Campylobacter jejuni: report of 2 cases and review of the literature. 1. Clin. Rheumatol. 9:362369. Costas, M., R. J. Owen, and P. H. J. Jackman. 1987. Classification of Campylobacter sputorum and allied campylobacters based on numerical analysis of electrophoretic patterns. Syst. Appl. Microbiol. 9:125-131. Da Silva Tatley, F., A. J. Lastovica, and L. M. Steyn. 1992. Plasmid profiles of “Campylobacter upsaliensis” isolated from blood cultures and stools of paediatric patients. J. Med. Microbiol. 37% 14. David, J., R. M. Nasser, J. W. Goldberg, K. D. Reed, and M. D. Earll. 2005. Bilateral prosthetic knee infection by Campylobacter fetus. J. Arthroplasty 20:401-405. Davies, A. P., C. J. Beghart, and S. A. Meric. 1984. Campylobacter associated chronic diarrhea in a dog. Am. J. Vet. Med. 184:469471. Dediste, A., A. Aeby, A. Ebraert, L. Vlaes, R. Tridiani, 0. Vandenberg, J.-M. Devaster, P. Vandamme, and J.-P. Butzler. 1998. Arcobacter in stools: clinical features, diagnosis and antibiotic susceptibility. p. 436-439. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. Institute of Child Health, Cape Town, South Africa. Devlin, H. R., and L. McIntyre. 1983. Campylobacter fetus subsp fetus in homosexual males. J. Clin. Microbiol. 18:999-1000. Dickgiesser, N., G. Kasper, and W. Kihm. 1983. Campylobacter fetus ssp. fetus bacteraemia: a patient with liver cirrhosis. lnfection 11:288. Donachie, S. P., J. P. Bowman, S. L. On, and M. Alam. 2005. Arcobacter halophilus sp. nov., the first obligate halophile in the genus Arcobacter. Int. J. System. Evol. Microbiol. 55:12711277. Dronda, F., I. Garcia-Arata, E. Navas, and L. de Rafael. 1998. Meningitis in adults due to Campylobacter fetus subspecies fetus. Clin. Infect. Dis. 27:906-907. Edmonds, P., C. M. Patton, T. J. Barrett, G. K. Morris, A. G. Steigerwalt, and D. J. Brenner. 1985. Biochemical and genetic characteristics of atypical Campylobacter fetus subsp. fetus strains isolated from humans in the United States. J. Clin. Microbiol. 21:936-940. Edmonds, P., C. M. Patton, P. M. Griffin, T. J. Barett, G. P. Schmid, C. N. Baker, M. A. Lambert, and D. J. Brenner. 1987. Campylobacter hyointestinalis associated with human gastrointestinal disease in the United States. J. Clin. Microbiol. 25:685691. Ellis, W. A., S. D. Neill, J. J. O’Brien, H. W. Ferguson, and J. Hanna. 1977. Isolation of spirillium/vibrio-like organisms from bovine fetuses. Vet. Rec. 100:45 1-452. Engberg, J., D. D. Bang, R. Aabenhus, F. M. Aarestrup, V. Fussing, and P. Gerner-Smidt. 2005. Campylobacter concisus: an evaluation of certain phenotypic and genotypic characteristics. Clin. Microbiol. Infect. 11:288-295. Engberg, J., S. L. On, C. S. Harrington, and P. Gerner-Smidt. 2000. Prevalence of Campylobacter, Arcobacter, Helicobacter, and Sutterella spp. in human fecal samples as estimated by a reevaluation of isolation methods for campylobacters.J. Clin Microbiol. 3 8:286-29 l. Etoh, Y., F. E. Dewhirst, B. J. Paster, A. Yamamoto, and N. Goto. 1993. Campylobacter showae sp. nov., isolated from the human oral cavity. lnt. J. Syst. Bacteriol. 43:631-639. Evans, T. G., and D. Riley. 1992. Campylobacter laridis colitis in a human immunodeficiency virus-positive patient treated with a quinolone. Clin. Infect. Dis. 15:172-173.
CHAPTER 7
-
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
Farrugia, D. C., S . J. Eykyn, and E. G. Smyth. 1994. Campylobacter fetus endocarditis: two case reports and review. Clin. Infect. Dis. 18:443-446. Fennell, C. J., P. A. Totten, T. C. Quinn, C. L. Patton, K. K. Holmes, and W. E. Stamm. 1984. Characterization of Campylobacter-like organisms isolated from homosexual men.]. Infect. Dis. 14958-66. Fennell, C. L., A. M., Rompalo, P. A. Totten, K. L. Bruch, B. M. Flores, and W. E. Stamm. 1986. Isolation of “Campylobacter hyointestinalis” from a human. J. Clin. Microbiol. 24:146-148. Fkrnandez, H., U. Fagundes, S . Ogatha, U. F. Neto, and U. S . Ogatha. 1997. Acute diarrhoea associated with Campylobacter jejuni subsp. doylei in Sao Paulo, Brazil. Infect. Dis. J. 16:10981099. (Letter.) Fick, R. B., R. Isturiz, and E. C. Cadman. 1979. Campylobacter fetus septic arthritis: report of a case. Yale J. Biol. Med. 52:339344. Figura, N., P. Gugliemetti, A. Zanchi, N. Partini, D. Armellini, P. F. Bayeli, M. Bugnoli, and S . Verdiani. 1993. Two cases of Campylobacter mucosalis enteritis in children.]. Clin. Microbiol. 3 1:727-728. Florent, A. 1959. Les deux vibriosis gknitales: la vibriose due 2 V. foetus venerialis et la vibriose d’origine intestinale due 2 V. foetus intestinalis. Meded. Veeartsenijsch.Rijksuniv. Gent 3:l-60. Flores, B. M., C. L. Fennell, K. K. Holmes, and W. E. Stamm. 1985. In vitro susceptibility of Campylobacter-like organisms to twenty antimicrobial agents. Antimicrob. Agents Chemother. 28: 188-191. Flores, B. M., C. L. Fennell, L. Kuller, M. A. Brondson, W. R. Morton, and W. Stamm. 1990. Experimental infection of pigtailed macaques (Macaca nemestrina) with Campylobacter cinaedi and Campylobacter fennelliae. Infect. Immun. 58:39473953. Flores, B. M., C. L. Fennell, and W. E. Stamm. 1989. Characterization of Campylobacter cinaedi and Campylobacter fennelliae antigens and analysis of human immune response. 1.Infect. Dis. 159~635-640. Font, C., A. Cruceta, A. Moreno, 0. Miro, B. Coll-Vinent, M. Almela, and J. Mensa. 1997. [A study of 30 patients with bacteremia due to Campylobacter spp.] Med. Clin. (Barc.). 108: 336-340. (In Spanish.) Foster, G., B. Holmes, A. G. Steigenvalt, P. A. Lawson, P.Thorne, D. E. Bryer, H. M. Ross, J. Xerry. P. M. Thompson, and M. D. Collins. 2004. Campylobacter insulaenigrae sp. nov., isolated from marine mammals. Int. I. Syst. Evol. Microbiol. 54:23692373. Fouts, D. E., E. F. Mongodin, R. E. Mandrel], W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S . C. Daugherty, R. J. Dodson, A. S . Durkin, R. Madupu, S . A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms form the genomes of multiple Campylobacter species. PLoS Biol. 3(1):e15. Fox, J. O., K. 0. Maxwell, N. S . Taylor, C. D. Runsick, P. Edmonds, and D. J. Brenner. 1989. “Campylobacter upsaliensis” isolated from cats as identified by DNA relatedness and biochemical features. 1. Clin. Microbiol. 27:2376-2378. Fujihara, N., S . Takakura, T. Saito, Y. Iinuma, and S . Ichiyama. 2006. A case of perinatal sepsis by Campylobacter fetus subsp. fetus infection successfully treated with carbapenem-case report and literature review. 1.Infect. 53:e199-e202. Garcia, M. M., M. D. Eaglesome, and C. Rigby. 1983. Campylobacters important in veterinary medicine. Vet. Bull. 53:793818.
143
Gaudreau, C., and F. Lamonthe. 1992. Campylobacter upsaliensis isolated from a breast abscess.]. Clin. Microbiol. 30:1354-1356. Gebhart, C. J., P. Edmonds, G. E. Ward, H. J. Kurtz, and D. J. Brenner. 1985. “Campylobacter hyointestinalis” sp. nov.: a new species of Campylobacter found in the intestines of pigs and other animals. 1. Clin. Microbiol. 21:715-720. Gebhart, C. J., C. L. Fennell, M. P. Murtaugh, and W. E. Stamm. 1989. Campylobacter cinaedi is the normal intestinal flora in hamsters. J. Clin. Microbiol. 27:1692-1694. Gebhart, C. J., M. P. Murtaugh, G. F. Lin, and G. E. Ward. 1990. Species-specific DNA probes for Campylobacter species isolated from pigs with proliferative enteritis. Vet. Microbiol. 24:367379. Gebhart, C. J., G. E. Ward, K. Chang, and H. J. Kurtz. 1983. Campylobacter hyointestinalis (new species) isolated from swine with lesions of proliferative ileitis. Am. J. Vet. Res. 44:361-367. Glunder, G., and S . Petermann. 1989. The occurrence and characterization of Campylobacter spp. in silver gulls ( L a m argentutus), three toed gulls (Rissa tridactyla), and house sparrows (Passer domesticus). Zentralbl. Veterinamed. B. 36:123-130. Goossens, H., B. A. J. Giesendorf, P. Vandamme, L. Vlaes, C. Van den Borre, A. Koeken, W. G. V. Quint, W. Blomme, P. Hanicq, D. S. Koster, H. Hofstra, J.-P. Butzler, and J. Van der Plas. 1995. Investigation of an outbreak of Campylobacter upsaliensis in day care centers in Brussels: analysis of relationships among isolates by phenotypic and genotypic typing methods. 1.Infect. Dis. 172:1298-1305. Goossens, H., B. Pot, L. Vlaes, C. Van den Borre, R. Van den Abbeele, C. Van Naelten, J. Levy, H. Gogniau, P. Marbehant, J. Verhoef, K. Kersters, J. P. Butzler, and P. Vandamme. 1990. Characterization and description of “Campylobacter upsaliensis” isolated from human feces. J. Clin. Microbiol. 28:10391046. Goossens, H., L. Vlaes, and J. P. Butzler, A. Adnet, P. Hanicq, S . N’Jufom, D. Massart, G . DeSchrijver, and W. Blomme. 1991. Campylobacter upsaliensis associated with canine infections. Lancet 337: 1486-1487. Gorkiewicz, G, G. Feieri, R. Zechner, and E. L. Zechner. 2002. Transmission of Campylobacter hyointestinalis from a pig to a human. 1. Clin. Microbiol. 40:2601-2605. Goudswaard, J., L. Sabbe, and W. te Winkel. 1995. Reactive arthritis as a complication of Campylobacter lari enteritis. 1.Infect. 31:171. (Letter.) Grayson, M. L., W. Tee, and B. Dwyer. 1989. Gastroenteritis associated with Campylobacter cinaedi. Med. ]. Aust. 150:214. Greig, A., D. Hanslo, E. Le Roux, and A. Lastovica. 1993. In-vitro activity of eight antimicrobial agents against pediatric isolates of C. concisus and C. mucosalis. Acta Gasteroenterol. Belg. 56(SuppI.):12. Gude, A., T. J. Hillman, C. R. Helps, V. M. Allen, and J. E. Corry. 2005. Ecology of Arcobacter species in chicken rearing and processing. Lett. Appl. Microbiol. 41:82-87. Guerrant, R. L., R. G. Lahita, W. C. Winn Jr., and R. B. Roberts. 1976. Campylobacteriosis in man: pathogenic mechanisms and review of 91 bloodstream infections. Am. 1.Med. 65584-592. Gurgan, T., and K. S . Diker. 1994. Abortion associated with Campylobacter upsaliensis.]. Clin. Microbiol. 32:93-94. Hald, B., and M. Madsen. 1997. Healthy puppies and kittens as carriers of Campylobacter spp., with special reference to Campylobacter upsaliensis. ]. Clin. Microbiol. 35:335 1-3352. Han, X. Y., J. J. Tarrand, and D. C. Rice. 2005. Oral Campylobacter species involved in extraoral abscess: a report of three cases.]. Clin. Microbiol. 43:2513-2515. Hanninen, M. L., L. Sarelli, A. Sukura, S . L. On, C. S . Harrington, P. Matero, and V. Hirvela-Koski. 2002. Carnpylobacter hyoin-
144
LASTOVICA AND ALLOS
testinalis subsp. hyointestinalis, a common Campylobacter species in reindeer. J. Appl. Microbiol. 92:717-723. Harmon, K. M., and I. V. Wesley. 1997. Multiplex PCR for the identification of Arcobacter and differentiation of Arcobacter butzleri from other arcobacters. Vet. Microbiol. 58:215-278. Harvey, S. M., and J. R. Greenwood. 1983. Relationships among catalase-positive campylobacters determined by deoxyribonucleic acid-deoxyribonucleic acid hybridization. Int. J. Syst. Bacteriol. 33:275-284. Heng, A. E., C. De Champs, B. Souweine, L. Guy, J. Sirot, and P. Deteix. 2002. Campylobacter fetus bacteraemia in a renal graft recipient. Nephrol. Dial. Transplant. 17:689-690. Herve, J., N. Aissa, P. Legrand, M. Sorkine, M. J. Calmette, A. Santin, E. Roupie, and B. Renaud. 2004. Campylobacter fetus meningitis in a diabetic adult cured by imipenem. Eur. J. Clin. Microbiol. Infect. Dis. 23:722-724. Hill, J. E., A. Paccagnella, K. Law, L. Melito, D. L.Woodward, L. Price, A.,H. Leung, L.-K. Ng, S. M. Hemmingsen, and S. H. Goh. 2006. Identification of Campylobacter spp. and discrimination from Helicobacter and Arcobacter spp. by direct sequencing of PCR-amplified cpn60 sequences and comparison to cpnDB, a chaperonin reference sequence database. J. Med. Microbiol. 55 :393-3 99. Ho, H. T., L. J. Lipman, and W. Gaastra. 2006a. Arcobacter, what is known and unknown about a potential foodbourne zoonotic agent! Vet. Microbiol. 115:l-13. Ho, T. K., L. J. Lipman, L. van der Graaf-van Bloois, M. van Bergen, and W. Gaastra. 2006b. Potential routes of acquisition of Arcobacter species by piglets. Vet. Microbiol. 114:123-133. Holst, E., C. Schalin, and P.-A. MHrdh. 1987. Isolation of Campylobacter spp. from the vagina, p. 167-168. In B. Kaijser and E. Falsen (ed.), Campylobacter N . University of Goteborg, Goteborg, Sweden. Houf, K., S. L. On, T. Coenye, J. Mast, J. Van Hoof, and P. Vandamme. 2005. Arcobacter cibarius sp. nov., isolated from broiler carcasses. Int. J. System. Evol. Microbiol. 55:713-717. Houf, K., and R. Stephan. 2007. Isolation and characterization of the emerging foodbourne pathogen Arcobacter from human stool. J. Microbiol. Methods 68:408-413. Hsueh, P.-R., L.-J. Teng, P.-C. Yang, S.-K. Wang, S.-C. Chang, S.-W. Ho. W.-C. Hsieh, and K.-T. Luh. 1997. Bacteraemia caused by Arcobacter cryaerophilus 1B. J. Clin. Microbiol. 35: 489-49 1. Hum, S., K. Quinn, J. Brunne, and S. L. On. 1997. Evaluation of a PCR assay for the identification and differentiation of Campylobacter fetus subspecies. Aust. Vet. J. 75:827-831. Hurtado, A, and R. J. Owen. 1997. A molecular scheme based on 23s rRNA gene polymorphisms for rapid identification of Campylobacter and Arcobacter species. ]. Clin. Microbiol. 35:24012404. Inglis, G. D., T. A. McAllister, H. W. Busz, L. J. Yanke, D. W. Morck, M. E. Olson, and R. R. Read. 2005. Effects of subtherapeutic administration of antimicrobial agents to beef cattle on the prevalence of antimicrobial resistance in Campylobacter jejuni and Campylobacter hyointestinalis. Ape. Environ. Microbiol. 71:3872-278 1. International Union of Microbiological Societies. 1991. Validation of the publication of new names and new combinations previously published outside the IJSB. Int. J. Syst. Bacteriol. 41:331. Istivan, T. S., P. J. Coloe, B. N. Fry, P. Ward, and S. C. Smith. 2004. Characterization of a haemolytic phospholipase (A2)activity in clinical isolates of Campylobacter concisus. J. Med. Microbiol. 53:483-493. Iwai, T., Y. Inoue, M. Umeda, Y. Huang, N. Kurihara, M. Koike, and I. Ishikawa. 2005. Oral bacteria in the occluded arteries of patients with Buerger disease. J. Vasc. Surg. 42:107-115.
Jacob, J., D. Woodward, I. Feuerpfeil, and W. M. Johnson. 1998. Isolation of Arcobacter butzleri in raw water and drinking water treatment plants in Germany. Zentralbl. Hyg. Umweltmed. 201: 189-198. Jensen, A. N., M. T. Andersen, A. Dalsgaard, D. L. Baggesen, and E. M. Nielsen. 2005. Development of real-time PCR and hybridization methods for detection and identification of thermophilic Campylobacter spp. in pig faecal samples. J. Appl. Microbiol. 99:292-300. Jirapongsananuruk, O., K. Wanotayan, W. Phongsamart, K. Chokephaibulkit, N. Visitsunthorn, V. Luangwedchakarn, N. Vanprapar, and P. Vichyanond. 2006. Recurrent Campylobacter lari bacteremia in X-linked agammaglobulinemia: a case report and review. Asian Pac. J. Allergy Immunol. 24:171-174. Johnson, C. C., and S. M. Finegold. 1987. Uncommonly encountered motile, anaerobic gram-negative bacilli associated with infection. Rev. Infect. Dis. 9:1150-1162. Johnson, C. C., J. F. Reinhardt, M. A. C. Edelstein, M. E. Mulligan, W. L. George, and S. M. Finegold. 1985. Bacteroides gracilis, an important anaerobic bacterial pathogen. J. Clin. Microbiol. 22:799-802. Johnson, W. M., and H. Lior. 1986. Cytotoxic and cytotonic factors produced by Campylobacter jejuni, Campylobacter coli and Campylobacter laridis. J. Clin. Microbiol. 24:275-28 1. Kabeya, H., S. Maruyama, Y. Morita, T. Ohsuga, S. Ozawa, Y. Kobayashi, M. Abe, Y. Katsube, and T. Mikami. 2004. Prevalence of Arcobacter species in retail meats and antimicrobial susceptibility of the isolates in Japan. Int. J. Food Microbiol. 90: 303-308. Kalkay, M. N., Z. S. Ayanian, E. A. Lehaf, and A. Baldi. 1983. Campylobacter-induced toxic megacolon. Am. J. Gastroenterol. 78557-559. Kanj, S. S., G. F. Araj, A. Taher, and L. B. Relier. 2001. Campylobacter fetus pericarditis in a patient with bets thalassemia case report and review of the literature. J. Clin. Microbiol. 7: 5 10-5 13. Kazawala, R. R., S. F. H. Jiwa, and A. E. Nkya. 1993. The role of management systems in the epidemiology of thermophilic campylobacters among poultry in the eastern zone of Tanzania. Epidemiol. Infect. 110:273-278. Kemper, C. A., P. Mickelson, A. Morton, B. Walton, and S. C. Deresinski. 1993. Helicobacter (Campylobacter) fennelliae-like organisms as an important but occult phase of bacteraemia in a patient with AIDS. J. Infect. 26:97-101. Kiehlbauch, J. A., D. J. Brenner, D. N. Cameron, A. G. SteigerWalt, J. M. Markowski, C. N. Baker, C. M. Patton, and I. K. Wachsmuth. 1995. Genotypic and phenotypic characterization of Helicobacter cinaedi and Helicobacter fennelliae strains isolated from humans and animals. J. Clin. Microbiol. 33:29402947. Kiehlbauch, J. A., D. J. Brenner, M. A. Nicholson, C. N. Baker, C. M. Patton, A. G. Steigerwalt, and I. K. Wachsmuth. 1991a. Campylobacter butzleri sp. nov. isolated from humans and animals with diarrheal illness. J. Clin. Microbiol. 29:376-3 85. Kiehlbauch, J. A., B. D. Plikaytis, B. Swaminathan, C. N. Cameron, and I. K. Wachsmuth. 1991b. Restriction fragment length polymorphisms in the ribosomal genes for specific identification and subtyping of aerotolerant Campylobacter species. J. Clin. Microbiol. 29:1670-1676. Klein, B. S, J. M. Vergeront, M. J. Blaser, P. Edmonds, D. J. Brenner, D. Janssen, and J. P. Davis. 1986. Campylobacter infection associated with raw milk: an outbreak of gastroenteritis due to Campylobacter jejuni and thermotolerant Campylobacter fetus subsp. fetus. JAMA 255:361-364. Klena, J. D., C. T. Parker, K. Knibb, J. C. Ibbitt, P. M. Devane, S. T. Horn, W. G. Miller, and M. E. Konkel. 2004. Differen-
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
tiation of Campylobacter coli, Campylobacter jejuni, Campylobacter lari, and Campylobacter upsaliensis by a multiplex PCR developed from the nucleotide sequence of the lipid A gene IpxA. J. Clin. Microbiol. 425549-5557. Laatu, M., H. Rautelin, and M. L. Hanninen. 2005. Susceptibility of Campylobacter hyointestinalis subsp. hyointestinalis to antimicrobial agents and characterization of quinolone-resistant strains. J. Antimicrob. Chemother. 55:182-187. Lammerding, A. M., J. E. Harris, H. Lior, D. E. Woodward, L. Cole, and C. A. Muckle. 1996. Isolation method for recovery of Arcobacter butzleri from fresh poultry and poultry products, p. 329-323. In D. G. Newell, J. M. Ketley and R. A. Feldman (ed.), Campylobacters, Helicobacters and Related Organisms. Plenum Press, New York, NY. La Scola, B., S. Chambourlier, and P. Bouillot. 1998. Campylobacter fetus ssp. fetus brain abscess. J. Infect. 37:309-310. (Letter.) La Scolea, L. J. 1985. Campylobacter fetus subsp. fetus meningitis in a neonate. Clin. Microbiol. Newslett. 7: 125-126. Lastovica, A. J. 1996. CampylobacterlHelicobacter bacteraemia in Cape Town, South Africa, 1977-1995, p. 475-479. In D. G. Newell, J. M. Ketley and R. A. Feldman (ed.), Campylobacters, Helicobacters and Related Organisms. Plenum Press, New York,
NY. Lastovica, A. J. 2006. Emerging Campylobacter spp.: the tip of the iceberg. Clin. Microbiol. News 28:49-55. Lastovica, A. J., and E. Le Roux. 2001. Efficient isolation of Campylobacter upsaliensis from stools. J. Clin. Microbiol. 39:422423. Lastovica, A. J., E. Le Roux, and M. Jooste. 1991. “Campylobacter upsaliensis” isolated from vervet monkeys. Microb. Ecol. Health Dis. 4(Suppl.):S87. Lastovica, A. J., E. Le Roux, and J. L. Penner. 1989. “Campylobacter upsaliensis” isolated from the blood cultures of pediatric patients. J. Clin. Microbiol. 27:657-659. Lastovica, A. J., E. Le Roux, R. Warren, and H. Klump. 1993. Clinical isolates of Campylobacter mucosalis. J. Clin. Microbiol. 31:2835-2836. Lastovica, A. J., and H. Smuts. 2007. A molecular study of Helicobacter fennelliae isolated from the stools and blood cultures of paediatric patients in Cape Town, South Africa (abstr). Abst. 107th Gen. Mtng. Am. SOC.Microbiol., Toronto, Ontario, Canada, 21 to 25 May 2007. Laughton, B. E., D. A. Druckman, A. Vernon, T. C. Quinn, B. F. Polk, J. F. Modlin, R. H. Yolken, and J. G. Bartlett. 1988. Prevalence of enteric pathogens in homosexual men with and without acquired immunodeficiency syndrome. Gastroenterology94: 984-993. Laughton, B. E., A. A. Vernon, D. A. Druckman, R. Fox, T. C. Quinnn, B. F. Polk, and J. G. Bartlett. 1988. Recovery of Campylobacter species from homosexual men. J. Infect. Dis. 158: 464-467. Lauwers, S., T. Devreker, R. Van Etterijck, J. Breynaert, A. Van Zeebroeck, L. Smekens, K. Kersters, and P. Vandamme. 1991. Isolation of Campylobacter concisus from human faeces. Microb. Ecol. Health Dis. 4(Suppl.):S91. Lawrence, R., A. F. Nibbe, and S. Lewis. 1971. Lung abscess secondary to Vibrio fetus malabsorption syndrome and acquired agammaglobulinemia. Chest 60: 191-194. Lawson, A. J., S. L. On, J. M. Logan, and J. Stanley. 2001. Campylobacter hominis sp nov., from the human gastrointestinal tract. Int. J. Syst. Evol. Microbiol. 51:651-660. Lawson, G. H. K., and A. C. Rowland. 1984. Campylobactersputorum subsp. mucosalis, p. 207-225. In J.-P. Butzler (ed.), Campylobacter Infection in Man and Animals. CRC Press, Boca Raton, FL.
145
Lehner, A., T. Tasara, and Stephan, R. 2005. Relevant aspects of Arcobacter spp. as potential foodborne pathogen. Int. J. Food Microbiol. 102: 127-1 35. Lentzsch, P., B. Rieksneuwohner, L. H. Wieler, H. Hotzel, and I. Moser. 2004. High-resolution genotyping of Campylobacter upsaliensis strains originating from three continents. J. Clin. Microbiol. 42:344 1-3448. Lindblom, G. B., M. Johny, K. Khalil, K. Mazhar, G. M. RuizPalacios, and B. Kaijser. 1990. Enterotoxigenicity and frequency of Campylobacter jejuni, C. coli and C. laridis in human and animal isolates from different countries. FEMS Microbiol. Lett. 54:163-167. Linton, D., R. J. Owen, and J. Stanley. 1996. Rapid identification by PCR of the genus Campylobacter and five Campylobacter species enteropathogenic for man and animals. Res. Microbiol. 147:707-718. Loades, C. J., L. E. Reiman, and C. W. Keevil. 2005. Development of an improved medium for the recovery of Campylobacter species avoiding common selective antibiotics, abstr. D-198. Abstr. 105th Gen. Mtng. Am. SOC.Microbiol., Washington, DC. Logan, J. M., J. A. Burens, D. Linton, A. J. Lawson, and J. Stanley. 2000. Campylobacter lanienae sp. nov., a new species isolated from workers in an abattoir. Int. J. Syst. Evol. Microbiol. 50: 865-8 72. Macfarlane, S., E. Furrie, G. T. Macfarlane, and J. F. Dillon. 2007. Microbial colonization of the upper gastrointestinal tract in patients with Barrett’s esophagus. Clin. Infect. Dis. 45:29-38. Marshall, S. M., P. L. Melito, D. L. Woodward, W. M. Johnson, F. G . Rodgers, and M. R Mulvey. 1999. Rapid identification of Campylobacter, Arcobacter, and Helicobacter isolates by PCRrestriction fragment length polymorphism analysis of the 16s rRNA Gene. J. Clin. Microbiol. 37:4158-4160. Maruyama, M., T. Tanaka, Y. Katsube, H. Nakanishi, and M. Nukina. 1990. Prevalence of thermophilic campylobacters in crows (Cowus lavaillantii, Cowus corone) and serogroups of the isolates. Jpn. J. Vet. Sci. 52:1237-1244. Matsheka, M. I., B. G. Elisha, A. J. Lastovica, and S. L. On. 2002. Genetic heterogeneity of Campylobacter concisus determined by pulsed field gel electrophoresis based macrorestriction profiling. FEMS Microbiol. Lett. 211:17-22. Matsheka, M. I., A. J. Lastovica, H. Zappe, and B. G. Elisha. 2006. The use of (GTG) oligonucleotide as an RAPD primer to type Campylobacter concisus. Lett. Appl. Microbiol. 42:600605. McFaydean, J., and S. Stockman. 1913. Report of the Departmental Committee Appointed by the Board of Agriculture and Fisheries to Inquire into Epizootic Abortion. Appendix to part 111: Abortion in Sheep, p. 1-29. His Majesty’s Stationery Office, London, United Kingdom. McGechie, D. B., T. B. Teoh, and V. W. Bamford. 1982. Campylobacter enteritis in Hong Kong and Western Australia, p. 1921. In D. G. Newell (ed.), Campylobacter: Epidemiology, Pathogenesis and Biochemistry. MTP Press, Lancaster, United Kingdom. MCgraud, F., and F. Bonnet. 1986. Unusual campylobacters in human feces. J. Infect 12:275-276. Mtgraud, F., D. Chevrier, N. Desplaces, A. Sedalfian, and J. L. Guesdon. 1988. Urease-positive thermophilic Campylobacter (Campylobacter laridis variant) isolated from an appendix and from human feces. J. Clin. Microbiol. 26:lOSO-1051. Mendelson, M. H., P. Nicholas, M. Malowany, and S. Lewin. 1986. Subdural empyema caused by Campylobacter fetus ssp. fetus. J. Infect Dis. 153:1183-1184. (Letter.) Miki, K., R Maekura, T. Hiraga, A. Hirotani, H. Hashimoto, S. Kitada, M. Miki, K. Yoshimura, N. Naka, M. Motone, T. Fujikawa, S. Takashima, R. Kitazume, H. Kanzaki, S. Nakatani,
146
LASTOVICA AND ALLOS
H. Watanuki, 0. Tagusari, J. Kobayashi, and M. Ito. 2005. Infective tricuspid valve endocarditis with pulmonary emboli caused by Campylobacter fetus after tooth extraction. Intern. Med. 44:1055-1059. Miller, V. A., R. Jenson, and J. J. Gilroy. 1959. Bacteremia in pregnant sheep following oral administration of Vibrio fetus. Am. J. Vet. Res. 20:677-679. Miller, W. G., S. L. On, G. Wang, S. Fontanoz, A. J. Lastovica, and R. E. Mandrell. 2005. Extended multilocus sequence typing system for Campylobacter coli, C. lari, C. upsaliensis, and C. helveticus. J. Clin. Microbiol. 43:23 15-2329. Miller, W. G., C. T. Parker, S. Heath, and A. J. Lastovica. 2007. Identification of genomic differences between Campylobacter jejuni subsp. jejuni and C. jejuni subsp. doylei at the nap locus leads to the development of a C. jejuni subspeciation multiplex PCR method. BMC Microbiol. 7:11. Minet, J., B. Grosbois, and F. MCgraud. 1988. Campylobacter hyointestinalis: an opportunistic enteropathogen?J. Clin. Microbiol. 26:2659-2660. Modolo, J. R., and R. Giuffrida. 2004. Campylobacter upsaliensis isolated from young dogs with and without diarrhea. Rev. SOC. Bras. Med. Trop. 37:72-73. Monno, R., M. Rendina, G. Ceci, C. Rizzo, I. Luzzi, A. Francavilla, G. Rizzo, and E. Ierardi. 2004. Campylobacter fetus bacteremia in an immunocompromised patient: case report and review of the literature. New Microbiol. 27:28 1-285. Montero, A., X. Corbella, J. A. L6pez, M. Santin, and I.-H. Bal16n. 1997. Campylobacter fetus-associated aneurysms: report of a case involving the popliteal artery and review of the literature. Clin. Infect. Dis. 24:1019-1021. Mooney, A., C. Byrne, M. Clyne, K. Johnson-Henry, P. Sherman, and B. Bourke. 2003. Invasion of human epithelial cells by Campylobacter upsaliensis. Cell Microbiol. 5:835-847. Mooney, A., M. Clyne, T. Curran, D. Doherty, B. Kilmartin, and B. Bourke. 2001. Campylobacter upsaliensis exerts a cytolethal distending toxin effect on HeLa cells and T lymphocytes. Microbiology 147:735-743. Moore, J. E., M. D. Barton, I. S. Blair, D. Corcoran, J. S. Dooley, S. Fanning, I. Kempf, A. J. Lastovica, C. J. Lowery, M. Matsuda, D. A. McDowell, A. McMahon, B. C. Millar, J. R. Rao, P. J. Rooney, B. S. Seal, W. J. Snelling, and 0. Tolba. 2006. The epidemiology of antibiotic resistance in Campylobacter. Microbes Infect. 8 :l 955- 196 6. Morey, F. 1996. Five years of Campylobacter bacteraemia in Central Australia, p. 491-494. In D. G. Newell, J. M. Ketley, and R. A. Feldman (ed.), Campylobacters, Helicobacters and Related Organisms. Plenum Press, New York, NY. Morooka, T., A. Umeda, M. Fujita, H. Matano, S. Fujimoto, K. Yukitake, Z. Amako, and T. Oda. 1996. Epidemiologic application of pulsed-field gel electrophoresis to an outbreak of Campylobacter fetus meningitis in a neonatal intensive care unit. Scand. J. Infect. Dis. 28:269-270. Musmanno, R. A., M. Russi, N. Figura, P. Guglielmetti, A. Zanchi, R. Signori, and A. Rossolini. 1998. Unusual species of campylobacters isolated in the Siena Tuscany area, Italy. New Microbiol. 21:15-22. Nachamkin, I., C. Stowell, D. Skalina, A. M. Jones, R. M. Roop, and R. M. Smibert. 1984. Campylobacter laridis causing bacteraemia in an immunosuppressed patient. Ann. Intern. Med. 101: 55-57. Neuzil, K., E. Wang, D. Haas, and M. J. Blaser. 1994. Persistence of Campylobacter fetus bacteraemia associated with absence of opsonizing antibodies. J. Clin. Microbiol. 32:1718-1720. Ng, V. L., W. K. Hadley, C. L. Fennell, B. M. Flores, and W. E. Stamm. 1987. Successive bacteremias with “Campylobacter cin-
aedi” and “Campylobacter fennelliae” in a bisexual male. J. Clin. Microbiol. 25:2008-2009. Ohya, T., and M. Nakazawa. 1992. Production and some properties of cytotoxins produced by Campylobacter species isolated from proliferative enteropathy in swine. J. Vet. Med. Sci. 54: 1031-1033. On, S. L. 1996. Identification methods for campylobacters, helicobacters and arcobacters. Clin. Microbiol. Rev. 9:405-422. On, S. L., B. Bloch, B. Holmes, B. Hoste, and P. Vandamme. 1995a. Campylobacter hyointestinalis subsp. lawsonii subsp. nov., isolated from the porcine stomach, and an emended description of Campylobacter hyointestinalis. Int. 1.Syst. Bacteriol. 45~767-774. On, S. L., F. Ridgewell, B. Cryan, and B. S. Azadian. 1992. Isolation of Campylobacter sputorum biovar sputorum from an axillary abscess. J. Infect. 24:175-179. On, S. L. W., H. I. Atabay, J. E. L. Corry, C. S. Harrington, and P. Vandamme. 1998. Emended description of Campylobacter sputorum and revision of its infrasubspecific (biovar) divisions, including C. sputorum biovar paraureolyticus, a urease-producing variant from cattle and humans. Int. 1. Syst. Bacteriol. 48: 195-206. On, S. L. W., R. J. Owen, A. J. Lastovica, M. Costas, and B. Lopez-Urquijo. 1991. Taxonomic study of Helicobacter (Campylobacter) fennelliae from clinical material by numerical analysis of one-dimensional electrophoretic protein patterns. Microb. Ecol. Health Dis. 4(Suppl.):S103. On, S. L. W., A. Stacey, and J. Smyth. 1995b. Isolation of Arcobacter butzleri from a neonate with bacteraemia. J. Infect. 31: 225-227. Orlicek, S. L., D. F. Welch, and T. L. Kuhls. 1993. Septicemia and meningitis caused by Helicobacter cinaedi in a neonate. J. Clin. Microbiol. 3 1569-57 1. Otero, J. de, C. Pigrau, M. Buti, and R Bartolome. 1994. Isolation of Campylobacter fetus subspecies fetus from a gluteal abscess. Clin. Infect. Dis. 19557-558. Owen, R. J., and J. Hernandez. 1990. Occurrence of plasmids in “Campylobacter upsaliensis” (catalase negative or weak group) from geographically diverse patients with gastroenteritis or bacteremia. Eur. J. Epidemiol. 6: 111-1 17. Owen, R. J., A. Beck, and P. Borman. 1985. Restriction endonuclease digest patterns of chromosomal DNA from nitratenegative Campylobacter jejuni-like organisms. Eur. J. Epidem. 13281-287. Owen, R. J., D. D. Morgan, M. Costas, and A. J. Lastovica. 1989. Identification of “Campylobacter upsaliensis” and other catalasenegative campylobacters from pediatric blood cultures by numerical analysis of electrophoretic protein patterns. FEMS Microbiol. Lett. 58:145-150. Oyarzabal, 0. A., I. V. Wesley, K. M. Harmon, L. SchroederTucker, J. M. Barbaree, L. H. Lauerman, S. Backert, and D. E. Comer. 1997. Specific identification of Campylobacter fetus by PCR targeting variable regions of the 16s rRNA. Vet. Microbiol. 58:61-71. Park, C. E., and G. W. Sanders. 1992. Occurrence of thermotolerant campylobacters in fresh vegetables sold at farmer’s outdoor markets and supermarkets. Can. J. Microbiol. 38:313-3 16. Parker, C. T., W. G. Miller, S. T. Horn, and A. J. Lastovica. 2007. Common genomic features of Campylobacter jejuni subsp doylei strains distinguish them from C. jejuni subsp. jejuni. BMC Microbiol. 7 5 0 . Patton, C. M., N. Shaffer, P. Edmonds, T. J. Barrett, M. A. Lambert, C. Baker, D. M. Perlman, and D. J. Brenner. 1989. Human disease associated with “Campylobacter upsaliensis” (catalasenegative or weakly positive Campylobacter species) in the United States. J. Clin. Microbiol. 27:66-73.
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
Popovic-Uroic, T., C. M. Patton, M. A. Nicholson, and J. A. Kiehlbauch. 1990. Evaluation of indoxylacetate hydrolysis test for rapid differentiation of Campylobacter, Helicobacter, and Wolinella species. J. Clin. Microbiol. 28:2335-2339. Preston, M. A., A. E. Simor, S. L. Walmsley, S. A Fuller, A. J. Lastovica, K. Sandstedt, and J. L. Penner. 1990. In vitro susceptibility of “Campylobacter upsaliensis” to twenty-four antimicrobial agents. Eur. 1. Clin. Microbiol. Infect. Dis. 9:822-824. Quinn, T. C., S. E. Goodell, C. L. Fennell, S. P. Wang, M. D. Schufler, K. K. Holmes, and E. W. Stamm. 1984. Infections with Campylobacter jejuni and Campylobacter-like organisms in homosexual men. Ann. Intern. Med. 101:187-192. Rams, T. E., D. Feik, and J. Slots. 1993. Campylobacter rectus in human periodontitis. Oral Microbiol. lmmunol. 8:230-235. Rao, G. G., Q. N. Karim, A. Maddocks, R J. Hillman, J. R. W. Harris, and A. J. Pinching. 1990. Campylobacter fetus infections in two patients with AIDS.]. Infect. 20:170-172. Rennie, R. P., D. Strong, D. E. Taylor, S. M. Salama, C. Davidson, and H. Tabor. 1994. Campylobacter fetus diarrhea in a Hutterite colony: epidemiological observations and typing of the causative organism. J. Clin. Microbiol. 32:721-724. Rice, E. W., M. R. Rodgers, I. V. Wesley, C. H. Johnson, and S. A. Tanner. 1999. Isolation of Arcobacter butzleri from ground water. Lett. Appl. Microbiol. 28:31-35. Rodriguez-Avial, I., C. Rodriguez-Avial, 0. Lopez, E. Culebras, and J. J. Picazo. 2006. In vitro activity of tigecycline (GAR-936) and other antimicrobials against tetracycline- and ciprofloxacinresistant Campylobacter clinical isolates. Int. J. Antimicrob. Agents 27:3 03-3 06. Romero, S., J. R. Archer, M. E. Hamacher, S. M. Bologna, and R. F. Schell. 1988. Case report of an unclassified microaerophilic bacterium associated with gastroenteritis. J. Clin. Microbiol. 26: 142-143. Russell, R. G., J. A. Kiehlbauch, C. J. Gebhart, and L. J. De Tolla. 1992. Uncommon Campylobacter species in infant Macaca nemestrina monkeys housed in a nursery. J. Clin. Microbiol. 30: 3024-3027. Sakran, W., R Raz, Y. Levi, R. Colodner, and A. Koren. 1999. Campylobacter bacteremia and pneumonia in two splenectomized patients. Eur. J. Clin. Microbiol. Infect. Dis. 18:496-498. Salama, S . M., M. M. Garcia, and D. E. Taylor. 1992a. Differentiation of the sub-species of Campylobacter fetus by genomic sizing. Int. J. Syst. Bacteriol. 42:446-450. Salama, S . M., H. Tabor, M. Richter, and D. E. Taylor. 1992b. Pulsed-field gel electrophoresis for epidemiological studies of Campylobacter hyointestinalis isolates. 1. Clin. Microbiol. 30: 1982-1984. Samie, A., C. L. Obi, L. J. Barrett, S. M. Powell, and R. L. Guerrant. 2007. Prevalence of Campylobacter species, Helicobacter pylori and Arcobacter species in stool samples from the Venda region, Limpopo, South Africa: studies using molecular diagnostic methods. J. Infect. 54:558-566. Sandstedt, K., and J. Ursing. 1991. Description of Campylobacter upsaliensis sp. nov. previously known as the CNW group. Syst. Appl. Microbiol. 14:39-48. Sandstedt, K., J. Ursing, and M. Walder. 1983. Thermotolerant Campylobacter with no or weak catalase activity isolated from dogs. Curr. Microbiol. 8:209-213. Sauerwein, R. W., J. Bisseling, and A. M. Horrevorts. 1993. Septic abortion associated with Campylobacter fetus subsp. fetus infection: case report and review of the literature. Infection 21:331333. Schrader, J. A., R. A. Venezia, and J. Tillotson. 1991. Ureasepositive thermophilic Campylobacter isolated from human blood. Clin. Microbiol. Newslett. 13:158-160.
147
Schultz, F., A. Bagon, W. Muller, and H. Hotzel. 2006. Identification of Campylobacter fetus subspecies by phenotypic differentiation and PCR. ]. Clin. Microbiol. 44:2019-2024. Sekizuka, T., T. Yokoi, 0. Murayama, B. C. Millar, J. Moore, and M. Matsuda. 2005. A newly constructed primer pair for the PCR amplification, cloning and sequencing of the flagellin (flaA) gene from isolates of urease-negative Campylobacter lari. Antonie van Leeuwenhoek 88:113-120. Serino, L., and M. Virji. 2002. Genetic and functional analysis of the phosphorylcholine moiety of commensal Neisseria lipopolysaccharide. Mol. Microbiol. 43:437-448. Skirrow, M. B., and J. Benjamin. 1980. “1001” campylobacters from man and animals.]. Hyg. 85:427-442. Skirrow, M. B., D. M. Jones, E. Sutcliffe, and J. Benjamin. 1993. Campylobacter bacteraemia in England and Wales, 198 1-1991. Epidemiol. Infect. 110567-573. Smibert, R. M. 1984. Genus Campylobacter, p. 111-118. In N. R. Kreig, and H. G. Holt (ed.), Bergey’s Manual of Systematic Bacteriology, vol. 1. Williams and Wilkins, Baltimore, MD. Smith, T. 1919. The etiological relation of spirilla pibrio fetus) to bovine abortion.]. Exp. Med. 30:313-322. Snelling, W. J., M. Matsuda, J. E. Moore, and J. S. G. Dooley. 2006. Under the microscope: Arcobacter. Lett. Appl. Microbiol. 42:7-14. Solnick, J. V. 2003. Clinical significance of Helicobacter species other than Helicobacter pylori. Clin. Infect. Dis. 36:349-354. Sorlin, P., P. Vandamme, J. Nortier, B. Hoste, C. Rossi, S. Pavlof, and M. J. Struelens. 1999. Recurrent “Flexispira rappini” bacteremia in an adult patient undergoing hemodialysis: case report. 1. Clin. Microbiol. 37:1319-1323. Stamp, S., 0. Varoli, F. Zanetti, and G. De Luca. 1993.Arcobacter cryaerophilus and thermophilic campylobacters in a sewage treatment plant in Italy: two secondary treatments compared. Epidemiol. Infect. 110:633-639. Stanley, J., A. P. Burnens, D. Linton, S. L. W. On, M. Costas, and R. J. Owen. 1992. Campylobacter helveticus sp. nov., a new thermophilic species from domestic animals: characterization, and cloning of a species-specific DNA probe. J. Gen. Microbiol. 138:2293-2303. Stanley, J., D. Linton, A. P. Burnens, F. E. Dewhirst, S. L. W. On, A. Porter, R. J. Owen, and M. Costas. 1994. Helicobacter pullorum sp. nov.-genotype and phenotype of a new species isolated from poultry and human patients with gastroenteritis. Microbiology 140:3441-3449. Stanley, J., D. Linton, A. P. Burnens, F. E. Dewhirst, R. J. Owen, A. Porter, S. L. W. On, and M. Costas. 1993. Helicobactercanis sp. nov., a new species from dogs: an integrated study of phenotype and genotype.]. Gen. Microbiol. 139:2495-2504. Steele, T. W., J. A. Lamer, and N. Sangster. 1985a. Nitratenegative Campylobacter-like organisms. Lancet i:294. Steele, T. W., and R J. Owen. 1988. Campylobacter jejuni subspecies doylei (subsp. nov.), is a subspecies of nitrate-negative campylobacters isolated from clinical specimens. lnt. J. Syst. Bacteriol. 38:3 16-3 18. Steele, T. W., N. Sangster, and J. A. Lamer. 1985b. DNA relatedness and biochemical features of Campylobacter spp. isolated in central and south Australia. J. Clin. Microbiol. 22:71-74. Steinkraus, G. E., and B. D. Wright. 1994. Septic abortion with intact fetal membranes caused by Campylobacter fetus subsp. fetus. 1. Clin. Microbiol. 32:1608-1609. Sylvester, F. A., D. Philpott, B. Gold, A. Lastovica, and J. F. Forstner. 1996. Adherence to lipids and intestinal mucin by a recently recognized human pathogen, Campylobacter upsaliensis. Infect. Immun. 64:4060-4066. Takatsu, M., S. Ichiyama, T. Nada, Y. Iinuma, H. Toyoda, Y. Fukuda, and N. Nakashima. 1997. Campylobacter fetus subsp.
148
LASTOVICA AND ALLOS
fetus cholecystitis in a patient with advanced hepatocellular carcinoma. Scand. J. Infect. Dis. 29:197-198. Tanner, A. C. R., S . Badger, C. H. Lai, M. A. Listgarten, R A. Visconti, and S . S . Socransky. 1981. Wolinella gen. nov., Wolinella succinogenes (Vibrio succinogenes Wolin et al.) comb. nov., and description of Bacteroides gracilis sp. nov., Wolinella recta sp. nov., Campylobacter concisus sp. nov., and Eikenella corrodens from humans with periodontal disease. Int. J. Syst. Bacteriol. 31:432-435. Tanner, A. C. R., J. L. Dzink, J. L. Ebersole, and S . S . Socransky. 1987. Wolinella recta, Campylobacter concisus, Bacteroides gracilis, and Eikenella corrodens from periodontal lesions, J. Periodont. Res. 22:327-330. Tanner, A. C. R., M. A. Listgarten, and J. L. Ebersole. 1984. Wolinella cuwa sp. nov.: “Vibriosuccinogenes” of human origin. Int. J. Syst. Bacteriol. 34:275-282. Targan, S . R., A. W. Chow, and L. B. Guze. 1976. Spontaneous peritonitis of cirrhosis due to Campylobacter fetus. Gastroenterology 71:3 11-3 13. Targan, S . R., A. W. Chow, and L. B. Guze. 1977. Campylobacter fetus associated with pulmonary abscess and empyema. Chest 71: 105-108. Tarrand, J. J., and D. H. M. Groschel. 1982. Rapid, modified oxidase test for oxidase-variable bacterial isolates. J. Clin. Microbiol. 16:772-774. Tauxe, R V., C. M. Patton, P. Edmonds, T. J. Barrett, D. J. Brenner, and P. A. Blake. 1985. Illness associated with Campylobacter laridis, a newly recognized Campylobacter species. J. Clin. Microbiol. 21:222-225. Taylor, D. E., K. Hiratsuka, and L. Mueller. 1989. Isolation and characterization of catalase-negative and catalase-weak strains of Campylobacter species, including “Campylobacter upsaliensis”, from humans with gastroenteritis. J. Clin. Microbiol. 27:20422045. Taylor, D. N., J. A. Kiehlbauch, W. Tee, C. Pitarangsi, and P. Echeverria. 1991. Isolation of group 2 aerotolerant Campylobacter species from Thai children with diarrhea. J. Infect. Dis. 163: 1062-1067. Taylor, D. N., K. T. McDermott, J. R. Little, J. G. Wells, and M. J. Blaser. 1983. Campylobacter enteritis from untreated water in the Rocky Mountains. Ann. Intern. Med. 99:38-40. Taylor, N. S., G. Zhongming, Z. Shen, F. E. Dewhist, and J. G. Fox. 2003. Cytolethal distending toxin: a potential virulence factor for Helicobacter cinaedi. 1.Infect. Dis. 188:1892-1897. Tee, W., K. Leder, E. Karroum, and M. Dyall-Smith. 1998a. “Flexispira rappini” bacteraemia in a child with pneumonia. J. Clin. Microbiol. 36:1679-1682. Tee, W., M. Luppino, and S . Rambaldo. 1998b. Bacteremia due to Campylobacter sputorum biovar sputorum. Clin. Infect. Dis. 27: 1544-1545. Totten, P. A., C. L. Fennell, F. C. Tenover, J. M. Wezenberg, P. L. Perine, W. E. Stamm, and K. K. Holmes. 1985. Campylobacter cinaedi (sp. nov.) and Campylobacter fennelliae (sp. nov.): two new Campylobacter species associated with enteric disease in homosexual men. J. Infect. Dis. 151:131-139. Tremblay, C., C. Gaudreau, and M. Lorange. 2003. Epidemiology and antimicrobial susceptibilities of 111 Campylobacter fetus subsp. fetus strains isolated in Quebec, Canada, from 1983 to 2000. J. Clin. Microbiol. 41:463-466. Tresierra-Ayala, A., M. E. Bendayan, A. Bernuy, G. Pereyra, and H. Fernandez. 1994. Chicken as potential contamination source of Campylobacter lari in Iquitos, Peru. Rev. Inst. Med. Trop. Sao Paul0 36:497-499. Trivett-Moore, N. L., W. D. Rawlinson, M. Yuen, and G. L. Gilbert. 1997. Helicobacter westmeadii sp. nov., a new species iso-
lated from blood cultures in two AIDS patients. J. Clin. Microbiol. 35:1144-1150. Uckay, I., J. Garbino, P.-Y. Dietrich, B. Ninet, P. Rohner, and V. Jacomo. 2006. Recurrent bacteremia with Helicobacter cinaedi: case report and review of the literature. BMC Infect. Dis. 6:86. Vandamme, P., M. I. Daneshvar, F. E. Dewhirst, B. J. Paster, K. Kersters, H. Goossens, and C. W. Moss. 1995. Chemotaxonomic analyses of Bacteroides gracilis and Bacteroides ureolyticus and reclassification of B. gracilis as Campylobacter gracilis comb. nov. Int. J. Syst. Bacteriol. 45:145-152. Vandamme, P., and J. De Ley. 1991. Proposals for a new family, Campylobacteraceae.Int. J. Syst. Bacteriol. 41:45 1-455. Vandamme, P., E. Falsen, B. Pot, B. Holste, K. Kersters, and J. De Ley. 1989. Identification of EF group 22 campylobacters from gastroenteritis cases as Campylobacter concisus. J. Clin. Microbiol. 27: 1775-1782. Vandamme, P., E. Falsen, R. Rossau, B. Hoste, R. Tygat, and J. De Ley. 1991. Revision of Campylobacter, Helicobacter and Wolinella taxonomy: emendation of generic descriptions and proposal of Arcobacter gen. nov. Int. J. Syst. Bacteriol. 41:88103. Vandamme, P., P. Pugina, G. Benzi, R. Van Etterijck, L. Vlaes, K. Kersters, J.4’. Butzler, H. Lior, and S . Lauwers. 1992a. Outbreak of recurrent abdominal cramps associated with Arcobacter butzleri in an Italian school. J. Clin. Microbiol. 30:2335-2337. Vandamme, P., M. Vancanneyt, B. Pot, L. Mels, B. Holste, D. Dewettinck, L. Vlaes, C. Van den Borre, R. Higgins, J. Hommez, K. Kersters, J.4’. Butzler, and H. Goossens. 1992b. Polyphasic taxonomic study of the emended genus Arcobacter with Arcobacter butzleri comb. nov. and Arcobacter skirrowii sp. nov., an aerotolerant bacterium isolated from veterinary specimens. Int. J. Syst. Bacteriol. 42:344-356. Vandenberg, O., A. Dediste, K. Houf, S . Ibekwem, H. Souayah, S . Cadranel, N. Douat, G. Zissis, J. P. Butzler, and P. Vandamme. 2004. Arcobacter species in humans. Emerg. Infect. Dis. 10~1863-1867. Vandenberg, O., K. Houf, N. Douat, L. Vlaes, P. Retore, J. P. Butzler, and A. Dediste. 2006. Antimicrobial susceptibility of clinical isolates of non-jejunilcoli campylobacters and arcobacters from Belgium. J. Antimicrob. Chemother. 57:908-913. Van Doorn, L. J., B. A. Giesendorf, R. Bax, B. A. Van der Zeijst, P. Vandamme, and W. G. Quint. 1997. Molecular discrimination between Campylobacter jejuni, Campylobacter coli, Campylobacter lari and Campylobacter upsaliensis by polymerase chain reaction based on a novel putative GTPase gene. Mol. Cell. Probes 11:177-1 85. Van Doorn, L. J., W. G. Verschuuren-Van Haperen, A. Van Belkum, H. P. Endtz, J. S . Vliegenthart, P. Vandamme, and W. G. Quint. 1998. Rapid identification of diverse Campylobacter lari strains isolated from mussels and oysters using a reverse hybridization line probe assay. J. Appl. Microbiol. 84545550. Van Dyke, T. E., V. R. Dowell, S . Offenbacher, S . Synder, and W. Hersh. 1986. Potential role of micro-organisms isolated from periodontal lesions in the pathogenesis of inflammatory bowel disease. Infect. Immun. 53:671-677. Vargas, J., J. E. Corzo, M. J. Perez, F., Lozano, and E. Martin. 1992. Bacteriemia por Campylobacter e infeccion por VIH. Enfern. Infecc. Microbiol. Clin. 10:155-157. VQon, M., and R. Chatelain. 1973. Taxonomic study of the genus Campylobacter Sebald and VCron and designation of the neotype strain for the type species, Campylobacter fetus (Smith and Taylor) Sebald and VCron. Int J. Syst. Bacteriol. 23:122-134. Viejo, G., B. Gomez, D. De Miguel, A. Del Valle, L. Otero, and P. De La Iglesia. 2001. Campylobacter fetus subspecies fetus bac-
CHAPTER 7
CLINICAL SIGNIFICANCE OF CAMPYLOBACTER AND RELATED SPECIES
teremia associated with chorioamnionitis and intact fetal membranes. Scand. J. Infect. Dis. 33:126-127. Walder, M., K. Sandstedt, and J. Ursing. 1983. Phenotypic characteristics of thermotolerant Campylobacter from human and animal sources. Curr. Microbiol. 9:291-296. Wang, W. L., and M. J. Blaser. 1986. Detection of pathogenic Campylobacter species in blood culture systems. 1. Clin. Microbiol. 23:709-714. Wens, R., M. Dratwa, C. Potvliege, C. Hansen, C. Tielemans, and F. Collart. 1985. Campylobacter fetus peritonitis followed by septicaemia in a patient on continuous ambulatory dialysis. J. Infect. 10:249-25 1. Wesley, I. V., A. L. Baetz, and D. J. Larson. 1996. Infection of cesarean-derived colostrum-deprived 1 day-old piglets with Arcobacter butzleri, Arcobacter cryaerophilus and Arcobacter skirrowii. Infect. Immun. 64:2295-2299. Wesley, I. V., R. D. Wesley, M. Cardella, F. E., Dewhirst, and B. J. Paster. 1991. Oligodeoxynucleotide probes for Campylo-
149
bacter fetus and Campylobacter hyointestinalis based on 16s rRNA sequences. J. Clin. Microbiol. 29:1812-1817. Wetsch, N. M., K. Somani, G. J. Tyrrell, C. Gebhart, R. J. Bailey, and D. E. Taylor. 2006. Campylobacter cuwus-associated hepatic abscesses: a case report. ]. Clin. Microbiol. 44:1909-1911. Wilcox, C. M., B. A. Byford, C. E. Forsmark, W. K. Hadley, J. P. Cello, and M. A. Jacobson. 1990. Campylobacter-like organisms are uncommon pathogens in patients infected with the human immunodeficiency virus. J. Clin. Microbiol. 28:2370-2371. Wybo, I., J. Breynaert, S . Lauwers, F. Lindenburg, and K. Houf. 2004. Isolation of Arcobacter skirrowii from a patient with chronic diarrhea. J. Clin. Microbiol. 42:1851-1852. Yamashita, K., Y. Aoki, and K. Hiroshima. 1999. Pyrogenic vertebral osteomyelitis caused by Campylobacter fetus subspecies fetus. A case report. Spine 24582-584. Zanetti, F., 0. Varoli, S . Stampi, and G. De Luca. 1996. Prevalence of thermophilic Campylobacter and Arcobacter butzleri in food of animal origin. lnt. J. Food Microbiol. 33:315-321.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 8
Burden of Illness of Campylobacteriosis and Sequelae U R E M0LBAK AND
ARIEHAVELAAR
single measure. Burden of disease studies include elements of epidemiology, disease modeling, risk assessment, and burden projections; the latter help policy makers learn where to target preventive strategies and what to expect in terms of future disease burden.
The burden of disease can be defined as the summary of morbidity and mortality associated with different acute and chronic manifestations of Campylobacter infections. Such health effects vary widely in severity and timescale but can be translated to a single summary measure. Burden of disease measures offer a more comprehensive insight into the impact of illness on a population than traditional measures such as the number of fatal cases or the number of life-years lost as a result of premature mortality. Reliable estimates of burden of illness are important to advise risk managers and decision makers on several aspects of public health policy. These include comparison of the public health impact of different diseases, which is an important aspect of priority setting for intervention (Kemmeren et al., 2006; Vijgen et al., 2007). Furthermore, product standards for the microbiological safety of food and water can be defined on a common basis. An example of this application is the World Health Organization (WHO) Guidelines for Drinking-Water Quality (2004), where a health outcome target for all microbial hazards of 1 disability-adjusted life-year (DALY) per 1 million person-years is suggested as a reference level of risk. This level is similar to the generally accepted risk for genotoxic carcinogens of per lifetime. A further application is the balancing of competing risks, as demonstrated by Havelaar et al. (2000), who demonstrated that the positive health effects of inactivation by ozone of Cryptosporidium oocysts in drinking water outweighed the negative health effects of bromate that is produced as a by-product. Disease burden estimates are an important factor in cost-utility analyses to compare various preventive and curative interventions. Burden of disease studies capitalize primarily on existing information and translate these data into a
HOW IS BURDEN OF ILLNESS MEASURED? Most of the information collected by public health laboratories and through other types of disease surveillance and health statistics cannot be directly translated into policy. Health statistics are often fragmented, and in particular for surveillance data, outcomes of disease are rarely reported. Thus, public health laboratories report the annual number of culture-confirmed cases of Campylobacter infections to the public health authority. However, the actual numbers of infections in the community are not known, nor are the different outcomes (duration of illness, admission to the hospital, and long-term complications) of the illness. Moreover, traditional health statistics use measures such as incidence of cultureconfirmed infections per 100,000, which do not permit direct comparisons of the cost-effectiveness of different interventions. Against this background, there was a need to develop a single metric that can be used to compare burden of illness and the cost-effectiveness of different interventions between different populations and between different diseases. To meet this requirement, the DALY measure has become increasingly popular. The DALY methodology has been described by Murray and Lopez (1996) and was adopted by WHO for its reporting on health information in the late 1990s. DALYs express the years of life lost to premature death (YLL) and the years lived with disability (YLD)
-
Klre Mslbak Department of Epidemiology, Statens Serum Institut, Copenhagen, Denmark. Arie Havelaar Laboratory for Zoonoses and Environmental Microbiology, National Institute for Public Health and the Environment, Bilthoven, and Division of Veterinary Public Health Institute for Risk Assessment Sciences, Utrecht University, Utrecht, The Netherlands.
151
152
MIZILBAK AND HAVELAAR
for varying degrees of severity, making time itself the common metric for death and disability.
DALY = YLL
+ YLD
YLL is the number of years of life lost to mortality, and YLD is the number of years lived with a disability, weighted with a factor between 0 and 1 for the severity of the disability. The YLL due to a specific disease in a specified population is calculated by summation of all fatal cases (d) due to the health outcomes (1) of a specific disease, each case multiplied by the expected individual life span (e) at the age of death: YLL
d, x el
= 1
A high number of YLL can be associated with a high number of fatal cases, with deaths of young children, with a high assumed life expectancy, or a combination of these factors. YLD is calculated as the product of the duration of the illness (t) and the disability weight (w) of a specific disease, accumulated over all cases (n) and all health outcomes (I):
A high number of YLD can be associated with a high incidence of morbidity, a high disability weight, a long duration of illness, or a high prevalence of (chronic) sequelae after primary illness. DALYs can be seen as health gap measures, equating to years of healthy life lost. Disadvantages of the DALY approach include the need for strong value judgments on disability and age, thus placing emphasis on death and morbidity in young adulthood. There are other ways to measure burden of illness than the DALY metric (Gold et al., 2002). Whereas DALY measures health losses, the qualityadjusted life-year (QALY) is a measure of the current health state. DALYs are typically used in public health research, whereas QALYs are used in clinical settings to indicate the potential benefit of treatment. However, under the same assumption, one averted DALY equals one gained QALY. Burden of illness may also be measured in terms of cost of illness. Cost of illness can be calculated as the sum of direct health care costs, direct non-health care costs, and indirect non-health care costs. Direct health care costs include medical services, etc.; direct non-health care costs includes costs of informal care, transport, additional diapers, etc.; and indirect non-
health care costs are the value of production lost to society. The production loss can stem from temporary absence, long-term disability, or premature death. The cost of illness methodology and related monetary approaches such as willingness to pay are based on country- and time-specific cost estimates. They are therefore less consistent than DALY if the aim is to compare burden of illness between different geographical areas and over time.
CHALLENGES IN THE MEASUREMENT OF BURDEN OF ILLNESS A careful assessment of the burden of illness due to Campylobacter should in principle be based on a detailed knowledge of the age-specific frequency of Campylobacter-associated illness in a given community and a detailed understanding of the frequency and the severity of all possible outcomes associated with the infection. The outcomes must include consequences in both the short and long term. In reality, not all these data will be available, and a calculation of burden of illness will be subject to many uncertainties and assumptions. In a pilot study of disease burden of infectious diseases (including campylobacteriosis) in Europe (Van Lier and Havelaar, 2007), major limitations in data availability were as follows: inconsistent information on morbidity, mortality, or both; limited information about patients’ age distribution; no reporting of the incidence of complications and chronic sequelae; and no consistent set of disability weights available. Major limitations with regard to data quality were as follows: no information on underascertainment of morbidity and mortality; no information on possible variation between countries of the duration, severity, and rate of complications and chronic sequelae; and differences between figures obtained from different sources. For a valid estimation of the burden of illness, it is a major challenge to overcome some of these limitations; selected aspects of this problem will be discussed below.
UNDERASCERTAINMENT OF MORBIDITYTHE PYRAMID OF SURVEILLANCE Official numbers of Campylobacter-infections are commonly derived from laboratory-based surveillance, where clinical microbiology laboratories report positive findings to national reference centers. These data are pivotal to measure trends over time but do not reflect the burden of illness. The sensitivity of surveillance differs among countries (chapter
CHAPTER 8
9). The reported incidence is a composite measure of several factors. These factors include the true incidence of Campylobacter infections, health careseeking behavior of patients with gastroenteritis, and the likelihood that the physician will request a stool culture. Furthermore, access to laboratories and the microbiological methods in place vary, and so does the completeness in reporting findings to the public health authorities. Thus, the official numbers therefore represent only the tip of the surveillance pyramid (Fig. 1).Although this top is likely to include the most severe cases, there is a significant amount of illness that is not recorded in the official system. In population-based studies in England and Wales (Wheeler et al., 1999) and in The Netherlands (De Wit et al., 2001), the true incidence of disease was investigated. The studies suggest that for every reported Campylobacter infection, between 7.6 and 19 persons actually fell ill (Table 1). In the United States, an underreporting factor of 38 was assumed on the basis of studies with Salmonella infections (Mead et al., 1999). Part of the difference between the three studies may be the result of different methodology, but it is interesting that there is less variation between estimated incidences in the general population than the reported figures indicate. This suggests that the sensitivity of the surveillance is a critical factor when attempts are made to compare numbers between countries. Flint et al. (2005) reviewed these key national studies and the international efforts that are providing the necessary information and technical re-
BURDEN OF ILLNESS O F CAMPYLOBACTERIOSIS
153
Table 1. Reported and estimated numbers of Cumpylobacterassociated gastroenteritis per 100,000 population No. of cases:
Country England United States The Netherlands
Reported
Estimated
Reference
110 24 36
870 916 685
Wheeler et al. (1999) Mead et al. (1999) Van Pelt et al. (2003)
sources to derive national, regional, and global burden of disease estimates. Integrated food chain surveillance, where community-based, etiology-specific disease estimates are obtained, is subject to several types of bias, is costly, and requires good collaboration with the study population. Seroepidemiology may provide an attractive alternative, in particular if serum samples from a random sample of the general population can be obtained for this purpose. The major limitation of seroepidemiology is that it measures the incidence of Campylobacter-related seroreactions, and some of these may be reactions to previous exposures that took place long ago. Hence, data on the disease-toinfection ratio are needed to translate the seroepidemiology estimates to estimates of morbidity. On the basis of serum samples from Danish blood donors, the yearly incidence of Campylobacter infections was determined to be 138 per 1,000 population (Simonsen et al., 2007). This figure is 249 times
Figure 1. Surveillance pyramid. Only a fraction of the population that is exposed to Campylobucter will become infected, and among those who are infected, only a fraction will experience symptoms. Furthermore, only a fraction of these patients will seek medical care, and only a fraction of those who seek care (whether at the hospital or at their general practitioner’s office) will have the organism culture confirmed and reported, and thus figure in the official statistics. The tip of the pyramid represents the most severe cases.
154
M0LBAK AND HAVELAAR
higher than the reported incidence. In The Netherlands, the seroconversion rate indicates at least one exposure every 6 to 7 years during childhood (Ang et al., 2007). These high incidences are the same order of magnitude as the results from a risk assessment study, in which the total exposure of the Dutch population to Campylobacter was estimated to average 1 CFU per week. Although this may seem to be a low exposure, application of a dose-response model (Teunis and Havelaar, 2000) results in a predicted estimate of approximately one infection (which may be asymptomatic) per 2 person-years (Evers et al., in press).
THE PROBLEM OF ASCERTAINMENT OF OUTCOMES To assess the burden of disease, the most important or relevant disease outcomes need to be defined. This can be illustrated by an outcome tree, which is a qualitative representation of the progres-
sion of a disease in time. For Campylobacter, main outcomes as defined by Mangen et al. (2005) include gastroenteritis, Guillain-Barrt syndrome, reactive arthritis, inflammatory bowel disease, and death (Fig. 2 and Table 2). It is obvious that constructing such trees requires a simplification of the real world. For example, an outcome such as sepsis (without death) is not included in the tree because preliminary calculations show that this outcome would not substantially add to the burden. On the other hand, the inclusion of inflammatory bowel disease was made on the basis of the observation of a statistical relationship by Helms et al. (2003). In the absence of sufficient mechanistic supportive evidence, such an inclusion might be disputed. New evidence may require additions to the tree, and currently, sufficient evidence for inclusion of irritable bowel syndrome appears to be available (Smith and Bayles, 2007). On the basis of the scope of relevant adverse health outcomes, it is often necessary to collate information from a variety of different sources in order to obtain data on frequency and duration of these
Reactive arthritis
Figure 2 . Example of an outcome tree for Campylobacter-associated gastroenteritis and sequelae. R, recovery. From Van Lier and Havelaar (2007).
CHAPTER 8
BURDEN OF ILLNESS OF CAMPYLOBACTERIOSIS
155
Table 2. Incidence (cases per year) of Campylobacter gastroenteritis and associated sequelae, The Netherlands, 2004 (population 16 million)“ Incidence estimate (cases per year) Infection Gastroenteritis Without medical care See general practitioner Admitted to hospital Fatal Sequelae Guillain-Barrt syndrome Reactive arthritis Inflammatory bowel disease
Most likely
Low estimate
High estimate
59,000 45,000 14,000 570 25
25,000 19,000 5,000 500 18
140,000 110,000 33,000 650 34
60 1,000 22
40 430 17
85 2,600 29
“From Kemmeren et al. (2006).
outcomes and disability weights. For example, in the project Priority Setting of Foodborne Pathogens in The Netherlands, disability weights were obtained from the Dutch Public Health Status and Forecast studies, other studies from The Netherlands, and the international literature (Kemmeren et al., 2006). In Denmark, a series of registry studies has contributed to a better understanding of the severity of Carnpylobacter infections; these estimates have proved important for improving burden of disease estimates. Data on outcomes were obtained by linking different national registries, including the surveillance registry for gastrointestinal pathogens, the Danish civil registry, and the national hospital discharge registry. Of 16,180 Campylobacter patients, 190 (1.2%) died within 1 year after infection; this was 2.3 times higher than the general Danish population (matched by age and geography). After adjustment for comorbidity, this relative mortality declined to 1.9. The estimate was highest in the acute phase 30 days after diagnosis (relative risk, 5.0) but remained significantly increased up to 1 year after the initial diagnosis. On the basis of a calculation of the attributable risk among exposed patients, it can be estimated that the fatality rate of culture-confirmed Carnpylobacter infection is 0.5%, a figure that is higher than, for example, the FoodNet case fatality rate for Campylobacter, which is 0.10% (Mead et al., 1999). This may suggest that the burden of illness has been somehow underestimated in studies that have ignored mortality as an outcome of campylobacteriosis. Of 17,99 1 patients with Campylobacter, 2,221 (12%) had been admitted to the hospital. Although most (1,937 patients) had a discharge diagnosis of gastroenteritis, other outcomes included irritable bowel syndrome (161 patients), inflammatory bowel disease (72 patients), invasive illness (41 patients), reactive arthritis (22 patients), gastrointestinal compli-
cations (16 patients), and Guillain-Barr6 syndrome (6 patients) (Helms et al., 2006). The burden of illness was measured in terms of numbers of days in the hospital, and the analysis shows that 1,000 Campylobacter patients result in some 700 bed-days (Table 3). In another study, it was shown that antimicrobial drug resistance (fluoroquinolone or macrolide resistance) was associated with an excess mortality compared with infections with drug-sensitive Campylobacter (Helms et al., 2005). This indicates that the burden of illness may increase by increasing levels of drug resistance. A similar observation has been made for duration of gastrointestinal illness among individuals infected with fluoroquinolone-resistant Campylobacter (reviewed by Mlirlbak, 2005). However, excess morbidity and mortality associated with drug resistance in Campylobacter remains a subject of debate (Wassenaar et al., 2007).
THE IMPACT OF IMMUNITY O N THE MEASUREMENT OF BURDEN OF ILLNESS One of the challenges in the measurement of burden of illness is to identify methods to include contribution from immunity. An exposure to Cam-
Table 3. Duration of hospital stay per 1,000 patients with Campylobacter infections Hospital stay
.
Hospital days per 1,000 patients
...
Total.. . . . . . . . . . . . . , , ,., Attributed tv gastroenteritis. . . . Attributed to complications. . . . Attributed to chronic sequelae..
714 days 530 days 30 days 150 days
“Analyses of the national patients registry in Denmark, 1991 to 1999, based on 17,991 patients. From Helms et al. (2006).
156
M0LBAK AND HAVELAAR
pylobacter may give rise to an infection that may lead to clinical symptoms or that may be asymptomatic. All things equal, the likelihood that exposure to a certain dose of Campylobacter results in an infection depends on the immunity of the host, and the immunity of the host will also modify the clinical course of infection. In other words, in a population with frequent exposure to Campylobacter, the disease-toinfection ratio will be lower than in an immunologically naive population where a larger proportion of infections will be symptomatic. This concept is corroborated by several lines of evidence. (i) In developing countries, where Campylobacter is ubiquitous, rates of asymptomatic carriers are high, and clinical disease is primarily seen in young children. This may suggest that immunity toward clinical illness is widespread, although strain differences between different areas of the world might contribute (Coker et al., 2002; Oberhelman and Taylor, 2000). (ii) In industrialized countries, chronic exposure to Campylobacter may lead to immunity against clinical disease, as discussed in chapter 9. Indeed, as also reviewed, some case-control studies of sporadic illness suggest that repeated exposure to high-risk food is associated with reduced risk of illness. It has also been found in some studies that chicken eaten outside of the home is a more important risk factor than chicken eaten in the home. These counterintuitive findings corroborate that notion that immunological factors mitigate the risk of clinical disease. (iii) Environmental risk factors for Campylobacter infections may be more important in children than in adults. This may point toward the notion that adults living in a rural community with frequent exposure to Campylobacter may be relatively spared clinical illness, whereas young, disease-naive children become sick (Ethelberg et al., 2005). (iv) Finally, patients with hypo- or agammaglobulinemia, patients who have a defect in humoral immunity, and patients infected with human immunodeficiency virus and who have compromised cell-mediated immunity experience more severe illness than individuals with normal immune response (Perlman et al., 1988; Skirrow and Blaser, 2000). It is beyond the scope of this chapter to discuss these immunological aspects in detail. The point is, however, that the relation between the overall exposure to Campylobacter in a given country and the burden of illness may be nonlinear. In particular, for countries with high rates of diarrheal illnesses or a high risk of frequent exposures to Campylobacter, the demonstration of Campylobacter in the stool of a patient with gastrointestinal symptoms may not be of etiological relevance, but merely a coincidental finding. Hence, for developing countries in particu-
lar, the methodology to determine morbidity from Campylobacter infections needs further development.
DISABILITY WEIGHTS Disability weights reflect the health impact of a condition, and they are based on the preferences of a panel of judges. Disability weights typically range between 0, reflecting the best possible health state, and 1, reflecting the worst possible health state. For example, for Guillain-BarrC syndrome, the severity score may range between 0.01 and 0.94 depending on the functional status of the patient. The assignment of disability weights is a controversial issue. Ideally, disability weights that are based on preferences of the general public are used in burden of disease studies aimed to inform policy making at the national or international level. Disability weights that are based on elicitation panels consisting of laypersons are increasingly becoming available. Previous work has depended on panels of medical professionals. Preferences of patients who actually have the disease are biased because of coping behavior. The international transferability of disability weights is of concern. A study in Western Europe (Schwarzinger et al., 2003) concluded that there was “a reasonably high level of agreement on disability weights in Western European countries with the visual analogue scale (VAS) and time trade-off (TTO) methods, but a lower level of agreement with the person trade-off (PTO) method.” Ustun et al. (1999) concluded that ranking of weights is relatively stable across countries, but differences are large enough to cast doubt on the universality of experts’ judgments of disability weights. Another study concluded that “Meaningful differences exist in directly elicited TTO valuations of EQ-5D health states between the US and UK general populations” (Johnson et al., 2005). Several preference measurement methods are available for panel elicitation, including the standard gamble (SG), TTO, PTO, and VAS. All methods give different results (VAS > TTO > PTO > SG), but they are highly correlated. The SG and VAS are not considered informative because they are only sensitive to severe disease (SG) or very sensitive to mild diseases (VAS), leading to compression at either end of the scale. Additionally, the VAS is not choice based because it does not allow a trade-off. The TTO and PTO methods are generally used. For chronic diseases, most descriptions are based on the impact of a disease in the course of a year. However, many infectious diseases have a rapid course, and consequently the disability weight can be assessed by fo-
CHAPTER 8
...................................................
.............
..............
"
.................
BURDEN O F ILLNESS OF CAMPYLOBACTERIOSIS
.l"."....."............".................
.................. .............................."............ "
"
157
........................... . .................................... "
T
Catnpylobacter STEC 0157
Sahnella
Norovhs
Rotavims
Iisteria
Tomplasm Cryptosporidium
Giardia
(tong.)
I El Undiscounted 0 Discounted (4%)I Figure 3. Disease burden of seven enteric pathogens in The Netherlands, 2004. From Kemmeren et al. (2006) and Vijgen et al. (2007). Bars indicate most likely estimate; error bars, minimum and maximum values. STEC, Shiga-toxin producing Escherichia coli; cong., congenital.
cusing on the phase of acute disease only (period profile) or by focusing on a year in which an episode of acute illness is experienced (annual profile; when using an annual profile, the duration of the illness is 1 year by definition). Both methods have been used in disease burden studies. In practice, large differences may be found between these two methods for diseases that have a high incidence but low severity (e.g., norovirus-associated gastroenteritis). For such diseases, the use of annual profiles may lead to high estimates of disease burden. One proposed solution is to introduce a threshold that is based on the median weight of the respondents in an elicitation study (J. Haagsma et al., unpublished data).
ESTIMATED BURDEN OF CAMPYLOBACTER INFECTIONS AND THE RELATIVE CONTRIBUTION OF ACUTE INFECTIONS VERSUS POSTINFECTIVE COMPLICATIONS Figure 3 shows the estimated burden of seven enteric pathogens in The Netherlands. Campylobacter causes the largest burden of these pathogens, followed by Salmonella and Toxoplasma. It is noteworthy that the burden of enteric viruses is lower, despite their higher incidence. Viral infections such as norovirus or rotavirus are usually associated with a relatively mild clinical course, a low case-fatality ratio, and the absence of complications. Figure 4 shows
Giardia Cryptosporidium Tomplasma (cong.) Listena Rotavirus Norovirus Salmonella STEC 0157
Campylobacter 0
DALYs per year
200
400
600
800
loo0
1200
1400
Figure 4. Breakdown of disease burden of seven enteric pathogens in The Netherlands, 2004. From Kemmeren et al. (2006) and Vijgen et al. (2007). STEC, Shiga toxin-producing Escherichia coli.
158
M0LBAK AND HAVELAAR
__I
45 240 x
5 35
a w 30 0 0
2
25
3.-
20
0
Z 15
,” 10 cd
8
5 0
Campylobacter Sabnella
Norovirus Undiscounted
Rotavirus Cryptosporidium
Giardia
Discounted (4%) I
Figure 5. Cost of illness of seven enteric pathogens in The Netherlands, 2004. From Kemmeren et al. (2006) and Vijgen et al. (2007). Bars indicate most likely estimate; error bars, minimum and maximum values.
that for Campylobacter, the largest contribution to the burden is from complications, in particular Guillain-Barrt syndrome. Mortality also causes a relatively high burden, while the contribution of acute gastroenteritis is relatively low. Data for the cost of illness are shown in Figs. 5 and 6. The enteric viruses have the largest economic impact because their high incidence leads to considerable productivity losses (indirect non-health care costs). The costs of campylobacteriosis are somewhat lower than the viruses but still significant. Direct health care costs (mainly associated with Guillain-
Barrt syndrome) are an important factor in these costs. Table 4 shows that the burden of campylobacteriosis is similar to that of other important bacterial infections such as meningitis, bacterial sexually transmitted infections, and tuberculosis but lower than AIDS, human immunodeficiency virus, and influenza. Collectively, the burden of infectious diseases is lower than that of chronic diseases such as cardiovascular disease and cancer. This does not imply that infectious diseases are unimportant but that in a country such as The Netherlands, infectious disease
Giardia
Rotavirus
Norovirus
Sahnella
0 5 Cost of illness (million €per year)
10
15
20
25
30
Figure 6. Breakdown of cost of illness of seven enteric pathogens in The Netherlands, 2004. From Kemmeren et al. (2006). DHC, direct health care costs; DNHC, direct non-health care costs; INHC, indirect non-health care costs.
CHAPTER 8
BURDEN OF ILLNESS OF CAMPYLOBACTERIOSIS
159
Table 4. Disease burden of Cumpylobacter infections in comparison with selected other diseases in The Netherlands (population 16 million)" Amount of lost DALYs per yr
Infectious diseases
Noninfectious diseases
>100,000
Acute lower respiratory tract infections (e.g., pneumonia and acute bronchi[oli]tis) Influenza HIVIAIDS, upper respiratory tract infections, stomach ulcers Campylobacter infections, bacterial meningitis, bacterial sexually transmitted infections, tuberculosis Shiga toxin-producing Escherichiu coli 0157
30,000 to 100,000 10,000 to 30,000 3,000 to 10,000 1,000 to 3,000
<1,000
Cardiovascular disease, cancer, depression, diabetes, alcohol dependency Traffic accidents, breast cancer, suicide Epilepsy, multiple sclerosis Ulcers of the stomach and duodenum Hip fractures
"From Havelaar et al. (2005).
parties (equity). Risk assessment, epidemiology, and economic analysis are all tools that can be used to aid risk management in such decisions. In The Netherlands, this interdisciplinary approach, which involved different scientific institutes, was applied to the control of Campylobacter on broiler meat. Estimates were made of the potential benefits and the cost-utility ratio of a large number of possible interventions to decrease human exposure to Campylobacter by consumption of chicken meat. For this purpose, a farm-to-fork risk assessment model was combined with economic analysis and epidemiological data (Havelaar et al., 2007; Katsma et al., 2007; Nauta et al., 2007; Mylius et al., 2007). Chicken
control is relatively effective. Maintaining the current level of protection is essential, and cost-effective interventions to further reduce the current burden are available (Figs. 7 and 8).
INTEGRATING BURDEN ESTIMATES WITH RISK ASSESSMENT AND ECONOMICS: THE CARMA PROJECT When evaluating an intervention or policy, decision makers need to determine its ability to positively impact public health (effectiveness)at a reasonable cost (efficiency) in a fair manner for all affected
^ ^
farm hygiene phage therapy red. faecal leakage
i i
decont. scald tank : decont. carcass dip : decont. dip and spray
i
crust freezing
i
irradiation : product freezing :
- 20%
0%
20%
40%
60%
80%
100%
Figure 7. Risk reduction of campylobacteriosis by interventions in the broiler meat chain. From Havelaar et al. (2007).
160
M0LBAK AND HAVELAAR
farm tygiene (opbmisbc scenario) phaw tkrapy red. fecal leakage dewnt scald tank decort. rip
decont dip and spray crust freezing irradiation pro&ct freezing prepaed meat hane hezing kitchen tygiene I
-lw
0
100
2w
30
-
400
500
Cost-utility ratio (x 1000 E per DALY averted)
Figure 8. Cost-utility ratio of different interventions to reduce contamination of broiler meat with Cumpylobacter. Data are presented when effects on Dutch consumers only are considered and when effects on all consumers (including exports from The Netherlands) are considered. Only products from flocks that tested positive by PCR are treated. From Havelaar et al. (2005). NL, The Netherlands.
meat consumption accounts for 20 to 40% of all cases of illness in The Netherlands. Reduction of contamination at broiler farms could be efficient in theory. However, it is unclear which hygienic measures need to be taken, and the costs can be high. The experimental treatment of colonized broiler flocks with bacteriophages has proven to be effective in laboratory-scale studies and could also be cost-efficient if confirmed in practice. Because a major decrease of infections at the broiler farm is not expected in the short term, additional measures in the processing plant were also considered. At this moment, guaranteed Campylobucter-free chicken meat at the retail level is not realistic. The most promising interventions in the processing plant include limiting fecal leakage during processing, separating contaminated and noncontaminated flocks (scheduling), and decontaminating a contaminated flock. New, faster, and more sensitive test methods to detect Cumpylobucter colonization in broilers flocks are a prerequisite for successful scheduling scenarios. Other methods to decrease the contamination of meat of colonized flocks, such as freezing and heat treatment, are more expensive and less effective than chemical decontamination. On the basis of the risk models developed for the project, quantitative criteria for Cumpylobacter
on broiler meat are being developed and related to the appropriate level of protection, as defined by Codex Alimentarius Commission (2005).
PERSPECTIVES A N D FUTURE RESEARCH The burden of Cumpylobucter infections and other food-borne disease is not well defined in many countries or regions or on a global level. WHO, in conjunction with other national public health agencies, is coordinating a number of international activities designed to assist countries in the strengthening of disease surveillance and to determine the burden of acute gastroenteritis. This activity is important to enable decision makers to make priorities between different interventions to improve food safety. WHO has recently developed a strategic framework for the estimation of global burden of foodborne diseases from all causes and is in the process of establishing the Foodborne Disease Burden Epidemiology Reference Group. This group will include experts from various disciplines, including microbiology, parasitology, and virology; burden of disease methodologies; microbiological and chemical risk assessment (including toxicological) and source attribution; disease and exposure modeling; statistics and
CHAPTER 8
geographic information systems; clinical medicine and nutrition; and food protection, policy, and regulation. The main aim of the Foodborne Disease Burden Epidemiology Reference Group will be to arrive at DALY estimates for food-borne illness (e.g., bacteria, viruses, parasites, fungi, chemicals). This activity will be closely coordinated with similar WHO initiatives in other domains (such as the Child Health Epidemiology Reference Group) and will also fit in a recent initiative to update the global burden of disease estimates. For developing countries, where deaths from diarrheal illness remain a leading cause of mortality, there is still a lack of data on the relative role of Campylobacter compared with better-described agents such as rotavirus, enterotoxigenic Escherichia coli, and Shigella spp. In addition, postinfective complications have largely been described in studies from industrialized countries. Although these complications also occur in developing countries, they have not been quantified. In the coming years, this project will be an area of intense work. In industrialized countries and countries in the transition phase, recent studies have shed light on the burden of illness. Metrics that allow a comparison of the burden of illness between different countries and over time have contributed to this, but many important areas of research still remain. One particular area of further work is vulnerability and burden of illness in certain population groups at risk. India, as a country in a transitional phase, has a growing population of persons with diabetes. This is one example of a large population group at a high risk of illness and complications from illness. Unless steps are taken to secure safe food, burden of illness may increase as population demographics change. Acknowledgments. We thank Marie-JosCe Mangen, Jeanet Kemmeren, Sylvia Vijgen, and Juanita Haagsma for their contributions to the Dutch disease burden studies; Gerhard Falkenhorst, Morten Helms, and Jacob Simonsen for the Danish studies; and Wilfrid van Pelt and Rob Tauxe for helpful discussions when drafting this chapter. The European Network of Excellence MED-VET-NET supported our work.
REFERENCES Ang, C. W., W. Van Pelt, P. Herbrink, J. Keijser, Y. T. H. P. Van Duynhoven, and C. E. Visser. 2007. Sero-epidemiologyindicates frequent and repeated exposure to Campylobacter during childhood. Zoonoses Public Health 5450. Codex Alimentarins Commission. 2005. Proposed Draft Principles and Guidelines for the Conduct of Microbiological Risk Management (MRM). Codex Committee on Food Hygiene, 37th session, 14 to 19 March 2005. Alinorm 05-28/13, appendix 3, 67-81. Codex Alimentarius Commission, Rome.
BURDEN OF ILLNESS OF CAMPYLOBACTERIOSIS
161
Coker, A. O., R D. Isokpehi, B. N. Thomas, K. 0. Amisu, and C. L. Obi. 2002. Human campylobacteriosis in developing countries. Emerg. Infect. Dis. 8:237-244. De Wit, M. A. S., M. P. G. Koopmans, L. M. Kortbeek, W. J. Wannet, J. Vinjt, F. Van Leusden, A. I. M. Bartelds, and Y. T. H. P. van Duynhoven. 2001. Sensor, a population-based cohort study on gastroenteritis in The Netherlands: incidence and etiology. Am. 1.Epidemiol. 154:666-674. Ethelberg, S., J. Simonsen, P. Gerner-Smidt, K. E. Olsen, and K. Molbak. 2005. Spatial distribution and registry-based casecontrol analysis of Campylobacter infections in Denmark, 19912001. Am. 1.Epidemiol. 162:1008-1015. Evers, E. G., H. J. Van Der Feld-Klerx, M. J. Nauta, J. F. Schijven, and A. H. Havelaar. In press. Campylobacter source attribution by exposure assessment. Int. 1.Risk Assess. Manage. Flint, J. A., Y. T. Van Duynhoven, F. J. Angulo, S. M. DeLong, P. Braun, M. Kirk, E. Scallan, M. Fitzgerald, G. K. Adak, P. Sockett, A. Ellis, G. Hall, N. Gargouri, H. Walke, and P. Braam. 2005. Estimating the burden of acute gastroenteritis, foodborne disease, and pathogens commonly transmitted by food: an international review. Clin. Infect. Dis. 41:698-704. Gold, M. R., D. Stevenson, and D. G. Fryback. 2002. HALYS and QALYS and DALYS, oh my: similarities and differences in summary measures of population health. Annu. Rev. Public Health 23~115-134. Havelaar, A. H., A. E. M. De Hollander, P. F. M. Teunis, E. G. Evers, H. J. Van Kranen, J. F. M. Versteegh, J. E. M. Van Koten, and W. Slob. 2000. Balancing the risks and benefits of drinking water disinfection: disability adjusted life-years on the scale. Environ. Health Perspect. 108:3 15-321. Havelaar, A. H., M. J. Mangen, A. A. de Koeijer, M. J. Bogaardt, E. G. Evers, W. F. Jacobs-Reitsma, W. van Pelt, J. A. Wagenaar, G. A. de Wit, H. van der Zee, and M. J. Nauta. 2007. Effectiveness and efficiency of controlling Campylobacter on broiler chicken meat. Risk Anal. 272331-844. Havelaar, A. H., M. J. Nauta, M.-J. J. Mangen, A. G. De Koeijer, M.-J. Bogaardt, E. G. Evers, W. F. Jacobs-Reitsma, W. Van Pelt, J. A. Wagenaar, G. A. De Wit, and H. Van der Zee. 2005. Costs and Benefits of Controlling Campylobacter in The NetherlandsIntegrating Risk Analysis, Epidemiology and Economics. National Institute for Public Health and the Environment, Bilthoven. Helms, M., J. Simonsen, and K. Melbak. 2006. Foodborne bacterial infection and hospitalization: a registry-based study. Clin. Infect. Dis. 42:498-506. Helms, M., J. Simonsen, K. E. Olsen, and K. Melbak. 2005. Adverse health events associated with antimicrobial drug resistance in Campylobacter species: a registry-based cohort study. 1. Infect. Dis. 191:1050-1055. Helms, M., P. Vastrup, P. Gerner-Smidt, and K. Melbak. 2003. Short and long term mortality associated with foodborne bacterial gastrointestinal infections: registry based study. BM] 326: 357. Johnson, J. A., N. Luo, J. W. Shaw, P. Kind, and S. J. Coons. 2005. Valuations of EQ-5D health states: are the United States and United Kingdom different? Med. Care 43:221-228. Katsma, W. E., A. A. De Koeijer, W. F. Jacobs-Reitsma, M. J. Mangen, and J. A. Wagenaar. 2007. Assessing interventions to reduce the risk of Campylobacter prevalence in broilers. Risk Anal. 27:863-876. Kemmeren, J. M., M.-J. J. Mangen, Y. T. H. P. Van Duynhoven, and A. H. Havelaar. 2006. Priority Setting of Foodborne Pathogens-Disease Burden and Costs of Selected Enteric Pathogens. National Institute for Public Health and the Environment, Bilthoven. Mangen, M.-J. J., A. H. Havelaar, and K. J. Poppe. 2005. Controlling Carnpylobacter in The Chicken Meat Chain-Estimation
162
MaLBAK AND HAVELAAR
of Intervention Costs. Agricultural Economics Institute, The Hague. Mead, P. S., L. Slutsker, V. Dietz, L. F. Mccaig, J. S. Bresee, C . Shapiro, P. M. Griffin, and R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg. Infect. Dis. 5:607625. Murray, C . J. L., and A. D. Lopez. 1996. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Harvard School of Public Health, Cambridge, MA, on behalf of the World Health Organization and the World Bank. Mylius, S. D., M. J. Nauta, and A. H. Havelaar. 2007. Crosscontamination during food preparation: a mechanistic model applied to chicken-borne Campylobacter. Risk Anal. 27:803-8 13. Melbak, K. 2005. Human health consequences of antimicrobial drug-resistant Salmonella and other foodborne pathogens. Clin. Infect. Dis. 41:1613-1620. Nauta, M. J., W. F. Jacobs-Reitsma, and A. H. Havelaar. 2007. A risk assessment model for Campylobacter in broiler meat. Risk Anal. 27:845-861. Oberhelman, R. A., and D. N. Taylor. 2000. Campylobacter infections in developing countries, p. 139-154. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Perlman, D. M., N. M. Ampel, R. B. Schifman, D. L. Cohn, C . M. Patton, M. L. Aguirre, W. L. Wang, and M. J. Blaser. 1988. Persistent Campylobacter jejuni infections in patients infected with the human immunodeficiency virus (HIV). Ann. Intern. Med. 108540-546. Schwarzinger, M., M. E. Stouthard, K. Burstrom, and E. Nord. 2003. Cross-national agreement on disability weights: the European Disability Weights Project. Popul. Health Metr. 1:9. Simonsen, J., K. Krogfelt, K. Melbak, and G. Falkenhorst. 2007. Sero-epidemiology as a tool for estimating the incidence of Campylobacter infections, abstr. RR04. 3rd Med-Vet-Net Annu. Sci. Mtng.
Skirrow, M. B., and M. J. Blaser. 2000. Clinical aspects of Campylobacter infection, p. 69-88. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Smith, J. L., and D. Bayles. 2007. Postinfectious irritable bowel syndrome: a long-term consequence of bacterial gastroenteritis. J. Food Prot. 70:1762-1769. Teunis, P. F. M., and A. H. Havelaar. 2000. The beta-Poisson dose-response model is not a single-hit model. Risk Anal. 20: 513-520. Van Lier, E. A., and A. H. Havelaar. 2007. Disease Burden o f Infectious Diseases in Europe: A Pilot Study. National Institute for Public Health and the Environment, Bilthoven. Van Pelt, W., M. A. de Wit, W. J. Wannet, E. J. Ligtvoet, M. A. Widdowson, and Y. T. van Duynhoven. 2003. Laboratory surveillance of bacterial gastroenteric pathogens in The Netherlands, 1991-2001. Epidemiol. Infect. 130:431-441. Vijgen, S. M. C., M. J. M. Mangen, L. M. Kortbeek, Y. T. H. P. Van Duynhoven, and A. H. Havelaar. 2007. Diseuse Burden and Related Costs of Cyptosporidiosis and Giardiasis in The Netherlands. National Institute for Public Health and the Environment, Bilthoven. Wassenaar, T. M., M. Kist, and A. de Jong. 2007. Re-analysis of the risks attributed to ciprofloxacin-resistant Campylobacter jejuni infections. Int. J. Antimicrob. Agents 30:195-201. Wheeler, J. G., D. Sethi, J. M. Cowden, P. G. Wall, L. C. Rodrigues, D. S. Tompkins, M. J. Hudson, and P. J. Roderick. 1999. Study of infectious intestinal disease in England: rates in the community, presenting to general practice, and reported to national surveillance. The Infectious Intestinal Disease Study Executive. BMJ 318:1046-1050. World Health Organization. 2004. Guidelines for Drinking-Water Quality, 3rd ed., vol. 1, Recommendations. World Health Organization, Geneva. Ustun, T. B., J. Rehm, S. Chatterji, S. Saxena, R. Trotter, R. Room, and J. Bickenbach. 1999. Multiple-informant ranking of the disabling effects of different health conditions in 14 countries. WHO/NIH Joint Project CAR Study Group. Lancet 354: 111-1 15.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 9
Epidemiology of Campylobacter jejuni Infections in Industrialized Nations CHRISTINE
K.
OLSON, STEEN
ETHELBERG, WILFRID VAN PELT, AND ROBERTv. TAUXE
Campylobacter organisms have long been recognized as a cause of diarrhea in cattle and of septic abortion in both cattle and sheep. More recently, they have been recognized as an important cause of human illness. Campylobacter organisms may have been what Escherich described as motile Vibrio-like organisms in the stools of infants with diarrhea in Germany in 1881 (Escherich, 1881). However, in 1962, Campylobacter, then still known as “related Vibrio,” was described as a rare and opportunistic human pathogen that was isolated from blood cultures of immunocompromised humans (King, 1962). It was not until 1972 that the most commonly identified species, Campylobacter jejuni, was first isolated from human diarrheal stools by applying a filtration technique that was originally developed for veterinary diagnosis (Dekeyser et al., 1972). The subsequent development of selective Campylobacter stool culture media led to the recognition that C. jejuni was a common cause of human diarrheal illness everywhere, often more common even than Salmonella. Since then, knowledge of the epidemiology of the infections this organism causes has been steadily accumulating. The majority of these infections are sporadic; C. jejuni is far less often recognized as a cause of outbreaks than is Salmonella. As a result, much of what we know has come from the study of sporadic cases with casecontrol methodologies and other epidemiological techniques. Campylobacter is a common cause of diarrheal illness in the developing world, particularly in the first few years of life (Coker et al., 2002). However, the scope of this chapter is restricted to the industrialized world. An excellent chapter in the previous
edition of this book summarized well the available information on Campylobacter in the developing world, and the interested reader is referred to that publication for further information (Oberhelman and Taylor, 2000).
INFECTIOUS DOSE AND IMMUNITY The infectious dose of C. jejuni for humans appears to be low. Robinson (1981) reported his own illness that followed ingestion of 500 organisms. Human volunteer studies at the University of Maryland have shown that half the subjects became infected after ingesting 800 organisms, the lowest dose tested. Approximately half also became ill after exposure to 90,000 organisms, although curiously, illness was less frequent at higher doses (Black et al., 1988). The variation in illness attack rates could reflect prior immunity. Data suggest that in a previously naive population, high rates of illness could well follow low doses (Teunis and Havelaar, 2000; Teunis et al., 2005). Protective immunity to Campylobacter is well documented after chronic exposure to a risky food, such as raw milk (Blaser et al., 1987; Jones et al., 1981). Persons who had already been infected once did not develop illness a second time on rechallenge with the same strain in volunteer trials (Black et al., 1988). In Scotland, older persons tend to be infected with less common serotypes, possibly indicating that repeated exposure may immunize a fraction of the population to the most common serotypes (Miller et al., 2005). Persons with AIDS are at increased risk of acquiring Campylobacter infections (Angulo and
Christine K. Olson Enteric Diseases Epidemiology Branch, Centers for Disease Control and Prevention, Atlanta, GA 30333. Steen Wilfrid van Ethelberg * Departments of Epidemiology and Bacteriology, Statens Serum Institut, Copenhagen, DK-2300, Denmark. Pelt Department of Epidemiology and Surveillance, Rijksinstituut voor Volksgezondheid en Milieu, 3721 MA Bilthoven, The Netherlands. Robert V. Tauxe Division of Foodborne, Bacterial and Mycotic Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30333.
163
164
OLSON ET AL.
Swerdlow, 1995). The reported incidence of laboratory-confirmed campylobacteriosis in persons with AIDS in Los Angeles County between 1983 and 1987 was 519 per 100,000, much higher than the reported rate in the general population (Sorvillo et al., 1991). Host factors may play a role in the risk of development of C. jejuni bacteremia. In the same Los Angeles study, patients with AIDS were more likely to develop bacteremia from Campylobacter than were similar patients without AIDS who had Campylobacter infection (Sorvillo et al., 1991). However, no association with invasive disease was seen in a study of 3 0 human immunodeficiency virus-infected patients in a study at a London hospital between 1986 and 1991 (Nelson et al., 1992). National Campylobacter Surveillance in the United States Passive surveillance The Centers for Disease Control and Prevention (CDC) began national Campylobacter surveillance in 1982 using weekly written reports from state public health laboratories (Riley and Finch, 1985; Tauxe et al., 1988). In the early 1990s, this was converted to an electronic data transmission that used the Public Health Laboratory Information System (Bean et al., 1992). The data reported from the state public health laboratories include the Campylobacter species isolated, the date of reporting to the CDC, the age and sex of the person from whom the organism was isolated, the person’s county of residence, and the clinical source of the culture. Unlike Salmonella or Shigella, however, public health laboratories do not serotype Campylobacter routinely, and states vary widely in whether they require Campylobacter isolates to be sent to public health laboratories after being isolated in a clinical laboratory. In many states, few isolates are referred from clinical laboratories to state public health laboratories outside of an outbreak setting, so the number of isolates reported through this system greatly underestimates the number of diagnosed cases. For example, in 2005, 7,578 isolates of Campylobacter were reported through this system, one-fifth of the number of Salmonella isolates reported through this system, although the actual incidence of diagnosed infections identified by active surveillance was similar between the two organisms. Requirements for notification of diagnosed cases of Campylobacter infections have also been highly variable, and notification does not happen at all in many states. Until the establishment of the Foodborne Disease Active Surveillance System (FoodNet), accurate estimates of the burden of Campylobacter disease in the United States have not been possible (Scallan, 2007).
Active surveillance In 1996, the CDC, together with the U.S. Department of Agriculture’s Food Safety and Inspection Service, the U.S. Food and Drug Administration, and the state health departments of the Emerging Infections Program sites, developed FoodNet to better determine the frequency and severity of foodborne disease (Angulo et al., 1997). FoodNet is an active population-based surveillance system that began conducting surveillance in 1996 in five sites: Minnesota, Oregon, and selected counties in California, Connecticut, and Georgia. Addition of five more sites (Maryland, New Mexico, Tennessee, and selected counties in Colorado and New York) and expansion of some of the initial sites mean that the current population of the FoodNet surveillance area is 44.9 million persons (15% of the U.S. population) and is relatively closely representative of the U.S. population, except for underrepresentation of the Hispanic population (9% of FoodNet site population versus 14% of the U.S. population) (Scallan, 2007). Public health officials from each site regularly contact microbiology laboratories in their surveillance catchment area to collect information on all culture-confirmed cases of the foodborne pathogens under surveillance and transmit it to CDC. Information is collected on culture-confirmed cases of Campylobacter, Cryptosporidium, Cyclospora, Listeria monocytogenes, Salmonella, Shiga toxin-producing Escherichia coli including 0157:H7, Shigella, Vibrio, and Yersinia infections. The system is active because public health staff reach out to the clinical laboratories to gather reports, rather than waiting passively for reports to arrive. Virtually all clinical laboratories in the FoodNet catchment area routinely culture diarrheal stool specimens for Campylobacter (Angulo et al., 1997). FoodNet thereby provides timely data on the incidence of diagnosed Campylobacter infections. Trends The number of illnesses from Campylobacter reported in the FoodNet population has increased from 1996 through 2005 as more FoodNet sites have been added and the population expanded. However, the crude incidence rate of Campylobacter infection has declined from 23.6 per 100,000 population in 1996 to 12.7 in 2005, with a mean over that 10-year period of 16.3 infections per 100,000 persons (Fig. l). In 2005, Campylobacter had the second highest incidence rate of all pathogens under surveillance in FoodNet after Salmonella infections, which were 14.5 per 100,000 (CDC, 2006). Because the incidence was affected by the variation among sites as
CHAPTER 9
0
C. TETUNI INFECTIONS IN INDUSTRIALIZED NATIONS
___~
I -
1997
1996
1398
1999
2000
165
2001
2002
2003
2004
2005
Year Figure 1. Reported incidence of Cumpylobucter infections in FoodNet, by month and year, 1996 to 2005.
new sites were added, the overall trend is best estimated by a multivariable model that accounts for those differences. By means of a negative binomial Poisson model, the incidence in FoodNet declined by approximately 30% in 2005 compared with the baseline period of 1996 to 1998 (Fig. 2), with most of
Relative Rate (logscale)
the decline occurring before 2002 (CDC, 2006). This decline corresponds in time with the introduction of new meat and poultry safety regulations in the United States, and to an effort on the part of the industry to reduce contamination of poultry meat at slaughter through increased chlorination and other process
_ _
1.0
__ _
_
_
_
_
~ --
- --
0.8 0.7
0.5 0.4
I1
r---------i
1996-1998
1999
_____
2000
2001
2002
-
I
I
1-
2003
2004
2005
I
Year Figure 2. Relative incidence rate of Cumpylobucter infections in FoodNet through 2005, compared with baseline period 1996 to 1998, using a negative binomial Poisson regression model to account for effect of new sites added over time (CDC, 2006).
166
OLSON ET AL.
28.63
o ?n ..
.s
.c,
JV
ce
0 0
I
All
CA
CO
CT
GA MD FoodNe t s ite
MN
NM
NY
OR
TN
Figure 3. Annual incidence of Cumpylobacter infections, by site, FoodNet, 2005. CA, California; CO, Colorado; CT, Connecticut; GA, Georgia; MD, Maryland; MN, Minnesota; NM, New Mexico; NY, New York; OR, Oregon; TN, Tennessee.
changes, although the impact of these changes on campylobacteriosis is unproven (U.S. Department of Agriculture, 1996).
and thus may reflect real differences in infection rates, although no regional differences in risk factor exposures have been identified.
Regional differences in incidence
Age and sex distribution
Differences in Campylobacter infection rates vary substantially by region. In 2005, rates per 100,000 population demonstrated California, Colorado, and New Mexico to have the highest rates of infection: 28.6, 19.1, and 18.2 cases per 100,000 population, respectively, whereas rates in the Eastern states were substantially lower (Fig. 3 ) . These differences have not been shown to be related to regional differences in rates of culturing or reporting of cases
In 2005, for children younger than 1 year of age, the rate of Campylobacter infection was 27 per 100,000, a rate substantially higher than for other age groups (Fig. 4). The next highest rates in male subjects occurred among those 1 to 4 and 40 to 49 years old and in female subjects among those 1 to 4 and 20 to 29 years old. Rates of Campylobacter infection are in general approximately 20% higher in boys and men than in girls and women, although the
0
8 0c
15 -
4
10
0
I-
I . I 1
<1
1-4
5-9
10-19
20-29
Age group
---
*
Females
30-39
40-49
50-59
-Males
Figure 4 . Incidence of Campylobacter infections by age group and sex, FoodNet, 2005.
6Dt
CHAPTER 9
C. TETUNI INFECTIONS IN INDUSTRIALIZEDNATIONS
prominent peak in young adult men observed in the 1980s has become less evident (Tauxe et al., 1988). The incidence rate of Campylobacter infection was higher for boys and men in all age groups, except for the 20- to 29-year-old age group, in which the female peak rate was 13 per 100,000. Some of this difference may be due to different food consumption and handling practices between men and women, with men tending to engage in riskier practices than women (Yang et al., 1998). Seasonality
Campylobacter infections reported through FoodNet show a consistent seasonal pattern, with rates of infection beginning to rise in March, peaking in June and July, and declining slowly during the fall (Fig. 1).The lowest rates are in December, January, and February. This may be due to higher consumption of contaminated products during the summer months, suboptimal refrigeration or freezing of contaminated foods at that time, and other as yet unidentified factors. Burden of disease The true annual incidence of Campylobacter infection in the United States is much greater than that identified by surveillance based on laboratory-
167
confirmed cases. The so-called surveillance pyramid describes this, with a broad base representing the many infections that actually occur and a narrow peak representing those that are diagnosed and reported (Fig. 5). Only a fraction of those who are ill seek medical care, only a fraction of those who seek care have specimens cultured, and only a fraction of cultures yield the organism, even in the case of true infection. By surveying the population to determine the frequency of diarrheal illness, and by estimating each of the fractions and combining them for bloody and nonbloody diarrhea, it was estimated that Campylobacter caused illness in 34 persons for every reported case in 1998 (Samuel et al., 2004). A similar approach applied to Salmonella developed a multiplier of 38.6 for Salmonella (Voetsch et al., 2004). By assuming that the population in FoodNet sites is representative of the total U.S. population, we can estimate there were 36,624 isolations of Campylobacter in the United States in 2005 (12.7 per 100,000 X 2005 U.S. population), and then, by use of the multiplier of 34, estimate that there were a total of 1,250,000 illnesses that year, an incidence of 432 per 100,000. Such a high rate of infection sustained over many years may mean that a substantial fraction of the population has been infected. During 2005, the vast majority of Campylobacter isolates came from stool specimens, and less than 1% of Campylobacter isolations were from normally
Lab tests for organism
Specimen obtained Person seeks care
Person becomes ill
Illness in the general population Figure 5. Surveillance pyramid used to estimate the burden of food-borne diseases in FoodNet.
168
OLSON ET AL.
sterile sites. In 2005, the rate of hospitalization for patients with Campylobacter was 14% in FoodNet sites. The overall case fatality rate from FoodNet data for 1996 to 2005 was 0.11%, or 11 per 10,000 culture-confirmed cases. Applying this case fatality rate to the estimated 36,624 isolations of Campylobacter in the United States, it can be estimated that there were approximately 40 deaths in 2005 due to diagnosed Campylobacter infections. National Campylobacter Surveillance in Other Parts of the Developed World
overall incidence of 51.6 per 100,000 population, including only reporting countries (EFSA, 2006; Gauci and Ammon, 2007). The Czech Republic reported an exceptionally high incidence of 296 cases per 100,000 population. In other European countries that traditionally have well-developed laboratorybased surveillance systems, reported incidence rates range from 50 to 90 cases per 100,000. Australia has a similar level of reported infections, whereas the incidence rate in New Zealand is the highest reported by any country, 396 per 100,000 in 2003 (Anonymous, 2006b; Baker et al., 2007).
Results of national surveillance systems for human campylobacteriosis
Trends
In Europe, the European Centre of Disease Prevention and Control (ECDC) was established in 2005 and has begun to collect and collate national surveillance data. In its first comprehensive epidemiological report covering 2005, campylobacteriosis was the most frequently occurring enteric bacterial infection, second only to Chlamydia infection among all diseases reported (Gauci and Ammon, 2007). There were about 200,000 campylobacteriosis cases reported among 28 countries covered, resulting in an
Most developed countries with long-standing surveillance have experienced an increase in the number of reported cases during the last 25 years (Fig. 6), and several countries have experienced dramatic increases over this period. In some countries, there has been a steady increase during the entire period, whereas in other countries, the increase has primarily occurred in the 1990s. Denmark reported no increase between 1980 and 1990 but had a fourfold increase between 1992 and 2001 (Anonymous, 2006a). New
I
400
n i ” 2
-+.
Denmark & Wales +England
300
0
1
250
L 1, 200
0 C
$ 150
.-0
=
i
-Sweden -m-- New Zealand * Australia ---I- .ECDC
- -
100
50 0
Year 1983 1987 1991 1995 1999 2003
Year Figure 6 . Trends in reported incidence of Campylobacter infections for six countries and the European Union, 1979 to 2006. Data sources: Denmark, Statens Serum Institut (http: //www.mave-tarm.dk) and Statistics Denmark; Norway, Folkehelseinstituttet (http://www.msis.no) and Statistics Norway; Sweden, Smittskyddinstitutet (http://www.smittskyddsinstitutet.se/ statistik/), M. Lofdahl (personal communication), and Statistics Sweden; England and Wales, Health Protection Agency (http: //www.hpa.org.uk/infections/topics-az/campy/data-ew.hun); Australia, National Notifiable Diseases Surveillance System (http: // http: //www.health.gov.au/internet/wcms/publishing.nsf/ Content/ health-pubhlth-strateg-communic-index.htm) and J. Musto, NSW Department of Health (personal communication); NZ, M. Baker, University of Otago, Wellington (personal communication) and Statistics New Zealand; ECDC, European Centre of Disease Prevention and Control and A. Amato (personal communication).
CHAPTER 9
-
C. TETUNI INFECTIONS IN INDUSTRIALIZED NATIONS
Zealand had a nearly ninefold increase in incidence, from 14 to 120 cases per 100,000 population, between 1981 and 1990, and another threefold increase, from 128 cases to 396 cases from 1991 to 2003 (Baker et al., 2007; Brieseman, 1990). The early phase of this increase in Campylobacter incidence is likely due to increased physician awareness about this pathogen, increased culturing by laboratories, improved laboratory methods for detection, and improved reporting (Brieseman, 1990; Cowden, 1992; Kapperud and Aasen, 1992; Skirrow, 1991). However, the increase observed during the 1990s, when culturing for Campylobacter had become routine in many laboratories, more likely reflects a true increase in infections (World Health Organization WHO], 2001). Since the beginning of the 2000s, the reported number of infections has been stable or has moderately decreased in several countries (Fig. 6), and it is possible that we have now seen the peak of the epidemic of Campylobacter in part of the industrialized world. The mechanisms behind the long-term changes in the number of infections are likely multifactorial and are not understood in detail, although in recent years, consumption and handling of fresh poultry has been generally accepted as the main source of human infections (WHO, 2001). In accordance with this, control measures aimed at reducing the level of contaminated chicken meat have been enacted in some countries, including England and Wales and all five Scandinavian countries. These have included measures such as improved biosecurity barriers at chicken farms, increased hygiene at transport and slaughter, freezing of contaminated poultry, labeling of chicken products, use of leak-proof packaging, and campaigns targeting consumers. These initiatives have generally resulted in a reduction of contaminated poultry sold at the retail level and likely contributed to a recently observed reduction in the number of human infections, although a causal association has been difficult to measure and reductions have also been seen in countries where control programs have not been implemented (Borck et al., 2007; Hofshagen and Kruse, 2005; Stern et al., 2003). In Iceland, after an epidemic increase in Campylobacter infections in the late 1990s, an intensive prevention effort was begun in 2000 that included public education about poultry handling, on-farm biosecurity measures, and freezing carcasses of chickens from flocks known to have Campylobacter. As a result, farmers producing chicken without Campylobacter could sell them at higher prices, as consumers prefer fresh chicken (Stern et al., 2003). The following year, domestically acquired Campylobacter infec-
169
tions declined by 70% as farmers producing chicken without Campylobacter got the premium price for fresh chicken. The reduction in Campylobacter counts after freezing of poultry carcasses has been well documented (Georgsson et al., 2006). In 2003, a similar strategy-freezing all chicken from Campylobacter-positive flocks-was launched in Denmark, along with a multipronged effort to reduce contamination from farm to table. That year, a decrease in Campylobacter infections occurred and has continued, with a decrease of 28% in human cases between 2001 and 2006. This has been accompanied by a decrease from 42 to 30% in the proportion of Danish flocks that yield Campylobacter and from 47 to 12% in the proportion of broiler meat at retail that test positive for Campylobacter (Borck et al., 2007). In New Zealand, where such measures have not been introduced, the annual increase in culture-confirmed Campylobacter cases mirrored the 10-fold increase in the annual consumption of fresh chicken meat from 1981 to 2005, during which time there was little change in the consumption of frozen chicken (Baker et al., 2006). Comparison of disease incidence between countries Although Campylobacter is the most frequently reported cause of acute bacterial gastroenteritis in many developed countries around the world, substantial variation in incidence rates has been observed between countries (Gauci and Ammon, 2007; WHO, 2001). The differences in incidence rates can be difficult to compare because they reflect differences in surveillance that result from varying cultural norms for health care-seeking and diagnostic practices, and in the surveillance systems themselves, as well as reflecting real differences in infection rates (Gauci and Ammon, 2007; van Pelt et al., 2003a). For example, much of the difference in incidence rates between the otherwise comparable countries Denmark and The Netherlands is explained by the fact that all Danish laboratories report Campylobacter diagnoses to the national level, compared with 51% of the Dutch laboratories (Anonymous, 2006a; van Pelt et al., 2003a). In France, reporting is based on the Campylobacter isolates received at the national reference center, which are usually collected from patients who have been diagnosed in hospitals (Mtgraud, 1999). In the Czech Republic, the reported incidence is severalfold higher than in other European countries (Gauci and Ammon, 2007). This may be due at least partly to frequent culturing of patients, to a policy of active case-finding among contacts of index cases, and to efficient reporting. In many industrialized countries, laboratory-based surveillance programs include Cam-
170
OLSON ET AL.
pylobacter, but they vary in geographic coverage. Some report Campylobacter routinely from all jurisdictions, others are partial or sentinel systems, and for some, the degree of coverage is not reported. A number of European countries do not have routine culturing of diarrheal stool specimens or a national surveillance system for Campylobacter at all. Comparison of the fraction of cases that are captured by surveillance should help international comparisons, although these estimates are just beginning to be systematically developed (Flint et al., 2005). By use of a prospective sentinel clinic-based approach, in England and Wales, laboratory surveillance for Campylobacter has been estimated to capture 13% of all community Campylobacter cases compared with about 5% in The Netherlands (van Pelt et al., 2003a; Wheeler et al., 1999). Some of the difference between the United Kingdom and The Netherlands may be related to partial reporting in the latter country (see above). To be most useful, the multipliers should be estimated in the same way in each country. Differences in the way that partial reporting is accounted for may lead to different multipliers. Even modest variation in the case definition of gastroenteritis may halve or double the number of estimated cases of that condition (Majowicz et al., in press). Countries may also differ in the true incidence of infection and in the degree to which their populations are immune. At one extreme, populations in much of the developing world may be subject to such frequent and recurrent infections as young children that the infections in older children and adults are inapparent (Coker et al., 2002). In the future, serological surveys of populations may enable comparisons to be made. In The Netherlands, the rapid increase in seropositivity between ages 10 and 20 was used to estimate the number of infections that occurred in that age group. It was estimated that 30 to 50 infections may occur for each acute gastroenteritis illness diagnosed as Campylobacter (Nauta et al., 2005; van Pelt et al., 2005). In Denmark, a calculation for Salmonella enterica subsp. enterica serovar Enteritidis used antibody decay profiles after acute infection, combined with cross-sectional seroprevalence data, to estimate the true incidence of that infection (Simonsen et al., in press). In principle, the same can be done for Campylobacter (Strid et al., 2001). Because estimates based on serology capture asymptomatic infections as well as symptomatic ones, the multipliers estimated from seroprevalence data will be greater than the multipliers that are based on symptomatic disease alone. As the efficacy of serological methods for diagnosis of Campylobacter infection are further validated, including appropriate control sera, and as mathematical models advance,
the usefulness of seroepidemiology to study the incidence of Campylobacter infections will become clearer. Age and sex distribution The age distribution of infection in other industrialized nations is similar to that in the United States. Campylobacter affects all age groups, but in many countries, the infection has a bimodal age distribution, with one peak in small children and a second peak at between 15 and 44 years of age. By comparing available data from 14 European countries, the ECDC found the incidence in children under 5 years old to be 3 times higher than in other age groups and found a modest second peak in the 15-to-24 age group (Gauci and Ammon, 2007). A high incidence in small children was also noted for other enteric diseases, including Salmonella, which may be because parents are more likely to seek medical care for sick children and doctors will culture specimens taken from the children, as well as because a genuinely higher incidence in that age group actually exists. It has been suggested that in some countries, the peak in incidence in young adults may be related to their greater likelihood of foreign travel (Kapperud and Aasen, 1992; Stafford et al., 1996). The sex distribution of Campylobacter infection in developed countries in other parts of the world is also similar to the distribution in the United States. The incidence is 1.1to 1.5 times higher in boys and men compared with girls and women. This difference has been more pronounced in persons under 30 years old (Gauci and Ammon, 2007; Kapperud and Aasen, 1992; Skirrow, 1987). Seasonality
Many countries in temperate zones observe a sudden increase in Campylobacter infections in the spring, a well-defined summer peak, and a gradual decrease afterward (Kovats et al., 2005; Miller et al., 2004; Nylen et al., 2002; Skirrow, 1987). In England and Wales, the onset, size, and shape of the summer peak remained remarkably stable within regions over a 10-year period, although the size and shape of the summer peak differed across regions (Louis et al., 2005). When surveillance data from five countries in Europe were compared, seasonal variation became more pronounced with increasing latitude (Nylen et al., 2002). A study conducted in Norway found a similar south-to-north gradient within that country (Kapperud and Aasen, 1992). By comparing 11 European countries, Canada, Australia, and New Zealand, countries with milder winters have peaks of in-
CHAPTER 9
-
C. TElUNI INFECTIONS IN INDUSTRIALIZED NATIONS
et al., 2004). In Wales, the seasonal peak in chicken at retail was observed to occur 2 weeks before the peak in human cases (Meldrum et al., 2005). Temperature may play a role in Campylobacter sources in the farm environment that are seasonal, such as insects, rodents, or migratory birds (Jacobs-Reitsma et al., 1994). The consistent sudden increase at the same time each year suggests the presence of a seasonal environmental stimulus. Transmission by flies has been suggested as one such driver of the seasonality of campylobacteriosis in both humans and in poultry, through seasonally dependent fly prevalence (Ekdahl et al., 2005; Hald et al., 2004; Nichols, 2005). Seasonal shifts in vacation travel, and swimming and other exposures may also play a role. In Minnesota, winter spikes in fluoroquinolone-resistant infections were related to a surge in winter holiday travel to Latin America, where fluoroquinolone resistance was more common (Smith et al., 1999). In The Netherlands, a sudden dilution of foodborne Cumpylobacter with those from environmental source may explain why during the sudden increase in Campylobacter infections in the spring, fluoroquinolone resistance drops by 15% and then gradually increases again after the summer peak (Fig. 7) (van Pelt et al., 2007).
fection earlier in the year (Kovats et al., 2005). In New Zealand, the summer peak is pronounced and stable, whereas in Queensland, Australia, the seasonal trend varies from year to year and incidence is only modestly higher during warmer months (Kovats et al., 2005; Stafford et al., 1996). Analysis of infections in Swedish travelers shows a consistent summer peak in travelers returning from temperate zones, and a more variable or absent peak in travelers to the tropics (Ekdahl and Anderson, 2004). Reports from Israel and Hong Kong suggest that those countries experience a higher number of infections in winter and spring (Ho and Wong, 1985; Shmilovitz et al., 1982). Climatic factors such as temperature, sunlight, and humidity may drive the dynamics of Campylobacter transmission by affecting prevalence at the reservoir, and by affecting human behavior that increases exposure to Campylobucter, such as barbecuing meat or swimming in surface waters. In broilers tested at farm, slaughter, and retail, higher carriage rates or levels of contamination are found in the summer (Meldrum et al., 2005; Patrick et al., 2004; van Pelt et al., 2003b; Wilson, 2003). In Denmark and The Netherlands, similar seasonal trends are seen in both the percentage of infected broiler flocks and human infections (Fig. 7) (Patrick et al., 2004; van Pelt et al., 2003b). In Denmark, the seasonality of human infections correlated best with average daytime temperature and sunlight 4 weeks before the date the infection was reported (and thus approximately 1 to 2 weeks before exposure), and for broilers a time lag of 3 weeks before slaughter gave the best fit (Patrick
1-4
5-8
9-12
13-16
17-20
21-24
171
Burden of illness The burden of gastroenteritis caused by Campylobacter can be estimated by surveying the population to determine the frequency of recent gastro-
25-28
29-32
3336
37-40
41-44
45-48
49-52
4-weekly period (mean for 2002-2006)
Figure 7. Number of human infections with Cumpylobucter, proportion of isolates from humans resistant to ciprofloxacin, and prevalence of cecal carriage of Cumpylobucter in broiler flocks, by 4-week interval, The Netherlands, aggregated over the time period, 2002 to 2006 (van Pelt et al., 2007). Human surveillance coverage is approximately 50%.
172
OLSON ET AL.
enteritis and constructing a surveillance pyramid, as is done in FoodNet. It can also be measured more directly by monitoring a population prospectively, attempting to identify and culture all episodes of diarrhea, and constructing a surveillance multiplier from those data, as was done in the Infectious Intestinal Disease study in England and the SENSOR study in The Netherlands (de Wit et al., 2001; Wheeler et al., 1999). In these studies, the annual incidence of symptomatic Campylobacter infection in the population was estimated to be 7.6 cases for every one reported to national surveillance in England, and 19 cases for every one reported in The Netherlands (van Pelt et al., 2003a; Wheeler et al., 1999). Estimating the full public health burden due to Carnpylobacter includes hospital admissions, chronic sequelae, and mortality (see chapter 8 for a full description). In Denmark, it has been possible to estimate the frequency of outcomes by linking national registries of gastrointestinal pathogen isolation, hospital admissions and discharges, and deaths (Helms et al., 2003, 2006). Thus, 11% of patients with Campylobacter infection were hospitalized around the time of the acute illness with gastroenteritis or complications, and 0.3% died within a month of the diagnosis (Helms et al., 2006). The mortality in that month was 5 times greater than that observed in a similar population without Campylobacter, and the mortality was still slightly high up to a year after diagnosis (Helms et al., 2003). In Sweden, mortality in the month after domestically acquired Campylobacter infection was 3 times higher than in the population in general but was far lower in travel-associated infections, probably because travelers have fewer underlying health conditions in general (Ternhag et al., 2005). Common-Source Outbreaks Common-source outbreaks caused by Campylobacter are recognized infrequently, particularly in comparison with the frequency of sporadic campylobacteriosis. Unlike Salmonella, Cumpylobacter does not tolerate exposure to atmospheric oxygen or to drying, and does not multiply on foods left out for many hours (Bayliss et al., 2000; Blaser et al., 1980). This could explain the rarity of large outbreaks related to solid foods (Franco, 1988). Unlike Salmonella, serotyping and molecular subtyping methods have not yet been identified that are useful for routine public health surveillance, and outbreaks that are small or dispersed may be missed by current surveillance (Ethelberg et al., 2004; Gillespie et al., 2003). The low infectious dose, ease of cross-contamination,
multiple subtypes contaminating a food product, and relatively long incubation period may also obscure the source of a cluster of related illnesses (Gillespie et al., 2003). Common-source outbreaks in the United States In the United States, the CDC receives reports of foodborne and waterborne disease outbreaks, including those caused by Campylobacter species, as part of the national food- and waterborne disease outbreak surveillance systems (Dziuban et al., 2006; Liang et al., 2006; Lynch et al., 2006). The first reported outbreak was also the largest. It occurred in 1977 in Burlington, Vermont, when a contaminated community water supply affected an estimated 3,000 persons (Vogt et al., 1982). Between 1998 and 2004, a total of 102 outbreaks of Campylobacter infections were reported, or 14 outbreaks per year, affecting a total of 4,587 persons (Table 1).They represented approximately 1% of all foodborne and waterborne outbreaks between 1998 and 2002. Of these, 74 were caused by C. jejuni, one was caused by C. coli, and for 27 the species was not reported. Seven of the 12 waterborne outbreaks included other enteric pathogens in addition to Campylobacter. This was not the case for foodborne outbreaks. The reported sources of outbreaks have changed over time. Between 1978 and 1987, water and unpasteurized milk accounted for 56% of outbreaks, while between 1988 and 1996, other foods ac-
Table 1. Food- and waterborne outbreaks of Campylobacter infections reported in the United States, by vehicle, 1998-2004 Method of transmission Foodborne
Waterborne
Total
Vehicle Beef Dairy Eggs Game Pork Poultry Vegetables Fruit Finfish Shellfish Complex foods No specific food implicated Community water supply Non-community water supply Recreational water
No. of outbreaks
No. of outbreakassociated cases
1 25 0 1 0 11 5 1 0 1 21 24
15 404 0 2 0 71 471 14 0 2 591 405
3
143
7
2,443
2
26
102
4,587
Next Page CHAPTER 9
-
C. 7E7UNI INFECTIONS IN INDUSTRIALIZED NATIONS
counted for 83% of all outbreaks. From 1998 to 2004, 49% of all outbreaks were attributed to dairy products and drinking water, while 48% were attributed to other foods (44% of these were related to poultry). In that same time period, the two largest common-source foodborne Campylobacter outbreaks were a lettuce outbreak in 1998 in Minnesota (300 cases), and a tuna salad outbreak in a Washington state prison in 2002 (115 cases). Large waterborne outbreaks have occurred as a result of breaks in system integrity and disruptions in disinfection in municipal water systems, or when a nonchlorinated groundwater supply was contaminated by infiltration from the surface. In 2004, 1,450 people became ill with Campylobacter jejuni and other enteric pathogens after visiting a resort island in Lake Erie (O’Reilly et al., 2007). This outbreak occurred because of fecal contamination of the groundwater aquifer, probably through a combination of limestone sinkholes and poor sewage disposal practices (O’Reilly et al., 2007). Of 10 drinking water source outbreaks reported between 1998 and 2004, only one source (a community water source) was determined to have consistent disinfection. In 2006, the U.S. Environmental Protection Agency published new regulations that require disinfection of both ground and surface water before distributing it in community systems, which may decrease these outbreaks. Among foodborne outbreaks, milk has historically been the most common single food source, usually as raw or incompletely pasteurized milk. Milkor other dairy-associated outbreaks accounted for 38% of the foodborne outbreaks with reported food sources (Table 1).In the 1 9 8 0 ~ a ~common scenario for these outbreaks was a school field trip to a dairy farm where a taste of raw milk was part of the experience (CDC, 1986; Wood et al., 1992). In one outbreak with two different species of Campylobacter linked to raw milk, those with C. jejuni infections were older than those with C. fetus subsp fetus infections, although all consumed milk from the same bucket (Klein et al., 1986). Illness traced to the consumption of raw milk has been a continuing public health challenge, even when sanitary conditions on the farm are exemplary (CDC, 2002; Headrick et al., 1998). Milk can become contaminated by cattle feces or as a result of asymptomatic Campylobacter mastitis in dairy cattle (Hudson et al., 1984; Hutchinson et al., 1985; Sat0 et al., 2004). Raw goat milk has also been a source of an outbreak (Harris et al., 1987). Routine pasteurization of all milk and good dairy plant hygiene can prevent such outbreaks. A broad variety of other foods has been implicated in outbreaks. Campylobacter outbreaks have been associated with produce. Cantaloupe was re-
173
sponsible for an outbreak in Wisconsin in 1985 affecting 16 people and lettuce for an outbreak in Minnesota in 1998 affecting 300 people (Bowen et al., 2006; CDC, unpublished data). In an outbreak in Kansas in 1998, illness was related to eating either canned pineapple or gravy, which were likely to have been contaminated shortly before serving, perhaps by an ill food handler (Olsen et al., 2001). In one outbreak affecting at least 30 patrons of a restaurant, garlic butter used on sandwiches was implicated, perhaps after becoming contaminated in the kitchen (Berg et al., 1996). Campylobacter can survive in refrigerated butter for many days, and can survive the process of melting butter onto garlic bread (Zhao et al., 2000). Milk- and waterborne outbreaks are most likely to occur in the spring and fall but are uncommon in the summer, when sporadic cases are at their height. Foodborne outbreaks due to other vehicles increase in spring and continue steadily through December, without a summer peak. The difference between the seasonality of sporadic cases and that of outbreaks suggests that somewhat different ecological events drive them. In Wisconsin, the incidence of asymptomatic bovine carriage was higher in spring than in fall, possibly explaining the seasonal nature of milkassociated outbreaks (Sato et al., 2004). It is also possible that surface-water sources become contaminated during spring and fall bird migrations, leading to seasonal waterborne outbreaks. Common-source outbreaks in other developed nations Outbreaks reported from other developed nations are also less frequent than those caused by other bacterial foodborne pathogens; they are predominantly caused by contaminated poultry, drinking water, and milk. Starting in 2004, an annual report summarizing foodborne outbreaks in Europe has been produced by EFSA, the European Union Food Safety Agency. In 2005, reporting of foodborne outbreaks by European Union countries became mandatory, and 494 outbreaks of campylobacteriosis involving 2,478 individuals were reported (EFSA, 2006; Frost et al., 2002). The source of infection was reported for 107 outbreaks; 39 were caused by broiler or poultry meat, 5 by dairy products, 2 by drinking water, and the remaining by a variety of other foods. Although reporting from different countries varies in level of detail and 81% of outbreaks were reported by only two countries (Germany and Austria), this reporting system may be an increasingly important source of outbreak information in the future. In England and Wales in the second half of the
Previous Page 174
OLSON ET AL.
1990s, 50 Campylobacter outbreaks were reported, representing only 2% of all reported outbreaks of diarrheal disease (Frost et al., 2002). Route of transmission was reported to be foodborne in 70%, waterborne in 8%, and undetermined in 18%. Among 24 foodborne outbreaks where a vehicle of infection was found, poultry was implicated in 14 outbreaks, substantially more often than in foodborne outbreaks due to other pathogens (Frost et al., 2002). In Denmark, three outbreaks due to Campylobacter were reported over a 3-year period, and all were transmitted in workplace kitchens and found to be caused by cross-contamination from raw chicken to ready-to-eat foods in the kitchen (Lewis et al., 2007; Mazick et al., 2006). Cross-contamination in the kitchen may explain a prolonged Australian outbreak traced to cucumber served at a salad bar (Kirk et al., 1997) and an outbreak in England in 1993 associated with a melon-and-prawn appetizer (Pebody et al., 1997). An outbreak in a school in Spain was likely caused by custard contaminated via chicken in the kitchen, and in an Australian restaurant-associated outbreak, chicken was also found to be the source (Black et al., 2006; Jimenez et al., 2005). Raw or inadequately pasteurized milk is often associated with outbreaks. Outbreaks caused by contaminated milk have been seen regularly in Sweden and in the United Kingdom (Pebody et al., 1997; SMI, 1997). Raw milk may frequently contain Campylobacter. In a survey of retail raw milk in southwest England, 5.9% of samples yielded the organism (Humphrey and Hart, 1988). Waterborne outbreaks are often much larger than foodborne outbreaks, affecting several hundreds or thousand of individuals (Anonymous, 2000; Engberg et al., 1998; Kuusi et al., 2005; Maurer and Sturchler, 2000; Melby et al., 1991; Mentzing, 1981). The outbreaks typically result from weaknesses in water supply systems combined with special circumstances, and sometimes involve several pathogens. In Canada, a large outbreak of E. coli 0157:H7 and Campylobacter infections occurred when groundwater supplying the drinking water system of a small town became contaminated after a heavy rainfall (Anonymous, 2000; Clark et al., 2003). In a Danish town, nonchlorinated municipal drinking water supplied from groundwater was contaminated with sewage, leading to a 6-week outbreak with an attack rate of 87.5% (Engberg et al., 1998). A similar outbreak occurred 10 years later in another Danish town (Vestergaard et al., 2007). In a Swiss town, contamination of the groundwater by the sewage system was believed to be the cause of a large outbreak (Maurer and Sturchler, 2000). In a municipality in Finland, water main repair work led to the contamination of drinking water with sewage, sick-
ening an estimated 2,700 persons (Kuusi et al., 2005). Outbreaks have also been traced to drinking water supplied from surface water. In a Norwegian outbreak in 1990, a small unfenced reservoir may have been contaminated by runoff from a nearby sheep pasture after heavy rain (Melby et al., 1990). Campylobacter species are rapidly inactivated by chlorine, and failure to disinfect is a frequent contributing cause. An outbreak in northern Norway took place during 4 weeks in which the chlorination system underwent repairs. Surface water was thought to have been contaminated by sheep cattle or wild animals (Melby et al., 2000). Similar episodes may have contaminated spring water in two outbreaks in Christchurch, New Zealand (Brieseman, 1987; CDC, 1991). During investigation of an outbreak in England in 1993, a dead lamb was found in the water supply of a university dormitory (Pebody et al., 1997). Drinking water supplied from springs, boreholes, or private wells may frequently be contaminated with microorganisms and pose a particular risk of outbreaks. Half of the 25 reported outbreaks associated with private water supplies in England and Wales from 1970 to 2000 were caused by Campylobacter (Said et al., 2003). Two non-drinking-waterrelated outbreaks among participants in an annual bicycle race have been described in Norway. The affected races took place on rainy days, and illness was associated with mud splashes and completing the route at the time of day when it rained most heavily; contamination of the roadway with runoff from adjacent pastures was the presumed mechanism (Kapperud et al., 2000). It appears that the proportion of investigated outbreaks associated with chicken has increased during the last 15 years relative to outbreaks associated with milk or water. The reported chicken outbreaks appear typically to be point-source outbreaks with a setting in a restaurant, cafeteria, or other institution. The data suggest that some Campylobacter outbreaks linked to other foods are the result of cross-contamination from poultry products in kitchens, so that the proportion of outbreaks caused by chicken is underestimated. Campylobacter outbreaks are rarely associated with processed ready-to-eat foods. Sources of Sporadic Infections The vast majority of Campylobacter infections are not linked to outbreaks but occur as sporadic infections. As noted above, the predominant sources of sporadic infections may differ from those of outbreaks. Although it is often difficult to determine the source of an individual infection, epidemiologic investigations that use a case-control methodology can
CHAPTER 9
C. 7E7uNl INFECTIONS IN INDUSTRIALIZED NATIONS
identify exposures that are most likely to be associated with C. jejuni infection, from which sources can be inferred. Ten such published studies have been done in the United States, including two large nationwide studies (Table 2) (Deming et al., 1987; Effler et al., 2001; Engleberg et al., 1984; Friedman et al., 2004; Fullerton et al., 2007; Harris et al., 1986b; Hopkins et al., 1984; Hopkins and Scott, 1983; Potter et al., 2003; Saeed et al., 1993; Schmid et al., 1987). Twenty four similar investigations have been conducted in other developed nations (Table 3) (Adak et al., 1995; Cameron et al., 2004; Doorduyn et al., 2005; Eberhart-Phillips et al., 1997; Ethelberg et al., 2005; Ikram et al., 1994; Kapperud et al., 1992, 2003; Kist, 1982; Lighton et al., 1991; Neal and Slack, 1995, 1997; Neimann et al., 2003; Norkrans and Svedhem, 1982; Oosterom et al., 1984; Rodrigues et al., 2001; Salfield and Pugh, 1987; Schonberg-Norio et al., 2004; Schorr et al., 1994; Southern et al., 1990; Stafford et al., 2007; Studahl and Anderson, 2000; Tenkate and Stafford, 2001; Wingstrand et al., 2006). Although these studies differ in technique and in the array of hypotheses tested, they consistently indicate several dominant routes of transmission: consumption of and contact with poultry, transmission from pets and other animals, and drinking contaminated water and raw milk. Campylobacter is also a major cause of traveler’s diarrhea, probably because similar exposures occur while traveling.
The case-control method as used for sporadic cases Case-control studies of infections collect exposure information from persons with the illness and compare it with similar data from comparable persons without illness. Long established as a technique for outbreak investigations, it has also been applied successfully to sporadic cases. Well-conducted casecontrol studies can offer data to estimate the population attributable risk (PAR), the fraction of cases that can be attributed to a particular exposure (Schlesselman, 1982). However, even in large casecontrol studies, the PARS often only explain a small portion of cases. For example, in New Zealand, where the incidence of campylobacteriosis is the highest in the world and poultry is the dominant source identified, the PAR for poultry just reaches 50% (Eberhart-Phillips et al., 1997). Calculated PARS may be diminished by selective recall, poor memory, unobserved cross-contamination of multiple foods, and prior population immunity. For example, immune persons who are exposed may develop undetected inapparent infection; including them as controls may mask obvious risk factors and lower the
175
estimated PARS (Nauta et al., 2005; Swift and Hunter, 2004). It may be more appropriate to view the associations as general indicators of risk, rather than as precise measures of the magnitude of the risk. Case-control studies in the United States Six early studies were conducted in the 1980s, soon after the importance of Campylobacter infections was recognized (Table 2). In 1981, a large casecontrol study conducted in Seattle, Washington, showed that eating poultry, including chicken, turkey, and Cornish game hen, accounted for over 50% of cases (Harris et al., 1986b). Smaller numbers of cases were attributed to drinking raw milk, including both cow’s and goat’s milk, consuming raw fish or shellfish, having contact with pets, and drinking contaminated surface water, and 9% were related to exposure during overseas travel (Harris et al., 1986b; Saeed et al., 1993). That same year, a study conducted among Hopi and Navajo Native Americans identified ownership of farm animals as a risk factor (Engleberg et al., 1984). Also in 1981, a small study in Colorado identified drinking untreated water or raw milk, having contact with cats, and eating undercooked chicken as risk factors (Hopkins et al., 1984). In a second Colorado study conducted in 1983, handling raw chicken, rather than just eating it, emerged as a risk factor (Hopkins and Scott, 1983). In 1984, in a study conducted among students at a Georgia university, 70% of cases were attributed to eating chicken and 30% to having contact with cats (Deming et al., 1987). In rural parts of the United States, raw milk can be an important source of sporadic cases as well as of outbreaks. In 1983, a study conducted in Iowa identified drinking raw milk as the only risk factor, and this accounted for nearly half the cases (Schmid et al., 1987). In the last decade, four additional case-control studies of sporadic cases have yielded similar results. In a 1998 study conducted in Hawaii, which has the highest incidence of Campylobacter infections in the United States, eating poultry prepared by a commercial establishment was the principal risk factor exposure, and taking antibiotics before onset of illness was also independently associated with illness (Effler et al., 2001). In 1998-1999, FoodNet conducted a large multisite study (Friedman et al., 2004). In this study, 12% of cases were attributed to foreign travel, whereas among the domestically acquired cases, 30% were attributed to chicken or turkey prepared in restaurants and 21% to other meats. Smaller fractions were attributed to drinking surface water (3%) or raw milk (1.5%), eating raw seafood (3%), having contact with a puppy (5%), and having contact with farm
Table 2. Case-control studies of laboratory-confirmed sporadic campylobacteriosis within the United States No. of cases
No. of controls
Yr
Population
Country
% Travel
10 40
10 71
1981 1981
Native Americans Denver and Fort Collins, CO
USA USA
NR“ NR
10
15
1982
Larimer County,
USA
NR
46 218
46 526
1982-1983 1982-1983
Iowa Washington state HMO members
USA USA
NR 9 yo
45
45
1983-1984
USA
0%
211
21 1
1998
University, Georgia Hawaii‘
USA
Exclb
1,316
1,316
1998-1999
7 FoodNet sites
USA
12%
83
122
2000-2001
Michigan rural population
USA
NS
123
928
2002-2004
8 FoodNet sites; O-ll-monthold infants
USA
5Yo
co
“NR, not reported. bExcl, travel-associated cases excluded from study. ‘NS, reported as not significantly associated with illness.
Risk factors identified related to consumption of food or water Untreated water, undercooked chicken, raw milk Chicken preparation Raw milk Undercooked chicken, other meat, raw milk, mushrooms, untreated water, raw seafood Undercooked chicken Chicken from restaurant; antibiotic use; slightly: ham, turkey Poultry prepared at restaurant, other meats prepared at a restaurant, raw milk, raw seafood, untreated water
Untreated well water, contact with meat or poultry in shopping cart, fruit and vegetables at home
Animal,environmental
contact Farm animals Cats
Reference(s) Engleberg et al. (1984) Hopkins et al. (1984)
Hopkins and Scott (1983)
Animals with diarrhea
Schmid et al. (1987) Harris et al. (1986a, 1986b), Saeed et al. (1993)
Cats
Deming et al. (1987)
Puppies, chickens
Effler et al. (2001)
Farm animals, animal stool, puppies
Friedman et al. (2004)
Poultry husbandry; dose-response for contact with poultry Visiting or living on farm, puppies
Potter et al. (2003)
Fullerton et al. (2007)
CHAPTER 9
c. r E r u w INFECTIONS IN INDUSTRIALIZED NATIONS
animals (4%). Nested within this study, the antimicrobial resistance of the isolates was measured, permitting determination of risk factors for fluoroquinolone-resistant infections: eating poultry prepared at a commercial establishment and traveling abroad (Kassenborg et al., 2004). In a 2000-2001 study conducted in rural Michigan, living on a farm and specifically having contact with poultry was associated with illness, and the risk rose with the frequency of poultry contact (Potter et al., 2003). In 2002-2004, FoodNet conducted an eight-site study focusing on infants younger than 1 year (see below) (Fullerton et al., 2007). Case-control studies in other developed nations Among 24 studies conducted in Europe, Australia and New Zealand, consumption of poultry was identified as a risk factor in 16 studies (Table 3). Nine studies identified other meats as a risk factor, most often pork or a variety of barbecued meats. Five studies identified drinking untreated water as a risk factor, and in two, swimming in natural surface waters was a risk. Drinking raw milk was identified in three studies, in Sweden, Denmark, and New Zealand. One study in the United Kingdom identified occupational contact with raw meat as a risk factor for infection, 13 identified contact with pets or farm animals as a risk factor, and 10 identified foreign travel. Pediatric infections Four studies focused on pediatric infections. Three non-U.S. studies all identified contact with dogs as a risk factor; one also identified untreated water, grilled meat, and raw milk; and one curiously identified mayonnaise as the risky exposure (Carrique-Mas et al., 2005; Salfield and Pugh, 1987; Tenkate and Stafford, 2001). The U.S. study found that breast-feeding was highly protective, and that foreign travel accounted for 5% of infant cases (Fullerton et al., 2007). In that study, drinking well water and riding in a shopping cart next to raw meat or poultry were risk factors for infants younger than 7 months, while being on a farm, having a pet with diarrhea, or eating fruits and vegetables prepared in the home were risk factors for infants aged 7 to 11 months. Sporadic infections and poultry The studies above shed some epidemiologic light on how poultry can cause infection. Location and circumstances of preparation often play a role. Some studies find that chicken prepared in the home is a
177
risk factor, but more often it is chicken eaten outside of the home; eating chicken in the home may be protective. Indeed, the large U.S. study found no risk for eating chicken in general, but both strong risk and protective effects were found when that exposure was evaluated by location of consumption (Friedman et al., 2004). It is unclear whether this might reflect differences in the contamination of chicken sold for restaurant use compared with that sold in grocery stores, or differences in variation in preparation practices between homes and restaurants. In Denmark, eating fresh, unfrozen chicken meat was the dominant risk factor, while the risk from other, previously frozen, chicken was only borderline significant (Wingstrand et al., 2006). Eating chicken that is perceived to be undercooked is often a risk factor, although one that accounts for a relatively small number of cases. Because incompletely cooked chicken would typically still be cooked on the surface, this risk factor suggests that there is contamination of deep muscle. Such contamination may be common, although at low counts. One survey found 22% of chicken legs had detectable Campylobacter in deep tissues next to the bone, and a survey of breast meat found internal contamination in 20% (Luber and Bartelt, 2007; Scherer et al., 2006). In both studies, levels of surface contamination were far greater than those of deep tissue contamination, leading the investigators to conclude that cross-contamination from the surface of the meat is the risk. Inapparent bacteremia was recently documented in 5 to 11% of broilers, suggesting a mechanism for deep tissue contamination (Richardson, 2007). In two studies, the risk was associated with handling raw chicken rather than with eating it (Hopkins and Scott, 1983; Norkrans and Svedhem, 1982). Ingestion of a drop of heavily contaminated raw chicken juice could provide a dose of 500 organisms. One study in the United Kingdom measured lo3 to l o 6 Campylobacter organisms on the surface per fresh chicken carcass, and two recent studies from Germany found a median of 250 CFUs per gram of skin on chicken legs and 2,000 CFUs per gram of chicken breast fillet (Hood et al., 1988; Luber and Bartelt, 2007; Scherer et al., 2006). This means that reducing the number of organisms on each carcass might have an important impact on public health even if there is little change in the proportion of carcasses that are contaminated. Other observations underline the importance of the poultry reservoir for human Campylobacter infections. In 1999, an inadvertent experiment occurred in Belgium when locally produced poultry was withdrawn from the market for several months
Table 3. Case-control studies of sporadic campylobacteriosis in industrialized countries outside the United States No. of cases
No. of controls
Yr
Population
Country
% Travel
Risk factors identified related to consumption of food or water
114 55
90 14
1979-1980 1980
Freiburg Gothenburg
West Germany Sweden
NR" u p to 75%
Chicken consumption Chicken preparation
44
54
1982
Rotterdam
The Netherlands
NS
Chicken, pork, food from barbecue
15 52
90 103
1984-1985 1999-2000
1 city, <6 yr 3 counties
England Norway
NR Excl"
32
64
1990
Bridgend
Wales
NR
29
41
1990
Manchester
England
NS
598
598
1990-1991
Nationwide
NR
167
282
1991
7 laboratories
England and Wales Switzerland
100
100
1992-1993
Christchurch
New Zealand
NR
282
318
1993-1994
England
Yes
313
512
1994-1995
Nottingham, >18 yr old Nottingham, >18 yr old
England
Up to 25%
621
621
1994-1995
4 areas
New Zealand
5%
229
229
1993-1996
Nationwide
9%
101
198
1995
1 county
England and Wales Sweden
Up to 46%
Excl
Poultry, sausage, crosscontamination Milk from bird-pecked bottles Milk from bird-pecked bottles Profession working with raw meat; untreated water Poultry liver, foreign citizenship Poultry outside home, undercooked or barbecued poultry Chicken, preparing raw chicken Antisecretory drugs, diabetes mellitus, bird-pecked milk, poultry Chicken pieces, raw or undercooked chicken, chicken eaten outside of home, raw milk, drinking untreated rainwater Chicken in restaurant Raw milk; chicken; pork; barbecue
Animal / environmental contact
Puppies, dogs
Puppies Dogs
Reference(s) Kist (1982) Norkrans and Svedhem (1982) Oosterom et al. (1985) Salfield and Pugh (1987) Kapperud et al. (1992) Southern et al. (1990) Lighton et al. (1991)
Pet with diarrhea
Adak et al. (1995) Schorr et al. (1994) Ikram et al. (1994)
Puppies
Neal and Slack (1995)
Puppies
Neal and Slack (1997)
Puppies, calf feces
Eberhart-Phillips et al. (1997)
Rodrigues et al. (2001) Poultry, live or work on farm
Studahl and Andersson (2000)
282
319
1996-1997
9 counties
Denmark
10%
81
144
1996-1997
Australia
NS
212
422
1999-2000
Brisbane, Queensland; children <3 yr old 3 counties
Norway
Excl
126
270
Children <6 yr old
Sweden
Excl
107
178
Nationwide
Denmark
15%
881
833
8 states 2 5 yr old
Australia
3yo
100
137
2001-2002, 12 mo 2000-2001, 11 mo 2001-2002, 12 mo 2002,3 mo
3 areas
Finland
Excl
1,292 C. jejuni
3,409
2002-2003, 12 mo
Nationwide
The Netherlands
16%
121 C. coli
3,409
2002-2003, 12 mo
Nationwide
The Netherlands
29%
318,958
1991-2001
Nationwide registry
Denmark
Excl
22,066
“NR, not reported. *NS, reported as not significantly associated with illness. Excl, travel-associated cases excluded from study.
Organ meats, undercooked poultry, red meat from barbecue, grapes, raw milk, pork Mayonnaise
Kittens, cows, pigs
Neimann et al. (2003)
Puppies, chickens as pets
Tenkate and Stafford (2001)
Untreated water, barbecues, fresh poultry, undercooked pork Untreated water, grilled meat, raw milk Chicken bought fresh; other poultry borderline Chicken and undercooked chicken, organ meats Undercooked meat or poultry, untreated dugwell water Chicken, undercooked meat or seafood, eating at a restaurant, antacid drugs, exposure to raw meat Game, tripe or organ meats, undercooked meat, food from street stall, antacid drugs Possible drinking water association
Occupational exposure to animals
Kapperud et al. (2003)
Dogs
Carrique-Mas et al. (2005) Wingstrand et al. (2006)
Chickens, puppies
Stafford et al. (2007)
Swimming in natural waters
Schonberg-Norio et al. (2004)
Cats or dogs, ill person outside of household
Doorduyn et al. (2005)
Swimming
Doorduyn et al. (2005)
Rural housing type
Ethelberg et al. (2005)
180
OLSON ET AL.
because of dioxin contamination (Vellinga and Van Loock, 2002). Campylobacter infections decreased by 30%, in parallel with the decrease in sales of local poultry. In 2003, when avian influenza appeared in The Netherlands, poultry production decreased and reported Campylobacter infections also decreased, so that about 14% fewer cases than expected were reported in the next 9 months (van Pelt et al., 2004). Much has been learned about the ecology and transmission of Campylobacter in flocks (Newel1 and Fearnley, 2003). The organisms can rapidly spread horizontally in chickens, presumably through a common water source or fecal contact (Shane, 1991). The observation that source and potability of the drinking water of poultry is associated with flock-to-flock variation in infection in Norway, the United Kingdom, and Iceland suggests that environmental interventions at the flock level may be a useful part of prevention (Guerin et al., 2007; Kapperud et al., 1993; Pearson et al., 1993). Recent trials used insect screens on chicken houses and found that it reduced flock prevalence by up to 70%, suggesting that excluding flies may be another simple and promising control strategy (Hald et al., 2007).
Seattle, FoodNet, Swedish, Norwegian, and New Zealand studies (Carrique-Mas et al., 2005; Eberhart-Phillips et al., 1997; Friedman et al., 2004; Harris et al., 1986a; Kapperud et al., 2003), or in areas where raw milk consumption is common, such as Dubuque, Iowa (Schmid et al., 1987). Other foods have also been implicated, including consumption of sausages at a barbecue, or undercooked pork in Norway (Kapperud et al., 1992, 2003). Consumption of barbecued red meat in Denmark, pork or meat from a barbecue in Sweden, game in The Netherlands, other meats in restaurants in the FoodNet study, or organ meats in Australia may represent direct transmission from those animal reservoirs, as well as possible cross-contamination in the kitchen from poultry (Doorduyn et al., 2005; Friedman et al., 2004; Neimann et al., 2003; Stafford et al., 2007; Studahl and Anderson, 2000). A survey of fresh mushrooms obtained from retail stores found 1.5% of the 200 samples to be contaminated with C. jejuni (Doyle and Schoeni, 1986).
Sporadic infections and cross-contamination
The contribution of drinking water to the burden of sporadic cases may be substantial in many parts of the developed and developing world, and small and unprotected rural water supplies may be a particular problem. Even pristine mountain streams can be a source of infection for unwary backpackers, presumably as a result of contamination by the feces of wild birds (Taylor et al., 1983). In Norway, 53% of cases had exposure to untreated drinking water, either from private wells or drinking river or stream water while hiking (Kapperud et al., 2003). In Denmark, drinking water from a private well that had a bad taste or smell was associated with infection, and in a large registry-based study, the risk varied by the residential address of patients in a way suggestive of, although not proving, an effect of the water supply (Ethelberg et al., 2005; Neimann et al., 2003). In New Zealand, drinking rainwater collected from a roof where birds might roost was associated with infection (Eberhart-Phillips et al., 1997). Routine disinfection of drinking water can prevent these infections at extremely low cost (Mintz et al., 2001). For example, in rural Bolivia, a controlled trial of chlorination of drinking water in the homes reduced overall diarrheal rates in intervention households by 44% and reduced infections with Campylobacter, the most commonly identified pathogen in the control group, by 80% (Quick et al., 1996). Two studies, from Finland and The Netherlands, identified swimming in surface waters as a risk factor, although in Norway, swimming was protective (Doorduyn et al.,
A study in Seattle in 1986 identified an association between infection and not washing the kitchen cutting board with soap, a marker for cross-contamination (Harris et al., 1986b). Quantification studies of cross-contamination in simulated kitchens confirm that it happens easily. In one study, 1 to 3% of the campylobacters present on the surface of poultry parts were transferred to hands or utensils, and from there to other foods with equal or greater efficiency (Luber et al., 2006). In another, 29% of cooking sessions that started with contaminated chicken resulted in documented contamination of other foods or food contact surfaces (Redmond et al., 2004). Campylobacter has been shown to survive on the cut surfaces of inoculated fruit for up to 6 hours at room temperature (Castillo and Escartin, 1994). Contamination may also occur from the outside of a package before it reaches the kitchen, as indicated by the finding that riding in shopping carts with raw meat and poultry is a risk factor for infants (Fullerton et al., 2007). In a survey in the United Kingdom, approximately 3% of retail poultry packages were contaminated on the outside with Campylobacter, as were 24% of packages in New Zealand (Burgess et al., 2005; Wong et al., 2004). Sporadic infections and other foods Raw milk has been implicated in the largest studies, where there is sufficient power, such as the
Sporadic infections and water
CHAPTER 9
C. TETUNI INFECTIONS IN INDUSTRIALIZED NATIONS
2005; Kapperud et al., 2003; Schonberg-Norio et al., 2004). If infection spreads in poultry through drinking water, and if cattle are infected in the spring by drinking contaminated surface water, then the waterborne route may be a common underlying pathway linking waterborne, milkborne and poultry-associated campylobacteriosis in humans to an underlying annual cycle involving water and environmental contamination (Fig. 8).
nation routes to humans have been difficult to document. In the western United Kingdom, three studies have associated illness with drinking pasteurized milk from bottles that were pecked by jackdaws (Lighton et al., 1991; hordan et al., 1993; Southern et al., 1990). The birds learned to pierce the foil tops of bottles delivered to the front door and may have contaminated the milk directly. Secondary transmission
Sporadic infections and contact with animals Several studies identified contact with farm animals, especially poultry but also cattle and pigs, to be a risk factor for Campylobacter infection (Engleberg et al., 1984; Friedman et al., 2004; Neimann et al., 2003; Potter et al., 2003; Schonberg-Norio et al., 2004; Stafford et al., 2007). In rural Michigan, risk increased with the frequency of contact with chickens (Potter et al., 2003). Contact with pets has been demonstrated repeatedly as a risk factor. In Norway, fecal culture yielded Campylobacter jejuni in 3% of cats and 3% of dogs, and the isolation rate was higher in puppies than in adult dogs (Sandberg et al., 2002). The nature of the contact with an infected dog or cat that actually transmits the illness remains to be defined, as does the reason that pets become infected themselves. Anecdotally, contact can be quite casual, the animal need not be sick, and there need be no direct contact with animal feces. In addition to established reservoirs in poultry and other domesticated animals, wild birds often carry Campylobacter, particularly crows, gulls, and shoreline foraging birds (It0 et al., 1988; Kapperud and Rosef, 1983; Waldenstroem et al., 2002). Wild birds may contaminate streams and lakes, from which people may drink, as well as the water collected from rooftops in New Zealand, although direct contami-
Despite the low infectious dose and the abundance of organisms in feces of ill persons, secondary transmission from one human to another is unusual, and sustained transmission among humans has not been documented. In the large U.S. case-control study, preceding diarrheal illness in a household member was not significantly associated with illness (Friedman et al., 2004). In general, outbreaks in settings such as day care centers or mental institutions are rare or unheard of, although person-to-person transmission of Shigella, E. coli 0157, Giardia, hepatitis A virus, and other low-dose pathogens has been well documented in these settings. In Australia, attending a child care center and contact with diapered children were protective (Tenkate and Stafford, 2001). Infection rates in homosexual men have been reported to be high but do not appear to be higher than infection rates in heterosexual men of the same age, and transmission through sexual contact has not been conclusively demonstrated (Quinn et al., 1984). Contamination of food by an infected food handler has apparently occurred but is surprising by its rarity (Olsen et al., 2001). It would appear that the infectiousness of the organism for humans diminishes on passage through the human gut, in contrast to passage through bird or cow.
Poultry meat
(also turkeys)
Chicken Chicken
Human Human
Raw milk
v=Wild birds
d
181
3
Pets
Figure 8. Schema of transmission pathways for Campylobacterjejuni, 2007.
182
OLSON ET AL.
Sporadic infections and foreign travel Many countries identify travel-associated cases as a routine part of surveillance. Antecedent foreign travel has been reported in 10% of cases in Britain, 50% of cases in Norway, and 65% of cases in Sweden (Cowden, 1992; Kapperud and Aasen, 1992; Norkrans and Svedhem, 1982). In Swiss and Danish studies, ill travelers had more frequently been to the developing world and Southern Europe (Neimann et al., 2003; Schorr et al., 1994). In a study of Swedish travelers, the risk of travel-associated campylobacteriosis varied dramatically by travel destination, from 3 of 100,000 travelers to other Nordic destinations to 1.2% of travelers to the Indian subcontinent (Ekdahl and Andersson, 2004). Identifying travel as a risk factor does not identify the actual sources of infection. Travel in other countries is a complex exposure, including consumption of foods that may be unusual and water of unknown quality. A casecontrol study comparing the specific exposures of those who are ill with other travelers to similar destinations who remain well remains to be done. Infections with Campylobacter coli Most studies did not attempt to distinguish Campylobacter coli from C. jejuni and did not identify risk factors specific to the latter. However, in the United Kingdom, a case-case comparison suggested that there may be epidemiologic differences between the two closely related species, and in particular that drinking bottled water and eating piit6 were more common among those with C. coli infection than among those with C. jejuni infection (Gillespie et al., 2002). In The Netherlands, a case-control study of C. coli infections found that 29% were related to travel, nearly twice the frequency as C. jejuni, and linked C. coli infections specifically to eating game, organ meats, tripe, undercooked meat, or food from street stalls (Doorduyn et al., 2005). Little is known about the sources of sporadic infections with other Campylobacter species. Protective factors Many case-control studies have identified exposures associated with reduced risk of Campylobacter infection (Adak et al., 1995; Carrique-Mas et al., 2005; Doorduyn et al., 2005; Eberhart-Phillips et al., 1997; Effler et al., 2001; Friedman et al., 2004; Ikram et al., 1994; Kapperud et al., 2003; Neimann et al., 2003; Rodrigues et al., 2001; Stafford et al., 2007; Tenkate and Stafford, 2001; Wingstrand et al., 2006). Apparently protective factors may have been
identified in other studies as well, but not included in the publication. Factors associated with reduced risk include good kitchen hygiene and breast-feeding that are likely to reflect a true lower risk of infection. Other factors, such as consumption of certain food items (particularly produce, but also poultry, red meat, dairy products, nuts, chocolate, and fish), contact with animals, and swimming are more difficult to interpret. Some recurrent risky exposures may induce protective immunity (Adak et al., 1995; Carrique-Mas et al., 2005; Friedman et al., 2004; Neimann et al., 2003). This might explain why handling raw chicken and being in a profession with frequent animal contact were protective factors in a study in the United Kingdom (Adak et al., 1995). Repeated exposure to one source, such as poultry, might induce enough protective immunity to introduce a misclassification bias, because the exposure of some controls will not be likely to result in illness, which could lower the odds ratio even to below 1 (Swift and Hunter, 2004). Protective immunity does not provide an obvious explanation of why in the same case-control study, chicken appears to be protective if prepared at home and a risk factor if prepared in a restaurant, as demonstrated in several large studies (Adak et al., 1995; Eberhart-Phillips et al., 1997; Effler et al., 2001; Friedman et al., 2004; Ikram et al., 1994). Differential recall bias might play a role, if exposure in one location is more memorable than a similar exposure in another location. It is also possible that there are real protective effects associated with some foods, particularly produce. For example, in a recent large study in Australia, risk decreased as the number of types of uncooked vegetables consumed increased, independent of chicken consumption (Stafford et al., 2007). In New Zealand, the protective effect of salad increased with the frequency of consumption (Eberhart-Phillips et al., 1997). Some have speculated that high levels of plant antioxidants and carotenoids could boost general immunity to infection or inhibit bacterial growth (Kapperud et al., 2003; Neimann et al., 2003). Specific diets may change the intestinal microflora and alter the host susceptibility to infection (Cameron et al., 2004; Rodrigues et al., 2001) or reduce the severity of diarrhea, as has been shown for vitamin A in young children in developing countries (Barreto et al., 1994). Surveillance for Emerging Antimicrobial Resistance Although most infections resolve without specific treatment, antimicrobial treatment can be critical in invasive or severe infections. Fluoroquinolone agents like ciprofloxacin are commonly used for the
CHAPTER 9
C. TETUNI INFECTIONS IN INDUSTRIALIZED NATIONS
treatment of infections caused by Campylobacter. Antimicrobial resistance has been monitored around the world in several ways, and the frequency of resistance to various agents varies widely by location. In Australia, fluoroquinolone resistance in domestically acquired Campylobacter infections in humans has been largely absent, reported to be 2% of strains in 2002; it has been reported at substantially higher levels from Europe and Asia since the end of the 1980s (Gaunt and Piddock, 1996; Hoge et al., 1998; Piddock, 1995; Unicomb et al., 2003, 2006). In 19891990 in the United States, no C. jejuni or coli strains collected from 19 sentinel counties were resistant to fluoroquinolones (CDC, unpublished data). By 1997, when the National Antimicrobial Resistance Monitoring System for enteric bacteria first began monitoring Campylobacter in humans, 86% of tested strains were resistant to at least one antimicrobial, 24% were resistant to nalidixic acid, and 13% were resistant to ciprofloxacin. The frequency of resistance among C. jejuni strains increased to a peak of 20.7% in 2002 (CDC, 2007) (Fig. 9). In 1997, a review of the fluoroquinolone-resistant cases indicated that foreign travel was an important risk factor, but that most of the fluoroquinolone-resistant infections were acquired in the United States (Friedman et al., 1998). In a case-control study of fluoroquinolone-resistant infections conducted in 1998-1999, nested inside the larger study of all C. jejuni and C. coli infections, 42% of persons with fluoroquinolone-resistant infections reported preceding foreign travel, but only 9% of those with susceptible infections did (Kassenborg et al., 2004). Among the non-travel-associated cases, consumption of poultry in restaurants was the only identified risk factor. The resistance is associated with clinical impact: illness with fluoroquinolone-resistant strains lasted longer than did illnesses with suscepti-
10 Tetracycline
Erythromycin
I
Ciprofloxacin
60
89-90
97
98
99
00 01 02 03 04
Figure 9. Antimicrobial resistance in C. jejuni isolates, United States, 1997 to 2004 (CDC, 2007).
183
ble strains in two separate studies, although whether this was due to treatment failure or to a greater virulence of resistant strains or both remains uncertain (Neimann et al., 2001; Nelson et al., 2004). Reports from Europe suggested that the increase in fluoroquinolone-resistant Campylobacter isolations in humans was associated with fluoroquinolone use in poultry that has led to fluoroquinolone-resistant Campylobacter strains in poultry, which have then spread to humans via the food chain (Gaunt and Piddock, 1996; Piddock, 1995). Veterinary use of fluoroquinolones in poultry began elsewhere before 1995, when it was approved in the United States; it has never been permitted in Australia (Unicomb et al., 2003). A 1997 study conducted in Minnesota showed that 20% of 60 C. jejuni isolates obtained from raw chicken bought at grocery stores were ciprofloxacin resistant (Smith et al., 1999). In 2000, the U.S. Food and Drug Administration proposed withdrawing approval for fluoroquinolone use in poultry, and in 2005, this rule was adopted (Nelson et al., 2007). Judicious use of fluoroquinolones and other antimicrobial agents in human and veterinary medicine is essential to preserve the efficacy of these important chemotherapeutic agents.
SUMMARY Campylobacter is the most common bacterial enteric pathogen in many developed countries and is the second most common such pathogen in the United States. Although infection is not usually invasive, it is occasionally fatal. In the last 30 years, Campylobacter rates have risen markedly in many developed countries and have begun to decrease in some. Part of the increase may be due to improvements in detection and reporting, but it also reflects a true increase in infections. The burden of illness of Campylobacter is substantial, and infection may be universal. The vast majority of Campylobacter infections are sporadic individual infections and often affect young adults and young children. Consumption of untreated water, raw milk, or milk products and contact with pets are important sources of infection. However, the most frequently identified source of sporadic disease is consumption of raw or undercooked poultry. Outbreaks of Campylobacter in developed countries are predominantly caused by contaminated water, milk, poultry, and produce. Control of this infection thus depends on a multipronged prevention strategy, including public education to lower the risk of infection and cross-contamination in the kitchen, better slaughter hygiene to reduce the level of carcass contamination, and control measures on
184
OLSON ET AL.
farms to reduce infection of live animals. Routine treatment of drinking water and pasteurization of milk are core prevention strategies. Veterinary use of fluoroquinolones in poultry explains the recent rise in ciprofloxacin-resistant Campylobacter infections in humans in the United States and may explain much of the geographic variation in resistance around the world. Acknowledgments. We gratefully acknowledge the assistance of Tracy Ayers and Cherie Long. The findings and conclusions in this publication are those of the authors and do not necessarily represent the views of the CDC or other institutions.
REFERENCES Adak, G., J. Cowden, S. Nicholas, and H. Evans. 1995. The Public Health Laboratory Service national case-control study of primary indigenous sporadic cases of Campylobacter infection. Epidemiol. Infect. 115:15-22. Angulo, F., and the FoodNet Working Group. 1997. Foodborne diseases active surveillance network (FoodNet). Emerg. Infect. Dis. 3581-583. Angulo, F., and D. Swerdlow. 1995. Bacterial enteric infections in persons infected with human immunodeficiency virus. Clin. Infect. Dis. 21:S84-S93. Anonymous. 2006a. Annual report on zoonoses in Denmark, 2005. http: //www.dfvf.dk/Default.asp?ID=9606. Ministry of Family and Consumer Affairs. Anonymous. 2006b. Burden and causes of foodborne disease in Australia: annual report of the OzFoodNet network, 2005. Comm. Dis. lntell. 30:278-300. Anonymous. 2000. Waterborne outbreak of gastroenteritis associated with a contaminated municipal water supply, Walkerton, Ontario, May-June 2000. Can. Commun. Dis. Rep. 26:170173. Baker, M., E. Sneyd, and N. Wilson. 2007. Is the major increase in notified campylobacteriosis in New Zealand real? Epidemiol. Infect. 135:163-170. Baker, M., N. Wilson, R Ikram, S. Chambers, P. Shoemack, and G.Cook. 2006. Regulation of chicken contamination is urgently needed to control New Zealand’s serious campylobacteriosis epidemic. N. Z. Med. J. 119:U2264. Barreto, M., L. Santos, A. Assis, G . Farenzena, P. Santos, and R Fiaccone. 1994. Effects of vitamin A supplementation on diarrhoea and acute lower-respiratory-tract infections in young children in Brazil. Lancet 344:228-231. Bayliss, C., S. MacPhee, K. Martin, T. Humphrey, and R. Betts. 2000. Comparison of three enrichment media for the isolation of Campylobacter spp. from foods. 1.Appl. Microbiol. 895384891. Bean, N. H., S. M. Martin, and H. J. Bradford. 1992. PHLIS: an electronic system for reporting public health data from remote sites. Am. J. Public Health 82:1273-1276. Berg, D., T. Farley, L. McFarland, and T. Zhao. 1996. Campylobacter gastroenteritis associated with foodhandler contamination, Louisiana, abstr. K142. In 36th Intersci. Confi Antimicrob. Agents Chemother., New Orleans, LA, 15-16 September 1996. Black, A., M. Kirk, and G. Millard. 2006. Campylobacter outbreak due to chicken consumption at an Australian Capital Territory restaurant. Comm. Dis. Intell. 30:373-377. Black, R., M. Levine, M. Clements, T. Hughes, and M. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479.
Blaser, M., H. Hardesty, B. Powers, and W.Wang. 1980. Survival of Campylobacter fetus subsp. jejuni in biological milieus. 1. Clin. Microbiol. 27:309-3 13. Blaser, M., E. Sazie, and L. J. Williams. 1987. The influence of immunity on raw milk-associated Campylobacter infection. JAMA 257:43-46. Borck, B., C. Galliono, S. Nordentoft, S. Ethelberg, L. Boysen, and B. Helwigh. 2007. Effects of the Danish intervention strategies aimed at reducing Carnpylobacter in broiler meat, abstr. 0046.24th Int. Workshop Campylobacter Helicobacter Related Organisms, 2-5 September 2007. Zoonoses Public Health 54(Suppl. 1):14. Bowen, A., A. Fry, G. Richards, and L. Beuchat. 2006. Infections associated with cantaloupe consumption: a public health concern. Epidemiol. Infect. 134:675-685. Brieseman, M.1987. Town water supply as the cause of an outbreak of Campylobacter infection. N. Z. Med. J. 100:212-213. Brieseman, M. A. 1990. A further study of the epidemiology of Campylobacter jejuni infections. N. Z. Med. J. 103:207-209. Burgess, F., C. Little, G. Allen, K. Williamson, and R. Mitchell. 2005. Prevalence of Campylobacter, Salmonella, and Escherichia coli on the external packaging of raw meat. J. Food Prot. 68: 469-475. Cameron, S., K. Ried, A. Worsley, and D. Topping. 2004. Consumption of foods by young children with diagnosed Campylobacter infection-a pilot case-control study. Public Health Nutr. 7:85-89. Carrique-Mas, J., Y. Andersson, M. Hjertqvist, A. Svensson, A. Torner, and J. Giesecke. 2005. Risk factors for domestic sporadic campylobacteriosis among young children in Sweden. Scand. J. Infect. Dis. 37:lOl-110. Castillo, A., and E. Escartin. 1994. Survival of Campylobacter jejuni on sliced watermelon and papaya. J. Food Prot. 57:166168. Centers for Disease Control and Prevention. 1991. Campylobacter enteritis-New Zealand, 1990. Morb. Mortal. Wkly. Rep. 40: 116-123. Centers for Disease Control and Prevention. 1986. Campylobacter outbreak associated with raw milk provided on a dairy tourCalifornia. Morb. Mortal. Wkly. Rep. 35:3 11-3 12. Centers for Disease Control and Prevention. 2007. National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS), final report 2004. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. http: //www.cdc.gov/narms/NARMSAnnualReport2004.pdf. Centers for Disease Control and Prevention. 2002. Outbreak of Campylobacter jejuni infections associated with drinking unpasteurized milk procured through a cow-leasing programWisconsin, 2001. Morb. Mortal. Wkly. Rep. 51548-549. Centers for Disease Control and Prevention. 2006. Preliminary FoodNet data on the incidence of infection with pathogens transmitted commonly through food-10 states, 2005. Morb. Mortal. Wkly. Rep. 55 :392-395. Clark, C., L. Price, R. Ahmed, D. Woodward, P. Melito, F. Rodgers, F. Jamieson, B. Ciebin, A. Li, and A. Ellis. 2003. Characterization of waterborne outbreak-associated Campylobacter jejuni, Walkerton, Ontario. Emerg. Infect. Dis. 9:1232-1241. Coker, A. O.,R D. Isokpehi, B. N. Thomas, K. 0. Amisu, and L. Obi. 2002. Human campylobacteriosis in developing countries. Emerg. Infect. Dis. 8:237-243. Cowden, J. 1992. Campylobacter: epidemiological paradoxes. Br. Med. J. 305:132-133. de Wit, M., M. Koopmans, L. Kortbeek, L. Wannet, J. Vinje, F. van Leusden, A. Bartelds, and Y. van Duynhoven. 2001. SENSOR, a population-based cohort study on gastroenteritis in The
CHAPTER 9
-
C. 7E7UNI INFECTIONS IN INDUSTRIALIZED NATIONS
Netherlands: incidence and etiology. Am. J. Epidemiol. 154: 666-674. Dekeyser, P. J., M. Gossuin-Detrain, J. P. Butzler, and J. Sternon. 1972. Acute enteritis due to related Vibrio: first positive stool cultures. J. Infect. Dis. 125:390-392. Deming, M., R. Tauxe, P. Blake, and C. Patton. 1987. Campylobacter enteritis at a university: transmission from eating chicken and from cats. Am. J. Epidemiol. 126526-534. Doorduyn, Y., B. Breukink, W. van Brandhof, Y. van Duynhoven, J. Wagenaar, and W. van Pelt. 2005. Risk factors for endemic Campylobacter jejuni and Campylobacter coli infections in The Netherlands, abstr. Epi2. MedVetNet Gen. Sci. Mtng., Winchester, United Kingdom, 29 June-1 July 2005. Doyle, M., and J. Schoeni. 1986. Isolation of Campylobacter jejuni from retail mushrooms. Appl. Environ. Microbiol. 51:449-450. Dziuban, E., J. L. Liang, G . Craun, V. Hill, P. Yu, J. Painter, M. Moore, R. Calderon, S. Roy, and M. Beach. 2006. Surveillance for waterborne disease and outbreaks associated with recreational water-United States, 2003-4. Morb. Mortal. Wkly. Rep. Surueill. Summ. 55(SS-12):32-65. Eberhart-Phillips, J., N. Walker, N. Garrett, D. Bell, D. Sindair, W. Rainger, and M. Bates. 1997. Campylobacteriosis in New Zealand: results of a case-control study. J. Epidemiol. Community Health 51:686-691. Effler, P., M . 4 . Ieong, A. Kimura, M. Nakata, R. Burr, E. Cremer, and L. Slutsker. 2001. Sporadic Campylobacter jejuni infections in Hawaii: associations with prior antibiotic use and commercially prepared chicken. J. Infect. Dis. 183:1152-1155. EFSA. 2006. The community summary report on trends and sources of zoonoses, zoonotic agents, antimicrobial resistance and foodborne outbreaks in the European Union in 2005. European Food Safety Agency. EFSA J. 2006:94. Ekdahl, K., and Y. Andersson. 2004. Regional risks and seasonality in travel-associated campylobacteriosis. BMC Infect. Dis. 4 5 4 . Ekdahl, K., B. Normann, and Y. Andersson. 2005. Could flies explain the elusive epidemiology of campylobacteriosis? BMC Infect. Dis. 5:11. Engberg, J., P. Gerner-Smidt, F. Scheutz, N. Moller, S. On, and K. Mdbak. 1998. Water-borne Campylobacter jejuni infection in a Danish town-a 6-week continuous source outbreak. Clin. Microbiol. Infect. 4:648-656. Engleberg, N., A. Correa-Villasenor, C. North, T. Crow, J. Wells, and P. Blake. 1984. Campylobacter enteritis on Hopi and Navajo Indian reservations: clinical and epidemiologic features. West. J. Med. 141:53-56. Escherich, T. 1881. Beitrage zur Kenntniss der Darmbacterien. 111. Ueber das vorkommen von Vibrionen in Darmcanal und den Stuhlgaengen der Saeuglinge. Muench. Med. Wochenschr. 33: 815-817. Ethelberg, S., K. Olsen, P. Gerner-Smidt, and K. Melbak. 2004. Household outbreaks among culture-confirmed cases of bacterial gastrointestinal disease. Am. J. Epidemiol. 159:406-412. Ethelberg, S., J. Simonsen, P. Gerner-Smidt, K. Olsen, and K. Melbak. 2005. Spatial distribution and registry-based casecontrol analysis of Campylobacter infections in Denmark, 19912001. Am. J. Epidemiol. 162:1008-1015. Flint, J., Y. van Duynhoven, F. Angulo, S. DeLong, P. Braun, M. Kirk, E. Scallan, M. Fitzgerald, G. Adak, P. Sockett, A. Ellis, G. Hall, N. Gargouri, H. Walke, and P. Braam. 2005. Estimating the burden of acute gastroenteritis, foodborne disease, and pathogens commonly transmitted by food: an international review. Clin. Infect. Dis. 41:698-704. Franco, D. A. 1988. Campylobacter species: considerations for controlling a foodborne pathogen. Food Prot. 51:145-153. Friedman, C., R. Hoekstra, M. Samuel, R. Marcus, J. Bender, B. Shiferaw, V. Reddy, S. Ahuja, D. Helfrick, F. Hardnett, M. Car-
185
ter, B. Anderson, and R. Tauxe. 2004. Risk factors for sporadic Campylobacter infection in the United States: a case-control study in FoodNet sites. Clin. Infect. Dis. 38(Suppl. 3):S285S296. Friedman, C., S. Yang, J. Rocourt, K. Stamey, D. Vugia, R. Marcus, s. Segler, B. Shiferaw, and F. Angulo. 1998. Fluoroquinolone-resistant Campylobacter infections in the United States: a pilot case-control study in FoodNet sites, abstr. 545. 36th Annu. Mtng. Infect. Dis. SOC.Am., Denver, CO, November 1998. Frost, J., I. Gillespie, and S. O’Brien. 2002. Public health implications of Campylobacter outbreaks in England and Wales, 1995-9: epidemiological and microbiological investigations.Epidemiol. Infect. 128: 111-1 18. Fullerton, K., L. Ingram, T. Jones, B. Anderson, P. McCarthy, S. Hurd, B. Shiferaw, D. Vugia, N. Haubert, T. Hayes, S. Wedel, E. Scallan, 0. Henao, and F. Angulo. 2007. Sporadic Campylobacter infections in infants: a population-based surveillance case-control study. Pediatr. Infect. Dis. J. 26:19-24. Gauci, A., and A. Ammon. 2007. The First European Communicable Disease Epidemiological Report. European Centre of Disease Prevention and Control, Stockholm. Gaunt, P., and L. Piddock. 1996. Ciprofloxacin resistant Campylobacter spp. in humans: an epidemiological and laboratory study. J. Antimicrob. Chemother. 37:747-757. Georgsson, F., A. Porkelsson, M. Geirsdottir, J. Reiersen, and N. Stern. 2006. The influence of freezing and duration of storage on Campylobacter and indicator bacteria in broiler carcasses. Food Microbiol. 23 :677-68 3. Gillespie, I., S. O’Brien, G. Adak, C. Tam, J. Frost, F. Bolton, D. Tompkins, and the Campylobacter Sentinel Surveillance Scheme Collaborators. 2003. Point source outbreaks of Carnpylobacter jejuni infection-are they more common than we think and what might cause them? Epidemiol. Infect. 130:367-375. Gillespie, I., S. O’Brien, J. Frost, G. Adak, P. Horby, A. Swam, M. Painter, K. Neal, and the Campylobacter Sentinel Surveillance Scheme Collaborators. 2002. A case-case comparison of Campylobacter coli and Campylobacter jejuni infection: a tool for generating hypotheses. Emerg. Infect. Dis. 9:937-942. Guerin, M. T., W. Martin, J. Reiersen, 0. Berke, S. A. McEwen, J. R. Bisaillon, and R. Lowman. 2007. A farm-level study of risk factors associated with the colonization of broiler flocks with Campylobacter spp. in Iceland, 2001-2004. Acta Vet. Scand. 49: 18. Hald, B., H. Skovgird, D. Bang, K. Pedersen, J. Dybdahl, J. Jespersen, and M. Madsen. 2004. Flies and Campylobacter infection of broiler flocks. Emerg. Infect. Dis. 10:1490-1492. Hald, B., H. Sommer, and H. Skovgraad. 2007. Reductive effect of fly screens on the Campylobacter prevalence of broiler flocks in summer, abstr. P488. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Rotterdam, The Netherlands, 1 to 5 September 2007. Zoonoses Public Health 54(Suppl. 1):154155. Harris, N., T. Kimball, P. Bennett, Y. Johnson, D. Wakely, and C. Nolan. 1987. Campylobacter jejuni gastroenteritis associated with raw goat’s milk. Am. J. Epidemiol. 126:179-186. Harris, N., T. Kimball, N. Weiss, and C. Nolan. 1986a. Dairy products, produce and other non-meat food as possible sources of Campylobacter jejuni and Campylobacter coli enteritis. J. Food Prot. 49:347-351. Harris, N., N. Weiss, and C. Nolan. 1986b. The role of poultry and meats in the etiology of Campylobacter jejunilcoli enteritis. Am. J. Public Health 76:407-411. Headrick, M., S. Korangy, N. Bean, F. Angulo, S. Altekruse, M. Potter, and K. Klontz. 1998. The epidemiology of raw milkassociated foodborne disease outbreaks reported in the United States, 1973 through 1992. Am. J. Public Health 88:1219-1221.
186
OLSON ET AL.
Helms, M., J. Simonsen, and K. Melbak. 2006. Foodborne bacterial infection and hospitalization: a registry-based study. Clin. Infect. Dis. 42:498-506. Helms, M., P. Vastrup, P. Gerner-Smidt, and K. Melbak. 2003. Short and long term mortality associated with foodborne bacterial gastrointestinal infections: registry based study. Br. Med. J. 326:357. Ho, B., and W. Wong. 1985. A one-year survey of Campylobacter enteritis and other forms of bacterial diarrhea in Hong Kong. J. Hyg. 9 4 5 - 6 0 . Hofshagen, M., and H. Kruse. 2005. Reduction in flock prevalence of Campylobacter spp. in broilers in Norway after implementation of an action plan. J. Food Prot. 68:2220-2223. Hoge, C., J. Gambel, A. Srijan, C. Pitarangsi, and P. Echeverria. 1998. Trends in antibiotic resistance among diarrheal pathogens isolated in Thailand over 15 years. Clin. Infect. Dis. 26:341345. Hood, A. M., A. D. Pearson, and M. Shahamat. 1988. The extent of surface contamination of retailed chickens with Campylobacter jejuni serogroups. Epidemiol. Infect. 100:17-25. Hopkins, R. S., R. Olmsted, and G. R. h e . 1984. Campylobacter jejuni infection in Colorado: identified risk factors. Am. J. Public Health 74:249-250. Hopkins, R. S., and A. B. Scott. 1983. Handling raw chickens as a source for sporadic Campylobacter jejuni infections. J. Infect. Dis. 148:770. (Letter.) Hudson, P. J., R. L. Vogt, J. Brondum, and C. M. Patton. 1984. Isolation of Campylobacter jejuni from milk during an outbreak of campylobacteriosis.J. Infect. Dis. 150:789. Humphrey, T., and R. Hart. 1988. Campylobacter and Salmonella contamination of unpasteurized cows’ milk on sale to the public. 1.Appl. Bacteriol. 65:463-467. Hutchinson, D. N., F. J. Bolton, P. M. Hinchliffe, H. D. Dawkins, S. D. Horsley, E. G. Jessop, P. A. Robertshaw, and D. E. Counter. 1985. Evidence of udder excretion of Campylobacter jejuni as the cause o f milk-borne Campylobacter outbreak. J. Hyg. (Lond.)94:205-215. Ikram, R., S. Chambers, P. Mitchell, M. Brieseman, and 0. Ikram. 1994. A case control study to determine risk factors for Campylobacter infection in Christchurch in the summer of 1992-3. N.Z. Med. J. 107:430-432. Ito, K., Y. Kubokura, K.-I. Kaneko, Y. Totake, and M. Ogawa. 1988. Occurrence o f Campylobacter jejuni in free-living wild birds from Japan. J. Wildl. Dis. 24:467-470. Jacobs-Reitsma, W., N. Bolder, and R. Mulder. 1994. Cecal carriage of Campylobacter and Salmonella in Dutch broiler flocks at slaughter; a one year study. Poultry Sci. 73:1260-1266. Jimenez, M., P. Soler, J. Venanzi, P. Cante, C. Varela, and N. Martinez. 2005. An outbreak of Campylobacter jejuni enteritis in a school of Madrid, Spain. Eurosurveillance 10:118-121. Jones, D., D. Robinson, and J. Eldridge. 1981. Serological studies in two outbreaks of Campylobacter jejuni infection. J. Hyg. (Lond.)87:163-170. Kapperud, G., and S. Aasen. 1992. Descriptive epidemiology of infections due to thermotolerant Campylobacter spp. in Norway, 1979-1988. APMIS 100:883-890. Kapperud, G., G. Espeland, E. Wahl, A. Walde, H. Herikstad, S. Gustavsen, I. Tveit, 0. Natas, L. Bevanger, and A. Digranes. 2003. Factors associated with increased and decreased risk of Campylobacter infection: a prospective case-control study in Norway. Am. J. Epidemiol. 158:234-242. Kapperud, G., 0. L0m0, K. Styrmo, S. Gregusson, K. Melby, and T. Vardund. 2000. [Two outbreaks of Carnpylobacterinfection after bicycle races-dirty water splash from the wheels a source of infection?]. MSIS-Rapport 28:8-9.
Kapperud, G., and 0. Rosef. 1983. Avian wildlife reservoir of Carnpylobacterfetus subsp. jejuni, Yersinia spp., and Salmonella spp. in Norway. Appl. Environ. Microbiol. 45:375-380. Kapperud, G., E. Skjerve, N. Bean, S. Ostroff, and J. Lassen. 1992. Risk factors for sporadic Campylobacter infections: results of a case-control study in southeastern Norway. J. Clin. Microbiol. 30:3117-3121. Kapperud, G., E. Skjerve, L. Vik, K. Hauge, A. Lysaker, I. Aalmen, S. Ostroff, and M. Potter. 1993. Epidemiological investigations of risk factors for Campylobacter colonization in Norwegian broiler flocks. Epidemiol. Infect. 111:245-255. Kassenborg, H., K. Smith, D. Vugia, T. Rabatsky-Ehr, M. Bates, M. Carter, N. Dumas, M. Cassidy, N. Marano, R. Tauxe, and F. Angulo. 2004. Fluoroquinolone-resistant Campylobacter infections: eating poultry outside of the home and foreign travel are risk factors. Clin. Infect. Dis. 38:S279-S284. King, E. 0. 1962. The laboratory recognition of Vibrio fetus and closely related Vibrio species isolated from cases of human vibriosis. Ann. N. Y. Acad. Sci. 98:700-711. Kirk, M., R. Waddell, C. Dalton, A. Creaser, and R. Rose. 1997. A prolonged outbreak of Campylobacter infection at a training facility. Commun. Dis. Intell. 2157-61. Kist, M. 1982. Campylobacter enteritis: epidemiological and clinical data from recent isolations in the region of Freiburg, West Germany, p. 138-143. In D. G. Newell (ed.), Campylobacter: Epidemiology, Pathogenesis, and Biochemistry. MTP Press, Lancaster, United Kingdom. Klein, B., J. Vergeront, M. Blaser, P. Edmonds, D. Brenner, D. Janssen, and J. Davis. 1986. Campylobacter infection associated with raw milk: an outbreak of gastroenteritis due to Campylobacter jejuni and thermotolerant Campylobacter fetus subsp fetus.]AMA 255~361-364. Kovats, R., S. Edwards, D. Charron, J. Cowden, R. D’Souza, K. Ebi, C. Gauci, P. Gerner-Smidt, S. Hajat, S. Hales, G. Hernindez Pezzi, B. Kriz, K. Kutsar, P. McKeown, K. Mellou, B. Menne, S. O’Brien, W. van Pelt, and H. Schmid. 2005. Climate variability and campylobacter infection: an international study. Int. 1. Biometeorol. 49:207-214. Kuusi, M., J. P. Nuorti, M. L. Hanninen, M. Koskela, V. Jussila, E. Kela, I. Miettinen, and P. Ruutu. 2005. A large outbreak of campylobacteriosis associated with a municipal water supply in Finland. Epidemiol. Infect. 133593-601. Lewis, H., A. Mazick, K. Molbak, S. Ethelberg, and M. Lisby. 2007. Campylobacter outbreaks. Epi-News 2007, week 20121. Liang, J. L., E. Dziuban, G. Craun, V. Hill, M. Moore, R. Gelting, R. Calderon, M. Beach, and S. Roy. 2006. Surveillance for waterborne-disease outbreaks associated with drinking water and water not intended for drinking-United States, 20032004. Morb. Mortal. Wkly. Rep. Surveill. Summ. 55(SS-12):131. Lighton, L., E. Kanmarski, and D. Jones. 1991. A study of risk factors for Campylobacter infection in spring. Public Health 105:199-203. Louis, V., I. Gillespie, S. O’Brien, E. Russek-Cohen, A. Pearson, and R Colwell. 2005. Temperature-driven Campylobacter seasonality in England and Wales. Appl. Environ. Microbiol. 71:8592. Luber, P., and E. Bartelt. 2007. Enumeration of Campylobacter spp. on the surface and within chicken breast fillets. J. Appl. Microbiol. 102:3 13-3 18. Luber, P., S. Brynestad, D. Topsch, K. Sherer, and E. Bartelt. 2006. Quantification of Campylobacter species cross-contamination during handling of contaminated fresh chicken parts in kitchens. Appl. Environ. Microbiol. 72:66-70. Lynch, M., J. Painter, R. Woodruff, and C. Braden. 2006. Surveillance for foodborne disease outbreaks-United States, 19982002. Morb. Mortal. Wkly. Rep. 55:l-42.
CHAPTER 9
C. rE7UNI INFECTIONS IN INDUSTRIALIZED NATIONS
Majowicz, S., G. Hall, E. Scallan, G. Adak, C. Gauci, T. Jones, S. O’Brien, 0. Henao, P. Sockett, and the International Collaboration on “Burden of Illness” Studies. Standardized case definition for acute gastroenteritis. Epidemiol. Infect. [Online.] doi: 10.1017/S0950268807009375. Maurer, A. M., and D. Sturchler. 2000. A waterborne outbreak of small round structured virus, campylobacter and shigella coinfections in La Neuveville, Switzerland, 1998. Epidemiol. Infect. 125:325-332. Mazick, A., S. Ethelberg, E. M. Nielsen, K. Mslbak, and M. Lisby. 2006. An outbreak of Campylobacter jejuni associated with consumption of chicken, Copenhagen, 2005. Eurosurveillance 11: 137-139. MCgraud, F. 1999. Les infections b Campylobacter en France: bilan du Centre National de Rdfdrence. Bulletin Epidhiologique Annual 1997. Centre National de Reference, Bordeaux, France. Melby, K., 0. P. Dahl, L. Crisp, and J. L. Penner. 1990. Clinical and serological manifestations in patients during a waterborne epidemic due to Campylobacter jejuni. J. Infect. Dis. 21:309316. Melby, K., B. Gondrosen, S. Gregusson, H. Ribe, and 0. P. Dahl. 1991. Waterborne campylobacteriosis in northern Norway. Int. J. Food Microbiol. 12:151-156. Melby, K., J. Svendby, T. Eggebo, L. Holmen, B. Andersen, L. Lind, E. Sjogren, and B. Kaijser. 2000. Outbreak of Campylobacter infection in a subartic community. Eur. J. Clin. Microbiol. Infect. Dis. 19542-544. Meldrum, R., J. Griffiths, R. Smith, and M. Evans. 2005. The seasonality of human campylobacter infection and Campylobacter isolates from fresh retail chicken in Wales. Epidemiol. Infect. 133:49-52. Mentzing, L. 0. 198 1. Waterborne outbreaks of campylobacter enteritis in central Sweden. Lancet ii:352-354. Miller, G., G. Dunn, T. Reid, I. Ogden, and N. Strachan. 2005. Does age acquired immunity confer selective protection to common serotypes of Campybacter jejuni? BMC Infect. Dis. 6:l-5. Miller, G., G. Dunn, A. Smith-Palmer, A. Ogden, and N. Strachan. 2004. Human campylobacteriosis in Scotland: seasonality regional trends and bursts of infection. Epidemiol. Infect. 132: 585-593. Mintz, E., J. Bartram, P. Lochery, and M. Wegelin. 2001. Not just a drop in the bucket: expanding access to point-of-use water treatment systems. Am. J. Public Health 91:1565-1570. Nauta, M., W. Jacobs-Reitsma, E. Evers, W. van Pelt, and A. Havelaar. 2005. Risk assessment of Campylobacter in The Netherlands via chicken and other routes. RIVM Report no. 25091100612005. National Institute for Public Health and the Environment, Bilthoven, The Netherlands. Neal, K., and R. Slack. 1995. The autumn peak in campylobacter gastro-enteritis. Are the risk factors the same for travel- and U.K.-acquired campylobacter infections? J. Public Health Med. 17:98-102. Neal, K., and R. Slack. 1997. Diabetes mellitus, anti-secretory drugs and other risk factors for Campylobacter gastroenteritis in adults: a case-control study. Epidemiol. Infect. 119:307-311. Neimann, J., J. Engberg, K. Mslbak, and H. Wegener. 2003. A case-control study of risk factors for sporadic Campylobacter infections in Denmark. Epidemiol. Infect. 130:353-366. Neimann, J., K. Mslbak, J. Engberg, F. Aarestrup, and H. Wegener. 2001. Longer duration of illness among Campylobacter patients treated with fluoroquinolones, abstr. L-18. 11th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Freiburg, Germany. Int. J. Med. Microbiol. 291(Suppl. 31):108. Nelson, J., T. Chiller, J. Powers, and F. Angnlo. 2007. Fluoroquinolone-resistant Campylobacter species and the withdrawal
187
of fluoroquinolones from use in poultry; a public health success story. Clin. Infect. Dis. 44:977-980. Nelson, J., K. Smith, D. Vugia, T. Rabatsk-Ehr, S. Segler, H. Kassenborg, S. Zansky, K. Joyce, N. Marano, R. Hoekstra, and F. Angnlo. 2004. Prolonged diarrhea due to ciprofloxacin-resistant Campylobacter infection. J. Infect. Dis. 190:1150-1157. Nelson, M., D. Shanson, D. Hawkins, and B. Gazzard. 1992. Salmonella, Campylobacter and Shigella in HIV-seropositive patients. AIDS 6:1495-1498. Newell, D., and C . Fearnley. 2003. Sources of Campylobacter colonization in broiler chickens. Appl. Environ. Microbiol. 69: 4343-4351. Nichols, G. 2005. Fly transmission of Campylobacter. Emerg. Infect. Dis. 11:361-364. Norkrans, G., and A. Svedhem. 1982. Epidemiological aspects of Campylobacter jejuni enteritis. J. Hyg. (Lond.) 89:163-170. Nylen, G., F. Dunstan, S. Palmer, Y. Andersson, F. Bager, J. Cowden, G. Feierl, Y. Galloway, G. Kapperud, F. Megraud, K. Mslbak, L. Petersen, and P. Ruutu. 2002. The seasonal distribution of campylobacter infection in nine European countries and New Zealand. Epidemiol. Infect. 128:383-390. Oberhelman, R. A., and D. N. Taylor. 2000. Campylobacter infections in developing countries, p. 139-153. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. American Society for Microbiology, Washington, DC. Olsen, S., G. Hansen, L. Bartlett, C . Fitzgerald, A. Sonder, R. Manjrekar, T. Riggs, J. Kim, R. Flahart, G. Pezzino, and D. Swerdlow. 2001. An outbreak of Campylobacter jejuni infections associated with food handler contamination: the use of pulsed-field gel electrophoresis. J. Infect. Dis. 183:164-167. Oosterom, J., J. R. J. Banffer, S. Lauwers, and A. E. Busschbach. 1985. Serotyping of and hippurate hydrolysis by Campylobacter jejuni isolates from human patients, poultry and pigs in The Netherlands. Antonie Van Leeuwenhoek 51:65-70. Oosterom, J., C. H. Den Uyl, J. R. J. Banffer, and J. Huisman. 1984. Epidemiological investigations on Campylobacter jejuni in households with a primary infection. J. Hyg. (Lond.) 92:325332. O’Reilly, C., A. Bowen, N. Perez, J. Sarisky, C. Shepherd, M. Miller, B. Hubbard, M. Herring, s. Buchanan, C. Fitzgerald, V. Hill, M. Arrowood, L. Xiao, R. Hoekstra, E. Mintz, M. Lynch, and the Outbreak Working Group. 2007. A waterborne outbreak of gastroenteritis with multiple etiologies among resort island visitors and residents: Ohio, 2004. Clin. Infect. Dis. 44: 506-512. Patrick, M., L. Christiansen, M. Waino, S. Ethelberg, H. Madsen, and H. Wegener. 2004. Effects of climate on incidence of Campylobacter spp. in humans and prevalence in broiler flocks in Denmark. Appl. Environ. Microbiol. 70:7474-7480. Pearson, A., M. Greenwood, T. Healing, D. Rollins, M. Shahamat, J. Donaldson, and R. Colwell. 1993. Colonization of broiler chickens by waterborne Campylobacter jejuni. Appl. Environ. Microbiol. 59:9 8 7-996. Pebody, R., M. Ryan, and P. Wall. 1997. Outbreaks of Campylobacter infection: rare events for a common pathogen. Commun. Dis. Rep. CDR Rev. 7:R33-R37. Piddock, L. 1995. Quinolone resistance and Campylobacter spp. J. Antimicrob. Chemother. 36: 8 9 1-89 8. Potter, R., J. Kaneene, and W. Hall. 2003. Risk factors for sporadic Campylobacter jejuni infections in rural Michigan: a prospective case-control study. Am. J. Public Health 93:2118-2123. Quick, R., L. Venczel, 0. Gonzalez, E. Mintz, A. Highsmith, A. Espada, E. Damini, N. Bean, E. De Hannover, and R. Tauxe. 1996. Narrow-mouthed water storage vessels and in situ chlorination in a Bolivian community: a simple method to improve drinking water quality. Am. J. Trop. Med. Hyg. 54511-516.
188
OLSON ET AL.
Quinn, T., S. Goodell, C. Fennell, S. Wang, M. Schuffler, K. Holmes, and W. Stamm. 1984. Infections with Campylobacter jejuni and Campylobacter-like organisms in homosexual men. Ann. Intern. Med. 101:187-192. Redmond, E., C. Griffith, J. Slader, and T. Humphrey. 2004. Microbiological and observational analysis of cross contamination risks during domestic food preparation. Br. Food J. 106581597. Richardson, L. 2007. Isolation of Campylobacter from circulating blood of commercial broilers via venipucnture of exposedl unexposed brachial veins, abstr. P287. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Rotterdam, The Netherlands, 1 to 5 September 2007. Zoonoses Public Health 54(Suppl. 1):98. Riley, L. W., and M. J. Finch. 1985. Results of the first year of national surveillance of Campylobacter infections in the United States. J. Infect. Dis. 151:956-959. Riordan, T., T. Humphrey, and A. Fowles. 1993. A point source outbreak of Campylobacter infections related to bird-pecked milk. Epidemiol. Infect. 110:261-265. Robinson, D. 1981. Infective dose of Campylobacterjejuni in milk. Br. Med. J. Clin. Res. Ed. 282:1584. Rodrigues, L., J. Cowden, J. Wheeler, D. Sethi, P. Wall, P. Cumberland, D. Tompkins, M. Hudson, J. Roberts, and P. Roderick. 2001. The study of infectious intestinal disease in England: risk factors for cases of infectious intestinal disease with Campylobacter jejuni infection. Epidemiol. Infect. 127:185-193. Saeed, A., N. Harris, and R DiGiacomo. 1993. The role of exposure to animals in the etiology of Campylobacter jejunilcoli enteritis. Am. J. Epidemiol. 137:108-114. Said, B., F. Wright, G. L. Nichols, M. Reacher, and M. Rutter. 2003. Outbreaks of infectious disease associated with private drinking water supplies in England and Wales 1970-2000. Epidemiol. Infect. 130:469-479. Salfield, N., and E. Pugh. 1987. Campylobacter enteritis in young children living in households with puppies. Br. Med. J. 294:2122. Samuel, M., D. Vugia, S. Shallow, R. Marcus, S. Segler, T. McGivern, H. Kassenborg, K. Reilly, M. Kennedy, F. Angulo, and R. Tauxe. 2004. Epidemiology of sporadic Campylobacter infection in the United States and declining trend in incidence. Clin. Infect. Dis. 38(Suppl 3):S165-S174. Sandberg, M., B. Bergsjo, M. Hofshagen, E. Skjerve, and H. Kruse. 2002. Risk factors for Campylobucter infection in Norwegian cats and dogs. Prev. Vet. Med. 55:244-253. Sato, K., P. Bartlett, J. Kaneene, and F. Downes. 2004. Comparison of prevalence and antimicrobial susceptibilities of Campylobacter spp. isolates from organic and conventional dairy herds in Wisconsin. Appl. Environ. Microbiol. 70:1442-1447. Scallan, E. 2007. Activities, achievements, and lessons learned during the first 10 years of the Foodborne Diseases Active Surveillance Network: 1996-2005. Clin. Infect. Dis. 44:718-725. Scherer, K., E. Bartelt, C. Sommerfeld, and G. Hildebrandt. 2006. Quantification of Campylobacter on the surface and in the muscle of chicken legs at retail. J. Food Prot. 69:757-761. Schlesselman, J. J. 1982. Case-Control Studies: Design, Conduct, Analysis. Oxford University Press, New York. Schmid, G. P., R. E. Schaefer, and B. D. Plikaytis. 1987. A oneyear study of endemic campylobacteriosis in a midwestern city: association with consumption of raw milk. J. Infect. Dis. 156: 218-222. Schonberg-Norio, D., J. Takkinen, M. Hanninen, M. Katila, S. Kaukoranta, L. Mattila, and H. Rautelin. 2004. Swimming and Campylobacter infections. Emerg. Infect. Dis. 10: 1474-1477. Schorr, D., H. Schmid, H. Rieder, A. Baumgartner, H. Vorkauf, and A. Burnens. 1994. Risk factors for Campylobacter enteritis in Switzerland. Zentralbl. Hyg. Umweltmed. 196:327-337.
Shane, S. M. 1991. Environmental factors associated with Cumpylobacter jejuni colonization of poultry, p. 29-46. In S. M. Shane (ed.), Colonization Control of Human Bacterial Enteropathogens in Poultry. Academic Press, San Diego, CA. Shmilovitz, M., B. Kretzer, and N. Rotman. 1982. Campylobacter jejuni as an aetiological agent of diarrheal diseases in Israel. Isr. J. Med. Sci. 18:935-940. Simonsen, J., M. Strid, K. Mslbak, K. Krogfelt, A. Linneberg, and P. Teunis. Sero-epidemiology as a tool to study the incidence of Salmonella infections. Epidemiol. Infect. [Online.] doi:10.1017 /S0950268807009314. Skirrow, M. 1991. Epidemiology of Campylobacter enteritis. Int. J. Food Microbiol. 12:9-16. Skirrow, M. B. 1987. A demographic survey of Campylobacter, Salmonella and Shigella infections in England. A public health laboratory service survey. Epidemiol. Infect. 99:647-657. SMI. 1997. Reported Foodborne and Waterborne Outbreaks in Sweden. Swedish Institute of Infectious Disease Control, Stockholm. Smith, K., J. Besser, C. Hedberg, F. Leano, J. Bender, J. Wicklund, B. Johnson, K. Moore, and M. Osterholm. 1999. Quinoloneresistant infections of Campylobacter jejuni in Minnesota, 19921998. N. Engl. J. Med. 340:1525-1532. Sorvillo, F., L. Lieb, and S. Waterman. 1991. Incidence of campylobacteriosis among patients with AIDS in Los Angeles County. J Acquir. Immune Defic. Syndr. 4598-602. Southern, J., R. Smith, and S. Palmer. 1990. Public Health. Bird attack on milk bottles: possible mode of transmission of Campylobacter jejuni to man. Lancet 336:1425-1427. Stafford, R., P. Schluter, M. Kirk, A. Wilson, L. Unicomb, R. Ashbolt, and J. Gregory. 2007. A multi-centre prospective casecontrol study of campylobacter infection in persons aged 5 years and older in Australia. Epidemiol. Infect. 135:978-988. Stafford, R, T. Tenkate, and B. McCall. 1996. A five year review of Campylobacter infection in Queensland. Commun. Dis. Intell. 20~478-482. Stem, N. J., K. L. Hiett, G. A. Alfredsson, K. G . Kristinsson, J. Reiersen, H. Hardardottir, H. Briem, E. Gunnarsson, F. Georgsson, R. Lowman, E. Berndtson, A. M. Lammerding, G. M. Paoli, and M. T. Musgrove. 2003. Campylobacter spp. in Icelandic poultry operations and human disease. Epidemiol. Infect. 130: 23-32. Strid, M., J. Engberg, L. Larsen, K. Begtrup, K. Mslbak, and K. Krogfelt. 2001. Antibody responses to Campylobacter infections determined by an enzyme-linked immunosorbent assay: 2-year follow-up study of 210 patients. Clin. Diagn. Lab. Immunol. 8: 314-319. Studahl, A., and Y. Andersson. 2000. Risk factors for indigenous Campylobacter infection: a Swedish case-control study. Epidemiol. Infect. 125:269-275. Swift, L., and P. Hunter. 2004. What do negative associations between potential risk factors and illness in analytical epidemiological studies of infectious disease really mean? Epidemiol. Infect. 19:219-223. Tauxe, R., N. Hargrett-Bean, C. Patton, and I. Wachsmuth. 1988. Campylobacter isolates in the United States, 1982-1986. Morb. Mortal. Wkly. Rep. 37:l-13. Taylor, D. N., K. T. McDermott, J. R Little, J. G. Wells, and M. J. Blaser. 1983. Campylobacter enteritis from untreated water in the Rocky Mountains. Ann. Intern. Med. 99:38-40. Tenkate, T., and R. Stafford. 2001. Risk factors for campylobacter infection in infants and young children: a matched case-control study. Epidemiol. Infect. 127:399-404. Ternhag, A., A. Toerner, A. Svensson, J. Giesecke, and K. Ekdahl. 2005. Mortality following Campylobacter infection: a registrybased linkage study. BMC Infect. Dis. 5:l-5.
CHAPTER 9
C. 7E7UNI INFECTIONS IN INDUSTRIALIZED NATIONS
Teunis, P., and A. Havelaar. 2000. The beta Poisson dose-model is not a single-hit model. Risk Assess. 20513-520. Teunis, P., W. Van den Brandhof, M. Nauta, J. Wagenaar, H. Van den Kerkhof, and W. Van Pelt. 2005. A reconsideration of the Campylobacter dose-response relation. Epidemiol. Infect. 133583-592. Unicomb, L., J. Ferguson, T. Riley, and P. Collignon. 2003. Fluoroquinolone resistance in Campylobacter absent from isolates, Australia. Emerg. Infect. Dis. 9:1482-1483. Unicomb, L., J. Ferguson, R. Stafford, R. Ashbolt, M. Kirk, N. Becker, M. Patel, G. Gilbert, M. Valcanis, L. Mickan, and the Australian Campylobacter Subtyping Study Group. 2006. Lowlevel fluoroquinolone resistance among Campylobacter jejuni isolates in Australia. Clin. Infect. Dis. 42: 1368-1374. U.S. Department of Agriculture. 1996. Pathogen reduction: hazard analysis and critical control point (HACCP) systems; final rule. Fed. Regist. 6 1:3 8 805-3 89 89. U.S. Environmental Protection Agency. 2006. National Primary Drinking Water Regulations: Groundwater Rule: Proposed Rules. 40 CFR parts 141 and 142. Environmental Protection Agency, Washington, DC. http: //www.epa.gov/safewater/gwr.html. van Pelt, W., M. A. de Wit, W. J. Wannet, E. J. Ligtvoet, M. A. Widdowson, and Y. T. van Duynhoven. 2003a. Laboratory surveillance of bacterial gastroenteric pathogens in The Netherlands, 1991-200 1. Epidemiol. Infect. 130:43 1-441. van Pelt, W., Y. Doorduyn, E. Evers, W. Ang, Y. van Duynhoven, and A. Havelaar. 2005. Presented at the MedVetNet General Scientific Meeting. 29 June-1 July 2005, Winchester, United Kingdom. van Pelt, W., S. van der Heijden, and Y. van Duynhoven. 2007. Similarities and differences in seasonality of Campylobacter in broilers and humans, 1998-2006, The Netherlands, abstr. P110. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms, Rotterdam, The Netherlands, 1 to 5 September 2007. Zoonoses Public Health 54(Suppl. 1):50-51. van Pelt, W., G. Visser, and S. Valkenburgh. 2003b. Zoonoses and Zoonotic Agents in Humans, Food, Animals and Feed in The Netherlands, 2002. Inspectorate for Health Protection and Veterinary Public Health, the Hague, The Netherlands. van Pelt, W., W. Wannett, A. van de Giessen, D. Mevius, and Y. van Duynhoven. 2004. Trends in gastroenteritis in The Netherlands, 1996-2003. Infect. Bull. 15:335-341. Vellinga, A., and F. Van Loock. 2002. The dioxin crisis as experiment to determine poultry-related Campylobacter enteritis. Emerg. Infect. Dis. 8:19-22. Vestergaard, L. S., K. E. Olsen, R. Stensvold, B. E. Bottiger, M. Adelhardt, M. Lisby, L. Mork, and K. Mslbak. 2007. Outbreak
189
of severe gastroenteritis with multiple aetiologies caused by contaminated drinking water in Denmark, January 2007. Eurosurveillance 12:E070329 1. Voetsch, A., T. Van Gilder, F. Angulo, M. Farley, S. Shallow, R. Marcus, R Cieslak, V. C. Deneen, and R. Tauxe. 2004. FoodNet estimate of the burden of illness caused by nontyphoidal Salmonella infections in the United States. Clin. Infect. Dis. 38:S127-S134. Vogt, R. L., H. E. Sours, T. Barrett, R. A. Feldman, R. J. Dickinson, and L. Witherell. 1982. Campylobacter enteritis associated with contaminated water. Ann. Intern. Med. 96~292-296. Waldenstroem, J., T. Broman, I. Carlsson, D. Hasselquist, R. Achterberg, J. Wagenaar, and B. Olsen. 2002. Prevalence of Campylobacter jejuni, Campylobacter lari, and Campylobacter coli in different ecological guilds and taxa of migrating birds. Appl. Environ. Microbiol. 685911-5917. Wheeler, J., D. Sethi, J. Cowden, P. Wall, L. Rodrigues, D. Tompkins, M. Hudson, and P. Roderick. 1999. Study of infectious intestinal diseases in England: rates in the community, presenting to general practice, and reported to national surveillance. Br. Med. I. 318:1046-1050. World Health Organization. 2001. The Increasing Incidence of Human Campylobacteriosis: Report and Proceedings of a W H O Consultation of Experts. WHO/CDS/CSR/APH 2001.7. World Health Organization, Copenhagen. Wilson, I. 2003. Salmonella and Campylobacter contamination of raw retail chicken from different producers: a six-year survey. Epidemiol. Infect. 130:320-323. Wingstrand, A., J. Neimann, J. Engberg, E. Nielsen, P. GernerSmidt, H. Wegener, and K. Mslbak. 2006. Fresh chicken as main risk factor for campylobacteriosis,Denmark. Emerg. Infect. Dis. 12:280-285. Wong, T., R. Whyte, A. Cornelius, and J. Hudson. 2004. Enumeration of Campylobacter and Salmonella on chicken packs. Br. Food J. 106:651-662. Wood, R., K. MacDonald, and M. Osterholm. 1992. Campylobacter enteritis outbreaks associated with drinking raw milk during youth activities. A 10-year review of outbreaks in the United States. JAMA 268:3228-3230. Yang, S., M. Leff, D. McTague, K. Horvath, T. JacksonThompson, T. Murayi, G. Boeslager, T. Melnik, M. Gildermaster, D. Ridings, s. Altekruse, and F. Angulo. 1998. Multistate surveillance for food handling, preparation and consumption behaviors associated with foodborne diseases: 1995 and 1996 BRFSS food safety questions. Morb. Mortal. Wkly. Rep. CDC Surveill. Summ. 47(4):33-57. Zhao, T., M. Doyle, and D. Berg. 2000. Fate of Campylobacter jejuni in butter. J. Food Prot. 63:120-122.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chauter 10
Molecular Epidemiology of Campylobacter Species STEPHEN L. W. ON, NOELMCCARTHY, WILLIAMG. MILLER,AND BRENT J. GILPIN
demonstrated that genetic change in at least C. jejuni and C. coli is driven by recombination events that result in a weakly clonal population structure (Dingle et al., 2001a, 2005). This, alongside other evolutionary drivers including spontaneous mutation and plasmid acquisition and loss (Wassenaar et al., 2000), can make the definitive linkage between any two strains a difficult task because genotyping results of related strains may differ. Simply put, there are many human cases and many possible sources of infection from a genetically diverse organism. It is little wonder that the study of the epidemiology of Campylobacter is complex and challenging. The need for subtyping methods to study Campylobacter epidemiology became evident after the seminal studies of Butzler et al. (1973) and Skirrow (1977) first demonstrated their prevalence and significance. Some of the early phenotypic methods (biotyping, serotyping, phage typing) are still used today. However, although such methods are suited for large-scale screening, their limited discriminatory potential has proved insufficiently informative for resolving any but the most clear-cut issues. Moreover, although serotyping (notably the Penner scheme based predominantly on heat-stable antigens) had for many years been the main tool used to analyze Campylobacter, many laboratories no longer have either the expertise or the antisera to perform serotyping. Other laboratories (based in New Zealand and The Netherlands, for example) only have antisera to determine a limited range of serotypes. Since the 1980s, the increased interest in exploiting molecular methods to determine genetic polymorphisms between bacteria has driven developments in identification and subtyping of Campylobacter spp. with the aim of
There are three major obstacles to the tracing of sources of human campylobacteriosis: the scale of the problem, the possible route or routes of infection, and the population biology of the major pathogenic Campylobacter species. In developed countries with surveillance systems, Campylobacter is known to be the most frequently reported bacterial cause of human gastroenteritis, usually accounting for more cases of disease than Salmonella, Yersinia, and pathogenic Escherichia coli combined. With an incidence of up to 432 cases per 100,000 population (Institute of Environmental Science and Research, Ltd., 2007), subtyping strains to identify sources of infection demands a major effort. Moreover, the potential sources of infection are many. Campylobacter species are widespread in nature and may be found in an extensive range of terrestrial animals where they form part of their normal intestinal microflora. Chickens, turkeys, ducks, geese, ostriches, sheep, cattle, deer, goats, pigs, cats, dogs, rodents, wild birds, and even reptiles harbor Campylobacter species-sometimes different species andl or strains concurrently. Rodents as well as insects such as flies and beetles may also serve as unwitting couriers of the organism (Adhikari et al., 2004; Bates et al., 2004; Hald et al., 2004a), transferring strains to hitherto uncontaminated areas. Consumption of, or contact (direct or indirect, through exposure to aqua- or agricultural environments or contaminated surfaces) with the animal host can result in transmission of the organism to humans and subsequent infection. In addition, the genetic diversity of many campylobacters adds to the dilemma of making an informed decision as to the relatedness of any two isolates. Advances in population biology have
Stephen L. W. On Food Safety Programme, Institute of Environmental Science and Research Ltd., Christchurch, New ZeaNoel McCarthy Peter Medawar Building for Pathogen Research and Department of Zoology, Health Protection Agency land. William G. Miller Produce Safety and Microbiology Research and University of Oxford, Oxford OX1 3SY, United Kingdom. Unit, United States Department of Agriculture, Albany, CA 94710. Brent J. Gilpin Water Management Programme, Institute of Environmental Science and Research Ltd., Christchurch, New Zealand.
191
192
ONETAL.
improving our understanding of their epidemiology. Interestingly, although genetic methods almost invariably prove more discriminatory than phenotypic tools, the population genetics of the bacteria (which differ markedly between different Campylobacter species) that drive their evolution often mean that typing results prove as challenging to interpret as the broad picture of the epidemiology of the organisms they aim to elucidate. This chapter contains a summary of the various modern genotypic methods available for subtyping the various Campylobacter species, the use of genetic methods for investigating outbreaks of disease, the application and interpretation of large-scale techniques in broader epidemiological studies, and a perspective of future developments in the field. MOLECULAR TYPING METHODSA BRIEF OVERVIEW Our overview of genotyping methods is divided into two broad sections: the first describes methods that have principally been applied to comparatively small-scale studies, limited to investigations of relatively few strains and/or restricted geographical or chronological origin. Discussions of three methods (pulsed-field gel electrophoresis [PFGE], multilocus sequence typing [MLST], and amplified fragmentlength polymorphism [AFLP]) used to compare relatively large numbers of strains follow in a section aimed at summarizing our knowledge of Campylobacter molecular epidemiology in a broader perspective.
fla Typing Campylobacter coli and C. jejuni subsp. jejuni (hereafter C. jejuni) have two flagellin genes, designated flaA and flaB. Each gene is approximately 1.7 kb and is separated by an intergenic spacer region of -170 bp. These genes encode repeats of flagellin subunits that form the flagellum found on one or both ends of Campylobacter cells. The 5’ and 3’ regions of each gene are highly conserved, with considerable sequence variation in the region in between. A number of PCR primers have been designed that amplify specific regions of this gene cluster. Variability within each amplicon is identified by either digestion followed by restriction fragment length polymorphism (fa-RFLP) or direct sequencing (fa-amplified short variable regions [fla-SVR]). fla-RFLP Nachamkin et al. (1996) described primers that amplify 1,728 bp of the flagellin A gene @aA) and
flanking sequences. A number of restriction enzymes including AluI, DdeI, EcoRI, HinfI, and PstI have been used either alone or in combination to identify variability in fZa sequences. Digestion of the amplified fragment with DdeI is favored by most, with fragments between 100 and 1,500 bp commonly analyzed. fZaA-RFLP is relatively simple to perform, fairly inexpensive, and well suited to relatively highthroughput analysis. However, standardization of electrophoresis conditions remains an issue, although groups in both Europe and Australia (Djordjevic et al., 2007) have shown that high levels of interlaboratory standardization are possible. Moreover, because polymorphisms within a single genetic locus are being determined, its discriminatory power is inherently limited, and certainly less so than methods determining whole-genome polymorphisms. For example, in a survey of C. jejuni isolates from 100 different poultry samples, 73 different SmaI PFGE types were observed, but only 30 fZaA types (Han et al., 2007). In 179 C. jejuni strains from cattle in Turkey, just 23 different fZaA-RFLPtypes were observed, with 5 fZaAtypes in 30 C. coli strains (Acik and Cetinkaya, 2005). Djordjevic et al. (2007) reported that fZaARFLP was less discriminatory than MLST but could be used to predict MLST clonal complexes in many cases. However, fZaA-RFLP was the method best associated with epidemiology in the Walkerton outbreak in 2000 (Clark et al., 2005), supporting its value in outbreak investigations where rapid results, high throughput, and low cost are important factors for consideration of applicable typing methods. fla-SVR Sequencing Direct sequencing of PCR-amplified short variable regions (SVR) of either the flaA or flaB genes overcomes many of the reproducibility problems which flaA-RFLP can have. An online database has been established (http://hercules.medawar.ox.ac.uk/ flaA/), which as of July 2007 contained 965 fla SVR alleles 321 bp in length. Novel sequences not in the database can have an allele number assigned after submission and validation of DNA trace files. However, the diversity of fZaA-SVR types is less than that observed with fZaA-RFLP types. For example, of 66 Irish human and poultry isolates, 58 different banding profiles with fZaA PCR-RFLP were observed, but just 28 different fZaA-SVR types (Corcoran et al., 2006). Similarly, 20 fZuA-RFLP types were found among 84 human diarrheal strains, but only 15 fZaASVR types (O’Reilly et al., 2006). In contrast, Mellmann et al. (2004) found fZaB-SVR to be slightly less discriminatory than flaA-SVR, but better correlated with outbreak strains.
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
193
Genomic instability of C. jejuni, and in particular the flaA gene, has been described by a number of researchers (Harrington et al., 1997). Although this could potentially affect the reliability of flaA typing, in practice, this appears not to be the case. High levels of genomic instability may be restricted to particular strains or situations. Instability of fla genes does not appear an impediment to typing methods that are based on these genes. fla typing methods can be used as a first screen to identify isolates for further investigation by other typing methods. Even though several researchers have demonstrated by means of fla typing methods that outbreak strains can be distinguished from sporadic strains, this may not hold true if a large number of sporadic strains are examined within the same time frame.
genotypes to flaA typing, with some studies finding ribotyping to be more discriminatory (O’Reilly et al., 2006) and others finding it to be less discriminatory (de Boer et al., 2000). The RiboPrinter system requires a considerable capital outlay, and according to some (O’Reilly et al., 2006), the cost of testing each isolate is twice as expensive as SmaI PFGE and six times more expensive than flaA-RFLP. In combination with other subtyping methods, automated ribotyping will produce valid and reproducible results. However, the low level of diversity and relatively high costs of automated ribotyping do restrict its usefulness for the study of Campylobacter species.
Ribotyping and Riboprinting
Random amplification of polymorphic DNA (RAPD) has been applied to the analysis of campylobacters for at least the last 10 years. DNA is isolated from purified isolates and PCR reactions performed by using a single 10- to 15-bp primer. This primer anneals to multiple sites in the genome, which, when close enough together, will generate amplicons whose size depends on the position of the annealing sites. These amplicons can be visualized by standard electrophoresis on 1.5% (wt/vol) agarose gels (Hernandez et al., 1995). By means of this approach, Acik and Cetinkaya (2006) identified from 348 cattle and sheep C. jejuni and C. coli isolates a total of 87 different RAPD genotypes, each composed of two to eight fragments. Nielsen et al. (2000) performed RAPD analysis with three different fluorescently labeled primers. The resultant PCR products were then mixed and amplicons separated by using the fragment sizing capacity of a DNA sequencer. In an analysis of 80 unrelated strains, 56 genotypes were identified by RAPD analysis-6 more than were identified by SmaI PFGE. RAPD analysis is relatively inexpensive and has the potential to identify a high level of diversity among isolates, but its use has not become widespread because of the justified concern of many researchers that the diversity of patterns may in part arise from variation in the assay rather than the organism. To address this concern, many researchers use some form of standardized, premixed “Ready to Go” RAPD analysis beads to provide a consistent and uniform source of Taq polymerase, deoxynucleoside triphosphates, and buffer conditions. The fact that RAPD analysis requires no prior knowledge of any genome targets makes it particularly applicable to the analysis of the less frequently studied Campylobacter species. There have been few published examples demonstrating the value of RAPD analysis in out-
Ribotyping involves cleaving of whole genomic DNA with a cutting restriction enzyme (e.g., HindIII), subsequent hybridization of a labeled oligonucleotide probe derived from the ribosomal gene to fragments containing the gene, and visualization of the resulting labeled patterns. This technique is technically demanding and difficult to standardize, but a semiautomated system is available that overcomes these difficulties, permitting its wider use. With the Dupont Qualicon RiboPrinter microbial characterization system, bacterial suspensions from isolated bacteria are added to the RiboPrinter, which undertakes in an automated manner lysis of cells, DNA isolation, restriction enzyme digestion, electrophoresis, Southern blot analysis, hybridization with a chemiluminescently labeled 16s to 23s rRNA primer, detection of patterns, and finally comparison with previously analyzed isolates. The automation facilitates a high level of reproducibility with relatively low levels of technical input. Up to eight bacterial isolates can be tested at one time, with results available 8 h from sample input. New batches of isolates can be added every 2 to 3 h. The RiboPrinter has been applied to analysis of Campylobacter strains, most commonly by using the enzyme PstI, which does not cut within the Campylobacter 16s rRNA gene. Most Campylobacter strains have only three copies of the ribosomal genes, which results in the generation of only three to six bands by using PstI. Ribotyping is less discriminatory than PFGE with SmaI. For example, of 120 isolates from meats, 44 SmaI PFGE patterns were found, but only 22 ribotypes (Ge et al., 2006). Of 84 human isolates, O’Reilly et al. (2006) found 53 subtypes by PFGE with SmaI, 33 by MLST, and only 28 by ribotyping. Ribotyping appears to generate a similar number of
Random Amplification of Polymorphic DNA Analysis
194
ONETAL.
break investigations (Nielsen et al., 2000; Ono et al., 2003). In an outbreak situation, indistinguishable RAPD profiles should be generated, but interpreting the significance of any differences in patterns may be difficult. In common with PFGE and other bandbased methods, inferring phylogenetic relationships on the basis of similarity of RAPD patterns is a potentially misleading exercise, particularly because the risk of obtaining different profiles for the same strain upon reanalysis is comparatively high. RAPD analysis appears to have largely been supplanted by other methodologies by most researchers. Rep-PCR and ERIC-PCR Repetitive extragenic palindromic-PCR (RepPCR) targets conserved repetitive DNA sequences. Rep-PCR has been applied widely to a number of organisms, but there is only a single report of its application to campylobacters (Hiett et al., 2006). In a survey of 48 C. jejuni isolates, Rep-PCR performed using the Uprime DT primer set generated four to eight bands in the 400- to 5,000-bp range. Rep-PCR was less discriminating than flaA-SVR sequencing, with a diversity index of 0.8364 for Rep-PCR compared with 0.9394 for flaA-SVR (Hiett et al., 2006). Rep-PCR does not appear to offer significant advantages over the other, more widely used methods for genotyping campylobacters. Similar to Rep-motifs, enterobacterial repetitive intergenic consensus (ERIC) sequences may also be targeted by PCR assays and used for subtyping. Its application to 107 strains of diverse origin was used to support other genotypic data, suggesting a dichotomy of infectious potential (Moser et al., 2002). In a study of Campylobacter epidemiology in pigs, ERICPCR patterns suggested that transmission could be maternal, although because some patterns resembled those from distinct farms, this hypothesis was not unequivocal (Weijtens et al., 1997). Other researchers have elected to use ERIC-PCR primers with random primers, whereby outbreak strains yield identical or similar patterns that are distinguished from sporadic isolates, which yield unique profiles (Iriarte et al., 1996). Nonetheless, neither Rep- nor ERIC-PCR fingerprinting has proven especially popular with researchers, most likely because of the same pitfalls that are encountered with RAPD fingerprinting. High-Resolution Melt Analysis and Denaturing Gradient Gel Electrophoresis High-resolution melt (HRM) analysis is a postPCR analysis step that uses the fluorescence detectors of real-time PCR machines to characterize PCR prod-
ucts. After PCR amplification, the amplified DNA is denatured by slowly increasing the temperature in small increments (0.01 to 0.05OC). As the DNA denatures, intercalating DNA dyes such as SYBR green are released, causing a change in fluorescence. Precise measurements of the melting point peak and the shape of the curve allow the differentiation of amplicons with different sequences, even if they share the same melting point. HRM analysis could, in principle, be applied to any PCR amplicon with sufficient genetic variation, including fla loci described above. Price et al. (2007) described the use of HRM to analyze PCR amplicons derived from clustered, regularly interspaced short palindromic repeats (CRISPRs). CRISPRs are composed of near-perfect direct repeats interspersed with similarly sized nonrepetitive spacer sequences. New primers were described which amplified CRISPR PCR product from almost all Australian C. jejuni isolates tested, but not from any of the seven C. coli isolates tested (Price et al., 2007). By means of a SYBR green-based HRM real-time PCR assay, Price et al. (2007) were able to identify 36 CRISPR HRM genotypes among 84 strains on the basis of differences in the melt curve generated. Previous analysis of these strains had resolved 53 SmaI PFGE types, 32 MLST types, and 15 flaA SVR types (O’Reilly et al., 2006). HRM analysis of Campylobacter requires evaluation in additional laboratories. Providing that results are reproducible, it may be a useful first screening tool, and in combination with other methods, it provides sufficiently high levels of discrimination. Denaturing gradient gel electrophoresis is an alternative way that sequence polymorphisms in PCR products may be distinguished. Here, differences in the relative stability of DNA sequences are detected by electrophoresis in a denaturing chemical gradient. Separation can be improved by adding a polynucleotide tail, or GC clamp, to the target during the PCR. This approach has been used to type C. jejuni strains by examination of the flaA gene, but with rather poor results: just 13 types from 80 unrelated strains were obtained, compared with 40 types found by fla-RFLP analysis (Nielsen et al., 2000). Single Nucleotide Polymorphism, Binary Typing, and DNA Microarray Methods Genomic analysis of increasing numbers of strains of Campylobacter is identifying an increasing number of gene targets that can be used to develop genotyping assays. Single nucleotide polymorphisms (SNPs) are DNA sequence variations that occur in coding and noncoding sequences, and for typing pur-
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
poses are best when only two variations are observed (e.g., A or G). SNPs are typically detected by PCRbased assays specific for either the wild-type (0) or the SNP (1)allele. When multiple SNPs are analyzed in parallel, a binary typing system can emerge. Price et al. (2006b) designed seven SNP assays that were based on loci in the C. jejuni MLST database. This research group also developed a binary typing system that was based on the presence or absence of eight specific genes (Price et al., 2006a). Both of these assays utilize real-time PCR-based assays for detection. Among a set of 84 isolates, the binary typing assay identified 27 genotypes, whereas the SNP assay identified 20 genotypes. When combined, 37 genotypes could be resolved-more than obtained by MLST, but less than SmaI PFGE (Price et al., 2006a). Microarray-based technologies offer an alternative detection methodology that is based on hybridization of DNA extracted from isolates to arrays containing specific oligonucleotide arrays. Use of a first-generation whole-genome microarray that is based on the genome library produced during the sequencing of strain NCTC 11168 revealed all distinct isolates to be genetically different, but strains found to be more similar in AFLP profiling (On and Harrington, 2000) were also more similar in genome content in the microarray analysis, giving credibility to the epidemiological validity of this approach (On et al., 2006). A set of 70-mer probes has also been described that not only allows species identification, but through inclusion of probes targeting the lipooligosaccharide biosynthesis locus, also allows a limited genotypic classification (Quinones et al., 2007). With the addition of further probes, microarraybased typing systems may facilitate genotyping goals and permit characterization of virulence and other attributes. The identification of appropriate probes is an area of active research, but to be feasible on a routine basis, it will require technological advances to reduce the cost of the analysis.
PFGE PFGE is widely regarded as among the most discriminatory typing methods generally available for bacteria. It involves the digestion of intact genomic DNA (embedded in agarose to avoid shearing) with rare-cutting restriction enzymes that cut the DNA into relatively few fragments, in contrast with its predecessor, restriction enzyme analysis (e.g., Lind et al., 1996). These fragments are then separated in a specially designed coordinated pulsed electric field, such that the large fragments are gently oriented through the agarose gel matrix. In a typical 18- to 20-h electrophoresis run, the separation of fragments
195
in the 20,000- to 1,000,000-bp range is achieved. Widespread, worldwide adoption of the PulseNet protocol developed at the CDC (Ribot et al., 2001) has significantly improved the comparability of PFGE analysis. However, it remains a technically demanding method with a low throughput and is a relatively high consumable and capital item cost. Furthermore, the DNase activity of some strains needs to be deactivated to ensure that DNA samples do not degrade before electrophoresis. A sample preparation protocol has been described that successfully deactivates DNase (Gibson et al., 1994) but involves the pretreatment of bacterial cells with the toxic chemical formaldehyde. On occasion, strains will resist digestion with the endonuclease of choice (On et al., 1999; On and Vandamme, 1997; W. G. Miller, unpublished data). Site-specific DNA methylation and similar processes are likely to be responsible for such phenomena. In the case of C. jejuni, most researchers have used the enzyme SmaI, which typically generates 4 to 10 fragments (Cornelius et al., 2005; Devane et al., 2005). KpnI has been used as a secondary enzyme in a number of cases (Eyles et al., 2006; Gilpin et al., 2006; On et al., 1998), or even as the primary enzyme (Karenlampi et al., 2007a). SalI (Malik-Kale et al., 2007) and BamHI (On et al., 1998) have also been used. For C. jejuni, the number of fragments generated is generally Ken1 > BamHI > SmaI > SalI. Comparisons between large numbers of isolates have been facilitated by the widespread availability of software such as BioNumerics (Applied Maths, Ghent, Belgium), which are able to electronically compare electrophoresis patterns generated. BioNumerics converts band positions into a percentage of migration down the gel. Through use of a standard marker (in PulseNet protocols, Salmonella enterica subsp. enterica serovar Braenderup H9182), gels can be electronically stretched or compressed to adjust the gel so that the bands of the standard migrate in silico to the same common distance. This adjustment for small variations in PFGE patterns that can and do occur between gels and laboratories allows comparison of isolates run on different gels and in different laboratories. This normalization is of course not perfect, so software packages such as BioNumerics allow the setting of a tolerance (variation where two bands are considered the same, typically 1.0%) and optimization values (allowance for movement of whole pattern up and down a gel, typically 1.5%).Setting these values is critical because making these two parameters smaller will split isolates up, and making them bigger will group isolates with patterns that are actually different, especially because identification of common patterns or types is most often attained by use of clus-
196
ONETAL.
tering algorithms such as UPGMA (unweighted pair group with mathematical average). Different patterns are typically assigned a unique pattern designation by most researchers. There is no common pattern designation internationally for PFGE patterns, nor is there an agreed consensus of what constitutes a different PFGE pattern. Isolates within PulseNet USA are assigned a common pattern number, but other networks and laboratories use their own pattern designations and their own criteria for pattern assignments. In New Zealand, we have decided to take a conservative approach that errs on the side of grouping isolates rather than risking splitting isolates, which may only appear different because of uncorrected gel-to-gel artifacts. However, even strains that give indistinguishable patterns with one restriction enzyme on a single gel may be distinguished with the application of a different enzyme (On et al., 1998). Digestion with one enzyme is sufficient in many cases to show differences between isolates but is insufficient to demonstrate similarity between isolates. The limitation of one enzyme is well illustrated by examination of the most frequent SmaI pattern in the New Zealand databases, Sm0001, which consists of only five bands (Isolates A-I Fig. 1).Some variation is evident within this SmaI pattern, but this variation is insufficient to reproducibly differentiate the isolates by computer-based comparisons as described above. Digestion with KpnI, however, demonstrates the genetic diversity in this group, some of which do form clonal groups (isolates A to I; Fig. l), a wellestablished phenomenon (On et al., 1998). Thus, what is arguably more important than the enhanced discrimination this approach allows is the increased confidence in interpretation that digestion with two enzymes allows. Digestion with a second (or more)
. . 8 : :
enzyme, however, is a strong indicator of a close clonal relationship, and the implied phylogenetic similarity between two enzymes is supported by MLST analysis in a number of situations we have examined (B. J. Gilpin, unpublished data; S. L. W. On, unpublished data). Whenever the intention is to show that two isolates are indistinguishable, PFGE should include the use of two enzymes. Conversely, it must be noted that a clonal relationship may exist between strains with dissimilar PFGE profiles. Several different types of genetic phenomena, including point mutation, transformation, and plasmid acquisition or loss can lead to minor or major changes in profiles (reviewed by Wassenaar et al., 2000). AFLP Profiling First designed for high-resolution genotyping of complex plant genomes, this method was later adapted for discriminating bacteria. Conventional AFLP profiling involves digesting DNA with two enzymes, ligating restriction half-site-specific oligonucleotide adaptors on to the ends of the fragments, and then amplifying the fragments with PCR primers. Typically, one of the primers is labeled with a fluorophore, and hence only a subset of fragments is detected under electrophoresis (normally achieved by use of a sequencer or fragment analyzer). Patterns can be further simplified by using one or both of the PCR primers, which binds to the adaptor with a 1-base overlap (one of A, G, T, or C ) to increase the specificity of the amplification. Moreover, fragments that are too large will amplify poorly or not at all, and those that are too small will not be visualized. There are several variations of fluorescent AFLP analyses. The systems vary in their choice of restriction enzymes, the use of selective bases in the PCR
PFGE-Kpnl
. . I
S
D
I-----+
Figure 1. PFGE patterns of selected New Zealand C. jejuni isolates when digested with SmaI and KpnI. UPGMA (unweighted pair group with mathematical average) dendrograms generated by BioNumerics v4.5, with DICE coefficient, optimization of 1.0%. and tolerance of 1.5%.
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
reaction, or both (Duim et al., 1999; Kokotovic et al., 1999; Lindstedt et al., 2000; Siemer et al., 2005). The sensitivity of the methods is well established, as is their potency for identification of related (i.e., outbreak) strains. Comparisons with MLST show a good correlation in identification of phylogenetic relationships (Miller et al., 2005; S. L. W. On, unpublished data). As discussed below, two of the systems described have additional value in identification and typing of the full taxonomic range of Campylobacter species and of the related Arcobacter species also (Aabenhus et al., 2005; On and Harrington, 2000; On et al., 2003; S. L. W. On and B. L. Siemer, unpublished data). Because species-level identification of these taxa is often demanding (On, 2005), this is a useful adjunct. Fluorescent AFLP (fAFLP) systems can be technically demanding and can also require expensive equipment to run. For C. jejuni, the use of a single enzyme combined with one-base PCR selectivity has been used (Champion et al., 2002) to reduce the number of fragments to a number practical for detection in conventional electrophoresis. These authors found that single-enzyme AFLP had a similar discriminatory ability to PFGE with SmaI. At the Institute of Environmental Science and Research, we have further developed this approach to enable the identification and typing of a broad taxonomic range of Campylobacter and Arcobacter species, similar to previous fAFLP methods (Fig. 2). Given the clinical significance of most species (On, 2005), the need for relatively inexpensive methods to identify and genotype a wide taxonomic range remains a priority.
197
Multilocus Sequence Typing (MLST) MLST involves the sequencing of genes, typically those with housekeeping functions, whereby each sequence variant represents a distinct allelic type. Combining the analyses of several (typically seven) genes alongside appropriate computational analyses is a powerful tool for the study of population biology, epidemiology, and evolution of microbes that has expanded hugely (Maiden, 2006) since first described in 1998 (Maiden et al., 1998). A general rapid and ongoing decline in the cost of DNA sequencing, and more pertinently the inherent portability of this approach (unlike fragment-based methods, there is no requirement to standardize or normalize data once common gene targets are defined, making sharing and wide-scale data comparisons relatively straightforward) are resulting in an increasing use of MLST globally. This extensive use enhances the MLST potential to address questions regarding geographical, temporal, and host species-related variation in type, all issues central to source attribution of human cases. Moreover, the evaluation of any attempts to control Campylobacter in the food chain will be supported if studies of specific interventions can build on comparable reference data. Further useful features of MLST mainly center on the following: (i) the ability to base analyses on biological models that integrate the growing knowledge of the genetic processes and population biology that create diversity in the nucleotide sequence, (ii) exploitation of advances in analytical approaches that are being made to allow explicit model-based inference from multi-
1
I
C.~
~
I
~I
~
#
i
Figure 2. Novel AFLP method for use in conventional electrophoresis and detection systems aimed at identification and typing Campylobacter and Arcobacter species. Lanes 1 and 27 are 100-bp markers; lanes 2 to 7, A. butzleri strains; lanes 8 to 15, C. coli; lanes 16 to 19, C. jejuni subsp. doylei; lanes 20 to 25, C. jejuni subsp. jejuni. Lanes 9 to 11 and 21 to 22 represent well-characterized outbreak strains and are indistinguishable (E. Podivinsky, K. Thorn, and S. L. W. On, unpublished data).
198
ONETAL.
locus sequence data as a result of the enormous amounts of these data being generated in many different fields, and (iii) changing from a typing paradigm that considers whether isolates are identical to (here, assigned to the same sequence type, or ST) or different from one that considers how closely related they are, ancestrally and potentially ecologically. Evaluating the proportion of allelic variants shared between isolates allows their assignation to clonally related groups or clonal complexes from which an objective assessment of strain phylogeny can be made. Information on the seven alleles or 3,309 nucleotides can also be analyzed directly. Either of these approaches has the potential to be useful in source attribution (see below). Limitations of MLST include cost (despite methodological advances, the method remains comparatively expensive) and the degree of resolution that is possible with the quantity of data provided by seven highly conserved gene fragments. Indeed, some studies show the discriminatory index of MLST as somewhat lower than that offered by PFGE (Mellmann et al., 2004; Sails et al., 2003). However, where a group of isolates share a very rare MLST type, some, although not conclusive, evidence for increased likelihood of an epidemiological relationship is given. The role of MLST in understanding the population biology is considered elsewhere in this book (chapter 2).
PEAKS AND TROUGHS: MOLECULAR EPIDEMIOLOGY AND SEASONALITY OF INFECTION The seasonal peak of Campylobacter infection is well recognized and widely studied, but the cause remains uncertain with many theories, none of which have been fully tested (Ekdahl et al., 2005; Hearnden et al., 2003; Hudson et al., 1999; Miller et al., 2004; Nylen et al., 2002; Stanley et al., 1998a, 1998b). Use of genotyping could help answer this question by assessing the extent to which types vary by season and the likely origin of the types involved in the seasonal peak. The ideal data set would be longitudinally sampled isolates from humans and the range of possible sources of infection, which is not currently available. However, some useful data are already available or coming on stream. One study from Finland reported no evidence for different isolates identified from cases of infection observed during the seasonal peak compared with other times in the year (Karenlampi et al., 2007b). An initial report, with a small amount of data, on a United Kingdom study did not report on this outcome directly but displays data indicating that the three largest clonal complexes throughout
the year (ST-21, ST-45, and other/unassigned) all increased during the seasonal peak, in line with the overall increase of cases (Sopwith et al., 2006). Our preliminary analysis of a larger accruing data set also suggests that seasonal variation in humans, at least in the United Kingdom, may be due more to an expansion of the same types that are present during the rest of the year rather than the appearance of novel types (McCarthy et al., 2007). A related finding is that a sample of 725 retail poultry isolates from the United Kingdom showed similar types in spring and early summer compared with the rest of the year, suggesting that the origins of infection in poultry do not vary substantially at this time (Sheppard et al., 2007). The seasonal peak therefore appears most likely to be due to an increase in common types among the sources of Campylobacter infection rather than novel contributions, although this remains to be confirmed, as does the possible contribution of adaptation for environmental survivability to the prevalence of common clonal groups (On et al., 2006).
POINTING THE FINGER OR PASSING THE BUCK? ASSESSING THE CONTRIBUTION OF DIFFERENT SOURCES T O HUMAN CASES OF C. JEJUnrI AND C. COLI INFECTION Source tracking depends, in part, on an accurate estimation of the frequency of different subtypes in each host reservoir. In Salmonella, particular serotypes and phage subtypes are stably found in the same host (Hald et al., 2004b). The biology underlying this success is first that specific clones are well adapted to specific hosts and second that the serophage type is a stable and reliable indicator of a specific clone. For C. jejuni, host-associated markers are extremely rare. Genetic methods of discrimination show enormous diversity, with studies typically reporting about half as many genotypes as there are strains in the study (Dingle et al., 2001a, 2002; French et al., 2005; Hopkins et al., 2004; Manning et al., 2003; Schouls et al., 2003; Siemer et al., 2004). Many common genotypes are broadly distributed (Dingle et al., 2002; Siemer et al., 2004), whereas for rare genotypes, it is not possible to accurately estimate the relative frequency of genotypes in different host reservoirs. These difficulties have meant that although host associations have been identified for particular genotypes, no generally useable approach has been developed. However, several MLST-based studies have shown host association of some clonal complexes such as ST-61 complex (Colles et al., 2003; Dingle et al., 2002; French et al., 2005) and for isolates from these clonal complexes a strong prediction of source
CHAPTER 10
-
MOLECULAR EPIDEMIOLOGY O F CAMPYLOBACTER SPECIES
could be made on the basis of clonal complex membership. However, other clonal groups have a wide distribution (Dingle et al., 2002; Miller et al., 2006). When applied as a general approach to attribution of isolates among food animals, clonal complex information alone provides little improvement on random guessing or the use of ST (McCarthy et al., 2007). The poor performance of clonal complexes in supporting attribution of C. jejuni suggests that major clones are not, as a whole, highly adapted to individual host species among food production animals-at least not when considering adaptation at the whole genomic level. Similar observations on the relative specificity of a few genotypes, and the wide host range of others, have been made with other wholegenome typing methods including AFLP, PFGE, and DNA microarray analysis, where major food animal sources such as poultry and cattle are strongly implicated reservoirs of human infection but genotypes from wild bird isolates show little overlap with those from human isolates (On et al., 1998; Petersen et al., 2001; Siemer et al., 2004). Results of the microarraybased study were correlated with strain viability in aerobic conditions, suggesting that major clones are better adapted for environmental survival than those that occur in a more limited host range (On et al., 2006). In contrast, an analysis of 25 Guillain-Bard and Miller Fisher syndrome-related isolates showed that one clonal complex appeared to be overrepresented, but the isolates came from a wide range of clonal groupings (Dingle et al., 2001b); further studies are required to confirm the consistency or otherwise of phenotypic markers of environmental survival
199
and pathogenicity among major and minor clonal groups. A somewhat different picture is obtained when considering host adaptation from a phylogenetic perspective. With MLST, analysis at allele level is less prone to the problems of diversity affecting analysis of ST and can use differentiation between C. jejuni gene pools in different host species, even if this differentiation is not based on separation between host species at the clonal complex level. This allows more accurate assignment of isolates to correct host species with 80% accuracy in identifying whether C. jejuni isolates originated from ruminants (cattle and sheep) or chicken (McCarthy et al., 2007). An exploration of the biological processes underlying the greater accuracy from allele-based assignment used the ST-21 complex, a large clonal group that is widespread in many food animals and other sources. There are many STs within the complex, each differing from the central genotype at between one and three of the seven alleles. Prediction used only these discordant alleles (i.e., those that differed from the control genotype) and assigned ST-21 complex isolates to host species on the basis of the frequency of each isolate’s discordant alleles in reference populations composed of non-ST-21 complex isolates from these host species. In Fig. 3 , each box indicates an ST and the shading represents the isolate source with this ST (light for chicken, dark for ruminant and mixed if from both sources). Considering only discordant alleles with strong evidence of differential distribution between chicken and ruminant reference populations, the lines joining STs to ST21 indicate whether it had
Figure 3. Prediction of origin by using only alleles for which substantial reference information is available. Light lines indicate alleles different from ST-21 present mainly in chickens in the reference population (i.e., an allele that would predict chicken origin); dark lines indicate alleles present mainly in bovids (i.e., predicts bovid origin). Light boxes indicate STs found only in chickens, dark boxes indicate STs found only in bovids, and boxes with light and dark shading indicate STs found in bovids and chickens. Figure from McCarthy et al. (2007).
200
ONETAL.
a discordant allele typical of chicken (light line, e.g., the line going to ST-368) or ruminant (dark line, e.g., the line going to ST-43). Isolates with two such discordant alleles are separated from ST-21 by two lines. The presence of an allele typical of a particular host species reference population predicts the origin of that isolate rather well. All four isolates having two alleles, both of which are suggestive of either chicken or ruminant origin, are indeed from the predicted source. In one case, two potentially informative alleles gave conflicting information, one suggesting bovid origin and the other chicken. Isolates with this ST came from both sources. Of the 24 STs with only one informative allele available, 18 are correctly assigned and only 4 incorrectly assigned with two STs isolated from both chicken and ruminant sources. Formal statistical assessment of the overall predictive power of discordant alleles within the ST-21 complex shows that they far outperform any effect likely by chance alone (P < 0.0001) (McCarthy et al., 2007). This result shows that ST-21 complex isolates coming from chickens are from lineages that have imported genetic material from other, clonally distinct, chicken lineages. This effect identifies host association not on the basis of ancestral (clonal) relationships but on a direct effect of different lineages having shared a host species niche, and by extension the opportunity to share DNA by lateral gene transfer. An extension of this is that isolates giving rise to this signal (as either donor or recipient of recombinant DNA) must, on average, represent lineages that preferentially inhabited chicken for a considerable period of time. This suggests that a clonal model of host association may also be effective, but for more highly resolved clonal groupings than are currently identified. The molecular epidemiology of C. coli is less well studied, but more recent work offers some insights into host association. Miller et al. (2006) applied MLST to the investigation of host association in C. coli in North America by using 484 isolates from cattle, swine, chickens, and turkeys. Substantial diversity was seen, with many STs observed only once. The large majority of STs (and isolates) clustered into a single clonal grouping and the large majority of the rest into a second grouping on the basis of a minimum spanning tree and eBURST clustering. The larger grouping included isolates from all four hosts, and the smaller included mainly turkey and chicken. Large clonal groups are thus not predictive of host in C. coli either. However, individual allele frequencies at all seven loci showed clear differences between hosts that were more striking than in C. jejuni, and there were some sublineages within the main clonal group that appeared to be host associ-
ated. However, a comparison of U.S. with United Kingdom isolates (Miller et al., 2006) revealed that overall host association of alleles was largely consistent, but with some marked exceptions such as tkt allele 35, which was mainly found in swine isolates in the United States but mainly in chicken isolates in the United Kingdom. We have assigned 259 independent United Kingdom C. coli poultry isolates (N. McCarthy, unpublished data) by using the U.S. data set as a reference population and by using previously described assignment methods (McCarthy et al., 2007). Fifty percent of the C. coli isolates were assigned to a chicken origin, 28% to cattle, 13% to swine, and 10% to turkey. Although this result is a lot better than the 25% correct expected by chance-and it shows that host association is geographically robust even between continents-further work is needed before attributing unknown isolates to likely source on the basis of MLST data. An investigation of 174 C. coli strains by AFLP profiling found human C. coli isolates in Denmark to more closely resemble those recovered from chicken, turkey, and duck, despite an overrepresentation of porcine isolates (Siemer et al., 2005). A subsequent MLST-based study of 160 Danish C. coli strains similarly found 38% of human isolates shared the same STs as those from chickens; just 9% of human strains were assigned to STs also found in pigs, and these STs were also detected among poultry isolates (Litrup et al., 2007). Two of seven cattle strains in this study belonged to STs found in humans. Preliminary work that used MLST to compare C. coli strains from human infections and chicken, cattle, and pigs in the United States also suggests that a greater proportion of poultry isolates belong to the same STs found in humans, but that some strains from pigs and especially cattle are also assigned to STs represented in human diarrheal isolates (W. G. Miller, unpublished data). Current results suggest that the attribution of C. jejuni and C. coli isolates from human cases to their source will be best undertaken by means of a phylogenetic approach that uses analysis based on allele frequencies, or that follows robust definition of hostassociated sublineages within the main clonal groups. Larger, more widely sampled reference data, explicit consideration of temporal and geographical aspects, and better analytical models may also be needed. Nonetheless, the consideration of many studies undertaken to date, together with other quantitative risk assessment data (Allen et al., 2007; Luber and Bartelt, 2007; Meldrum et al., 2006; Wingstrand et al., 2006), clearly identify chicken as a major source of human infection in industrialized countries. It is the relative contribution of all other sources to the bur-
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
den of disease that is in need of most clarity; however, almost all other foods of animal origin (notably other poultry and cattle) are credible sources of infection. MOLECULAR TYPING AND OUTBREAK INVESTIGATIONS An outbreak is defined here as two or more cases of disease that arise from the exposure to a common source of contamination. In contrast with other enteric pathogens such as Salmonella species and pathogenic Escherichia coli, human outbreaks of campylobacteriosis appear relatively uncommon, although the difficulty in linking human cases with, for example, contaminated meat from a single supplier but distributed nationally is a clear obstacle in identifying an outbreak. What is incontrovertible is the fact that the definitive identification of the origin of the outbreak by any typing method requires that the type of an isolate from the source matches those of the outbreak isolates. With Campylobacter outbreaks (irrespective of the species), isolates are rarely recovered from the source of infection. With an incubation period of 2 to 5 days, this is hardly surprising: by the time the patient has manifested diarrhea, the source has most likely been consumed or discarded. Even if samples remain, recovery of Campylobacter from the environment is far from trivial, and because multiple types may be harbored in a single specimen, the isolation of the same type that was recovered from the patient is not guaranteed. Ergo, definitive identification of the source is rarely attained. There are a number of outbreaks documented where the use of molecular typing has definitively identified the source. These are summarized in Table 1. The number is small, but the number of similar outbreaks where phenotypic typing methods have de-
finitively kesolved the outbreak source is little better (On, 2001), proving the challenges posed in epidemiological investigations of Campylobacter outbreaks. There are, however, numerous descriptions of outbreaks whereby the source was not identified but where the typing data did at least alert the authorities to a common cause. Some of these are notable in involving unusual settings (school: Fayos et al., 1993; ostriches: Stephens et al., 1998), strain phenotypes (C. jejuni Penner serotype 55; Harrington et al., 1999), or species (C. fetus, Rennie et al., 1994; C. hyointestinalis subsp. hyointestinalis, Salama et al., 1992). In all but the former study, where plasmid profiling, ribotyping and RAPD were used, PFGE was the method applied. OUTBREAKS AND EVOLUTION: RECONSIDERING THE PARADIGM Given the particularly complex nature of Campylobacter epidemiology from a wider perspective, together with recent advances in, and the increased use of, molecular typing, it may be considered pertinent to review the core question posed in such studies. Rather than ask if cases were caused by isolates from a common source, it may be more prudent to consider whether the isolates from these patients shared a common ancestor within the past 2 to 6 months, if postulating a shared source such as a recent cycle of poultry from a particular producer. In studying C. coli in cattle, U.S. researchers followed up an observation that a single MLST type was identified 47 times over a wide area and in isolates 4 years apart by additionally sequencing three highly variable loci: all 47 isolates were identical at each of the 10 loci (Miller et al., 2006). We have made similar unpublished observations for C. jejuni in chicken where
Table 1. Summary of outbreaks in which the source of infection was definitively identified by use of genotyping" Source
201
Year
Method(s)
Reference
Water Cow milk and feces Water
1996 1996 1998
Cattle Infected food handler
2000 2001
Nosocomial infection Milk Water
2003 2003 2004
REA REA PFGE, rihotyping AFLP PFGE PFGE FlaA, SVR, FlaA sequencing PFGE MLST PFGE
Lind et al., 1996 Lind et al., 1996 Engberg et al., 1998 Kokotovic and On, 1999 Lehner et al., 2000 Olsen et al., 2001 Fitzgerald et al., 2001 Llovo et al., 2003 Sails et al., 2003 Kuusi et al., 2004
"An isolate from the implicated source of infection must have been available for comparison with an outbreak strain and a matching genotype must have been obtained.
202
ONETAL.
isolates with identical sequence at MLST plus additional antigen-endoding gene segments were present in different areas of the United Kingdom several years apart (N. McCarthy and colleagues, unpublished data). Sharing an identical genotype does not therefore of itself indicate a common source, or even a high probability of a common source, and should not be taken to indicate that isolates are epidemiologically related. The main contribution of MLST to outbreak detection to date, including more extensive typing including additional loci over the usual seven, is therefore not in allowing outbreak detection but in making explicit in population genetic terms how very loosely related apparently identical isolates may be. It should be noted that MLST plus two or more sequenced antigen-encoding genes is highly discriminatory, and where high-throughout sequencing facilities and appropriate personnel and financial resources are readily available, this combined approach is thus a highly informative tool in outbreak investigations.
GENOTYPING OF NON-JEjUnrr, NON-COLI CAMPYLOBACTER SPECIES Given the scale of the human problems caused by C. jejuni and C. coli, comparatively little attention has been given to methods for typing other Campylobacter species, although many of these are strongly implicated as human pathogens. The use of improved isolation and detection methods are likely to demonstrate the need for improved typing methods for at least some of these species. The taxonomic diversity of Campylobacter has led to most workers in this area using broad-range whole-genomic methods that may be applied without a detailed knowledge of the genome, although recent advances in genome sequencing have shifted this paradigm somewhat. The major methods are summarized below. PFGE
C. concisus Matsheka et al. (2002) examined 53 C. concisus strains, principally from cases of diarrhea in South African children, by means of NotI-based macrorestriction profiling. Fifty-one distinct patterns were obtained, indicating both the high discriminatory potential of the method and the genetic diversity of the species. Indeed, other studies indicate C. concisus represents a taxonomic complex of several genomospecies (Aabenhus et al., 2005). Formaldehyde pre-
treatment was required to obtain DNA suitable for analysis, suggesting a high level of DNase activity, a trait not detected with standard phenotypic methods. C . fetus
Salama et al. (1992) applied PFGE analysis to strains of C. fetus subsp. fetus, C. fetus subsp. venerealis, and C. fetus subsp. venerealis bv. intermedius (i-e., glycine-tolerant variants of subsp. venerealis). The authors distinguished most (but not all) strains assigned to these taxa by differences in their estimated genome sizes (1.1, 1.3, and 1.5 Mb, respectively). Such estimates are likely to have been compromised by the technical limitations of PFGE because the genome sequence of C. fetus subsp. fetus strain 82-40 is 1.77 Mb (NCBI accession no. NC008599). Hum et al. (1997) used visual comparison of C. fetus PFGE profiles to differentiate between the two subspecies; although generally effective, one strain yielded a profile so atypical that it could not be readily assigned to either subspecies. A later study (On and Harrington, 2000) used numerical analysis of PFGE DNA profiles to assign each of 31 strains into one of two major clusters, of which the first contained 19 strains clearly identified as C. fetus subsp. venerealis. The second cluster comprised 12 strains, of which 10 were unambiguously identified as C. fetus subsp. fetus. The remaining two strains were identified as C. fetus subsp. venerealis by either phenotypic or PCR methods, but not both. At higher similarity levels, clusters containing isolates from each of two countries were identified, suggesting that certain clones predominate in certain geographical regions. Results from these studies have indicated that PFGE profiling is an effective means of genotyping C. fetus because most unrelated strains can be readily distinguished. The validity of the method for outbreak investigations of this species has been proven in outbreaks of diarrheal illness in a closed community (Rennie et al., 1994) and nosocomially acquired meningitis in a neonatal intensive care unit (Morooka et al., 1996). In both reports, outbreak strains were readily distinguished from unrelated strains. Ichiyama et al. (1998) used PFGE profiling to show that isolates of C. fetus subsp. fetus from feces, blood, and, in one case, a cellulitis lesion from three patients differed from each other, but that strains found in each of the different sites from a given patient were the same. A detailed discussion about typing of C. fetus can be found in chapter 11.
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
C. hyointestinalis Twenty-two different SmaI-derived PFGE profiles were obtained in a study of 28 strains representing C. hyointestinalis subsp. hyointestinalis, C. hyointestinalis subsp. lawsonii, and two distinct but unclassified groups of C. hyointestinalis-like strains (On and Vandamme, 1997). In addition, certain features of the profiles obtained were taxon specific and could thus be used for identification purposes. In this study, restriction enzyme SalI provided discrimination equal to SmaI for PFGE profiling C. hyointestinalis subsp. hyointestinalis but was significantly less effective for typing the other taxa because the DNA of most strains resisted digestion (On and Vandamme, 1997). The converse was observed by Salama et al. (1992) in an epidemiological investigation of a family outbreak of C. hyointestinalis (probably subsp. hyointestinalis) because strain DNA was digested only with SalI, not SmaI. Three of the five outbreak strains yielded the same PFGE profile, and raw milk was considered the most likely vehicle of infection (Salama et al., 1992). It was suggested that multiple strains of C. hyointestinalis may have been present in the milk, accounting for the different patterns obtained. This hypothesis is given credence because a diversity of genotypes has been found among isolates from a single animal herd: 10 SmaI patterns and 11 KpnI patterns were identified among 24 isolates from farmed reindeer (Hanninen et al., 2002). Several genotypes were seen in each herd where more than one isolate was available; the largest herd sample contained nine different genotypes (combined SmaI and KpnI patterns) among 13 isolates.
C. upsaliensis A range of restriction enzymes suitable for PFGE profiling this species were tested by Bourke et al. (1996), who examined 20 strains representing each of the seven C. upsaliensis HL serogroups described thus far. Of the enzymes examined (including SalI, SacII, SmaI, NruI, NarI, RsrII, and BssHII), only XhoI was considered to yield suitable profiles for all strains examined. There is some evidence for clonality as indicated by serotype association of similar macrorestriction profiles (Bourke et al., 1996; Moser et al., 2001), although these studies used different restriction enzymes to reach this conclusion. The more conserved SmaI profiling was used by Hald et al. (2004a) to indicate that domestic pet dogs could harbor the same strain for up to 21 months.
C. sputorum
A study of 18 C. sputorum bv. paraureolyticus isolates obtained from a single dairy cow herd over
203
a l-year period indicated that the same SmaI-defined PFGE genotype persisted in some cows for up to 12 months (On et al., 1999). Moreover, differences between banding patterns obtained could be accounted for by up to two mutational events in restriction sites, suggesting the herd infection was clonal in origin. Inclusion of an additional C. sputorum bv. paraureolyticus strain, and five strains of C. sputorum bv. sputorum (all epidemiologically unrelated), and further PFGE profiling that used enzymes SmaI, SalI, KpnI, and BamHI led the authors to propose that each of the C. sputorum biovars studied represented distinct clonal lines of this species. Most notably, KpnI sites were absent (or methylated) in all C. sputorum bv. paraureolyticus strains and SalI PFGE profiles of this biovar comprised just three bands. In contrast, C. sputorum bv. sputorum strains were readily differentiated by these enzymes. The results also illustrated that PFGE profiling was a useful method for genotyping C. sputorum because a total of 11 different profiles were noted among 26 strains examined.
AFLP The use of two restriction enzymes that cleave genomic DNA frequently in AFLP analysis greatly facilitates its use to a wider taxonomic range of species than most other genotyping methods. Three combinations of restriction enzymes have been used in AFLP profiling campylobacters, and each has, to a greater or lesser extent, been found successful in typing a wide range of species. The HindIII-HhaI-based method (Duim et al., 1999) is known to genotype Campylobacter species of animal origin, but to our knowledge, its validity in typing other species that are putative human pathogens such as C. concisus and C. rectus has not been tested (Duim et al., 2001). It has, however, been used to confirm canine-to-human transmission of C. upsaliensis (Wolfs et al., 2001) and applied to examine the taxonomy of C. lari, where four genomospecies were delineated; of these, two groups demonstrated resistance to nalidixic acid, but the two remaining groups lacking this trait were not clearly differentiated in their ability to produce urease (Duim et al., 2004). Conversely, the genetic diversity revealed in C. fetus led the authors to conclude that although it is a useful adjunctive method for discriminating C. fetus subsp. fetus from C. fetus subsp. venerealis, this protocol was likely to be limited in epidemiological studies (Wagenaar et al., 2001). The BglII-CspVI-based protocol (Kokotovic et al., 1999) typed most Campylobacter species satisfactorily (On and Harrington, 2000; Siemer et al., 2004), as well as the related Arcobacter spp. (On et al., 2003), but the limited number of fragments ob-
204
ONETAL.
tained in profiles for C. helveticus and C. upsaliensis (On and Harrington, 2000) led this group to develop an improved protocol that used BspDI and MfeI (Siemer et al., 2005). Subsequently, the method has been found to successfully genotype Carnpylobacter and Arcobucter species and proved useful in investigating an outbreak of C. fetus abortion in Danish sheep (S. L. W. On, unpublished data). This method was also used to investigate 62 C. concisus strains, 60 of which were of Danish origin, and assign them to one of four major clusters that represented distinct genomospecies with statistically significant differences in host immune status, indicative of differences in pathogenic potential (Aabenhus et al., 2005). Given the correspondence between AFLP and MLST generally (e.g., Miller et al., 2005), the high level of genetic diversity (62 distinct types among 62 strains) suggests a panmictic population structure for C. concisus.
RAPD As described above, the RAPD principle lends itself to genetic analysis of different species, although it has not been widely used for this purpose, arguably because of the difficulties in obtaining stable markers. Matsheka et al. (2006) described the use of a (GTG), oligonucleotide for RAPD analysis of C. concisus isolates, identifying unique profiles for 86% of all the strains typed. A system first applied to C. jejuni (Nielsen et al., 2000) identified unique patterns for all 37 C. concisus isolates examined by Engberg et al. (2005) by means of the same approach.
MLST The availability of whole-genome sequences has played an important role in the ease by which PCR primers can be designed for enabling MLST analysis. This comparative genomics approach was instrumental in the development of an approach that allows MLST of the closely related C. jejuni, C. coli, C. lari, C. helveticus, and C. upsaliensis taxa (Miller et al., 2005). Carnpylobacter insulaenigrae is a recently identified member of this aforementioned clade of species (Foster et al., 2004). Isolated exclusively from marine mammals (Foster et al., 2004; Stoddard, 2005), its closest relative is C. lari. The examination of 71 putative C. insulaenigrae isolates for 12 phenotypic traits identified eight profiles, suggesting that this species is diverse phenotypically and also that some strains of the species could be misidentified as C. lari (Stoddard et al., 2007). By use of MLST, phylogenetic analysis of the 40 C. insulaenigrae STs indicated that all 71 isolates were members of a highly related clade, distinct from related species, such as C.
lari and C. upsaliensis (Stoddard et al., 2007). Interestingly, failure to detect the citrate synthase gene gltA in any of the C. insulaenigrae or C. lari strains, supports further the close association between the two species. In C. lari strain Rh42100, absence ofgltA was confirmed after closure of the genome sequence. Therefore, absence of gltA (and presumably other TCA cycle genes) is likely a genomic feature shared by C. insulaenigrae and C. lari and probably defines a common evolutionary origin for the two species. Differentiation of C. fetus subspecies usually relies on biochemical tests, especially growth in 1% glycine, whereby C. fetus subsp. fetus (which is associated with sporadic abortion in cattle and sheep) grows on 1% glycine medium while C. fetus subsp. venereulis (which causes infectious infertility in said animals) does not. However, the glycine tolerance test may not be robust (On, 1996; On and Harrington, 2001). Additional genotypic tests, such as PCR assays, 16s rDNA sequencing, AFLP, and PFGE profiling, have also proved unable to unequivocally distinguish between the two c. fetus subspecies (On and Harrington, 2001; van Bergen et al., 2005). Therefore, a novel MLST method for C. fetus was developed. Despite the highly clonal nature of C. fetus and minimal sequence diversity between the two subspecies, the C. fetus MLST method defined unambiguously a C. fetus subsp. venerealis-specific ST, ST-4 (van Bergen et al., 2005). Thus, MLST can be used to accurately classify C. fetus strains at the subspecies level and provides an attractive alternative to standard phenotypic testing. As described with C. coli, individual alleles or allelic profiles correlate in some instances with host. However, host association of alleles and STs has not been demonstrated to date in the non-jejuni, non-coli Campylobacter (Miller et al., 2005; Stoddard et al., 2007; van Bergen et al., 2005). For species such as C. lari, lack of host association might be attributed to the small number and low host diversity of strains typed. However, MLST host association in some species can be difficult to identify because the species (or subspecies) in question is itself strongly host associated or host restricted. An example of the former is C. helveticus, which is primarily associated with cats and is rarely isolated from dogs (Moser et al., 2001; Shen et al., 2001; Stanley et al., 1992), and examples of the latter include C. insulaenigrae, isolated exclusively from marine mammals, especially pinnipeds, and C. fetus subsp. venerealis, isolated almost exclusively from the bovine genital tract (Thompson and Blaser, 2000). Van Bergen et al. (2005) demonstrated that Campylobacter fetus ST-4 was restricted to bovine isolates; however, because ST-4 was the C. fetus subsp. venerealis ST, this apparent host restriction of
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
ST-4 was attributed instead to restriction of C. fetus subsp. venerealis to cattle. Similarly, little association between ST and the geographical region of isolation has been identified (Miller et al., 2005; van Bergen et al., 2005). Phylogenetic analysis of C. upsaliensis concatenated allele sequences identified two major clades, cluster I and cluster I1 (Fig. 4). Cluster I comprises strains primarily from South Africa and Belgium, while cluster I1 comprises mainly strains from the United States. However, attribution of these clusters to geographic association of C. upsaliensis STs may be premature. It is noteworthy that almost all of the cluster I strains were isolated by the Cape Town protocol filtration method (Lastovica, 2006), which uses passive filtration and antibiotic-free media, while the majority of cluster I1 strains were isolated on agar amended with cefoperazone or cephalothin. Thus, separation of the C. upsaliensis STs into two distinct clusters may re-
205
flect isolation methodology rather than the geographical origin of the strains. Additional investigations that used isolation methodologies and strains from diverse geographical origins will be necessary to determine whether C. upsaliensis STs are geographically associated. Separation of C. upsaliensis STs on the basis of isolation methodology might indicate that for some species, MLST STs can associate with phenotype. Indeed, cluster I strains, such as RM3195 (ST-5), are sensitive to cefoperazone (Fouts et al., ZOOS), suggesting that clusters I and I1 represent cefoperazonesensitive and -resistant strains, respectively. Although the Cape Town isolation protocol (based on filtration onto a nonselective blood agar medium in an atmosphere containing 7 to 10% H,) would isolate both cluster I and cluster I1 strains, use of media amended with cefoperazone would isolate necessarily only cluster I1 strains. The existence of an antibiotic-
I
I1
0.002
Figure 4. Dendrogram of Campylobacter upsaliensis STs. Allele sequences from each of the 66 C. upsaliensis STs were concatenated in the order adk-aspA-atpA-glnA-glyA-pgi-tkt and aligned by ClustalX. The dendrogram was constructed by the neighbor-joining algorithm and the Kimura two-parameter distance estimation method. Bootstrap values of >75%, generated from 500 replicates, are shown at the nodes. The scale bar represents substitutions per site. Numeric labels represent STs. Geographic source of the strains was, where indicated: South Africa (solid circles), Belgium (open triangles), United States (crosses), or United Kingdom, France, and Sweden (solid triangles). STs representing Zic+ strains are boxed in gray.
206
ONETAL.
sensitive subpopulation indicates that subpopulations of some species may be undetected, depending on the isolation method used, especially if the isolation method uses antibiotic resistance. Such undetected subpopulations would have important implications for epidemiology. In addition to the separation of the C. upsaliensis STs by antibiotic resistance, the C. upsaliensis STs segregate also by the presence or absence of another marker, namely the phosphorylcholine locus licABCD (Fouts et al., 2005). In Haemophilus, Neisseria, and Streptococcus spp., the lic locus is involved in phosphorylcholine synthesis and the transfer of phosphorylcholine to the lipooligosaccharide or teichoic and lipoteichoic acids, and is considered to be a putative virulence locus (Fischer, 2000; Tong et al., 2000). Significantly, all cluster I strains contain the lic locus, whereas the lic locus is absent in all but two of the cluster I1 strains. Therefore, the C. upsaliensis STs segregate not only by antibiotic resistance but also by the presence or absence of the putative lic virulence locus. A similar segregation of STs according to surface structure-associated loci was identified in C. fetus (van Bergen et al., 200.5), where serotype A and serotype B C. fetus strains were associated with distinct STs. Lateral gene transfer within Campylobacter has important implications for the epidemiology and indeed phylogeny of the genus. Detection of lateral transfer events between two species could imply that they share, however transiently, the same environmental niche or host. Also, although the MLST genes themselves are housekeeping genes and are not considered to be related to pathogenicity, identification of lateral gene transfer between two species would suggest that other genes, involved putatively in host range, colonization, pathogenicity, antibiotic resistance, etc., might be exchanged between the two species. Lateral gene transfer events cannot be detected by fragment-based typing methods such as PFGE and AFLP. However, the events identified by MLST would represent only a small fraction of the total: the MLST gene set of seven genes is <0.5% of the total number of genes in an average Campylobacter genome. Detection of lateral transfer by MLST is facilitated by the interspecies nucleotide sequence diversity of Campylobacter, typically >10% for any pairwise combination of genes from two different Campylobacter species; thus, sequences from two different species can be identified unequivocally within an alignment. Detection of lateral transfer is facilitated also by the presence of four common loci, i.e., atpA(uncA), glnA, glyA, and tkt, within the MLST typing methods for the seven Campylobacter species (Miller et al., 2005; Stoddard et al., 2007; van Ber-
gen et al., 2005). Two of these loci, atpA(uncA) and glyA, can be extended still further into additional Campylobacter species as a result of a high level of sequence conservation across the genus; for example, atpA(uncA) MLST alleles have been amplified and sequenced from multiple strains of 14 Campylobacter species (data not shown). Given the fact that C. jejuni and C. coli are isolated often from the same host or food source, it is not surprising that lateral gene transfer events between C. jejuni and C. coli have been detected by MLST (Miller et al., 200.5; Schouls et al., 2003). Similarly, it is not surprising that lateral gene transfer between C. upsaliensis and C. helveticus, found together in dogs and occasionally in cats, was identified (Miller et al., 200.5). Other examples of putative lateral gene transfer events exist: C. lari allele p g m l l is identical to C. jejuni allele pgm1 10 and similar to C. jejuni alleles pgm100, pgm108, and pgmZ09 (Miller et al., 2005); the C. showae type strain glyA allele is identical to C. jejuni glyA27 (data not shown); and the C. lari alleles atpA36 and atpA49 are similar to atpA alleles from C. lanienae and C. concisus, respectively. Although it is possible that these lateral transfer events reflect simple strain mistyping, divergent MLST alleles represent generally only one or two alleles out of the seven-gene set. For example, the C. showae type strain atpA allele is only 76.5% identical at the nucleotide level to its nearest C. jejuni homolog but is similar to atpA alleles from the genetically related species C. concisus and C. curuus, with whom a common ecological niche is shared (the human oral cavity). Given recent work that has demonstrated species including C. lari (principally associated with the enteric tract of birds and mammals) in oral rinses from cats and dogs (Petersen et al., 2007), these observations may represent additional evidence of the oral route of infection for wider consideration in Campylobacter epidemiology. The expansion of typing methods into other Campylobacter species and the typing of additional Campylobacter strains by means of existing methodology should provide further insights into the role of lateral gene transfer in Carnpylobacter epidemiology and evolution.
CONCLUSION Molecular epidemiological studies of Campylobacter species have matured considerably since the early applications of plasmid profiling and restriction enzyme analysis in homogeneous electric fields. The more recent advances have facilitated the development of large-scale databases of normalized genotypic
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
or sequence data that can be applied at national or international scales to obtain a broad view of epidemiology of the organisms under study. The true value of these standardized methods, and the corresponding databases, will become increasingly evident-and relevant-when their content truly reflects the geographical, temporal, and ecological diversity that is exhibited by Campylobacter species. The role of continuous long-term surveillance at a national level particularly is self-evident here. The intelligent use of the methods currently available will ensure that the next advances in understanding the sources of infection and thereby their relative risk to humans may be made, and consequently will inform the scientific community how best to reduce the burden of disease. There are early signs that some methods are yielding an insight into the underlying biology of major clones widespread in human disease, offering new options for better understanding-and thus controlling-the burden of campylobacteriosis. For acute outbreaks of disease, there remains a need for rapid, discriminatory, but inexpensive methods to help identify the infections source so that appropriate countermeasures can be taken. Combined with the appropriate expertise in epidemiology and public health response, molecular typing can play a vital role in limiting the spread of infection. It is important that each of the stakeholders in the public health system-from nurses to environmental health officers and beyond-understands his or her role in outbreak detection and ensures that the typing tools used can be used to best efficacy. The role of species other than C. jejuni and C. coli as causes of human disease has been contested for some time. Nonetheless, the sheer volume of evidence that has been brought to bear in this area should now convince even the most hardened skeptic that some of the other Carnpylobacter species at least represent important agents of human and animal disease. Advances in genotyping have facilitated the identification and epidemiological typing of these species such that researchers will be able to rapidly address the central epidemiological questions that will arise when the importance of any given taxon becomes evident. With the integration of statistics, epidemiology, biology, and a risk-based paradigm, genotyping campylobacters is about to come of age. All that is certain is that the discoveries will be as thrilling, challenging, and surprising as the developments leading up to them have been in the 25 or so years that genetic typing methods have been applied. REFERENCES Aabenhus, R., S. L. W. On, B. L. Siemer, H. Permin, and L. P. Andersen. 2005. Delineation of Campylobacter concisus geno-
207
mospecies by amplified fragment length polymorphism analysis and correlation of results with clinical data. J. Clin. Microbiol. 435091-5096. Acik, M. N., and B. Cetinkaya. 2005. The heterogeneity of Campylobacter jejuni and Campylobacter coli strains isolated from healthy cattle. Lett. Appl. Microbiol. 41:397-403. Acik, M. N., and B. Cetinkaya. 2006. Random amplified polymorphic DNA analysis of Campylobacter jejuni and Campylobacter coli isolated from healthy cattle and sheep. J. Med. Microbiol. 55:331-334. Adhikari, B., J. H. Connolly, P. Madie, and P. R. Davies. 2004. Prevalence and clonal diversity of Campylobacter jejuni from dairy farms and urban sources. N. Z. Vet. J. 52:378-383. Allen, V. M., S. A. Bull, J. E. Corry, G. Domingue, F. Jorgensen, J- A. Frost, R. Whyte, A. Gonzalez, N. Elviss, and T. J. Humphrey. 2007. Campylobacter spp. contamination of chicken carcasses during processing in relation to flock colonisation. Int. J. Food Microbiol. 11354-61. Bates, C., K. L. Hiett, and N. J. Stern. 2004. Relationship of Campylobacter isolated from poultry and from darkling beetles in New Zealand. Avian Dis. 48:138-147. Bourke, B., P. M. Sherman, D. Woodward, H. Lior, and V. L. Chan. 1996. Pulsed-field gel electrophoresis indicates genotypic heterogeneity among Campylobacter upsaliensis strains. FEMS Microbiol. Lett. 14357-61. Butzler, J. P., P. Dekeyser, M. Detrain, and F. Dehaen. 1973. Related Vibrio in stools. J. Pediatr. 82:493-495. Champion, 0. L., E. L. Best, and J. A. Frost. 2002. Comparison of pulsed-field gel electrophoresis and amplified fragment length polymorphism techniques for investigating outbreaks of enteritis due to campylobacters. J. Clin. Microbiol. 40:2263-2265. Clark, C., L. Bryden, W. Cuff, P. Johnson, F. Jamieson, B. Ciebin, and G. Wang. 2005. Use of the Oxford multilocus sequence typing protocol and sequencing of the flagellin short variable region to characterize isolates from a large outbreak of waterborne Campylobacter sp. strains in Walkerton, Ontario, Canada. J. Clin. Microbiol. 43:2080-2091. Colles, F. M., K. Jones, R. M. Harding, and M. C. Maiden. 2003. Genetic diversity of Campylobacter jejuni isolates from farm animals and the farm environment. Appl. Environ. Microbiol. 69: 7409-7413. Corcoran, D., T. Quinn, L. Cotter, P. Whyte, and S. Fanning. 2006. Antimicrobial resistance profiling and flu-typing of Irish thermophillic Cumpylobucter spp. of human and poultry origin. Lett. Appl. Microbiol. 43560-565. Cornelius, A., C. Nicol, and J. Hudson. 2005. Campylobacterspp. in New Zealand raw sheep liver and human campylobacteriosis cases. Int. J. Food Microbiol. 99:99-105. de Boer, P., B. Duim, A. Rigter, J. van Der Plas, W. F. JacobsReitsma, and J. A. Wagenaar. 2000. Computer-assisted analysis and epidemiological value of genotyping methods for Campylobacter jejuni and Campylobacter coli. J. Clin. Microbiol. 38: 1940-1946. Devane, M. L., C. Nicol, A. Ball, J. D. Klena, P. Scholes, J. A. Hudson, M. G. Baker, B. J. Gilpin, N. Garrett, and M. G. Savill. 2005. The occurrence of Campylobacter subtypes in environmental reservoirs and potential transmission routes. J. Appl. Microbiol. 98:980-990. Dingle, K. E., F. M. Colles, D. Falush, and M. C. Maiden. 2005. Sequence typing and comparison of population biology of Campylobacter coli and Campylobacter jejuni. J. Clin. Microbiol. 43: 340-347. Dingle, K. E., F. M. Colles, R. Ure, J. A. Wagenaar, B. Duim, F. J. Bolton, A. J. Fox, D. R. Wareing, and M. C. Maiden. 2002. Molecular characterization of Campylobacter jejuni clones: a ba-
208
ONETAL.
sis for epidemiologic investigation. Emerg. Infect. Dis. 8:949955. Dingle, K. E., F. M. Colles, D. R. Wareing, R. Ure, A. J. Fox, F. E. Bolton, H. J. Bootsma, R. J. Willems, R Urwin, and M. C. Maiden. 2001a. Multilocus sequence typing system for Campylobacter jejuni. J. Clin. Microbiol. 39:14-23. Dingle, K. E., N. Van Den Braak, F. M. Colles, L. J. Price, D. L. Woodward, F. G. Rodgers, H. P. Endtz, A. Van Belkum, and M. C. Maiden. 2001b. Sequence typing confirms that Campylobacter jejuni strains associated with Guillain-Barre and MillerFisher syndromes are of diverse genetic lineage, serotype, and flagella type. 1. Clin. Microbiol. 39:3346-3349. Djordjevic, S. P., L. E. Unicomb, P. J. Adamson, L. Mickan, and R. Rios. 2007. Clonal complexes of Campylobacter jejuni identified by multilocus sequence typing are reliably predicted by restriction fragment length polymorphism analyses of the flaA gene.]. Clin. Microbiol. 45:102-108. Duim, B., P. A. Vandamme, A. Rigter, S. Laevens, J. R. Dijkstra, and J. A. Wagenaar. 2001. Differentiation of Campylobacter species by AFLP fingerprinting. Microbiology 147:2729-2737. Duim, B., J. A. Wagenaar, J. R. Dijkstra, J. Goris, H. P. Endtz, and P. A. Vandamme. 2004. Identification of distinct Campylobacter lari genogroups by amplified fragment length polymorphism and protein electrophoretic profiles. Appl. Environ. Microbiol. 70:18-24. Duim, B., T. M. Wassenaar, A. Rigter, and J. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with amplified fragment length polymorphism fingerprinting. Appl. Environ. Microbiol. 65:23692375. Ekdahl, K., B. Normann, and Y. Andersson. 2005. Could flies explain the elusive epidemiology of campylobacteriosis? BMC Infect. Dis. 5:ll. Engberg, J., D. D. Bang, R. Aabenhus, F. M. Aarestrup, V. Fussing, and P. Gerner-Smidt. 2005. Campylobacter concisus: an evaluation of certain phenotypic and genotypic characteristics. Clin. Microbiol. Infect. 11:288-295. Engberg, J., P. Gerner-Smidt, F. Scheutz, E. Msller Nielsen, S. L. W. On, and K. Mslbak. 1998. Water-borne Campylobacter jejuni infection in a Danish town-a 6-week continuous source outbreak. Clin. Microbiol. Infect. 4:648-656. Eyles, R. F., H. J. Brooks, C. R. Townsend, G. A. Burtenshaw, N. C. Heng, R W. Jack, and P. Weinstein. 2006. Comparison of Campylobacter jejuni PFGE and Penner subtypes in human infections and in water samples from the Taieri River catchment of New Zealand. J. Appl. Microbiol. 101:18-25. Fayos, A., R. J. Owen, J. Hernandez, C. Jones, and A. Lastovica. 1993. Molecular subtyping by genome and plasmid analysis of Campylobacter jejuni serogroups 01 and 0 2 (Penner) from sporadic and outbreak cases of human diarrhoea. Epidemiol. Infect. 111:415-427. Fischer, W. 2000. Phosphocholine of pneumococcal teichoic acids: role in bacterial physiology and pneumococcal infection. Res. Microbiol. 151:421-427. Fitzgerald, C. L. 0. Helsel, M. A. Nicholson, S. J. Olsen, D. L. Swerdlow, R. Flahart, J. Sexton, and P. I. Fields. 2001. Evaluation of methods for subtyping Campylobacter jejuni during an outbreak involving a food handler. J. Clin. Microbiol. 39:23862390. Foster, G., B. Holmes, A. G. Steigerwalt, P. A. Lawson, P. Thorne, D. E. Byrer, H. M. Ross, J. Xerry, P. M. Thompson, and M. D. Collins. 2004. Campylobacter insulaenigrae sp. nov., isolated from marine mammals. Int J Syst Evol Microbiol54:2369-2373. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. Dehoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sul-
livan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol 3x15. French, N., M. Barrigas, P. Brown, P. Ribiero, N. Williams, H. Leatherbarrow, R. Birtles, E. Bolton, P. Fearnhead, and A. Fox. 2005. Spatial epidemiology and natural population structure of Campylobacter jejuni colonizing a farmland ecosystem. Environ. Microbiol 7:1116-1126. Ge, B., W. Girard, S. Zhao, S. Friedman, S. A. Gaines, and J. Meng. 2006. Genotyping of Campylobacter spp. from retail meats by pulsed-field gel electrophoresis and ribotyping.]. Appl. Microbiol. 100: 175-184. Gibson, J., K. Sutherland, and R. Owen. 1994. Inhibition of DNAse activity in PFGE analysis of DNA from Campylobacter jejuni. Lett. Appl. Microbiol. 19:357-358. Gilpin, B., A. Cornelius, B. Robson, N. Boxall, A. Ferguson, C. Nicol, and T. Henderson. 2006. Application of pulsed-field gel electrophoresis to identify potential outbreaks of campylobacteriosis in New Zealand. 1. Clin. Microbiol. 44:406-412. Hald, B., K. Pedersen, M. Waino, J. C. Jorgensen, and M. Madsen. 2004a. Longitudinal study of the excretion patterns of thermophilic Campylobacter spp. in young pet dogs in Denmark. J. Clin. Microbiol. 42:2003-2012. Hald, B., H. Skovgard, D. D. Bang, K. Pedersen, J. Dybdahl, J. B. Jespersen, and M. Madsen. 2004b. Flies and Campylobacter infection of broiler flocks. Emerg. Infect. Dis. 10:1490-1492. Han, K., S. S. Jang, E. Choo, S. Heu, and S. Ryu. 2007. Prevalence, genetic diversity, and antibiotic resistance patterns of Campylobacter jejuni from retail raw chickens in Korea. Int. J. Food Microbiol. 11450-59. Hanninen, M. L., L. Sarelli, A. Sukura, S. L. On, C. S. Harrington, P. Matero, and V. Hirvela-Koski. 2002. Campylobacter hyointestinalis subsp. hyointestinalis, a common Campylobacter species in reindeer. J. Appl. Microbiol. 92:717-723. Harrington, C. S., F. M. Thomson-Carter, and P. E. Carter. 1997. Evidence for recombination in the flagellin locus of Campylobacter jejuni: implications for the flagellin gene typing scheme. J. Clin. Microbiol. 35:2386-2392. Harrington, C. S., F. M. Thomson-Carter, and P. E. Carter. 1999. Molecular epidemiological investigation of an outbreak of Campylobacter jejuni identifies a dominant clonal line within Scottish serotype HSSS populations. Epidemiol. Infect. 122:367-375. Hearnden, M., C. Skelly, R. Eyles, and P. Weinstein. 2003. The regionality of campylobacteriosis seasonality in New Zealand. Int. 1.Environ. Health Res. 13:337-348. Hernandez, J., A. Fayos, M. A. Ferrus, and R. J. Owen. 1995. Random amplified polymorphic DNA fingerprinting of Campylobacter jejuni and C. coli isolated from human faeces, seawater and poultry products. Res. Microbiol. 146:685-696. Hiett, K. L., B. S. Seal, and G. R. Siragusa. 2006. Campylobacter spp. subtype analysis using gel-based repetitive extragenic palindromic-PCR discriminates in parallel fashion to flaA short variable region DNA sequence analysis. J. Appl. Microbiol. 101: 1249-1258. Hopkins, K. L., M. Desai, J. A. Frost, J. Stanley, and J. M. Logan. 2004. Fluorescent amplified fragment length polymorphism genotyping of Campylobacter jejuni and Campylobacter coli strains and its relationship with host specificity, serotyping, and phage typing. J. Clin. Microbiol. 42:229-235. Hudson, J. A., C. Nicol, J. Wright, R. Whyte, and S. K. Hasell. 1999. Seasonal variation of Campylobacter types from human cases, veterinary cases, raw chicken, milk and water. J. Appl. Microbiol. 87:115-124.
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
Hum, S., K. Quinn, J. Brunner, and S. L. W. On. 1997. Evaluation of a PCR assay for identification and differentiation of Campylobacter fetus subspecies. Aust. Vet. J. 75:827-831. Ichiyama, S., S. Hirai, T. Minami, Y. Nishiyama, S. Shimizu, K. Shimokata, and M. Ohta. 1998. Campylobacter fetus subspecies fetus cellulitis associated with bacteremia in debilitated hosts. Clin. Infect. Dis. 27:252-255. Institute of Environmental Science and Research Ltd. 2007. Notifiable and Other Diseases in New Zealand Annual Report 2006. Institute of Environmental Science and Research Limited. Wellington, New Zealand. Iriarte, M. P., and R. J. Owen. 1996. Repetitive and arbitrary primer DNA sequences in PCR-mediated fingerprinting of outbreak and sporadic isolates of Campylobacter jejuni. FEMS lmmunol. Med. Microbiol. 15:17-22. Karenlampi, R., H. Rautelin, and M. L. Hanninen. 2007a. Evaluation of genetic markers and molecular typing methods for prediction of sources of Campylobacter jejuni and C. coli infections. Appl. Environ. Microbiol. 73:1683-1 685. Karenlampi, R., H. Rautelin, D. Schonberg-Norio, L. Paulin, and M. L. Hannineu. 2007b. Longitudinal study of Finnish Campylobacter jejuni and C. coli isolates from humans, using multilocus sequence typing, including comparison with epidemiological data and isolates from poultry and cattle. Appl. Environ. Microbiol. 73:148-155. Kokotovic, B., and S. L. On. 1999. High-resolution genomic fingerprinting of Campylobacter jejuni and Campylobacter coli by analysis of amplified fragment length polymorphisms. FEMS Microbiol. Lett. 173:77-84. Kuusi, M., P. Klemets, I. Miettinen, I. Laaksonen, H. Sarkkinen, M.-L. Hanninen, H. Rautelin, E. Kela, and J. P. Nuorti. 2004. An outbreak of gastroenteritis from a non-chlorinated community water supply. 1. Epidemiol. Community Health 58:273277. Lastovica, A. J. 2006. Emerging Campylobacter spp., the tip of the iceberg. Clin. Microbiol. News 28:49-56. Lehner, A. C. Schneck, G . Feierl, P. Pless, A. Deutz, E. Brandl, and M. Wagner. 2000. Epidemiologic application of pulsed-field gel electrophoresis to an outbreak of Campylobacter jejuni in an Austrian youth centre. Epidemiol. Infect. 125: 13-16. Lind, L., E. Sjogren, K. Melby, and B. Kaijser. 1996. DNA fingerprinting and serotyping of Campylobacter jejuni isolates from epidemic outbreaks. 1. Clin. Microbiol. 34:892-896. Lindstedt, B. A., E. Heir, T. Vardund, K. K. Melby, and G. Kapperud. 2000. Comparative fingerprinting analysis of Campylobacter jejuni subsp. jejuni strains by amplified-fragment length polymorphism genotyping. J. Clin. Microbiol. 38:3379-3387. Litrup, E., M. Torpdahl, and E. M. Nielsen. 2007. Multilocus sequence typing performed on Campylobacter coli isolates from humans, broilers, pigs and cattle originating in Denmark.!. Appl. Microbiol. 103:210-218. Llovo, J., E. Mateo, A. Muiioz, M. Urquijo, S. L. W. On, and A. Fernandez-Astorga. 2003. Molecular typing of Campylobacter jejuni isolates involved in a neonatal outbreak indicates nosocomial transmission. J. Clin. Microbiol. 41:3926-3928. Luber, P., and E. Bartelt. 2007. Enumeration of Campylobacter spp. on the surface and within chicken breast fillets. 1. Appl. Microbiol. 102:3 13-3 18. Maiden, M. C. 2006. Multilocus sequence typing of bacteria. Annu. Rev. Microbiol. 60561-588. Maiden, M. C., J. A. Bygraves, E. Feil, G. Morelli, J. E. Russell, R. Urwin, Q. Zhang, J. Zhou, K. Zurth, D. A. Caugant, I. M. Feavers, M. Achtman, and B. G. Spratt. 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.
209
Malik-Kale, P., B. H. Raphael, C. T. Parker, L. A. Joens, J. D. Klena, B. Quinones, A. M. Keech, and M. E. Konkel. 2007. Characterization of genetically matched isolates of Campylobacter jejuni reveals that mutations in genes involved in flagellar biosynthesis alter the organism’s virulence potential. Appl. Environ. Microbiol. 73 :3 123-3 136. Manning, G., C. G. Dowson, M. C. Bagnall, I. H. Ahmed, M. West, and D. G. Newell. 2003. Multilocus sequence typing for comparison of veterinary and human isolates of Campylobacter jejuni. Appl. Environ. Microbiol. 69:6370-6379. Matsheka, M. I., B. G. Elisha, A. L. Lastovica, and S. L. On. 2002. Genetic heterogeneity of Campylobacter concisus determined by pulsed field gel electrophoresis-based macrorestriction profiling. FEMS Microbiol. Lett. 211:17-22. Matsheka, M. I., A. J. Lastovica, H. Zappe, and B. G. Elisha. 2006. The use of (GTG)5 oligonucleotide as an RAPD primer to type Campylobacter concisus. Lett. Appl. Microbiol. 42:600605. McCarthy, N. D., F. M. Colles, K. E. Dingle, M. C. Bagnall, G. Manning, M. C. Maiden, and D. Falush. 2007. Host-associated genetic import in Campylobacter jejuni. Emerg. Infect. Dis. 13: 267-272. Meldrum, R. J., R. M. Smith, and I. G . Wilson. 2006. Three-year surveillance program examining the prevalence of Campylobacter and Salmonella in whole retail raw chicken. J. Food Prot. 69: 928-931. Mellmann, A., J. Mosters, E. Bartelt, P. Roggentin, A. Ammon, A. W. Friedrich, H. Karch, and D. Harmsen. 2004. Sequencebased typing of flaB is a more stable screening tool than typing of @A for monitoring of Campylobacter populations. 1. Clin. Microbiol. 42:4840-4842. Miller, G., G. M. Dunn, A. Smith-Palmer, I. D. Ogden, and N. J. Strachan. 2004. Human campylobacteriosis in Scotland: seasonality, regional trends and bursts of infection. Epidemiol. Infect. 132~585-593. Miller, W. G., M. D. Englen, S. Kathariou, I. V. Wesley, G . Wang, L. Pittenger-Alley, R. M. Siletz, W. Muraoka, P. J. FedorkaCray, and R. E. Mandrell. 2006. Identification of host-associated alleles by multilocus sequence typing of Campylobacter coli strains from food animals. Microbiology 152:245-255. Miller, W. G., S . L. On, G. Wang, S. Fontanoz, A. J. Lastovica, and R. E. Mandrell. 2005. Extended multilocus sequence typing system for Campylobacter coli, C. lari, C. upsaliensis, and C. helveticus. 1. Clin. Microbiol. 43:23 15-2329. Morooka, T., A. Umeda, M. Fujita, H. Matano, S. Fujimoto, K. Yukitake, K. Amako, and T. Oda. 1996. Epidemiologic application of pulsed-field gel electrophoresis to an outbreak of Campylobacter fetus meningitis in a neonatal intensive care unit. Scandinavian 1.Infect. Dis. 28:269-270. Moser, I., P. Lentzsch, B. Rieksneuwoehner, P. Schwerk, and L. H. Wieler. 2002. High resolution genotyping of Campylobacter jejuni strains by macrorestriction analysis with XhoI and polymerase chain reaction targeting enterobacterial repetitive intergenic consensus sequences: can we predict the zoonotic potential of strains? Epidemiol. Infect. 129:435-443. Moser, I., B. Rieksneuwohner, P. Lentzsch, P. Schwerk, and L. H. Wieler. 2001. Genomic heterogeneity and 0-antigenic diversity of Campylobacter upsaliensis and Campylobacter helveticus strains isolated from dogs and cats in Germany. J. Clin. Microbiol. 39:2548-2557. Nachamkin, I., H. Ung, and C. M. Patton. 1996. Analysis of HL and 0 serotypes of Campylobacter strains by the flagellin gene typing system. J. Clin. Microbiol. 34:277-28 1. Nielsen, E. M., J. Engberg, V. Fussing, L. Petersen, C. H. Brogren, and S. L. On. 2000. Evaluation of phenotypic and genotypic
210
ONETAL.
methods for subtyping Campylobacter jejuni isolates from humans, poultry, and cattle. J. Clin. Microbiol. 38:3800-3810. Nylen, G., F. Dunstan, S. R. Palmer, Y. Andersson, F. Bager, J. Cowden, G. Feierl, Y. Galloway, G. Kapperud, F. Megraud, K. Molbak, L. R. Petersen, and P. Ruutu. 2002. The seasonal distribution of Campylobacter infection in nine European countries and New Zealand. Epidemiol. Infect. 128:383-390. Olsen, S. J., G. R. Hansen, L. Bartlett, C. Fitzgerald, A. Sonder, R. Manjrekar, T. Riggs, J. Kim, R. Flahart, G. Pezzino, and D. L. Swerdlow. 2001. An outbreak of Campylobacter jejuni infections associated with food handler contamination: the use of pulsed-field gel electrophoresis. J. Infect. Dis. 183:164-167. On, S. L. W. 1996. Identification methods for campylobacters, helicobacters, and related organisms. Clin Microbiol. Rev. 9:405422. On, S. L. W. 2005. Taxonomy, phylogeny, and methods for the identification of Campylobacter species, p. 13-42. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Press, Norfolk, United Kingdom. On, S. L. W. 2001. What have we learned about Campylobacter coliljejuni from bacteriological typing studies? p. 53-57 In World Health Organization Department of Communicable Disease Surveillance and Response, The Increasing Incidence of Human Campylobacteriosis. Report and Proceedings of a WHO Consultation of Experts, 21 to 25 November 2000, Copenhagen, Denmark. World Health Organization, Geneva. On, S. L. W., H. I. Atabay, and J. E. Corry. 1999. Clonality of Campylobacter sputorum bv. paraureolyticus determined by macrorestriction profiling and biotyping, and evidence for longterm persistent infection in cattle. Epidemiol. Infect. 122:175182. On, S. L. W., N. Dorrell, L. Petersen, D. D. Bang, S. Morris, S. J. Forsythe, and B. W. Wren. 2006. Numerical analysis of DNA microarray data of Campylobacter jejuni strains correlated with survival, cytolethal distending toxin and haemolysin analyses. lnt. J. Med. Microbiol. 296:353-363. On, S. L. W., and C. S. Harrington. 2000. Identification of taxonomic and epidemiological relationships among Campylobacter species by numerical analysis of AFLP profiles. FEMS Microbiol. Lett. 193:161-169. On, S. L. W., and C. S. Harrington. 2001. Evaluation of numerical analysis of PFGE-DNA profiles for differentiating Campylobacter fetus subspecies by comparison with phenotypic, PCR and 16s rDNA sequencing methods. J. Appl. Microbiol. 90:285-293. On, S. L. W., C. S. Harrington, and H. I. Atabay. 2003. Differentiation of Arcobacter species by numerical analysis of AFLP profiles and description of a novel Arcobacter from pig abortions and turkey faeces. J. Appl. Microbiol. 95:1096-1105. On, S. L. W., E. M. Nielsen, J. Engberg, and M. Madsen. 1998. Validity of Sma-defined genotypes of Campylobacter jejuni examined by Sull, Kpnl,and BamHl polymorphisms: evidence of identical clones infecting humans, poultry, and cattle. Epidemiol. Infect. 120:231-237. On, S. L. W., and P. A. Vandamme. 1997. Identification and epidemiological typing of Campylobacter hyointestinalis subspeciies by phenotypic and genotypic methods and description of novel subgroups. Syst. Appl. Microbiol. 20:238-247. Ono, K., T. Kurazono, H. Niwa, and K. Itoh. 2003. Comparison of three methods for epidemiological typing of Campylobacter jejuni and C. coli. Curr. Microbiol 47:364-371. O’Reilly, L. C., T. J. Inglis, and L. Unicomb. 2006. Australian multicentre comparison of subtyping methods for the investigation of Campylobacter infection. Epidemiol. Infect. 134:768779. Petersen, L., E. M. Nielsen, J. Engberg, S. L. On, and H. H. Dietz. 2001. Comparison of genotypes and serotypes of Campylobacter
jejuni isolated from Danish wild mammals and birds and from broiler flocks and humans. Appl. Environ. Microbiol. 67:31153121. Petersen, R. F., C. S. Harrington, H. E. Kortegaard, and S. L. W. On. 2007. A PCR-DGGE method for detection and identification of Campylobacter, Helicobacter, Arcobacter and related Epsilobacteria and its application to saliva samples from humans and domestic pets. J. Appl. Microbiol. 103:2601-2615. Price, E. P., F. Huygens, and P. M. Giffard. 2006a. Fingerprinting of Campylobacter jejuni by using resolution-optimized binary gene targets derived from comparative genome hybridization studies. Appl. Environ. Microbiol. 72: 7793-78 03. Price, E. P., H. Smith, F. Huygens, and P. M. Giffard. 2007. Highresolution DNA melt curve analysis of the clustered, regularly interspaced short-palindromic-repeat locus of Campylobacter jejuni. Appl. Environ. Microbiol. 73:343 1-3436. Price, E. P., V. Thiruvenkataswamy, L. Mickan, L. Unicomb, R. E. Rios, F. Huygens, and P. M. Giffard. 2006b. Genotyping of Campylobacter jejuni using seven single-nucleotide polymorphisms in combination with flaA short variable region sequencing. J. Med. Microbiol. 55:1061-1070. Quinones, B., C. T. Parker, J. M. Janda, Jr., W. G. Miller, and R. E. Mandrell. 2007. Detection and genotyping of Arcobacter and Campylobacter isolates from retail chicken samples by use of DNA oligonucleotide arrays. Appl. Environ. Microbiol. 73: 3645-3655. Rennie, R. P., D. Strong, D. E. Taylor, S. M. Salama, C. Davidson, and H. Tabor. 1994. Campylobacter fetus diarrhea in a Hutterite colony: epidemiological observations and typing of the causative organism. 1. Clin. Microbiol. 32:721-724. Ribot, E. M., C. Fitzgerald, K. Kubota, B. Swaminathan, and T. J. Barrett. 2001. Rapid pulsed-field gel electrophoresis protocol for subtyping of Campylobacter jejuni. J. Clin. Microbiol. 39: 1889-1894. Sails, A. D., B. Swaminathan, and P. I. Fields. 2003. Utility of multilocus sequence typing as an epidemiological tool for investigation of outbreaks of gastroenteritis caused by Campylobacter jejuni. J. Clin. Microbiol. 41:4733-4739. Salama, S. M., M. M. Garcia, and D. E. Taylor. 1992. Differentiation of the subspecies of Campylobacter fetus by genomic sizing. Int. J. Syst. Bacteriol. 42:446-450. Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin.Microbiol. 41:15-26. Shen, Z., Y. Feng, F. E. Dewhirst, and J. G . Fox. 2001. Coinfection of enteric Helicobacter spp. and Campylobacter spp. in cats. J. Clin. Microbiol39:2 166-2 172. Sheppard, S., N. D. McCarthy, M. C. Maiden, C. Little, R. Elson, R. J. Owen, and R. J. Meldrum. 2007. Multilocus sequence typing and analysis of Campylobacter isolates from the current retail poultry surveys (Project code B15011). Study report to the Foods Standards Agency Siemer, B. L., C. S. Harrington, E. M. Nielsen, B. Borck, N. L. Nielsen, J. Engberg, and S. L. On. 2004. Genetic relatedness among Campylobacter jejuni serotyped isolates of diverse origin as determined by numerical analysis of amplified fragment length polymorphism (AFLP) profiles. J. Appl. Microbiol. 96:795-802. Siemer, B. L., E. M. Nielsen, and S. L. On. 2005. Identification and molecular epidemiology of Campylobacter coli isolates from human gastroenteritis, food, and animal sources by amplified fragment length polymorphism analysis and Penner serotyping. Appl. Environ. Microbiol. 71:1953-1958.
CHAPTER 10
MOLECULAR EPIDEMIOLOGY OF CAMPYLOBACTER SPECIES
Skirrow, M. B. 1977. Campylobacter enteritis: a “new” disease. Br. Med.]. 2:9-11. Sopwith, W., A. Birtles, M. Matthews, A. Fox, S. Gee, M. Painter, M. Regan, Q. Syed, and E. Bolton. 2006. Campylobacter jejuni multilocus sequence types in humans, northwest England, 20032004. Emerg. Infect. Dis. 12:1500-1507. Stanley, J., A. P. Burnens, D. Linton, S. L. On, M. Costas, and R. J. Owen. 1992. Campylobacter helveticus sp. nov., a new thermophilic species from domestic animals: characterization, and cloning of a species-specific DNA probe. J Gen Microbiol 138(Pt. 11):2293-2303. Stanley, K. N., J. S. Wallace, J. E. Currie, P. J. Diggle, and K. Jones. 1998a. The seasonal variation of thermophilic campylobacters in beef cattle, dairy cattle and calves. J. Appl. Microbiol. 85:472-480. Stanley, K. N., J. S. Wallace, J. E. Currie, P. J. Diggle, and K. Jones. 1998b. Seasonal variation of thermophilic campylobacters in lambs at slaughter. J. Appl. Microbiol. 84:1111-1116. Stephens, C. P., S. L. On, and J. A. Gibson. 1998. An outbreak of infectious hepatitis in commercially reared ostriches associated with Campylobacter coli and Campylobacter jejuni. Vet. Microbiol. 61:183-1 90. Stoddard, R. A. 2005. Salmonella and Campylobacter spp. in Northern Elephant Seals, California. Emerg. Infect. Dis. 11: 1967-1969. Stoddard, R. A., W. G. Miller, J. E. Foley, J. Lawrence, F. M. Gulland, P. A. Conrad, and B. A. Byrne. 2007. Campylobacter insulaenigrae Isolates from Northern Elephant Seals (Mirounga angustirostris) in California. Appl Environ Microbiol 73 :17291735. Thompson, S. A., and M. J. Blaser. 2000. Pathogenesis of Campylobacter fetus infections, p. 321-347. In I. Nachamkin and
211
M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Tong, H. H., L. E. Blue, M. A. James, Y. P. Chen, and T. F. DeMaria. 2000. Evaluation of phase variation of nontypeable Haemophilus influenzae lipooligosaccharide during nasopharyngeal colonization and development of otitis media in the chinchilla model. Infect. Immun. 68:4593-4597. van Bergen, M. A., K. E. Dingle, M. C. Maiden, D. G. Newell, L. van der Graaf-Van Bloois, J. P. van Putten, and J. A. Wagenaar. 2005. Clonal nature of Campylobacter fetus as defined by multilocus sequence typing.]. Clin. Microbiol 435888-5898. Wagenaar, J. A., M. A. van Bergen, D. G. Newell, R. GrogonoThomas, and B. Duim. 2001. Comparative study using amplified fragment length polymorphism fingerprinting, PCR genotyping, and phenotyping to differentiate Campylobacter fetus strains isolated from animals. J. Clin. Microbiol. 39:2283-2286. Wassenaar, T. M., S. L. W. On, and R Meinersmann. 2000. Genotyping and the consequences of genetic instability, p. 369380 In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Weijtens, M. J., J. van der Plas, P. G. Bijker, H. A. Urlings, D. Koster, J. G. van Logtestijn, and J. H. Huis in’t Veld. 1997. The transmission of Campylobacter in piggeries; an epidemiological study.]. Appl. Microbiol. 83:693-698. Wingstrand, A., J. Neimann, J. Engberg, E. M. Nielsen, P. GernerSmidt, H. C. Wegener, and K. Molbak. 2006. Fresh chicken as main risk factor for campylobacteriosis, Denmark. Emerg. Infect. Dis. 12~280-285. Wolfs, T. F., B. Duim, S. P. Geelen, A. Rigter, F. Thomson-Carter, A. Fleer, and J. A. Wagenaar. 2001. Neonatal sepsis by Campylobacter jejuni: genetically proven transmission from a household puppy. Clin. Infect. Dis. 32:E97-E99.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 11
Isolation, Identification, Subspecies Differentiation, and Typing of Campylobacter fetus MARCEL
A. P.
VAN
BERGEN,Jos P. M.
VAN PUTTEN, KATEE. AND JAAPA. WAGENAAR
DINGLE,MARTIN J. BLASER,
as well as humans (Dennis, 1967; Harvey and Greenwood, 1985; Meinershagen et al., 1965; Tu et al., 2004; Watson et al., 1967). Carrier states of other animals probably play a role in transmission. C. fetus subsp. fetus is generally orally transmitted by the ingestion of food or water contaminated by feces, aborted fetuses, membranes, or discharges (Garcia et al., 1983).
Campylobacter fetus is recognized as an important veterinary and human pathogen (Vandamme, 2000). It is associated with genital infections in livestock, resulting in different clinical presentations including abortion and infertility. The species is divided into two subspecies: C. fetus subsp. venerealis and C. fetus subsp. fetus. C. fetus subsp. venerealis is the causative agent of bovine genital campylobacteriosis, also known as bovine venereal campylobacteriosis, bovine vibriosis, infectious infertility, and enzootic sterility (Dekeyser, 1984). It is mainly restricted to the bovine genital tract (Table 1);however, in rare cases, C. fetus subsp. venerealis has been isolated from other hosts, including humans (Eaglesome and Garcia, 1992; Salama et al., 1992). C. fetus subsp. venerealis is mainly transmitted venereally both naturally or by artificial insemination with contaminated semen (Stegenga and Terpstra, 1949). The main differences between C. fetus subsp. fetus and C. fetus subsp. venerealis are the clinical manifestation and the host and niche where the bacteria reside. C. fetus subsp. fetus causes abortion in cattle and sheep, and sporadic infections in humans. Sheep may serve as the primary reservoir of C. fetus subsp. fetus (Garcia et al., 1983), whereas other animals probably serve as secondary (less important) reservoirs (Dennis, 1967; Meinershagen et al., 1965; Watson et al., 1967) (Table 1). C. fetus subsp. fetus can be isolated from different animal species, including cattle, sheep, goats, pigs, horses, fowl, and reptiles,
HISTORIC CLASSIFICATION AND NOMENCLATURE OF C. FETUS The bacterial species now known as Campylobacter fetus was first identified in 1913 and named Vibrio fetus. It was known as a causative agent of abortions in cattle and sheep (McFadyean and Stockman, 1913). Florent (1959) observed that venereally transmitted enzootic infertility in cattle was caused by a V. fetus variant that he designated V. fetus venerealis, whereas sporadic abortions in cattle were caused by another variant of intestinal origin named V. fetus intestinalis. He described several biochemical tests to differentiate the two variants. V. fetus venerealis did not produce H,S and would not grow in a semisolid medium containing 1%glycine, whereas V. fetus intestinalis produced H,S and was resistant to glycine, as earlier described by Lecce (1958; Florent, 1959). On the basis of these clinical and biochemical observations, the species was split into two subspecies.
Marcel A. P. van Bergen and Jaap A. Wagenaar Division of Infectious Diseases, Animal Sciences Group, OIE Reference Laboratory Jos for Campylobacteriosis and WHO Collaboration Centre for Campylobacter, P.O. Box 65, 8200 AB Lelystad, The Netherlands. P. M. van Putten and Jaap A. Wagenaar Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, OIE Reference Laboratory for Campylobacteriosis and WHO Collaboration Centre for Campylobacter, P.O. Box 80.165,3508 Kate E. Dingle Department of Microbiology, John Radcliffe Hospital, Nuffield Department of TD Utrecht, The Netherlands. Clinical Sciences, Oxford University, OX3 9DU Oxford, United Kingdom. Martin J. Blaser Departments of Medicine and Microbiology, New York University School of Medicine, New York NY 10016.
213
214
VAN
BERGEN ET AL.
Table 1. Overview of the subspecies and their host preference and their clinical importance“ _ _ _ _ _ _ ~ ~
Species
Disease and/or commensal status
Principal host@)
C. fetus subsp. venerealis
Cattle
C. fetus subsp. fetus
Sheep Cattle Humans Cattle/sheep
Bovine genital campylobacteriosis: infertility, early embryonic death, and occasional abortion Ovine genital campylobacteriosis: outbreaks of abortion Sporadic abortions Sporadic infections mainly in immunocompromised people Commensal in the intestinal tract
“From Quinn et al. (1994).
In 1973, the results of a taxonomic study of a newly defined genus, “Campylobacter,y’ were published by Vtron and Chatelain (1973). They proposed that Vibrio fetus intestinalis should be renamed Campylobacter fetus subspecies fetus, and Vibrio fetus venerealis should be renamed Campylobacter fetus subspecies venerealis. An H,S-positive intermediate of C. fetus subsp. venerealis was designated Campylobacter fetus subspecies venerealis biovar intermedius (Vtron and Chatelain, 1973). An overview of the historic classifications and nomenclature of C. fetus is shown in Table 2. The current recognized nomenclature for Campylobacteraceae was officially accepted in 1980, published in the approved list of bacterial names (Skerman et al., 1980), and in accordance with that used in Bergey’s Manual of Systematic Bacteriology (Smibert, 1984). Both the nomenclature and diagnostic methods for C. fetus have changed over time, complicating the interpretation of previous studies and their comparison with more recent work (Penner, 1988). C. fetus is the earliest Campylobacter identified, and a considerable number of studies, especially those concerning pathogenesis, were performed in the 1950s to 1970s. At that time, identification, subspecies differentiation, and typing results were based on phenotypic methods that are limited in both reliability and interpretation.
rooka et al., 1996; Rennie et al., 1994). Reliable confirmation of the diagnosis of a C. fetus infection requires the detection of the bacterium by direct or indirect methods. This can be done by culturing, serological testing, immunofluorescence testing, or using a DNA-based assay. Recently, a DNA-based detection assay (TaqMan PCR) was described for the detection of C. fetus subsp. venerealis in crude clinical extracts (McMillen et al., 2006). However, this PCR uses a target sequence, which may give false results as discussed in the subspecies differentiation section. Serological and immunofluorescence testing detection methods also have been described (Ardrey et al., 1972; Bokkenheuser, 1972; Brooks et al., 2004; Camper0 et al., 2005; Hum et al., 1991; OIE, 2004), but culture has significant advantages because isolates can then be used in further tests and characterized more fully in terms of their antimicrobial susceptibility, subspecies differentiation, and epidemiological typing. Therefore, the first step is usually isolation, followed by identification and possibly subspecies differentiation, typing, or both. However, if different methods are used, they may in some cases give contradictory results (van Bergen et al., 2005a, 2005b). Recently, several additional molecular methods have been developed and assessed in comparison with the generally accepted methods. Practical considerations on the use of the best available methods for C. fetus diagnosis and epidemiology are described below.
C. FETUS DIAGNOSTICS Isolation, identification, subspecies differentiation, and to a lesser extent typing are important in C. fetus diagnostics. Screening programs aiming to control bovine genital campylobacteriosis and limit the economic impact are performed by veterinary health services. Veterinary clinical laboratories use these methods for confirmation of the isolated causative agent from clinical samples (e.g., aborted fetuses). Among medical clinical microbiology laboratories, reliable diagnosis of C. fetus infections is important to ensure correct antimicrobial treatment and to trace outbreaks and sources of infection (Mo-
ISOLATION OF C. FETUS
C. fetus subsp. venerealis detection is performed on specimens submitted to the laboratory for veterinary health screening programs or on samples from suspected clinical cases. For bovine samples, sampling methods are described in the chapter entitled “Bovine Genital Campylobacteriosis” in the Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (OIE, 2004). Because C. fetus subsp. fetus is not a microorganism for which notification is required, no pre-
CHAPTER 11
C. FETUS ISOLATION. IDENTIFICATION. AND TYPING
215
scribed methods for its isolation are described. However, in practice, the same methods as for C. fetus subsp. venerealis are used to culture C. fetus subsp. fetus from the genital tract and from aborted fetuses. The latter are the most common type of veterinary clinical sample, whereas human specimens usually include, among others, blood or fecal samples. Collection of veterinary or clinical samples, treatments applied (such as the use of filters), and the choice of transport and isolation media all affect C. fetus isolation rates (Atabay and Corry, 1998; Garcia et al., 1984; Hum et al., 1994; Lander, 1990; Monke et al., 2002). These factors are critically important because successful isolation of this fastidious bacterium requires not only its survival and propagation, but also inhibition of contaminating biota. When confirmation of the diagnosis of C. fetus infection by bacteriological culture is sought, their failure to survive suboptimal conditions should be borne in mind. If the sample is unlikely to be contaminated with other microorganisms, nonselective blood agar plates can be used; however, selective plates (without cephalosporins) must be used for fecal samples. Unlike thermophilic campylobacters, C. fetus can grow at 25 and 37"C, and although thermotolerant strains have been described, they do not generally survive at 42°C (Barrett et al., 1988; Firehammer and Berg, 1965; Harvey and Greenwood, 1983; Leaper and Owen, 1981; Roop et al., 1984). Recent evaluation of a global diverse C. fetus collection showed that the majority of C. fetus subsp. fetus isolates (from recent decades) do grow at 42°C (C. Fitzgerald and M. van Bergen, unpublished data), whereas C. fetus subsp. venerealis isolates generally do not grow at 42°C. Visible growth is usually obtained in 48 to 72 h, longer than most of the other campylobacters (Smibert, 1984).
s
s
s IDENTIFICATION AND CONFIRMATION OF C. FETUS Phenotypic Identification
d
ci
The accepted phenotypic identification scheme for C. fetus is described in Bergey's Manual of Systematic Bacteriology (Smibert, 1984). Traditional biochemical assays (Holt et al., 1994) have the advantages of simplicity and discriminatory power. Conventional assays are also relatively cheap because they do not require sophisticated hardware. However, reliance on this approach can lead to misidentifications (Hum et al., 1997; On, 1996). A 3-day incubation period under standardized conditions of constant atmosphere and temperature is also required
216
VAN
BERGEN ET AL.
for a reliable outcome. Two immunological approaches have been described that can be used. Immunofluorescence testing (OIE, 2004) can be a valuable tool, but it requires antiserum that is difficult to obtain commercially and that may cross-react with other Campylobacter spp. Antigen capture immunological assays that use C. fetus-specific monoclonal or polyclonal antibodies are useful for identification (Ardrey et al., 1972; Brooks et al., 2004; Camper0 et al., 2005; OIE, 2004). Molecular Identification The C. fetus genome has a mean G+C content of 33 to 36% (Charlier et al., 1974; Owen, 1983; Viron and Chatelain, 1973), varying from 1.1 to 1.6 Mb in size (Salama et al., 1992), although a larger genome of 2.0 Mb has also been described (Fujita and Amako, 1994). A variety of methods for the molecular identification of C. fetus have been developed that monitor either cellular protein or fatty acid composition, or that detect C. fetus by DNA amplification, sometimes together with DNA sequencing. Catalase-positive C. fetus strains (Ferguson and Lambe, 1984) can be identified by whole-cell protein analysis by polyacrylamide gel electrophoresis. This provides reproducible banding patterns of soluble protein extracts. Numerical analysis of the protein profiles then allows a C. fetus cluster, separate from other Campylobacter species to be identified (Vandamme et al., 1990). Most Campylobacter species, including C. fetus and C. jejuni, can be distinguished by analysis of the cellular fatty acid composition (Blaser et al., 1980; Vandamme, 2000). This parameter is stable, provided the analysis is performed on isolates cultured under standardized conditions. DNA-based methods include 16s and 23s rDNA sequencing (Moureau et al., 1989), amplified fragmentlength polymorphism (AFLP) (Duim et al., 2001; Wagenaar et al., 2001), and multilocus sequence typing (MLST) (van Bergen et al., 2005a). The latter allowed clear differentiation of C. fetus from other species and showed that it is more homogenous (Fig. 1). The majority of the above methods are sensitive and can be used to identify C. fetus. However, several are laborious, time-consuming, and relatively expensive and therefore are not well suited to routine diagnostic laboratories. PCR is another DNA-based method that is simple, robust, and quick. A number of PCR-based assays with various target genes including the 16s ribosomal gene-specific region (Linton et al., 1996; Oyarzabal et al., 1997) and others (Casadtmont et al., 1998; Hum et al., 1997) have been developed. PCR probe-based methods are also avail-
able to detect C. fetus (Blom et al., 1995; Logan et al., 2001; Wesley et al., 1991). The combination of PCR and nonradioactive microplate hybridization is convenient for rapid C. fetus identification (Casadtmont et al., 2000), and also a PCR-enzyme linked immunosorbent assay has been described (Metherell et al., 1999). PCR followed by restriction enzyme analysis has differentiated C. fetus from other organisms, but some cross-reaction with other campylobacters was observed (Eaglesome et al., 1995). Other restriction enzyme analysis methods revealed many similarities in the patterns obtained for all strains of C. fetus (Collins and ROSS,1984). The PCR described by Hum et al. (1997) currently appears the most evaluated and is therefore the most useful method for C. fetus identification (Muller et al., 2003; Newel1 et al., 2000a; On and Harrington, 2001; Schulze et al., 2006; van Bergen et al., 2005a, 2 0 0 5 ~Wagenaar ; et al., 2001; Willoughby et al., 2005). SUBSPECIES DIFFERENTIATION OF C . FETUS Identification of C. fetus is relatively straightforward, but reliable subspecies differentiation is much more difficult. The two subspecies are closely related genetically and hence are indistinguishable by DNADNA hybridization (Harvey and Greenwood, 1983; Roop et al., 1984; Thompson et al., 1988). Even so, several phenotypic, animal, and genotypic assays have been evaluated for their ability to differentiate C. fetus subsp. venerealis and C. fetus subsp. fetus. Phenotypic Subspecies Differentiation The currently recommended World Organization for Animal Health assay to discriminate the two subspecies is the 1% glycine tolerance test (C. fetus subsp. venerealis negative; C. fetus subsp. fetus positive) (Lecce, 1958). However, doubts exist as to the stability of this marker because tolerance may be phage mediated or may occur by point mutation (Chang and Ogg, 1970, 1971). The biochemical H,S test on sensitive medium is also described as able to differentiate C. fetus subsp. fetus (positive) from C. fetus subsp. venerealis (negative). C. fetus subsp. venerealis biovar intermedius strains were so called because of their positive reaction in this assay (Vtron and Chatelain, 1973). Other methods have been described but did not provide definitive discrimination. Parameters assayed include growth rate, size, and type of the spiral, aerotolerance, and growth at 42°C (Dekeyser, 1984; George et al., 1978; Karmali et al., 1981). A list of seven apparent key phenotypic tests that derived from probability matrices (including selenite reduc-
C. FETUS ISOLATION. IDENTIFICATION.AND TYPING
CHAPTER 11
217
C-76 C-116 ~-118--
. lati
h-7 h-6 h-2 v h-8
C.jejuni
I I
0.02
U-6
C.upsaliensis
h
f-3 f-2 f- 1
C.fetus Figure 1. Lack of genetic diversity in the C. fetus glnA locus compared with other Curnpylobucter species. Radial neighborjoining tree comparing 477 nt sequences at the glnA locus of C. fetus with those of five other Cumpylobucterspecies. Alleles are numbered as in the MLST databases at http: //pubmlst.org/and prefixed by a letter to indicate the species. Bootstrap values are shown. Reprinted from the Journal of Clinical Microbiology (van Bergen et al., 2005a).
tion, growth at 42"C, 1% glycine tolerance, susceptibility for metronidazole and cefoperazone, basic fuchsin, and KMnO,) proved indicative for subspecies differentiation but revealed different classifications per test (On and Harrington, 2001). Susceptibility to cephalosporins-cefoxitin and cefamandole (Varga et al., 1990), cefaperazone, and cephalothin (Harvey and Greenwood, 1983) and penicillin (Dekeyser, 1984)-has been described as discriminatory, but again, contradictory results have been noted (Tremblay and Gaudreau, 1998; Tremblay et al.,
2003), suggesting that antibiotic resistance is not a useful marker for subspecies. Subspecies Differentiation Using Animal Models Animal models for C. fetus infections are used primarily in pathogenesis and vaccine studies (Bryner et al., 1978; Grogono-Thomas et al., 2003; Schurig et al., 1973). However, one study has investigated the use of animal models for subspecies differentiation, on the premise that C. fetus subsp. venerealis and C.
218
VAN
BERGEN ET AL.
fetus subsp. fetus have differences in their preference for host and niche (Bryner et al., 1971). Isolates of C. fetus subsp. fetus should be able to persist in the gallbladder and intestines, whereas C. fetus subsp. venerealis, because of its niche restriction, should not be able to colonize these sites. The method was evaluated for cattle, sheep, rabbits, guinea pigs, and mice, and the latter were concluded to be most suitable. The use of animals for subspecies differentiation purposes is not practical and has never been publicly evaluated. One study found the assay not to be reproducible (van Bergen, 2005). Molecular Subspecies Differentiation Only a few molecular tools to discriminate C. fetus subsp. fetus and C. fetus subsp. venerealis have been developed to date. Analysis of a specific AFLP region provided the first reliable molecular method to differentiate C. fetus subsp. fetus, C. fetus subsp. venerealis, and C. fetus subsp. venerealis biovar intermedius (Fig. 2) (van Bergen et al., 2005a; Wagenaar et al., 2001). This was confirmed by MLST data, which demonstrate an evolutionary relationship among strains consistent with the divergence of the two subspecies (van Bergen et al., 2005a). However, MLST does not separate C. fetus subsp. venerealis biovar intermedius from C. fetus subsp. venerealis (van Bergen et al., 2005a) because isolates of both groups are ST-4 (or, exceptionally, ST-7 or ST-12) (Fig. 3). The use of pulsed-field gel electrophoresis (PFGE) for subspecies differentiation has been described; the resulting banding patterns enable genome sizing and provide data for numerical analysis. However, neither is optimal for routine application. Although numerical analysis of PFGE DNA profiles is effective, it requires sophisticated software and a database (On and Harrington, 2001). Genome sizing is less useful because inconclusive results are often obtained (On and Harrington, 2001; Salama et al., 1992; Vargas et al., 2003). Routine diagnostic laboratories would be best served by a simple PCR for differentiation of the subspecies. Several PCRs have been developed and claimed to be subspecies specific (Hum et al., 1997; Tu et al., 2005; van Bergen et al., 2 0 0 5 ~ ;Wang et al., 2002). The multiplex PCR described by Hum et al. (1997) is currently the most cited. It amplifies a C. fetus-specific DNA fragment approximately 200 bp smaller than the 960 bp described in the original publication (Muller et al., 2003; Wagenaar et al., 2001; Willoughby et al., 2005), as well as a 142-bp fragment thought to be specific to C. fetus subsp. venerealis. C. fetus subsp. venerealis biovar intermedius strains were not evaluated in the study of Hum and
colleagues, but results presented by van Bergen et al. (2005a) show that C. fetus subsp. venerealis biovar intermedius isolates (identified by AFLP) are either C. fetus subsp. fetus or C. fetus subsp. venerealis in the PCR of Hum and colleagues (van Bergen et al., 2005a). Comparison of AFLP and MLST data confirm that this PCR can give false-positive and falsenegative reactions (van Bergen et al., 2005a), as previously shown in comparisons with other assays (Hum et al., 1997; Willoughby et al., 2005). The PCR described by Wang et al. (2002) was thought to amplify only a C. fetus subsp. fetusspecific product, but these results were obtained for a small number of strains. More recent evaluation that used a larger number of strains yielded both false-positive and false-negative reactions (van Bergen et al., 2 0 0 5 ~Willoughby ; et al., 2005). The PCR that gives data most consistent (140 of 140) with C. fetus subsp. venerealis subspecies as defined by AFLP was described by van Bergen et al. (2005~).However, C. fetus subsp. venerealis biovar intermedius, as defined by AFLP, is not identified by this PCR. The random amplification of polymorphic DNA (RAPD) PCRs described by Tu et al. (2005) were evaluated with a small number of C. fetus subsp. venerealis strains, so further evaluation with a larger group of strains is required. In conclusion, AFLP and MLST are currently the best methods for definitive subspecies differentiation. AFLP and a combination of MLST with the PCR as described by van Bergen et al. allow differentiation of C. fetus subsp. venerealis and C. fetus subsp. venerealis biovar intermedius. For laboratories without molecular facilities, unfortunately, there are no alternatives to the unreliable 1% glycine tolerance test. Subspeciation Justified? A subspecies is by definition a group of organisms that are genetically closely related but that diverge in phenotype (Staley and Krieg, 1984; Wayne et al., 1987). The current classification of C. fetus into the two subspecies-C. fetus subsp. venerealis and C. fetus subsp. fetus-meets this definition and was therefore officially accepted in 1980 (Skerman et al., 1980). Ever since, doubts have been cast whether subspeciation is justified (Harvey and Greenwood, 1983; Penner, 1988), especially because the outcome of the 1% glycine tolerance assay appears unreliable as well as sensitive to transduction and point mutation (Chang and Ogg, 1970, 1971). Therefore, research has focused on identification of additional differences between the subspecies. Berg et al. (1971), who combined serotyping and biochemical results, reported strain characteristics that crossed the classical subspecies barriers. Currently, the clearest di-
CHAPTER 11
C. FETUS ISOLATION. IDENTIFICATION. AND TYPING
Figure 2. Dendrogram showing the AFLP banding patterns of 69 C. fetus strains. Cluster analysis was based on the similarity levels among bands in region 841 to 879 of the banding patterns (arrow). The different clusters of C. fetus subsp. fetus and C. fetus subsp. venerealis are indicated. The percentage of genetic similarity among banding patterns is shown. Reprinted from the Journal of Clinical Microbiology (Wagenaar et al., 2001).
219
220
VAN
BERGEN ET AL.
Figure 3. Congruence of sequence type (ST), sup type, and subspecies. Analysis of concatenated MLST sequences by split decomposition. The three STs identified in association with C. fetus subsp. venereulis are indicated by the dotted line, the remainder being C. fetus subsp. fetus. The correlation among ST and sup type is marked; solid line, supA; dashed line, supB. A radial neighbor-joining tree constructed by using the same data (inset) is shown to indicate the corresponding treelike phylogenies obtained by both methods. This supports the idea that C. fetus subsp. fetus and C. fetus subsp. venereulis evolved recently, with little (if any) evidence of recombination, and that genetic changes have accumulated by the vertical transmission of point mutations, yielding a clonal structure to the population. Reprinted from the ]ournu1 of Clinical Microbiology (van Bergen et al., 2005a).
chotomy in C. fetus is the division into the LPS/sap type A or B (Dworkin et al., 1995; Moran et al., 1994; Perez-Perez, 1986; Tu et al., 2001). C. fetus subsp. fetus can be either type A or By whereas C. fetus subsp. venerealis is always type A, confirming the crossing of the classical subspecies barrier. The fact that the serotype A C. fetus subsp. fetus and C. fetus subsp. venerealis strains are on the same branch suggests that their differentiation occurred after the type A-type B split (Tu et al., 2001) (Fig. 4). Because different results have been obtained with MLST and AFLP compared with the currently applied classical phenotypic and PCR analysis, it seems appropriate to reconsider the basis of C. fetus subspecies classification.
C. fetus subsp. venerealis Biovar Intermedius: to Which Subspecies Do These Strains Belong? A group of strains that reacted positively in the “sensitive” H,S test, and whose characteristics resembled those of C. fetus subsp. venerealis, was designated as C. fetus subsp. venerealis biovar intermedius (VCron and Chatelain, 1973). Taxa below the rank of subspecies are not covered by the rules of the Bacteriological Code (1990 revision, http: //www. bacterio.cict.fr/) but were introduced to determine infrasubspecific subdivisions. Biovar intermedius is therefore not recorded in the approved list of bacterial names (Skerman et al., 1980), but it is mentioned in Bergey’s Munuul (Smibert, 1984). It is not
CHAPTER 11
Ancient
C. FETUS ISOLATION. IDENTIFICATION. AND TYPING
221
I I
I I
i Ancestral Mammal type A C. fetus (Mammal 16s rRNA sequence)
- - I I.. - I - -
1-----------I
Modern
presented as a separate biovar in most of the C. fetus studies described to date, probably because of the lack of obvious reference strains, its questionable clinical relevance, and its incomplete original description. Yet it has been reported that C. fetus subsp. venerealis intermedius strains cover a substantial number of a diverse collection of C. fetus strains (van Bergen et al., 2005a). The existence of biovar intermedius was initially proposed on the basis of its identical properties to C. fetus subsp. venerealis, except for a single biochemical trait (positive in the H,S assay) (Vtron and Chatelain, 1973), and it was distinguished from C. fetus subsp. fetus only on the basis of a different result in the glycine tolerance test. As a result of the discrepancies described for the biological niche of this biovar (genital and/or intestinal) (Bryner et al., 1964; Elazhary, 1968; Garcia et al., 1983; VCron and Chatelain, 1973) and the ambiguous interpretation of the subspecies differentiation assays, it has been difficult to decide whether these strains are more closely related to C. fetus subsp. venerealis or C. fetus subsp. fetus. Vkron and Chatelain (1973) initially described this group as a different subspecies (subsp. intermedius Elazhary), but they
h
TypeB
eventually opposed the biovar because of the few available differentiating tests and the ecological niche shared with C. fetus subsp. venerealis. Recent AFLP data (van Bergen et al., 2005a) support the definition of the biovar intermedius as a separate subspecies, according to the definition of a subspecies (genetically homolog, with a phenotypic difference). However, MLST data showed that these strains were the same sequence type as C. fetus subsp. venerealis isolates and therefore not allow to differentiate these strains from C. fetus subsp. venerealis (van Bergen et al., 2005a). Better evidence that biovar intermedius is or is not a separate group of strains should be obtained by detailed genetic analyses. Differentiation of C. fetus Isolates of Classical and Reptile Origin
A group of C. fetus isolates of reptile origin has been identified containing isolates that are genetically distinct from the classical C. fetus isolates. A reptile strain isolated from a turtle was suggested to be the cause of an acute diarrheal illness in a child (Harvey and Greenwood, 1985); however, no causative agent
222
VAN
BERGEN ET AL.
was isolated from the child. The first confirmed human infection with a C. fetus strain with markers of reptile origin was described by Tu et al. in 2004. To date, several reptile and human isolates have been identified (C. Fitzgerald, personal communication). To further investigate the genetic diversity among C. fetus strains of different origins, Tu et al. (2005) performed multiple genetic analyses, including RAPD, DNA-DNA hybridization, and sequence analysis of 16s rRNA. All RAPD primers used distinguished C. fetus of classical and reptile origin. DNA-DNA hybridization demonstrated substantial genomic homology differences between strains of mammal and reptile origin. Sequence analysis of 16s rRNA also showed that the reptile strains form a distinct phylotype between mammalian C. fetus and Cumpylobucter hyointestinalis (Tu et al., 2001). Isolates from mammals and reptiles formed two distinct tight phylogenetic clusters that were well separated (Tu et al., 2001). In total, these data suggest that C. fetus subsp. fetus strains of reptile and mammal origin have genetic divergence more extensive than between the two subspecies and than between the type A and type B strains (Fig. 4) (Tu et al., 2005). Currently, a C. fetus MLST scheme is being applied to the previously described reptile isolates and human isolates with reptile markers, as well as to newly identified organisms.
TYPING OF C. FETUS Although identification and subspecies differentiation of C. fetus is performed routinely, typing is generally used only for epidemiological purposes, outbreak identification being the most important. Typing by Phenotypic Methods Phenotyping of C. fetus has been confined to serotyping. Several serotyping schemes have been described on the basis of different isolated antigens (e.g., heat-labile antigen, heat-stable antigen, 0 antigen, and whole-cell antigen). However, the discriminatory power of serotyping is low because of the small number of serotypes. The currently accepted serotyping scheme is based on the heat-stable 0 antigens and consists of only two serotypes, A and B (Berg et al., 1971; Morgan, 1959). The different serotypes correlate with the surface array protein (sap) types. The sup typing scheme is confined to two types, A and B (Yang et al., 1992), and is thus considered to not be very informative. Phage typing has been proposed as a practical supplemental typing
method for C. fetus (Bryner et al., 1973) but has never been developed. The same is true for flagellin typing. Although C. fetus has polar flagella at one or both ends, no flagellin-based typing (flu typing) system has been developed. Typing by Molecular Methods The introduction of molecular methods has allowed new approaches for typing of C. fetus because they allow further discrimination within the subspecies. MLST results showed that C. fetus has a more clonal structure compared with most other species within the genus Campylobucter. This clonality limits differentiation within the C. fetus species. Nevertheless, a PCR-based method focusing on the sap genes was developed, which classified strains into type A or B (Tu et al., 2001). This classification correlated with MLST types (Fig. 3) (van Bergen et al., 2005a). One study that used restriction enzyme analysis showed patterns that could be divided into a limited number of different groups (Collins and ROSS,1984). Typing of strains on the basis of the diversity of the molecular weight of the Sap protein also has been reported, but because frequent shifts in the molecular weight of Sap occur during in vitro and in vivo passage, this typing scheme is not reliable (Fujimoto et al., 1991; Wang et al., 1993). Southern blot hybridization with a sup probe appeared useful for typing and involved the use of a supA probe in combination with PFGE (Fujita et al., 1995b), the whole supA gene as a probe (Denes et al., 1997), and the use of a supB2 probe (Casademont et al., 1998). PFGE was found to be suitable in several epidemiological studies and is currently the best available method for veterinary and human applications (Fujita et al., 1995a; Mannering et al., 2004; Morooka et al., 1996; Newel1 et al., 2000b; Rennie et al., 1994; Vargas et al., 2003). Other molecular methods such as AFLP (Wagenaar et al., 2001) and MLST (van Bergen et al., 2005a) appeared less useful as a result of limited discriminatory power compared with PFGE. For future epidemiological studies, c. fetus microarrays may prove to be a good alternative to detect differences between strains. However, this awaits completion of C. fetus whole-genome sequences. Currently, the genome of one human C. fetus subsp. fetus isolate (82-40) has been deposited in GenBank.
CONCLUSIONS For species identification, the specific PCR described by Hum et al. (1997) is the best for routine use. For subspecies differentiation, AFLP is the best
CHAPTER 11
C. FETUS ISOLATION. IDENTIFICATION. AND TYPING
because C. fetus subsp. fetus, C. fetus subsp. venerealis, and C. fetus subsp. venerealis intermedius strains can be distinguished (van Bergen et al., 2005a). Alternatively, MLST in combination with a C. fetus subsp. venerealis-specific PCR (van Bergen et al., 2005c) may be applied. Currently, the best method for typing purposes is PFGE. REFERENCES Ardrey, W. B., P. Armstrong, W. A. Meinershagen, and F. W. Frank. 1972. Diagnosis of ovine vibriosis and enzootic abortion of ewes by immunofluorescence technique. Am. 1.Vet. Res. 33: 2535-2538. Atabay, H. I., and J. E. L. Corry. 1998. The isolation and prevalence of campylobacters from dairy cattle using a variety of methods. J. Appl. Microbiol. 84:733-740. Barrett, T. J., C. Patton, and G. K. Morris. 1988. Differentiation of Campylobacter species using phenotypic characterization. Microbiology 19:96-102. Berg, R. L., J. W. Jutila, and B. D. Firehammer. 1971. A revised classification of Vibrio fetus. Am. J. Vet. Res. 32:ll-22. Blaser, M. J., C. W. Moss, and R. E. Weaver. 1980. Cellular fatty acid composition of Campylobacter fetus. 1.Clin. Microbiol. 11: 448-45 1. Blom, K., C. M. Patton, M. A. Nicholson, and B. Swaminathan. 1995. Identification of Campylobacter fetus by PCR DNA probe method. J. Clin. Microbiol. 33:1360-1362. Bokkenheuser, V. 1972. Vibrio fetus infection in man: a serological test. Infect. Immun. 5:222-226. Brooks, B. W., J. Devenish, C. L. Lutze-Wallace, D. Milnes, R H. Robertson, and G. Berlie-Surujballi. 2004. Evaluation of a monoclonal antibody-based enzyme-linked immunosorbent assay for detection of Campylobacter fetus in bovine preputial washing and vaginal mucus samples. Vet. Microbiol. 103:77-84. Bryner, J. H., P. C. Estes, J. W. Foley, and P. A. O’Berry. 1971. Infectivity of three Vibrio fetus biotypes for gallbladder and intestines of cattle, sheep, rabbits, guinea pigs, and mice. Am. ]. Vet. Res. 32:465-470. Bryner, J. H., J. W. Foley, W. T. Hubbert, and P. J. Matthews. 1978. Pregnant guinea pig model for testing efficacy of Campylobacter fetus vaccines. Am. ]. Vet. Res. 39:119-121. Bryner, J. H., A. H. Frank, and P. A. O’Berry. 1962. Dissociation studies of Vibrios from the bovine genital tract. Am. /. Vet. Res. 23 :32-4 1. Bryner, J. H., P. A. O’Berry, and A. H. Frank. 1964. Vibrio infection of the digestive organs of cattle. Am. ]. Vet. Res. 25: 1048-1050. Bryner, J. H., A. E. Ritchie, G. D. Booth, and J. W. Foley. 1973. Lytic activity of vibrio phages on strains of Vibrio fetus isolated from man and animals. Appl. Microbiol. 26:404-409. Campero, C. M., M. L. Anderson, R. L. Walker, P. C. Blanchard, L. Barbano, P. Chiu, A. Martinez, G. Combessies, J. C. Bardon, and J. Cordeviola. 2005. Immunohistochemical identification of Campylobacter fetus in natural cases of bovine and ovine abortions. J. Vet. Med. B Infect. Dis. Vet. Public Health 52:138-141. Casadkmont, I., C. Bizet, D. Chevrier, and J.-L. Guesdon. 2000. Rapid detection of Campylobacter fetus by polymerase chain reaction combined with non-radioactive hybridization using a oligonucleotide covalently bound to microwells. Mol. Cell. Probes 14:233-240. Casadtmont, I., D. Chevrier, and J. Guesdon. 1998. Cloning of a sapB homologue (sapB2) encoding a putative 112-kDa Campylobacter fetus S-layer protein and its use for identification and
223
molecular genotyping. FEMS Immunol. Med. Microbiol. 21: 269-281. Chang, W., and J. E. Ogg. 1971. Transduction and mutation to glycine tolerance in Vibrio fetus. Am. J. Vet. Res. 32:649-653. Chang, W., and J. E. Ogg. 1970. Transduction in Vibrio fetus. Am. 1.Vet. Res. 31:919-924. Charlier, G., P. Dekeyser, A. Florent, R. Strobbe, and J. De Ley. 1974. DNA base composition and biochemical characters of Campylobacter strains. Antonie Van Leeuwwenhoek 40: 145151. Collins, D. M., and D. E. Ross. 1984. Restriction endonuclease analysis of Campylobacter strains with particular reference to Campylobacter fetus ss. fetus. 1.Med. Microbiol. 18:117-124. Dekeyser, J. 1984. Bovine genital campylobacteriosis, p. 181-191. In J. P. Butzler (ed.), Campylobacter Infection in Man and Animals. CRC Press, Boca Raton, FL. Denes, A. S., C. L. Lutze-Wallace, M. L. Cormier, and M. M. Garcia. 1997. DNA fingerprinting of Campylobacter fetus using cloned constructs of ribosomal RNA and surface array protein genes. Vet. Microbiol. 54:185-193. Dennis, S . M. 1967. The possible role of the raven in the transmission of ovine vibriosis. Aust. Vet. /. 43:45-48. Duim, B., P. A. Vandamme, A. Rigter, S . Laevens, J. R. Dijkstra, and J. A. Wagenaar. 2001. Differentiation of Carnpylobacter species by AFLP fingerprinting. Microbiology 147:2729-2737. Dworkin, J., M. K. Tummuru, and M. J. Blaser. 1995. Segmental conservation of sapA sequences in type B Campylobacter fetus cells. J. Biol. Chem. 270:15093-15101. Eaglesome, M. D., and M. M. Garcia. 1992. Microbial agents associated with bovine genital tract infections and semen. I. Brucella abortus, Leptospira, Campylobacter fetus, and Tritrichomonas foetus. Vet. Bull. 62:743-775. Eaglesome, M. D., M. I. Sampath, and M. M. Garcia. 1995. A detection assay for Campylobacter fetus in bovine semen by restriction analysis of PCR amplified DNA. Vet. Res. Commun. 19: 253-263. Elazhary, M. 1968. An assay of isolation and identification of some animal vibrios and of elucidation of their pathological significance. Med. Veeartschool Rijksuniv Gent 12: 1-80. Ferguson, D. A., and D. W. Lambe. 1984. Differentiation of Campylobacter species by protein banding patterns in polyacrylamide slab gels. 1.Clin. Microbiol. 20:453-460. Firehammer, B. D., and R. L. Berg. 1965. The use of temperature tolerance in the identification of Vibrio fetus. Am. J. Vet. Res. 26~995-997. Florent, A. 1959. Les deux vibriosis ginitales: la vibriose due Q V. fetus venerealis et la vibriose d’origine intestinale due Q V. fetus intestinalis. Med. Veeartschool Rijksuniv. Gent 3: 1-60. Fujimoto, S., A. Takade, K. Amako, and M. J. Blaser. 1991. Correlation between molecular size of the surface array protein and morphology and antigenicity of the Campylobacter fetus S layer. Infect. Immun. 59:2017-2022. Fujita, M., and K. Amako. 1994. Localization of the sapA gene on a physical map of Campylobacter fetus chromosomal DNA. Arch. Microbiol. 162:375-80. Fujita, M., S . Fujimoto, T. Morooka, and K. Amako. 1995a. Analysis of strains of Campylobacter fetus by pulsed-field gel electrophoresis.]. Clin. Microbiol. 33:1676-1678. Fujita, M., T. Morooka, S. Fujimoto, T. Moriya, and K. Amako. 1995b. Southern blotting analyses of strains of Campylobacter fetus using the conserved region of sapA. Arch. Microbiol. 164: 444-447. Garcia, M. M., M. D. Eaglesome, and C. Rigby. 1983. Campylobacters important in veterinary medicine. Vet. Bull. 53:793818.
224
VAN
BERGEN ET AL.
Garcia, M. M., R. B. Stewart, and G. M Ruckerbauer. 1984. Quantitative evaluation of a transport-enrichment medium for Campylobacter fetus. Vet. Rec. 115:434-436. George, H. A., P. S. Hoffman, R. M. Smibert, and N. R Krieg. 1978. Improved media for growth and aerotolerance of Campylobacter fetus. J. Clin. Microbiol. 8:36-41. Grogono-Thomas, M. J. B., M. Ahmadi, and D. G. Newell. 2003. Role of S-layer protein antigenic diversity in the immune responses of sheep experimentally challenged with Campylobacter fetus subsp. fetus. Infect. Immun. 71:147-154. Harvey, S., and J. R. Greenwood. 1985. Isolation of Campylobacter fetus from a pet turtle. J. Clin. Microbiol. 21:260-261. Harvey, S., and J. R. Greenwood. 1983. Relationships among catalase-positive Campylobacters determined by deoxyribonucleic acid-deoxyribonucleic acid hybridization. lnt. J. Syst. Bacteriol. 33:275-284. Holt, J. G., N. R Krieg, P. H. A. Sneath, J. T. Staley, and S. T. Williams (ed.). 1994. Bergey’s Manual of Determinative Bacteriology, 9th ed., vol. 1. Lippincott, Williams and Wilkins, Baltimore, MD. Hum, S., J. Brunner, A. McInnes, G. Mendoza, and J. Stephens. 1994. Evaluation of cultural methods and selective media for the isolation of Campylobacter fetus subsp venerealis from cattle. Aust. Vet. J. 71:184-186. Hum, S., K. Quinn, J. Brunner, and S. L. W. On. 1997. Evaluation of a PCR assay for identification and differentiation of Campylobacter fetus subspecies. Aust. Vet. J. 759327-83 1. Hum, S., L. R. Stephens, and C. Quinn. 1991. Diagnosis by ELISA of bovine abortion due to Campylobacter fetus. Aust. Vet. J. 68: 272-5. Karmali, M. A., A. K. Allen, and P. C. Fleming. 1981. Differentiation of catalase-positive campylobacters with special reference to morphology. Int. J. Syst. Bacteriol. 31:64-71. Lander, K. P. 1990. The application of a transport and enrichment medium to the diagnosis of Campylobacter fetus infections in bulls. Br. Vet. J. 146:334-340. Leaper, S., and R. J. Owen. 1981. Identification of catalaseproducing Campylobacter species based on biochemical characteristics and on cellular fatty acid composition. Curr. Microbiol. 6:31. Lecce, J. G. 1958. Some biochemical characteristics of Vibrio fetus and other related Vibrios isolated from animals. J. Bacteriol. 76: 312-316. Linton, D., R. J. Owen, and J. Stanley. 1996. Rapid identification by PCR of the genus Campylobacter and of five Campylobacter species enteropathogenic for man and animals. Res. Microbiol. 147:707-718. Logan, J. M. J., K. J. Edwards, N. A. Saunders, and J. Stanley. 2001. Rapid identification of Campylobacter spp. by melting peak analysis of biprobes in real-time PCR. J. Clin. Microbiol. 39:2227-2232. Mannering, S. A., D. M. West, S. G. Fenwick, R. M. Marchant, N. R. Perkins, and K. O’Connell. 2004. Pulsed-field gel electrophoresis typing of Campylobacter fetus subsp. fetus isolated from sheep abortions in New Zealand. N. Z. Vet. J. 52:358-363. McFadyean, J., and S. Stockman. 1913. Appendix to part 111, Abortion in sheep. In Report of the Departmental Committee Appointed by the Board of Agriculture and Fisheries to Inquire into Epizootic Abortion. Her Majesty’s Stationery Office, London. McMillen, L., G. Fordyce, V. J. Doogan, and A. E. Lew. 2006. Comparison of culture and a novel 5’ Taq nuclease assay for direct detection of Campylobacter fetus subsp. venerealis in clinical specimens from cattle. 1. Clin. Microbiol. 44:93 8-945. Meinershagen, W. A., D. G. Waldhalm, F. W. Fank, and L. H. Scrivner. 1965. Magpies as a reservoir of infection for ovine vibriosis. J. Am. Vet. Med. Assn. 147:843-845.
Metherell, L. A., J. M. J. Logan, and J. Stanley. 1999. PCRenzyme-linked immunosorbent assay for detection and identification of Campylobacter species: application to isolates and stool samples. J. Clin. Microbiol. 37:433-435. Mitscherlich, E., and B. Liess. 1958. Die serologische differzierung von Vibrio fetus-Stammen. Dtsch. Tierartzl. Wochenschr. 65:25, 36-39. Mohanty, S. B., G. J. Plumer, and J. E. Faber. 1962. Biochemical and colonial characteristics of some bovine vibrios. Am. J. Vet. Res. 23554-557. Monke, H. J., B. C. Love, T. E. Wittum, D. R. Monke, and B. A. Byrum. 2002. Effect of transport enrichment medium, transport time, and growth medium on the detection of Campylobacter fetus subsp. venerealis.J. Vet. Diagn. Invest. 14:35-39. Moran, A. P., D. T. O’Malley, T. U. Kosunen, and I. M. Helander. 1994. Biochemical characterization of Campylobacter fetus lipopolysaccharides. Infect. Immun. 62:3922-3929. Morgan, W. J. 1959. Studies on the antigenic structure of Vibrio fetus. J. Comp. Pathol. 69:125-140. Morooka, T., A. Umeda, M. Fujita, H. Matano, S. Fujimoto, K. Yukitake, K. Amako, and T. Oda. 1996. Epidemiologic application of pulsed-field gel electrophoresis to an outbreak of Campylobacter fetus meningitis in a neonatal intensive care unit. Sand. J. Infect. Dis. 28:269-270. Moureau, P., I. Derclaye, D. Gregoire, M. Janssen, and G . R. Cornelis. 1989. Campylobacter species identification based on polymorphism of DNA encoding rRNA. J. Clin. Microbiol. 27: 1514-1517. Muller, W., H. Hotzel, and F. Schulze. 2003. Identification and differentiation of Campylobacter fetus subspecies by PCR. Dtsch. Tierarztl. Wochenschr. 11055-59. Newell, D. G., B. Duim, M. A. P. van Bergen, R. GrogonoThomas, and J. A. Wagenaar. 2000a. Speciation, subspeciation and subtyping of Campylobacter spp. associated with bovine infertility and abortion. Cattle Pract. 8:421-425. Newell, D. G., J. A. Frost, B. Duim, J. A. Wagenaar, R. H. Madden, J. van der Plas, and s. L. W. On. 2000b. New developments in the subtyping of Campylobacter species, p. 27-44. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. OIE. 2004. Chapter 2.3.2. Bovine Genital Campylobacteriosis, p. 439-450. In OIE (ed.), Manual of Diagnostic Tests and Vaccines for Terrestrial Animals, 5th ed. World Organisation for Animal Health, Paris. On, S . L. W. 1996. Identification methods for Campylobacters, Helicobacters and related organisms. Clin. Microbiol. Rev. 9: 405-422. On, S . L. W., and C. S. Harrington. 2001. Evaluation of numerical analysis of PFGE-DNA profiles for differentiating Campylobacter fetus subspecies by comparison with phenotypic, PCR and 16s rDNA sequencing methods.]. Appl. Microbiol. 90:285-293. Owen, R. J. 1983. Nucleic acids in the classification of Campylobacters. Eur J. Clin. Microbiol. 2:367-377. Oyarzabal, 0. A., I. V. Wesley, K. M. Harmon, L. SchroederTucker, J. M. Barbaree, L. H. Lauerman, S. Backert, and D. E. Conner. 1997. Specific identification of Campylobacter fetus by PCR targeting variable regions of the 16s rDNA. Vet. Microbiol. 58:6 1-71. Penner, J. L. 1988. The genus Campylobacter: a decade of progress. Clin. Microbiol. Rev. 1:157-172. Perez-Perez, G. I., M. J. Blaser, and J. H. Bryner. 1986. Lipopolysaccharide structures of Campylobacter fetus are related to heat-stable serogroups. Infect. Immun. 51:209-212. Plastridge, W. N., L. F. Williams, and D. G. Trowbridge. 1964. Antibiotic sensitivity of physiological groups of microaerophilic vibrios. Am. 1. Vet. Res. 25:1295-1299. Quinn, P. J., M. E. Carter, M. E. Markey, and G. R. Carter. 1994. Campylobacter species, p. 268-272. In P. J. Quinn, M. E. Car-
CHAPTER 11
C. FETUS ISOLATION. IDENTIFICATION. AND TYPING
ter, M. E. Markey, and G. R. Carter (ed.), Clinical Veterinary Microbiology. Mosby, London. Rennie, R. P., D. Strong, D. E. Taylor, S. M. Salama, C. Davidson, and H. Tabor. 1994. Campylobacter fetus diarrhea in a Hutterite colony: epidemiological observations and typing of the causative organism. J. Clin. Microbiol. 32:721-724. Roop, R. M., 11, R. M. Smibert, J. L. Johnson, and N. R. Krieg. 1984. Differential characteristics of catalase-positive Campylobacters correlated with DNA homology groups. Can. J. Microbiol. 30:938-951. Salama, S. M., M. M. Garcia, and D. E. Taylor. 1992. Differentiation of the subspecies of Campylobacter fetus by genomic sizing. Int. J. Syst. Bacteriol. 42:446-450. Schulze, F., A. Bagon, W. Muller, and H. Hotzel. 2006. Identification of Campylobacter fetus subspecies by phenotypic differentiation and PCR. J. Clin. Microbiol. 44:2019-2024. Schurig, G. D., C. E. Hall, K. Burda, L. B. Corbeil, J. R. Duncan, and A. J. Winter. 1973. Persistent genital tract infection with Vibrio fetus intestinalis associated with serotypic alteration of the infecting strain. Am. J. Vet. Res. 34:1399-1403. Skerman, V. B. D., V. McGowan, and P. H. A. Sneath. 1980. Approved list of bacterial names. Int. J. Syst. Bact. 30:225-420. Smibert, R. M. 1974. Genus I1 Campylobacter Sebald and Veron 1963, 907, p. 207-212. In R. E. Buchana and N. E. Gibbons (ed.), Bergey’s Manual of Determinative Bacteriology, 8th ed. Williams and Wilkins, Baltimore. Smibert, R M. 1984. Genus I1 Campylobacter Sebald and VCron 1963, p. 111-118. In N. R. Krieg and H. G. Holt (ed.), Bergey’s Manual of Systematic Bacteriology, vol. 1. Williams and Wilkins, Baltimore, MD. Staley, J. T., and N. R. Krieg. 1984. Classification of procaryotic organisms: an overview, p. 1-4. In N. R. Krieg and J. G. Holt (ed.), Bergey’s Manual of Systematic Bacteriology, vol. 1. Williams and Wilkins, Baltimore, MD. Stegenga, T., and J. I. Terpstra. 1949. Over Vibrio fetus infecties bij het rund en “enzootische” steriliteit. Tijdschr. Diergeneesk. 74~293-296. Thompson, L. M., 111, R. M. Smibert, J. L. Johnson and N. R. Krieg. 1988. Phylogenetic study of the genus Campylobacter. lnt. J. Syst. Bacteriol. 38:190-200. Thompson, S. A., and M. J. Blaser. 2000. Pathogenesis of Campylobacter fetus infections, p. 321-347. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd. ed. American Society for Microbiology, Washington, D. C. Tremblay, C., and C. Gaudreau. 1998. Antimicrobial susceptibility testing of 59 strains of Campylobacter fetus susbp. fetus. Antimicrob. Agents Chemother. 42:1847-1849. Tremblay, C., C. Gaudreau, and M. Lorange. 2003. Epidemiology and antimicrobial susceptibilities of 111 Campylobacter fetus subsp. fetus strains isolated in Quebec, Canada, from 1983 to 2000. J. Clin. Microbiol. 41:463-466. Tu, 2.-C., F. E. Dewhirst, and M. J. Blaser. 2001. Evidence that the Campylobacter fetus sap locus is an ancient genomic constituent with origins before mammals and reptiles diverged. Infect. Immun. 69:2237-2244. Tu, 2.-C., G. Zeitlin, J-P. Gagner, T. Keo, B. A. Hanna, and M. J. Blaser. 2004. Campylobacter fetus of reptile origin as a human pathogen. J. Clin. Microbiol. 42:4405-4407. Tu, Z. C., W. Eisner, B. N. Kreiswirth, and M. J. Blaser. 2005. Genetic divergence of Campylobacter fetus strains of mammal and reptile origins. J. Clin. Microbiol. 43:3334-3340. van Bergen, M. A. P. 2005. Subspecies differentiation and typing of Campylobacter fetus. Ph.D. thesis. Utrecht University, Utrecht, The Netherlands. van Bergen, M. A. P., K. E. Dingle, M. C. Maiden, D. G. Newell, L. van der Graaf-Van Bloois, J. P. M. van Putten, and J. A.
225
Wagenaar. 2005a. Clonal nature of Campylobacter fetus as defined by multilocus sequence typing. J. Clin. Microbiol. 43: 5 888-9 8. van Bergen, M. A. P., S. Linnane, J. P. M. van Putten, and J. A. Wagenaar. 2005b. Global detection and identification of Campylobacter fetus subsp. uenerealis. Rev. Sci. Tech. 24:1017-26. van Bergen, M. A. P., G. Simons, L. van der Graaf-van Bloois, J. P. M. van Putten, J. Rombout, I. Wesley, and J. A. Wagenaar. 2005c. Amplified fragment length polymorphism based identification of genetic markers and novel PCR assay for differentiation of Campylobacter fetus subspecies. J. Med. Microbiol. 54: 1 217-24. Vandamme, P. 2000. Taxonomy of the family Campylobacteraceae, p. 3-26. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Vandamme, P., B. Pot, E. Falsen, K. Kersters, and J. De Ley. 1990. Intra- and interspecific relationships of veterinary Campylobacters revealed by numerical analysis of electrophoretic protein profiles and DNA:DNA hybridizations. Syst. Appl. Microbiol. 13:295-3 03. Varga, J., B. Mezes, L. Fodor, and I. Hajtos. 1990. Serogroups of Campylobacter fetus and Campylobacter jejuni isolated in cases of ovine abortion. Zbl. Vet. Med. B 37:148-152. Vargas, A. C., M. M. Costa, M. H. Vainstein, L. C. Kreutz, and J. P. Neves. 2003. Phenotypic and molecular characterization of bovine Campylobacter fetus strains isolated in Brazil. Vet. Microbiol. 93:121-132. VCron, M., and R. Chatelain. 1973. Taxonomic study of the genus Campylobacter Sebald and VCron and designation of the neotype strain for the type species, Campylobacter fetus (Smith and Taylor) Sebald and VCron. Int. J. Syst. Bacteriol. 23:122-134. Wagenaar, J. A., M. A. P. van Bergen, D. G. Newell, R. GrogonoThomas, and B. Duim. 2001. Comparative study using amplified fragment length polymorphism fingerprinting, PCR genotyping, and phenotyping to differentiate Campylobacter fetus strains isolated from animals. J. Clin. Microbiol. 39:2283-2286. Wang, E., M. M. Garcia, M. S. Blake, 2. H. Pei, and M. J. Blaser. 1993. Shift in S-layer protein expression responsible for antigenic variation in Campylobacter fetus. J. Bacteriol. 175:49794984. Wang, G., C. G. Clark, T. M. Taylor, C. Pucknell, C. Barton, L. Price, D. L. Woodward, and F. G. Rodgers. 2002. Colony multiplex PCR assay for identification and differentiation of Campylobacter jejuni, C. coli, C. lari, C. upsaliensis and C. fetus subsp. fetus. J. Clin. Microbiol. 40:4744-4747. Watson, W. A., D. Hunter, and R. Bellhouse. 1967. Studies on vibrionic infection of sheep and carrion crows. Vet. Rec. 81:220226. Wayne, L. G., D. J. Brenner, R. R. Colwell, P. A. D. Grimont, P. Kandler, M. I. Krichevsky, L. H. Moore, W. E. C. Moore, R. G. E. Murray, E. Stackebrandt, M. P. Starr, and H. G. Triiper. 1987. Report of the ad hoc committee on reconciliation of approaches to bacterial systematics. Int. J. Syst. Bacteriol. 37: 463-464. Wesley, I. V., R. D. Wesley, M. Cardella, F. E. Dewhirst, and B. J. Paster. 1991. Oligodeoxynucleotide probes for Campylobacter fetus and Campylobacter hyointestinalis based on 16s rRNA sequences. J. Clin. Microbiol. 29:1812-1817. Willoughby, K., P. F. Nettleton, M. Quirie, M. A. Maley, G. Foster, M. Toszeghy, and D. G. Newell. 2005. A multiplex polymerase chain reaction to detect and differentiate Campylobacter fetus subspecies fetus and Campylobacter fetus subspecies uenerealis: use on UK isolates of Campylobacter fetus and other Campylobacter spp. J. Appl. Microbiol. 99:758-766. Yang, L. Y., Z. H. Pei, S. Fujimoto, and M. J. Blaser. 1992. Reattachment of surface array proteins to Campylobacter fetus cells. J. Bacteriol. 174: 1258-1267.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 12
Diagnosis and Antimicrobial Susceptibility of Campylobacter Speciest COLLETTEFITZGERALD, JEANWHICHARD, AND IRVINGNACHAMKIN
Campylobacter jejuni subsp. jejuni (referred to as C. jejuni throughout this chapter) and C. coli have been recognized since the late 1970s as important agents of gastrointestinal infections. Within the genus Campylobacter, C. jejuni and C. coli are the most common species associated with diarrheal illness in humans and cause clinically similar illnesses (Gupta et al., 2004) Although most laboratories do not routinely distinguish these organisms, 80 to 90% of Campylobacter infections in industrialized countries are thought to be due to C. jejuni and 5 to 10% due to C. coli when the diagnosis is performed solely on selective media. The distribution of species may be different in other parts of the world and when a nonselective isolation technique, such as the filter technique, is applied in conjunction with a selective medium (Allos et al., 1995; Le Roux and Lastovica, 1998; Taylor, 1992). C. jejuni and C. coli cause a spectrum of illnesses in humans. Fever, abdominal cramping, and diarrhea (with or without blood or fecal white cells) are prominent features of uncomplicated illness and may last for as little as a few days to more than 1 week. Campylobacter infections are generally self-limited with a relapse rate of 5 to 10% in untreated patients (Blaser, 1995). Patients with Campylobacter infection may manifest signs and symptoms of acute appendicitis and result in unnecessary surgery. Bacteremia, endocarditis, meningitis, urinary tract infection, and other extraintestinal diseases may result from Campylobacter infection (Blaser, 1995). Bacteremia occurs at an estimated rate of 1.5 per 1,000 intestinal infections, with the elderly having the highest rate (Sltirrow et al., 1993). Patients with immunodeficien-
cies, such as in patients with human immunodeficiency virus, may develop persistent diarrheal illness and bacteremia (Perlman et al., 1988). C. jejuni is now the most recognized antecedent cause of Guillain-BarrC syndrome, an acute paralytic disease of the peripheral nervous system, which is reviewed in chapter 13. Death after C. jejuni infection is rare but does occur (Font et al., 1997). CLASSIFICATION
Campylobacter, Arcobacter, and Sulfurospirillum are members of the family Campylobacteraceae (Vandamme et al., 1991, 1992; Vandamme and De Ley, 1991). An extensive review on the taxonomy of Campylobacter is in chapter 1. Specimen Considerations Fecal samples should routinely be submitted to the laboratory for isolation of Campylobacter species from patients with diarrheal illness. Although not ideal, rectal swabs are also acceptable for culture. Like the other common enteric pathogens (e.g., Salmonella and Shigella), Campylobacter infections are usually community acquired, and therefore, cultures for Campylobacter should not be routinely performed on hospitalized patients with diarrhea according to the 3-day rule (Hines and Nachamkin, 1996). Although culture of a single fecal specimen during acute illness has high sensitivity, two fecal samples may be required to rule out infection (Valenstein et al., 1996).
-
Collette Fitzgerald and Jean Whichard Enteric Diseases Laboratory Branch (Proposed), Centers for Disease Control and Prevention, Irving Nachamkin Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Atlanta, GA 30333. Medicine, Philadelphia, PA 19104-4283. +This chapter contains information presented in chapter 3 by Irving Nachamkin, Jorgen Engberg, and Frank Moller Aarestrup in the 2nd edition of this book.
227
228
FITZGERALD ET AL.
Numerous formulations of transport media have been used for Campylobacter, such as alkaline peptone water with thioglycolate and cystine (Wang et al., 1983), modified Stuart medium (Ah0 et al., 1988), and Cary-Blair medium (Cary and Blair, 1964). Buffered glycerol saline and Stuart medium do not have good performance characteristics (Wang et al., 1983). As a single transport medium, modified Cary-Blair medium containing reduced agar (1.6 g/liter) is quite useful and is suitable for Campylobacter and other enteric pathogens (Wang et al., 1983). If cultures are not performed immediately, specimens received in Cary-Blair medium should be stored at 4°C. Supplementation of Cary-Blair medium with laked sheep’s blood aids in prolonged storage of stool samples and recovery of C. jejuni (Wasfy et al., 1995).
Blood Samples Campylobacter species are occasionally isolated from blood cultures, primarily C. fetus, C. jejuni, and C. upsaliensis, but the optimal conditions for isolating Campylobacter from blood culture systems have not been evaluated to any great degree. Commercial blood culture systems such as the BACTEC system (BD, Sparks, MD) (aerobic bottles) and Septi-Chek system (BD, Sparks, MD) support the growth of the common Campylobacter species (Kasten et al., 1991; Lastovica et al., 1989; Wang and Blaser, 1986). Anaerobic broth or lysis centrifugation may not be as sensitive for the isolation of Campylobacter species from blood samples (Kasten et al., 1991). Direct Detection in Stool Samples A rapid and sensitive method for presumptive diagnosis of Campylobacter enteritis may be accomplished by performing Gram stain analysis of stool samples. The most commonly used Gram-stain procedure uses safranin counterstain, which does not visualize Campylobacter well; however, carbol fuchsin or 0.1% aqueous basic fuchsin counterstain works well for visualizing organisms in smears of stools or pure cultures (Park et al., 1983; Sazie and Titus, 1982; Wang and Murdoch, 2004). The sensitivity and specificity of stool Gram stain for detecting Campylobacter during the acute stage of illness is high, with sensitivity ranging from 66 to 94% and >95% specificity (Park et al., 1983; Sazie and Titus, 1982; Wang and Murdoch, 2004). Other methods have been used to directly detect organisms in fresh stool, such as phase contrast and dark-field microscopy; however, these methods are more technically de-
manding (Karmali and Fleming, 1979; Paisley et al., 1982). Some clinicians look for the presence of fecal white cells as an indicator of bacterial infection and may be present during Campylobacter infection, with positive results in 25 to 80% of culture-proven cases (Hines and Nachamkin, 1996). The presence of fecal white cells (or other white cell markers) may indicate a higher likelihood of bacterial infection; however, the lack of white cells (or white cell markers) does not discount the diagnosis (Valenstein et al., 1996). Further, there is no clinical evidence that management of bacterial infections with Campylobacter should be different on the basis of the presence or absence of fecal white cells. Thus, we do not recommend routine examination of stool samples for fecal leukocytes for predicting bacterial infection. Further, fecal white cell analysis should not be used for selective culturing for Campylobacter or other stool pathogens (Guerrant et al., 2001). Stool antigen tests for Campylobacter have been developed and are commercially available. The ProSpecT Carnpylobacter immunoassay (AlexonTrend, Inc., distributed through Remel) varies in sensitivity from 80 to 96% compared with culture and has a specificity of >97% (Dediste et al., 2003; Hindiyeh et al., 2000; Tolcin et al., 2000). There are some reported cross-reactions with C. upsaliensis, and antigen may be detected after cold storage for several days (Dediste et al., 2003; Endtz et al., 2000). Another commercial assay, Ridascreen Campylobacter (R-Biopharm, Darmstadt, Germany), was recently shown to have 69% sensitivity and 87% specificity in an analysis of 1,050 stool samples submitted for culture (Tissari and Rautelin, 2007). There is a need for molecular-based assays to detect Campylobacter in stool samples (Persson and Olsen, 2005). Commercial molecular systems are not yet available for detecting Campylobacter or for other enteric pathogens directly in fecal specimens. Laboratories continue to have difficulty isolating and identifying Campylobacter spp. in stool cultures, and molecular approaches may improve the ability of laboratories to diagnose these infections. PCR-based detection, identification to the species level, and typing of campylobacters directly from stool samples has been reported (Amar et al., 2007; Best et al., 2007; Linton et al., 1997b). Advantages of molecular approaches over culture include same-day detection, additional data regarding mixed infections, and uncommon Campylobacter species that are often missed when routine culture and procedures that are amenable to automation and high-throughput capabilities are used (Kulkarni et al., 2002; Lawson et al., 1999; Maher et al., 2003). However, PCR does not provide
CHAPTER 12
an isolate for further characterization such as susceptibility testing or subtyping. In addition, diagnostic laboratories are required to comply with clinical laboratory improvement amendments regulations in the United States, so it is important to ensure that quality standards are first established for use of molecular diagnostic assays to maintain the accuracy, reliability, and timeliness of patient test results.
CULTURE AND ISOLATION Enrichment A number of enrichment broths have been formulated to improve the recovery of Campylobacter from stool samples, including Preston broth (Bolton and Robertson, 1982), Campy-thio (Rubin and Woodard, 1983), and Campylobacter enrichment broth (Martin et al., 1983). It is unclear whether enrichment cultures are useful for primary isolation of Campylobacter during acute illness because of the high concentration of organisms present in the stool. However, enrichment cultures may enhance the yield of culture where low numbers of organisms are expected because of delays in transport of the sample to the laboratory. After the acute stage of disease when the concentration of organisms decrease, or in the investigation of Guillain-BarrC syndrome after acute Campylobacter infection, enrichment cultures may be necessary to recover Campylobacter (Nachamkin, 1997). Filtration In addition to direct plating of samples, filtration techniques were designed to isolate Campylobacter species including C. jejuni and C. coli as well as other Campylobacter species (Engberg et al., 2000; Goossens et al., 1986; Vandenberg et al., 2004) and Arcobacter spp. (Kiehlbauch et al., 1991; Vandenberg et al., 2004) that are susceptible to antibiotics present in most selective media (Vandenberg et al., 2004). Filtration methods should be used as an adjunct method in addition to direct plating for the assessment of clinical specimens. Because of the small size of Campylobacter, organisms pass through membrane filters with pore sizes of 0.45 to 0.65 pm. Wells et al. (1989) recommend cellulose acetate membrane filters (0.65-pm pore size) for routine use, and they are available commercially. Mixed cellulose ester membrane filters (0.60-pm pore size) were found to be optimal in the filtration method described by Lastovica (2006; see chapter 7). The method is easily performed by placing a sterile 0.65-pm cellulose acetate filter onto the
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
229
surface of an antibiotic-free agar medium, placing 10 to 15 drops of fecal suspension on the filter, and incubating the plate at 37°C for 60 min. The filter is removed and the plate incubated at 37°C under microaerobic conditions. Because many of the species not isolated on routine selective media require increased hydrogen environments, the filtration method is best used if an atmosphere containing increased hydrogen can be used. Filtration is a relatively insensitive method and requires that the stool samples contain -lo5 CFU/ml of Campylobacter in order to detect the organism (Goossens et al., 1990b).
ISOLATION PROCEDURES A microaerobic atmosphere containing approximately 5% 0,, 10% CO,, and 85% N, is required for optimal recovery of most Campylobacter species. Gas generator packs that generate suitable environments are available from commercial sources. Specialized incubators such as the tri-gas incubator and evacuation replacement procedures may also be used (Morris and Patton, 1985; Thompson et al., 1990). Candles jars should not be used for isolation and maintenance of Campylobacter species because of the suboptimal environment produced with this method (Luechtefeld et al., 1982). An increased hydrogen concentration in the incubation gas mixture is required to isolate several Campylobacter species, including C. sputorum, C. concisus, C. mucosalis, C. curvus, C. rectus, and C. hyointestinalis (Vandenberg et al., 2004). Unfortunately, commercial gas pack systems do not generate sufficient hydrogen for effective isolation of hydrogen-requiring Campylobacter species. An environment containing 10% CO,, 6%H2, and the balance N, should be sufficient for isolating hydrogen-requiring species (Bolton, 2001). Campylobacter jejuni and C. coli are readily isolated on a variety of selective media. Charcoal cefoperazone deoxycholate agar (Hutchinson and Bolton, 1984) and charcoal-based selective medium (Karmali et al., 1986) are blood-free media, and Campy-CVA medium (Reller et al., 1983) and Skirrow medium (Skirrow, 1977) are blood-containing media used for routine isolation procedures. A selective medium for isolation of C. upsaliensis (CAT medium) has also been described (Aspinall et al., 1996), but its effectiveness has not been regularly demonstrated (Engberg et al., 2000; Hindiyeh et al., 2000). C. upsaliensis may be isolated by other media or filtration methods. A hydrogen-enriched atmosphere may enhance growth of some strains (Goossens et al., 1990b; Lastovica and Skirrow, 2000).
230
FITZGERALD ET AL.
Several studies have shown that the use of a combination of media, including either charcoal cefoperazone deoxycholate agar or charcoal-based selective medium, achieves a higher yield of Campylobacter from stool samples (Endtz et al., 1991; Gun-Monro et al., 1987) than the use of only a single medium. As a single medium, charcoal cefoperazone deoxycholate agar medium was found to be the most sensitive for detecting C. jejuni and C. coli compared with Skirrow’s medium, CAT agar, and filtration technique (Engberg et al., 2000). Most selective media for isolation of Campylobacter contain antibiotics to inhibit the growth of normal enteric bacteria. Some antibiotics including cephalothin, colistin, and polymyxin B that are present in some formulations inhibit some strains of C. jejuni and C. coli and C. fetus subsp. fetus (Goossens et al., 1986; Ng et al., 1988). Cephalothin will inhibit C. jejuni subsp. doylei, C. upsaliensis, and A. butzleri, and thus, we do not recommend cephalothincontaining culture media such as Campy-BAP for isolation of Cumpylobacter and Arcobacter from fecal samples.
IDENTIFICATION
Cumpylobacter species are difficult to identify phenotypically because of relative biochemical inactivity, special growth requirements, and complex taxonomy. Numerous schemes are thus available for phenotypic and genotypic characterization of these bacterial species (On, 1996). Campylobacter colonies may have different appearances and depend on the type of medium used for primary isolation. Cumpylobucter spp. usually produce gray, flat, irregular, and spreading colonies on most isolation media and nonhemolytic colonies on blood-containing media. When utilizing freshly prepared media, spreading along the streak line is commonly seen. With decreased moisture content, reduced spreading is observed, and colonies may form a round, convex, and glistening appearance. Thus, for optimal isolation and identification of Campylobacter species, proper storage of media to control moisture content is essential. Arcobacter colonies are similar in appearance to Campylobucter (Vandamme, 2000; Vandamme et al., 1992). At primary isolation, colonies should be examined by Gram stain, and an oxidase test should be performed. Campylobucter species give a positive reaction with oxidase reagent. Colonies that exhibit positive oxidase reactions along with a typical Gram stain appearance can be presumptively reported as Cumpylobacter species subject to further testing with
the hippurate hydrolysis test (MacFaddin, 2000a). C. jejuni is distinguished from other species on the basis of the hippurate hydrolysis test. Oxidase-positive, curved (or S-shaped), gram-negative rods that grow at 42°C and are hippurate positive can be reported as C. jejuni subsp. jejuni, and further characterization is not routinely necessary. Not all strains of C. jejuni are hippurate positive by conventional biochemical assays and may require alternative assays, such as gas-liquid chromatography for detecting benzoic acid (liberated from hydrolysis of sodium hippurate) (Steinbruckner et al., 1999). Molecular detection of the hip0 (hippuricase) gene (Hani and Chan, 1995) or other C. jejuni-specific DNA markers (Table 1) by PCR may also be useful for identifying phenotypically hippurate-negative isolates (Totten et al., 1987), weakly positive isolates (Burnett et al., 2002a) and to clarify false-positive results for non-C. jejuni species (Denis et al., 1999). However, evaluation of C. jejuni-specific PCR tests has showed no single test to be entirely specific or sensitive; therefore, the use of more than one target for molecular identification of C. jejuni is recommended, and assay validation is important (Burnett et al., 2002b; On and Jordan, 2003). In addition, purified DNA rather than crude cell lysates may be needed for some PCR assays (Mohran et al., 1998; On and Jordan, 2003). C. coli is similar to C. jejuni biochemically except for hippuricase activity, for which C. coli is lacking. Molecular methods are recommended to accurately identify C. coli (Siemer et al., 2005), and most have proved both accurate and sensitive (On and Jordan, 2003). Disk antimicrobial susceptibility tests (nalidixic acid, cephalothin) have traditionally been used to aid in the identification of Cumpylobacter jejuni; however, disk tests are no longer appropriate for species identification because of increasing resistance to fluoroquinolones worldwide. The identification of species other than C. jejuni is difficult by using phenotypic characterization, and an algorithm for identification of the thermophilic Campylobucter spp. was recently published (Fitzgerald and Nachamkin, 2007). Phenotypic tests that may be useful for initial identification include temperature growth studies, catalase production, hippurate hydrolysis, indoxyl acetate hydrolysis (MacFaddin, 2000b), and production of H,S (Barrett et al., 1988). Some species of thermophilic Cumpylobucter may be differentiated by the indoxyl acetate hydrolysis test. Commercial systems for identifying Campylobacter species are available. In the United States, there are two immunologic reagents for culture identification: INDX Campy-JCL (Hardy Diagnostics and
CHAPTER 12
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
231
Table 1. Commonly used gene targets and PCR assays for genus and species differentiation of Campylobacter spp. Gene target
hipO" mapA" ceuE" , asp' nap locus (multiplex) Unknown (multiplex)" 1 6 s rRNA
Species detected
Reference(s)
C. jejuni C. jejuni C. coli C. coli C. jejuni subspecies C. jejuni and C. coli Genus Campylobacter C. upsaliensis, C. helveticus, C. fetus, C. hyointestinalis, and C. lari C. hominis C. lanienae
Linton et al. (1997a) Stucki et al. (1995) Gonzalez et al. (1997) Linton et al. (1997a) Miller et al. (2007) Vandamrne et al. (1997) Linton et al. (1996); Vanniasinkarn et al. (1999) Linton et al. (1996)'
C. gracilis C. rectus
Siqueira and Rocas (2003) Ashimoto et al. (1996); Siqueira and Rocas (2003) A1 Rashid et al. (2000) Maher et al. (2003)
Lawson et al. (1998) Inglis and Kalischuk (2003); Logan et al. (2000)
dYAb 16123s rRNAb GTPase' lpxA (multiplex) Unknown Unknown 23s rRNA GyrB/ hypothetical protein
C. jejuni, C. coli, C. lari, and C. upsaliensis Genus Campylobacter, C. jejuni, C. coli, C. upsaliensis, and C. lari C. jejuni, C. coli, C. lari, and C. upsaliensis C. jejuni, C. coli, C. lari, C. upsaliensis C. fetus C. fetus subsp. venerealis C. concisus C. concisus
van Doom et al. (1999) Klena et al. (2004) Hum et al. (1997) van Bergen et al. (2005) Bastyns et al. (1995) Matsheka et al. (2001)
PCR sensitivity and sensitivity evaluated by On and Jordan (2003). PCR probe-based methods. "PCR assays for C. upsuliensis and C. helveticus or C. fetus and C. hyointestinulis are duplex reactions that use a common forward primer and species-specific reverse primers.
Fisher Scientific) and Dry Spot Campylobacter Test Kit (Remel). INDX Campy-JCL does not differentiate between C. jejuni and C. coli (Nachamkin and Barbagallo, 1990). The Dry Spot Campylobacter latex test identifies, but does not differentiate, C. jejuni, C. coli, C. lari, or C. upsaliensis with variable results for C. fetus subsp. fetus (Oxoid USA, http:// www.oxoid.com/). Several studies have evaluated nucleic acid probes (Accuprobe, Gen-Probe Inc., San Diego, CA) for genus-level identification of Campylobacter and detection of C. jejuni subsp. jejuni, C. jejuni subsp. doylei, C. coli, and C. lari (PopovicUroic et al., 1991; Tenover et al., 1990); it may also hybridize with C. hyointestinalis strains (PopovicUroic et al., 1991). If phenotypic tests are inconclusive, immunologic and molecular probe identification may be useful for confirming Campylobacter to the genus level. The problems associated with the standard phenotypic methods for species identification of Cumpylobacter spp. have led to the development of a growing number of PCR assays for species-specific identification of Campylobacter. The 16s rRNA gene and 23s rRNA gene are two widely used targets for the design of species-specific tests; PCR assays based on these targets have been described for 12 different
Cumpylobacter species (On, 2005). Other gene targets used for species-specific PCR assays include gyrA (Menard et al., 2005), glyA (A1 Rashid et al., 2000), ceuE gene (Gonzalez et al., 1997), asp (Linton et al., 1997a), lpxA (Klena et al., 2004), and a GTPase gene (van Doorn et al., 1999). Subspecies identification by PCR within C. fetus (Hum et al., 1997; van Bergen et al., 2005) and C. jejuni (Miller et al., 2007) has also been described (see also chapter 11).Other assays involve restriction fragment analysis of PCRamplified regions of the 16s or 23s rRNA genes for differentiation of the thermotolerant Campylobacter spp. (Fermer and Engvall, 1999; Hurtado and Owen, 1997). These assays require further validation to fully determine their specificity and sensitivity (On, 2005). More recently, species-specific microarrays have been described for identification of several Campylobacter species including C. jejuni, C. coli, C. lari, and C. upsaliensis (Keramas et al., 2003; Quinones et al., 2007; Volokhov et al., 2003). Although these methods are promising tools for both identification and further genetic characterization of Campylobacter spp., the cost and limited availability of the technology in the clinical laboratory makes this approach not currently practical for routine application in the clinical setting.
232
FITZGERALD ET AL.
EPIDEMIOLOGIC TYPING SYSTEMS
Serology
Numerous typing systems have been developed to study the epidemiology of Campylobacter infection. They vary in complexity and ability to discriminate between strains. Some of the more common phenotypic methods used in studies include biotyping, phage typing, and serotyping, but the latter is the most commonly used phenotypic system (Newell et al., 2000; Patton and Wachsmuth, 1992). Originally described by Lior et al. (1982), the heat-labile system detects over 100 serotypes of C. jejuni, C. coli, and C. lari. The seroderminants of the heat-labile system include uncharacterized bacterial surface antigens and in some serotypes, flagellar antigens (Alm et al., 1991). The serotyping system described by Penner and Hennessy (1980), the heat-stable serotyping scheme, detects 60 types of C. jejuni and C. coli (Patton and Wachsmuth, 1992). The major serodeterminant of the heat-stable system is a Carnpylobacter capsular polysaccharide (Karlyshev et al., 2000). Because of the complexity of performing serotype analysis, only a few reference laboratories worldwide perform serotyping studies. A commercially produced serotyping kit is also available (Denka Seiken USA Inc., Campbell, CA). In recent years, numerous molecular subtyping systems have been developed, including restriction endonuclease analysis, ribotyping, PCR-based techniques, pulsed-field gel electrophoresis of macrorestricted chromosomal DNA, and amplified fragmentlength polymorphism (Newell et al., 2000; Patton and Wachsmuth, 1992). Rapid 1-day standardized PFGE protocols (Ribot et al., 2001) have facilitated the use of PFGE in the U.S. National Surveillance network, PulseNet, for food-borne pathogens (http:// www.cdc.gov/pulsenet/) (Olsen et al., 2001), but interpretation of PFGE profiles can be challenging (Fitzgerald et al., 2005). DNA sequence-based technology has emerged in recent years, and multilocus sequence typing has been useful for studies of the population structure and molecular epidemiology of C. jejuni (Dingle et al., 2001), although recognition of outbreaks caused by strains with common sequence types may be problematic (Clark et al., 2005; Sails et al., 2003). Genomotyping based on microarray technology has also been described for Campylobacter (Klena and Konkel, 2005). Unfortunately, a single method for determining strain relatedness and investigating outbreaks is not adequate, and multiple approaches may be required (Newell et al., 2000). A detailed discussion about typing systems and approaches to studying the epidemiology of Campylobacter can be found in chapter 10.
During the first few weeks of infection, fecal and serum immunoglobulin (Ig) A antibodies are apparent and then decrease precipitously, and as expected, serum IgG, I@, and IgA levels increase in response to infection (Black et al., 1992; Taylor et al., 1993). The performance characteristics of serum antibody assays are variable, and sensitivity, specificity, and predictive value will be population dependent. Patients with Campylobacter infection may have false-positive legionella antibody tests (Boswell and Kudesia, 1992). Serologic testing is not recommended for routine diagnosis of Campylobacter infection but appears to be useful for epidemiologic investigations (Taylor et al., 2004) (see chapter 6). Susceptibility Testing of Campylobacter The Clinical Laboratory Standards Institute (CLSI, formerly NCCLS) has published an agar dilution method for testing Campylobacter spp. and includes quality control (QC) ranges for several antimicrobial agents (CLSI, 2005; McDermott et al., 2005a). CLSI has also recently approved a disk diffusion and broth microdilution method. QC ranges for eight antimicrobial agents are given for C. jejuni ATCC 33560 (McDermott et al., 2005b) (CLSI, 2006). This publication also includes interpretive criteria for erythromycin, ciprofloxacin, tetracycline, and doxycycline. Tests can be incubated at 36°C for 48 h or 42°C for 24 h. In our experience, 36°C for 48 h results in optimal growth for test interpretation. CLSI recommends performing primary testing of clinical isolates with erythromycin and ciprofloxacin. The Etest (PDM Epsilometer, AB Biodisk, Solna, Sweden) method, which uses Mueller-Hinton agar with 5% sheep’s blood, has been shown to compare favorably with the agar dilution method (Baker, 1992).
In Vitro Susceptibility Profiles C. jejuni and C. coli are almost universally resistant to penicillins, cephalosporins (except a few third-generation cephalosporins), trimethoprim, sulfamethoxazole, rifampicin, and vancomycin. Originally described as highly susceptible to erythromycin, fluoroquinolones, tetracyclines, aminoglycosides, and clindamycin and moderately susceptible to chloramphenicol, cefotaxime, ceftazidime, and cefpirome, increasing resistance to a variety of antimicrobial agents is occurring with increased frequency (McNulty, 1987; Skirrow and Blaser, 1995).
CHAPTER 12
-
Macrolides and Clindamycin Since the recognition of Cumpylobucter enteritis in the 1970s, erythromycin has commonly been used for treating patients with uncomplicated enteritis (Elharrif et al., 1985; Karmali et al., 1981; Nolan et al., 1983; Vanhoof et al., 1978). Erythromycin resistance rates in C. jejuni range from 0 to 12% and are generally higher in C . coli, ranging from 0 to 50% (Gibreel and Taylor, 2006) (chapter 6). Trends over time for erythromycin resistance show stable and low rates in countries like Japan, Canada, Finland, and Denmark, whereas recent data from Thailand and Sweden document development of resistance (Harnett et al., 1995; Sjogren et al., 1997). It is unclear whether erythromycin resistance develops during therapy, although posttreatment resistant isolates have been described (Funke et al., 1994; Pitkanen et al., 1982; Snijders et al., 1997). The activity of clindamycin against C. jejuni is equivalent to that of erythromycin, and the drug may be a useful alternative for the treatment of serious infections in children. However, reported resistance rates are variable (CDC, 1998; Guevremont et al., 2006; Li et al., 1998; Papavasileiou et al., 2007; Reina et al., 1994; Taylor et al., 1987). Quinolones The fluoroquinolones have activity against most major pathogens causing bacterial enteritis. Early clinical trials of both community-acquired acute diarrhea and traveler’s diarrhea demonstrated good clinical response (Piddock, 1995; Wistrom and Norrby, 1995). Fluoroquinolones had good in vitro activity for all Cumpylobacter species as well as for members of the family of Enterobucteriuceue. Unfortunately , it soon became apparent that resistance in Campylobacter spp. can arise in vivo, sometimes even after just one or two administrations of fluoroquinolones (Adler et al., 1991). Additionally, an increasing number of reports claimed that fluoroquinoloneresistant strains were isolated from patients who had not received any medical treatment for their illness, suggesting that strains were already fluoroquinolone resistant before causing the infection (Chatzipanagiotou et al., 1993; Gaunt and Piddock, 1996; Li et al., 1998; Piddock, 1995; Reina et al., 1994; Sanchez et al., 1994). The 1990s were notable for a striking emergence of resistance among campylobacters and other enteric pathogens to nalidixic acid and fluoroquinolones. Before 1990, quinolone resistance among campylobacters was rarely reported. With the introduction of enrofloxacin and sarafloxacin (derivatives of cipro-
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
233
floxacin) in veterinary medicine, and less importantly, fluoroquinolones in human medicine in mainland Europe, a rapid emergence of quinolone resistance in Cumpylobucter isolates from patients was registered (Adler et al., 1991; Chatzipanagiotou et al., 1993; Rautelin et al., 1991; Reina et al., 1992). Surveillance data on resistance rates in human isolates from Asia soon indicated an equal and worrisome increase (Kuschner et al., 1995; Tee et al., 1995). Studies in the late 1990s from Taiwan, Thailand, and Spain showed rates of fluoroquinolone resistance in C. jejuni or C. jejuni and C. coli of 57, 84, and 88%, respectively (Gallardo et al., 1998; Hoge et al., 1998; Li et al., 1998). With the approval of quinolones for veterinary use in the United Kingdom and in the United States in late 1993 and 1995, respectively, reports from these areas showed increasing quinolone resistance profiles (CDC, 1998; Frost and Thwaites, 1998; Nachamkin, 1994). Antimicrobial resistance in animals and relevance to human infections are discussed in more detail in chapters 6 and 36. More recent studies indicate variable rates of ciprofloxacin resistance, in some cases higher among travel-associated infections. Ciprofloxacin resistance in France in 2004 was 25.3% (Gallay et al., 2007). During the same year, isolates collected in the United States exhibited 19% ciprofloxacin resistance, higher in C. coli (30.8%) than C. jejuni (18.1%) (CDC, 2007). The U.S. Food and Drug Administration proposed withdrawal of fluoroquinolone approval for use in poultry in 2000. A federal administrative law process upheld the proposed withdrawal, which took effect in September 2005 (Nelson et al., 2007). Australia continued to show low prevalence of ciprofloxacin resistance in 2001 to 2002-only 2% for domestically acquired cases (Unicomb et al., 2006). By contrast, 82% of isolates collected in Thailand in 2002 to 2003 were ciprofloxacin resistant (Boonmar et al., 2005). In Denmark, 26.7% of domestically acquired infections in 2005 were ciprofloxacin resistant; 7 of the 10 infections acquired abroad were resistant. The Netherlands documented 33% resistance among domestically acquired infections from 2002 to 2004, but 54% among travel-associated infections for the same time period (van Hees et al., 2007). Although ciprofloxacin resistance correlates well with nalidixic acid resistance (Boonmar et al., 2007; Kinana et al., 2006), some ciprofloxacin susceptibility among nalidixic acid-resistant isolates is commonly reported (Englen et al., 2007; Gallay et al., 2007; Jesse et al., 2006; McGill et al., 2006; National Antimicrobial Resistance Monitoring System, 2006; Varela et al., 2007) Occasionally ciprofloxacinresistant isolates are reported to be nalidixic acid susceptible (Bachoual et al., 2001; Reina et al., 1996).
234
FITZGERALD ET AL.
Induction of fluoroquinolone resistance during treatment is well recognized (Adler et al., 1991; Ellis et al., 1995; Segreti et al., 1992; Tee and Mijch, 1998). It has been predicted that in 10 to 20% of patients treated with a fluoroquinolone for Cumpylobacter enteritis, the Campylobacter strains will develop quinolone resistance (Piddock, 1997; Wistrom and Norrby, 1995). However, in a study by EllisPegler and colleagues (1995), between 18 and 28% of the patients in their prospective trial developed fluoroquinolone resistance. Development of resistance has been registered in short-term treatments, but prolonged therapy, i.e., in immunosuppressed patients, is a risk factor and has been associated with both clinical and bacteriological failures (Adler et al., 1991; Molina et al., 1995; Segreti et al., 1992; Tee and Mijch, 1998; Tee et al., 1995). In conclusion, the available information on fluoroquinolone resistance, frequent and rapid in vivo development of high-grade resistance and lack of clear clinical efficacy on community-acquired diarrhea suggest that fluoroquinolones may now be of limited use in the treatment of Campylobacter infections in many regions (Harnett et al., 1995; Hoge et al., 1998; Kuschner et al., 1995; Li et al., 1998; Mandal et al., 1984; Murphy et al., 1996; Pigrau et al., 1997; Rautelin et al., 1991; Ruiz et al., 1998; Sjogren et al., 1997). /.?-Lactam Agents With the exception of imipenem, the majority of C. jejuni and C. coli strains are resistant to p-lactam agents, principally penicillins and cephalosporins. However, they are moderately susceptible to cefotaxime, ceftazidime, and cefpirome (Van der Auwera and Scorneaux, 1985). p-Lactamase-producing strains of C. jejuni and C. coli are frequently identified (reviewed by Li et al., 2007). The chromosomal molecular class D p-lactamase of C. jejuni and C. cob seems to play a role only in resistance to amoxicillin, ampicillin, and ticarcillin, and a novel enzyme, OXA61, has also been described in a clinical C. jejuni isolate collected in 2001 (Alfredson and Korolik, 2005). With penicillin G, piperacillin, and cephalosporins, the mechanism of resistance in C. jejuni and C. coli is primarily considered to depend on their limited ability to bind to penicillin-binding proteins and permeability (Lachance et al., 1991, 1993; Tajada et al., 1996). Ampicillin is not recommended for use in the treatment of C. jejuni infections (Gaudreau and Gilbert, 1998). The reported rate of resistance to amoxicillin and clavulanic acid has been low (Gaudreau and Gilbert, 2003; Rodriguez-Avial et al., 2006), but data on the efficacy for treatment of Campylobacter
bacterernia are sparse (Pigrau et al., 1997). Imipenem may be useful for serious systemic infections in patients with renal damage, where the use of aminoglycosides may be problematic (Gaudreau and Gilbert, 2003; Skirrow and Blaser, 1995). Tetracycline Although tetracyclines have been suggested as an alternative choice in the treatment of C. jejuni and C. coli enteritis, large geographical differences on their susceptibility have been reported. Rates of resistance reported in the 1990s in Denmark were 0 to 11% (Aarestrup et al., 1997; Danish Zoonosis Centre, 1998), in Spain 25% (Gomez et al., 1995), in the United States 48% (CDC, 1998; Nachamkin, 1994), in Israel 70% (Schwartz et al., 1993), in Singapore 79% (Lim and Tay, 1992), and in Taiwan 85 to 95% (Li et al., 1998). In a Canadian study, Gaudreau and Gilbert (1998) noticed a significant increase in the resistance of C. jejuni to tetracycline from 19.1% in 1985 to 1986 to 55.7% in 1995 to 1997. In 2005, prevalence in Denmark was 16% for domestically acquired infections and 50% among those acquired abroad (Danish Zoonosis Centre, 2006). Forty-six percent of isolates in the United States were resistant in 2004, and 70% of clinical isolates collected in the Madrid area of Spain in 2003 were resistant (Rodriguez-Avial et al., 2006). Tetracyclines are ecologically disadvantageous drugs with a broad antibacterial spectrum, and their use is contraindicated in children. General use of tetracycline is not recommended but can be used in areas of low resistance to the agent, or better, after susceptibility testing of the clinical isolate in situations when other agents are contraindicated because of strain resistance or idiosyncratic responses in patients. Multidrug Resistance Multidrug resistance in Campylobacter appears to be occurring more frequently and poses the risk that an effective antimicrobial regimen to treat infection may be lacking. Hoge and colleagues (1998) from Thailand found 100% coresistance between azithromycin and ciprofloxacin in the last 2 years of a 15-year investigation (1981 to 1995). In addition, the level of tetracycline and ampicillin resistance in this area is now so high that these agents have no role in the treatment of Campylobacter or noncholera diarrhea in general. Chung-Chen and colleagues from Taiwan (Li et al., 1998) reported that concomitant resistance rates among their nalidixic acidresistant C. jejuni isolates from patients (exclusively children) were as follows: gentamicin 2%, erythro-
CHAPTER 12
mycin 12%, clindamycin 12%, tetracycline 97%, and ciprofloxacin 66%. A total of 100% of their human erythromycin-resistant C. jejuni isolates and 90% of their C. coli isolates were concomitantly resistant to clindamycin. Erythromycin and ciprofloxacin coresistance was found among C. jejuni collected in Quebec from 1998 to 2001, some of which were also resistant to tetracycline (Gaudreau and Michaud, 2003). Among ciprofloxacin-resistant C. jejuni collected in Finland from 1995 to 2000, Hakanen et al. (2003) found 3% coresistance to erythromycin or clindamycin, 68% coresistance to tetracycline, and 25% coresistance to ampicillin. Six of the eight erythromycin-resistant strains were ciprofloxacin resistant.
Campylobacter lari Until the 1990s, C. lari could easily be distinguished from other thermophilic species on the basis of intrinsic resistance to nalidixic acid (Nachamkin et al., 1984). However, since then, resistance to nalidixic acid and fluoroquinolones has emerged among C. jejuni and C. coli isolates, and nalidixic acidsensitive biovars of C. lari have been described (Megraud et al., 1988). Thus, more reliable tests are required to identify this species. The indication for treatment with antimicrobial agents is similar to that of C. jejuni and C. coli, and with the same classes of drugs, e.g., macrolides, aminoglycosides, imipenem, clindamycin, tetracycline, and fluoroquinolones according to susceptibility testing results (Bruneau et al., 1998; Chiu et al., 1995; Evans and Riley, 1992; Simor and Wilcox, 1987). As for the other thermophilic Campylobacter spp., fluoroquinolone resistance has also been described for this species (Evans and Riley, 1992). Thwaites and Frost (1999) reported no erythromycin resistance among C. lari isolated in England, but 100% resistance to nalidixic acid and ciprofloxacin.
Campylobacter upsaliensis No controlled trials of antibiotic treatment for C. upsaliensis enteritis have been performed, and experience with antimicrobial treatment is limited. The organism is typically sensitive to aminoglycosides, cephalosporins, tetracycline, and nalidixic acid and resistant to vancomycin, methicillin, piperacillin, and chloramphenicol (Bourke et al., 1998; Preston et al., 1990). C. upsaliensis is usually sensitive to erythromycin, but up to 10% resistance has been observed (Goossens et al., 1990a). Jenkin and Tee (1998) reported resistance to fluoroquinolones in three isolates from two human immunodeficiency virus-
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
235
seropositive patients, suggesting that routine susceptibility testing for this species, as for any other species in this genus, is advised. Vandenberg et al. (2006) found 12% nalidixic acid resistance, 6% ciprofloxacin resistance, and 13% erythromycin resistance among 85 C. upsaliensis isolated from humans in Belgium from 1995 to 2005.
Campylobacter fetus subsp. fetus Reported C. fetus cases most often involve bloodstream and extraintestinal infections that are often prolonged and relapsing. The prognosis depends largely on the severity of underlying illnesses. Immunocompetent patients with uncomplicated enteritis may not require antibiotics. Systemic infections require parenteral therapy, but erythromycin is not always effective and is no longer a recommended treatment for systemic C. fetus infections (Allos et al., 1995; Francioli et al., 1985; Kwon et al., 1994; Tremblay and Gaudreau, 1998). Appropriate antimicrobial therapy for serious infections due to fetus is not fully assessed for two reasons. First, C. fetus infection is rare, and second, many patients die despite intensive therapy as a result of associated predisposing factors (Goossens et al., 1989). However, patients with endovascular infections due to C. fetus require at least 4 weeks of treatment, and gentamicin with ampicillin seems to be the combination of choice (Blaser et al., 1978; Goossens et al., 1989).Imipenem may be a useful alternative (Dronda et al., 1998; Tremblay and Gaudreau, 1998). An abdominal aortic aneurism infected by C. fetus was successfully treated with surgery and ciprofloxacin (la Scola et al., 1998). A 2003 study of C. fetus subsp. fetus isolated from humans in Quebec found resistance to ciprofloxacin (5%), erythromycin (71%), and tetracycline (34%) (Tremblay et al., 2003). Most studies have used agar dilution to assess the susceptibility of C. fetus. However, Tremblay and Gaudreau (1998) compared minimal inhibitory concentration (MIC), disk diffusion, and Etest on 59 C. fetus subsp. fetus isolates. They found relative good agreement between the three methods for most antimicrobial agents. However, results for cefotaxime and erythromycin varied from 0 to 90% and 39 to 100% susceptible, respectively. In general, C. fetus were found to be susceptible to most antimicrobials, except tetracycline, with 27% resistant isolates.
c.
Arcobacter butzleri and A. ctyaerophilus Kiehlbauch and colleagues (1992) studied 78 human and animal isolates of Arcobacter butzleri and A. cyaerophilus for susceptibility to 22 antimicrobial
236
FITZGERALD ET AL.
agents. Most isolates were resistant to macrolides including erythromycin, cephalosporins except cefotaxime, ampicillin, ampicillin-sulbactam, clindamycin, chloramphenicol, and trimethoprim-sulfamethoxazole, whereas the fluoroquinolones, aminoglycosides, and one tetracycline (minocycline) demonstrated most activity. Fluoroquinolone resistance has also been reported (Lerner et al., 1994). Two studies have addressed MIC distributions and resistance prevalence among Arcobacter isolated in Belgium (Houf et al., 2004; Vandenberg et al., 2006). Houf et al. (2004) found that the nalidixic acid MICs for human isolates collected from 1995 to 2001 were 1 to 64 pg/ml, whereas all ciprofloxacin MICs were 50.25. Erythromycin MICs ranged from 2 to 32 pg/ml. The MIC distributions for poultry isolates were found to be similar, except that the ciprofloxacin MICs for 3 of the 68 A. butzleri isolates from poultry products were 16 pg/ml. Vandenberg et al. (2006) found no antimicrobial resistance among the 10 human clinical A. cryaerophilus isolates tested, whereas resistance was found among A. butzleri: nalidixic acid 18%, ciprofloxacin 3%, and erythromycin 21%. No tetracycline resistance was found among A. butzleri. C . concisus
Little information is available on the antimicrobial susceptibility of C. concisus. Greig and colleagues from South Africa have tested the MIC values of eight isolates and found that ciprofloxacin was the most active agent. All strains were also sensitive to tetracycline, ampicillin, and gentamicin. In contrast, all but one of the strains were resistant to erythromycin. Activity of cephalosporins was variable. Vandenberg et al. (2006) found 16 of 20 C. consisus isolates tested were resistant to nalidixic acid, but only one of these was resistant to ciprofloxacin. Acknowledgments. Use of trade names is for identification only and does not imply endorsement by the Public Health Service or by the U.S. Department of Health and Human Services.
REFERENCES Aarestrup, F. M., E. M. Nielsen, M. Madsen, and J. Engberg. 1997. Antimicrobial susceptibility patterns of thermophilic Campylobacter spp. from humans, pigs, cattle, and broilers in Denmark. Antimicrob. Agents Chemother. 41:2244-2250. Adler, M., H. Luthy, L. Martinetti, A. Burnens, and M. Altwegg. 1991. Development of resistance to quinolones in five patients with campylobacteriosis treated with norfloxacin or ciprofloxacin. Eur. J. Clin. Microbiol. Infect. Dis. 10:953-957. Aho, M., M. Kauppi, and J. Hirn. 1988. The stability of small number of campylobacteria in four different transport media. Acta Vet. Scand. 29:437-442. Alfredson, D. A., and V. Korolik. 2005. Isolation and expression of a nova1 molecular class D B-lactamase, OXA-61, from Cam-
pylobacter jejuni. Antimicrob. Agents Chemother. 49:25152518. Allos, B. M., A. J. Lastovica, and M. J. Blaser. 1995. Atypical campylobacters and related organisms, p. 849-865. In M. J. Blaser, P. D. Smith, J. I. Ravdin, H. B. Greenberg, and R. L. Guerrant (ed.), Infections of the Gastrointestinal Tract. Raven Press, New York. Alm, R. A., P. Guerry, M. E. Power, H. Lior, and T. J. Trust. 1991. Analysis of the role of flagella in the heat-labile Lior serotyping scheme of thermophilic campylobacters by mutant allele exchange. J. Clin. Microbiol. 29:2438-2445. A1 Rashid, S . T., I. Dakuna, H. Louie, D. Ng, P. Vandamme, W. Johnson, and V. L. Chan. 2000. Identification of Campylobacter jejuni, C. coli, C. lari, C. upsaliensis, Arcobacter butzleri, and A. butzleri-like species based on the glyA gene. J. Clin. Microbiol. 38:1488-1494. Amar, C. F., C. L. East, J. Gray, M. Iturriza-Comara, E. A. Maclure, and J. McLauchlin. 2007. Detection by PCR of eight groups of enteric pathogens in 4,627 faecal samples: reexamination of the English case-control Infectious Intestinal Disease Study (1993-1996). Eur. J. Clin. Microbiol. Infect. Dis. 26: 3 11-3 23. Ashimoto, A., C. Chen, I. Bakker, and J. Slots. 1996. Polymerase chain reaction detection of 8 putative periodontal pathogens in subgingival plaque of gingivitis and advanced periodontitis lesions. Oral Microbiol. Immunol. 11:266-273. Aspinall, S . T., D. R. Wareing, P. G. Hayward, and D. N. Hutchinson. 1996. A comparison of a new selective medium (CAT) with membrane filtration for the isolation of thermophilic campylobacters including Campylobacter upsaliensis. ]. Appl. Bacteriol. 80:645-650. Bachoual, R., S . Ouakdesselam, F. Mory, C. Lascols, C. J. Soussy, and J. Tankovic. 2001. Single or double mutational alterations of GyrA associated with fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli. Microb. Drug Resist. 7: 257-261. Baker, C. N. 1992. The E-test and Campylobacter jejuni. Diagn. Microbiol. Infect. Dis. 15:469-472. Barrett, T. J., C. M. Patton, and G. K. Morris. 1988. Differentiation of Campylobacter species using phenotypic characterization. Lab. Med. 19:96-102. Bastyns, K., S . Chapelle, P. Vandamme, H. Goossens, and R. De Wachter. 1995. Specific detection of Campylobacter concisus by PCR amplification of 23s rDNA areas. Mol. Cell. Probes 9:247250. Best, E. L., A. J. Fox,R. J. Owen, J. Cheesbrough, and F. J. Bolton. 2007. Specific detection of Campylobacter jejuni from faeces using single nucleotide polymorphisms. Epidemiol. Infect. 135: 839-846. Black, R. E., D. Perlman, M. L. Clements, M. M. Levine, and M. J. Blaser. 1992. Human volunteer studies with Campylobacter jejuni, p. 207-215. In I. Nachamkin, M. J. Blaser, and L. S . Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Blaser, M. J. 1995. Campylobacter and related species. p. 19481956. In G. L. Mandell, J. E. Bennett, and R. Dolin (ed.), Principles and Practice of Infectious Diseases. Churchill Livingstone, New York. Blaser, M. J., B. Cravens, B. W. Powers, and W. L. Wang. 1978. Campylobacter enteritis associated with canine infection. Lancet ii:979-981. Bolton, F. 2001. Methods for isolation of Campylobacter from humans, animals, food and water, p. 87-93. In World Health Organization (ed.), The Increasing Incidence of Human Campylobacteriosis. Report and Proceedings of a WHO Consultation of
CHAPTER 12
Experts. Copenhagen, Denmark, 21-25 November 2000. Department of Communicable Disease Surveillance and Response, World Health Organization, Geneva. Bolton, F. J., and L. Robertson. 1982. A selective medium for isolating Campylobacter jejunitcoli.]. Clin. Pathol. 35:462-467. Boonmar, S., Y.Morita, M. Fujita, L. Sangsuk, K. Suthivarakom, P. Padungtod, S. Maruyama, H. Kabeya, M. Kato, K. Kozawa, S. Yamamoto, and H. Kimura. 2007. Serotypes, antimicrobial susceptibility, and gyrA gene mutation of Campylobacter jejuni isolates from humans and chickens in Thailand. Microbiol. lmmunol. 51531-537. Boonmar, S., L. Sanqsuk, K. Suthivarakom, P. Padnnqtod, and Y. Morita. 2005. Serotypes and antimicrobial resistance of Campylobacter jejuni isolated from humans and animals in Thailand. S. Asian J. Trop. Med. Public Health 36:130-134. Boswell, T. C. J., and G. Kudesia. 1992. Serological crossreactions between Legionella pneumophila and Campylobacter in the indirect fluorescent antibody test. Epidemiol. Infect. 109: 291-295. Bourke, B., V. L. Chan, and P. Sherman. 1998. Campylobacter upsaliensis: waiting in the wings. Clin. Microbiol. Rev. 11:440449. Bruneau, B., L. Bnrc, C. Bizet, Z. Lambert, and C. Branger. 1998. Purulent pleurisy caused by Campylobacter lari. Eur. J. Clin. Microbiol. Infect. Dis. 17:185-188. Burnett, T. A., A. Hornitzky, P. Kuhnert, and S. P. Djordjevic. 2002a. Speciating Campylobacter jejuni and Campylobacter coli isolates from poultry and humans using six PCR-based assays. FEMS Microbiol. Lett. 216:201-209. Burnett, T.A., A. H. M. P. Kuhnert, and S. P. Djordjevic. 2002b. Campylobacter jejuni and Campylobacter coli isolates from poultry and humans using six PCR-based assays. FEMS Microbiol. Lett. 216:201-209. Cary, S. G., and E. B. Blair. 1964. New transport medium for shipment of clinical specimens. I. Fecal specimens. J. Bacteriol. 88~96-98. Centers for Disease Control and Prevention. 1998. National Antimicrobial Resistance Monitoring System NARMS-1997 Annual Report Revised. Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention. 2007. NARMS Human Isolates Final Report, 2004. Centers for Disease Control and Prevention, Atlanta, GA. Chatzipanagiotou, S., E. Papavasiliou, and L. Malamou. 1993. Isolation of Campylobacter jejuni strains resistant to nalidixic acid and fluoroquinolones from children with diarrhea in Athens, Greece. Eur. J. Clin. Microbiol. Infect. Dis. 12566-568. (Letter.) Chiu, C. H., C. Y. Kuo, and J. T. On. 1995. Chronic diarrhea and bacteremia caused by Campylobacter lari in a neonate. Clin. Infect. Dis. 21:700-701. Clark, C. G., L. Bryden, W. R. Cuff, P. L. Johnson, F. Jamieson, B. Ciebin, and G. Wang. 2005. Use of the oxford multilocus sequence typing protocol and sequencing of the flagellin short variable region to characterize isolates from a large outbreak of waterborne Campylobacter sp. strains in Walkerton, Ontario, Canada. J. Clin. Microbiol. 43:2080-2091. Clinical and Laboratory Standards Institute. 2005. Performance standards for antimicrobial susceptibility testing; 15th informational supplement. Clinical and Laboratory Standards Institute, Wayne, PA Clinical and Laboratory Standards Institute. 2006. Methods for antimicrobial dilution and disk susceptibility testing of infrequently isolated or fastidious bacteria; approved guideline. Clinical and Laboratory Standards Institute, Wayne, PA.
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
237
Danish Zoonosis Centre. 1998. DANMAP 97. Consumption ofAntimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria from Food Animals, Food and Humans in Denmark, p. 3-59. Danish Zoonosis Centre, Copenhagen. Danish Zoonosis Centre. 2006. DANMAP 2005. Use ofdntimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in Denmark. Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), Copenhagen, Denmark. http: //www.food.dtu.dk/Default.aspx?ID-l0436. Dediste, A., 0. Vandenberg, L. Vlaes, A. Ebraert, N. Douat, P. Bahwere, and J.4'. Butzler. 2003. Evaluation of the ProSpecT microplate assay for detecton of Campylobacter: a routine laboratory perspective. Clin. Microbiol. Infect. 9: 1085-1090. Denis, M., C. Soumet, K. Rivoal, G. Ermel, D. Blivet, G. Salvat, and P. Colin. 1999. Development of a m-PCR assay for simultaneous identification of Campylobacter jejuni and C. coli. Lett. Appl. Microbiol. 29:406-410. Dingle, K. E., F. M. Colles, D. R. Wareing, R. Ure, A. J. Fox, F. E. Bolton, H. J. Bootsma, R. J. Willems, R. Urwin, and M. C. Maiden. 2001. Multilocus sequence typing system for Campylobacter jejuni. J. Clin. Micriobol. 39: 14-23. Dronda, F., I. Garcia-Arata, E. Navas, and L. de Rafael. 1998. Meningitis in adults due to Campylobacter fetus subspecies fetus. Clin. Infect. Dis. 27:906-907. Elharrif, Z.,F. Megraud, and A. M. Marchand. 1985. Susceptibility of Campylobacter jejuni and Campylobacter coli to macrolides and related compounds. Antimicrob. Agents Cbemother. 28~695-697. Ellis-Pegler, R. B., L. K. Hyman, R. J. Ingram, and M. McCarthy. 1995. A placebo controlled evaluation of lomefloxacin in the treatment of bacterial diarrhoea in the community. J. Antimicrob. Chemother. 36:259-263. Endtz, H.P., C. W. Ang, N. Van Den Braak, A. Luijendijk, B. C. Jacobs, P. de Man, J. M. van Duin, A. van Belkum, and H. A. Verbrugh. 2000. Evaluation of a new commercial immunoassay for rapid detection of Campylobacter jejuni in stool samples. Eur. 1. Clin. Microbiol. Infect. Dis. 19:794-797. Endtz, H.P.,G. J. Ruijs, A. H. Zwinderman, T. van der Reijden, M. Biever, and R. P. Mouton. 1991. Comparison of six media, including a semisolid agar, for the isolation of various Campylobacter species from stool specimens. J. Clin. Microbiol. 29: 1007-1010. Engberg, J., S. L. W. On, C. S. Harrington, and P. Gerner-Smidt. 2000. Prevalence of Campylobacter, Arcobacter, Helicobacter, and Sutterella spp. in human fecal samples as estimated by a reevaluation of isolation methods for Campylobacter. J. Clin. Microbiol. 38:286-29 1. Englen, M. D.,A. E. Hill, D. A. Dargatz, S. R. Ladely, and P. Fedorka-Cray. 2007. Prevalence and antimicrobial resistance of Campylobacter in US dairy cattle. J. Appl. Microbiol. 102:15701577. Evans, T.G., and D. Riley. 1992. Campylobacter laridis colitis in a human immunodeficiency virus-positive patient treated with a quinolone. Clin. Infect. Dis. 15:172-173. (Letter.) Fermer, C., and E. 0. Engvall. 1999. Specific PCR identification and differentiation of the thermophilic campylobacters, Campylobacter jejuni, C. coli, C. lari, and C. upsaliensis.J. Clin. Microbiol. 37:3370-3373. Fitzgerald, C., and I. Nachamkin. 2007. Campylobacter and Arcobacter, p. 933-946. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. L. Landry, and M. A. Pfaller (ed.), Manual of Clinical Microbiology. ASM Press, Washington, DC. Fitzgerald, C., A. Sails, and B. Swaminathan. 2005. Genetic techniques: molecular subtyping methods, p. 271-293. In T.
238
FITZGERALD ET AL.
McMeekin (ed.), Detecting Pathogens in Food. Woodhouse Publishing, Cambridge, United Kingdom. Font, C., A. Cruceta, A. Moreno, 0. Miro, B. Coll-Vinent, M. Almela, and J. Mensa. 1997. Estudio de 30 pacientes con bacteriemia por Campylobacter spp. Med. Clin. (Barc.) 108:336340. Francioli, P., J. Herzstein, J. Grob, J. J. Valloton, G. Mombelli, and M. P. Glauser. 1985. Campylobacter fetus subspecies fetus bacteremia. Arch. Intern. Med. 145 :289-292. Frost, J. A., and R. T. Thwaites. 1998. Drug resistance in C. jejuni, C. coli and C. lari isolated from humans in Wales and North West England during 1997. Working Paper 20.10b. World Health Organization, Geneva. Funke, G., R. Baumann, J. L. Penner, and M. Altwegg. 1994. Development of resistance to macrolide antibiotics in an AIDS patient treated with clarithromycin for Campylobacter jejuni diarrhea. Eur. J. Clin. Microbiol. Infect. Dis. 13:612-615. Gallardo, F., J. Gascon, J. Ruiz, M. Corachan, M. T. de Anta, and J. Vila. 1998. Campylobacter jejuni as a cause of traveler’s diarrhea: clinical features and antimicrobial susceptibility.J. Travel Med. 5:23-26. Gallay, A., V. Prouzet-MaulCon, I. Kempf, P. Lehours, L. Labadi, C. Camou, M. Denis, H. de Valk, J. C. Desenclos, and F. Mtgraud. 2007. Campylobacter antimicrobial drug resistance among humans, broiler chickens, and pigs, France. Emerg. Infect. Dis. 13:259-266. Gaudreau, C., and H. Gilbert. 2003. Antimicrobial resistance of Campylobacter jejuni subsp. jejuni strains isolated from humans in 1998 to 2001 in Montreal, Canada. Antimicrob. Agents Chemother. 47:2027-2029. Gaudreau, C., and H. Gilbert. 1998. Antimicrobial resistance of clinical strains of Campylobacter jejuni subsp. jejuni isolated from 1985 to 1997 in Quebec, Canada. Antimicrob. Agents Chemother. 42:2 106-2108. Gaudreau, C., and S . Michaud. 2003. Cluster of erythromycinand ciprofloxacin-resistant Campylobacter jejuni subsp. jejuni from 1999 to 2001 in men who have sex with men, Quebec, Canada. Clin. Infect. Dis. 37:131-136. Gaunt, P. N., and L. J. Piddock. 1996. Ciprofloxacin resistant Campylobacter spp. in humans: an epidemiological and laboratory study. J. Antimicrob. Chemother. 37:747-757. Gibreel, A., and D. E. Taylor. 2006. Macrolide resistance in Campylobacter jejuni and Campylobacter coli. J. Antimicrob. Chemother. 58:243-255. Gomez, G., R. Cogollos, and J. L. Alos. 1995. Susceptibilities of fluoroquinolone-resistant strains of Campylobacter jejuni to 11 oral antimicrobial agents. Antimicrob. Agents Chemother. 39: 542-544. Gonzalez, I., K. A. Grant, P. T. Richardson, S . F. Park, and M. D. Collins. 1997. Specific identification of the enteropathogens Campylobacter jejuni and Campylobacter coli by using a PCR test based on the ceuE gene encoding a putative virulence determinant. J. Clin. Microbiol. 35:759-763. Goossens, H., H. Coignau, L. Vlaes, and J. P. Butzler. 1989. Invitro evaluation of antibiotic combinations against Campylobacter fetus. J. Antimicrob. Chemother. 24:195-201. Goossens, H., M. De Boeck, H. Coignau, L. Vlaes, C. Van den Borre, and J.-P. Butzler. 1986. Modified selective medium for isolation of Campylobacter spp. from feces: comparison with Preston medium, a blood-free medium, and a filtration system. J. Clin. Microbiol. 24:840-843. Goossens, H., B. Pot, L. Vlaes, C. Van den Borre, R. Van den Abbeele, C. Van Naelten, H. Cogniau, P. Marbehant, and J. Verhoef. 1990a. Characterization and description of “Campylobacter upsaliensis: isolated from human feces. J. Clin. Microbiol. 28:1039-1046.
Goossens, H., L. Vlaes, M. De Boeck, B. Pot, K. Kersters, J. Levy, P. de Mol, J. P. Butzler, and P. Vandamme. 1990b. Is “Campylobacter upsaliensis” an unrecognised cause of human diarrhoea? Lancet 335:584-5 86. Greig, A., D. Hanslo, E. Le Roux, and A. Lastovica. 1993. In-vitro activity of eight antimicrobial agents against paediatric isolates of C. concisus and C. mucosalis. Acta Gastroenterol. Belg. 56: 12. Guerrant, R. L.,T. Van Gilder, T. S. Steiner, N. M. Thielman, L. Slutsker, R. V. Tauxe, T. Hennessy, P. M. Griffin, H. DuPont, R. B. Sack, P. Tarr, M. Neil], I.Nachamkin, L. B. Reller, M. T. Osterholm, M. L. Bennish, and L. K. Pickering, for the Infectious Diseases Society of America. 2001. Practice guidelines for managing infectious diarrhea. Clin. Infect. Dis. 32:33 1-351. Guevremont, E., E. Nadeau, M. Sirois, and S. Quessy. 2006. Antimicrobial susceptibilities of thermophilic Campylobacter from humans, swine, and chicken broilers. Can. J. Vet. Res. 70:8186. Gun-Monro, J., R. P. Rennie, J. H. Thornley, H. L. Richardson, D. Hodge, and J. Lynch. 1987. Laboratory and clinical evaluation of isolation media for Campylobacter jejuni. J. Clin. Microbiol. 25:2274-2277. Gupta, A., J. M. Nelson, T. J. Barrett, R. V. Tauxe, S . P. Rossiter, C. R. Friedman, K. W. Joyce, K. E. Smith, T. F. Jones, M. A. Hawkins, B. Shiferaw, J. L. Beebe, D. J. Vugia, T. RabatskyEhr, J. A. Benson, T. P. Root, and F. J. Angulo, for the NARMS Working Group. 2004. Antimicrobial resistance among Campylobacter strains, United States, 1997-2001. Emerg. Infect. Dis. 10:1102-1109. Hakanen, A. J., M. Lehtopolku, A. Siitonen, P. Huovinen, and P. Kotilainen. 2003. Multidrug resistance in Campylobacter jejuni strains collected from Finnish patients during 1995-2000. J. Antimicrob. Chemother. 52:1035-1039. Hani, E., and V. L. Chan. 1995. Expression and characterization of Campylobacter jejuni benzoylglycine amidohydrolase (hippuricase) gene in Escherichia coli. J. Bacteriol. 177:2396-2402. Harnett, N., S. Mcleod, Y. A. Yong, C. Hewitt, M. Vearncombe, and C. Krishnan. 1995. Quinolone resistance in clinical strains of Campylobacter jejuni and Campylobacter coli. J. Antimicrob. Chernother. 36:269-270. (Letter.) Hindiyeh, M., S. Jense, S. Hohmann, H. Benett, C. Edwards, W. Aldeen, A. Croft, J. Daly, S . Mottice, and K. C. Carroll. 2000. Rapid detection of Campylobacter jejuni in stool specimens by an enzyme immunoassay and surveillance for Campylobacterupsaliensis in the greater Salt Lake City area. J. Clin. Micriobol. 38~3076-3079. Hines, J., and I. Nachamkin. 1996. Effective use of the clinical microbiology laboratory for diagnosing diarrheal diseases. Clin. Infect. Dis. 23:1292-1301. Hoge, C. W., J. M. Gambel, A. Srijan, C. Pitarangsi, and P. Echeverria. 1998. Trends in antibiotic resistance among diarrheal pathogens isolated in Thailand over 15 years. Clin. Infect. Dis. 26~341-345. Houf, K., L. A. Devriese, F. Haesebrouck, 0. Vandenberg, J.-P. Butzler, J. Van Hoof, and P. Vandamme. 2004. Antimicrobial susceptibility pattersn of Arcobacter butzleri and Arcobacter cryaerophibs strains isolated from humans and broilers. Microb. Drug Resist. 10:243-247. Hum, S., K. Quinn, J. Brunner, and S. L. On. 1997. Evaluation of a PCR assay for identification and differentiation of Campylobacter fetus subspecies. Aust. Vet. J. 75:827-831. Hurtado, A., and R. J. Owen. 1997. A molecular scheme based on 23s rRNA gene polymorphisms for rapid identification of Campylobacter and Arcobacter species. J. Clin. Microbiol. 35:24012404.
CHAPTER 12
Hutchinson, D. N., and F. J. Bolton. 1984. Improved blood free selective medium for the isolation of Campylobacter jejuni from faecal specimens. 1.Clin. Pathol. 37:956-957. Inglis, G. D., and L. D. Kalischuk. 2003. Use of PCR for direct detection of Campylobacter species in bovine feces. Appl. Environ. Microbiol. 69:3435-3447. Jenkin, G. A., and W. Tee. 1998. Campylobacter upsaliensisassociated diarrhea in human immunodeficiency virus-infected patients. Clin. Infect. Dis. 27:816-821. Jesse, T. W., M. D. Englen, L. G. Pittenger-Alley, and P. FedorkaCray. 2006. Two distinct mutations in gyrA lead to ciprofloxacin and nalidixic acid resistance in Campylobacter coli and Campylobacter jejuni isolated from chickens and beef cattle. /, Appl. Microbiol. 100:682-688. Karlyshev, A. V., D. Linton, N. A. Gregson, A. J. Lastovica, and B. W. Wren. 2000. Genetic and biochemical evidence of a Campylobacter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 35529-541. Karmali, M. A., S. De Grandis, and P. C. Fleming. 1981. Antimicrobial susceptibility of Campylobacter jejuni with special reference to resistance patterns of Canadian isolates. Antimicrob. Agents Chemother. 19593-597. Karmali, M. A., and P. C. Fleming. 1979. Campylobacter enteritis in children.]. Pediatr. 94527-533. Karmali, M. A., A. E. Simor, M. Roscoe, P. C. Flemming, S. S. Smith, and J. Lane. 1986. Evaluation of a blood-free, charcoal based, selective medium for the isolation of Campylobacter organisms from feces. J. Clin. Microbiol. 23:456-459. Kasten, M. J., F. Allerberger, and J. P. Anhalt. 1991. Campylobacter bacteremia: clinical experience with three different blood culture systems at Mayo Clinic 1984-1990. Infection 19:88-90. Keramas, G., D. D. Bang, M. Lund, M. Madsen, S. E. Rasmussen, H. Bunkenborg, P. Telleman, and C. B. Christensen. 2003. Development of a sensitive DNA microarray suitable for rapid detection of Campylobacter spp. Mol. Cell. Probes 17:187-196. Kiehlbauch, J. A., C. N. Baker, and I. K. Wachsmuth. 1992. In vitro susceptibilities of aerotolerant Campylobacter isolates to 22 antimicrobial agents. Antimicrob. Agents Chemother. 36:717722. Kiehlbauch, J. A., D. J. Brenner, M. A. Nicholson, C. N. Baker, C. M. Patton, A. G. Steigerwalt, and I. K. Wachsmuth. 1991. Campylobacter butzleri sp. nov. isolated from humans and animals with diarrheal illness. 1. Clin. Microbiol. 29:376-385. Kinana, A. D., E. Cardinale, F. Tall, I. Bahsoun, J. M. Sire, B. Garin, S. Breurec, C. S. Boye, and J. D. Perrier-Gros-Claude. 2006. Genetic diversity and quinolone resistance in Campylobacter jejuni isolates from poultry in Senegal. Appl. Environ. Microbiol. 72:3309-3313. Klena, J. D., and M. E. Konkel. 2005. Methods for epidemiologic analysis of Campylobacter jejuni, p. 165-179. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Klena, J. D., C. T. Parker, K. Knibb, J. C. Ibbitt, P. M. Devane, S . T. Horn, W. G. Miller, and M. E. Konkel. 2004. Differentiation of Campylobacter coli, Campylobacter jejuni, Campylobacter lari, and Campylobacter upsaliensis by a multiplex PCR developed from the nucleotide sequence of the lipid A gene 1pxA.J. Clin. Microbiol. 425549-5557. Kulkarni, S. P., S. Lever, J. M. Logan, A. J. Lawson, J. Stanley, and M. S. Shafi. 2002. Detection of campylobacter species: a comparison of culture and polymerase chain reaction based methods. 1. Clin. Pathol. 55:749-753. Kuschner, R. A., A. F. Trofa, R. J. Thomas, C. W. Hoge, C. Pitarangsi, s. Amato, R. P. Olafson, P. Echeverria, J. c. Sadoff, and D. N. Taylor. 1995. Use of azithromycin for the treatment of Campylobacter enteritis in travelers to Thailand, an area
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
239
where ciprofloxacin resistance is prevalent. Clin. Infect. Dis. 21: 536-541. Kwon, S. Y., D. H. Cho, S. Y. Lee, K. Lee, andY. Chong. 1994. Antimicrobial susceptibility of Campylobacter fetus subsp. fetus isolated from blood and synovial fluid. Yonsei Med. I. 35:314319. la Scola, B., S. Chambourlier, C. Mercier, and J. P. Casalta. 1998. Abdominal aortic aneurysm infected by Campylobacter fetus subsp. fetus. Clin. Microbiol. Infect. 4527-529. Lachance, N., C. Gaudreau, F. Lamothe, and L. A. Lariviere. 1991. Role of the B-lactamase of Campylobacter jejuni in resistance to beta-lactam agents. Antimicrob. Agents Chemother. 35: 813-818. Lachance, N., C. Gaudreau, F. Lamothe, and F. Turgeon. 1993. Susceptibilitiesof beta-lactamase-positive and -negative strains of Campylobacter coli to beta-lactam agents. Antimicrob. Agents Chemother. 37: 1174-1 176. Lastovica, A. J. 2006. Emerging Campylobacter spp.: the tip of the iceburg. Clin. Microbiol. News 28:49-55. Lastovica, A. J., E. Le Roux, and J. L. Penner. 1989. “Campylobacter upsaliensis”: isolated from blood cultures of pediatric patients. J. Clin. Microbiol. 27:657-659. Lastovica, A. J., and M. B. Skirrow. 2000. Clinical significance of Campylobacter and related species other than Campylobacter jejuni, p. 89-121. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter. ASM Press, Washington, DC. Lawson, A. J., D. Linton, and J. Stanley. 1998. 16s rRNA gene sequences of “Candidatus Campylobacter hominis,” a novel uncultivated species, are found in the gastrointestinal tract of healthy humans. Microbiology 144(Pt. 8):2063-2071. Lawson, A. J., J. M. J. Logan, G. L. O’Neill, M. Desai, and J. Stanley. 1999. Large scale survey of Campylobacter species in human gastroenteritis by PCR and PCR-enzyme linked immunosorbent assay. J. Clin. Microbiol. 37:3860-3864. Le Ronx, E., and A. J. Lastovica. 1998. The Cape Town protocol: how to isolate the most campylobacters for your dollar, pound, franc, yen, etc., p. 30-33. In A. L. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. Institute of Child Health, Cape Town. Lerner, J., V. Brumberger, and M. Preac. 1994. Severe diarrhea associated with Arcobacter butzleri. Eur. 1. Clin. Microbiol. lnfect. Dis. 13:660-662. Li, C. C., C. H. Chiu, J. L. Wu, Y. C. Huang, and T. Y. Lin. 1998. Antimicrobial susceptibilities of Campylobacter jejuni and coli by using E-test in Taiwan. Scand. J. Infect. Dis. 30:39-42. Li, X.-Z., M. Mehrotra, S. Ghimire, and L. Adewoye. 2007. Llactam resistance and L-lactamases in bacteria of animal origin. Vet. Microbiol. 121:197-2 14. Lim, Y. S., and L. Tay. 1992. A one-year study of enteric Campylobacter infections in Singapore. J. Trop. Med. Hyg. 95:119123. Linton, D., A. J. Lawson, R J. Owen, and J. Stanley. 1997a. PCR detection, identification to species level, and fingerprinting of Campylobacter jejuni and Campylobacter coli direct from diarrheic samples. J. Clin. Microbiol. 35:2568-2572. Linton, D., A. J. Lawson, R. J. Owen, and J. Stanley. 1997b. PCR detection, identification to species level, and fingerprinting of Campylobacter jejuni and Campylobacter coli direct from diarrheic samples.]. Clin. Microbiol. 35:2568-2572. Linton, D., R. J. Owen, and J. Stanley. 1996. Rapid identification by PCR of the genus Campylobacter and of five Campylobacter species enteropathogenic for man and animals. Res. Microbiol. 147~707-718. Lior, H., D. L. Woodward, J. A. Edgar, L. J. Laroche, and P. Gill. 1982. Serotyping of Campylobacter jejuni by slide agglutination based on heat-labile antigenic factors. J. Clin. Microbiol. 15: 761-768.
240
FITZGERALD ET AL.
Logan, J. M. J., A. Burnens, D. Linton, A. J. Lawson, and J. Stanley. 2000. Campylobacter lanienae sp. nov., a new species isolated from workers in an abattoir. Int. J. Sys. Evol. Microbiol. 50~865-872. Luechtefeld, N. W., L. B. Reller, M. J. Blaser, and W.-L. L. Wang. 1982. Comparison of atmospheres of incubation for primary isolation of Campylobacter fetus subsp. jejuni from animal specimens:S% oxygen versus candle jar. J. Clin. Microbiol. 155357. MacFaddin, J. F. 2000a. Hippurate hydrolysis test, p. 188-204. In J. F. MacFaddin (ed.), Biochemical Tests for Identification of Medical Bacteria. Lippincott Williams, and Wilkins, Philadelphia. MacFaddin, J. F. 2000b. Indoxyl substrate hydrolysis tests, p. 233238. In J. F. MacFaddin (ed.), Biochemical Tests for Identification of Medical Bacteria. Lippincott Williams, and Wilkins, Philadelphia. Maher, M., C. Finnegan, E. Collins, B. Ward, C. Carroll, and M. Cormican. 2003. Evaluation of culture methods and a DNA probe-based PCR assay for detection of Campylobacter species in clinical specimens of feces. J. Clin. Microbiol. 41:2980-2986. Mandal, B. K., M. E. Ellis, E. M. Dunbar, and K. Whale. 1984. Double-blind placebo-controlled trial of erythromycin in the treatment of clinical campylobacter infection. J. Antimicrob. Chemother. 13:6 19-623. Martin, W. T., C. M. Patton, G. K. Morris, M. E. Potter, and N. D. Puhr. 1983. Selective enrichment broth for isolation of Campylobacter jejuni. J. Clin. Microbiol. 17:853-855. Matsheka, M. I., A. J. Lastovica, and B. G. Elisha. 2001. Molecular identification of Campylobacter concisus. J. Clin. Microbiol. 39:3 684-3689. McDermott, P. F., S. M. Bodeis, F. M. Aarestrup, S. Brown, M. Traczewski, P. Fedorka-Cray, M. Wallace, I. A. Critchley, C. Thornsberry, S. Graff, R. Flamm, J. Beyer, D. Shortridge, L. J. Piddock, V. Ricci, M. M. Johnson, R. N. Jones, B. Reller, S. Mirrett, J. Aldrobi, R. Rennie, C. Brosnikoff, L. Turnbull, G. Stein, S. Schooley, R. A. Hanson, and R. D. Walker. 2005a. Development of a standardized susceptibility test for Campylobacter with quality-control ranges for ciprofloxacin, doxycycline, erythromycin, gentamicin and meropenem. Microb. Drug Resist. 10:124-13 1. McDermott, P. F., S . M. Bodeis-Jones, T. R. Fritsche, R N. Jones, and R. D. Walker. 2005b. Broth microdilution susceptibility testing of Campylobacter jejuni and the determination of quality control ranges for fourteen antimicrobial agents. J. Clin. Microbiol. 43:6136-6138. McGill, K., D. Cowley, L. Moran, P. Scates, A. O’Leary, R. H. Madden, C. Carroll, E. McNamara, J. E. Moore, S. Fanning, J. D. Collins, and P. Whyte. 2006. Antibiotic resistance of retail food and human Campylobacter isolates on the island of Ireland from 2001-2002. Epidemiol. Infect. 134:1282-1291. McNulty, C. A. 1987. The treatment of campylobacter infections in man. J. Antimicrob. Chemother. 19:281-284. Megraud, F., D. Chevrier, N. Desplaces, A. Sedallian, and J. L. Guesdon. 1988. Urease-positive thermophilic Campylobacter (Campylobacter laridis variant) isolated from an appendix and from human feces. J. Clin. Microbiol. 26:1050-1051. Menard, A., F. Dachet, V. Prouzet-Mauleon, M. Oleastro, and F. Megraud. 2005. Development of a real-time fluorescence resonance energy transfer PCR to identify the main pathogenic Campylobacter spp. Clin. Microbiol. Infect. 11:281-287. Miller, W. G., C. T. Parker, S. Heath, and A. J. Lastovica. 2007. Identification of genomic differences between Campylobacter jejuni subsp. jejuni and C. jejuni subsp. doylei at the nap locus leads to the development of a C. jejuni subspeciation multiplex PCR method. BMC Microbiol. 7:11.
Mohran, Z. S., R. R. Arthur, B. A. Oyofo, L. F. Peruski, M. 0. Wasfy, T. F. Ismail, and J. R. Murphy. 1998. Differentiation of Campylobacter isolates on the basis of sensitivity to boiling in water as measured by PCR-detectable DNA. Appl. Environ. Microbiol. 6 4 3 63-3 65, Molina, J., I. Casin, P. Hausfater, E. Giretti, Y. Welker, J. Decazes, V. Garrait, P. Lagrange, and J. Modai. 1995. Campylobacter infections in HIV-infected patients: clinical and bacteriological features. AIDS 9:881-885. Morris, G. K., and C. M. Patton. 1985. Campylobacter, p. 302308. In E. Lennette, A. Balows, W. J. Hausler Jr., and H. J. Shadomy (ed.), Manual of Clinical Microbiology. American Society for Microbiology, Washington, DC. Murphy, G. S., Jr., P. Echeverria, L. R Jackson, M. K. Arness, C. LeBron, and C. Pitarangsi. 1996. Ciprofloxacin- and azithromycin-resistant Campylobacter causing traveler’s diarrhea in U.S. troops deployed to Thailand in 1994. Clin. Infect. Dis. 22: 868-869. Nachamkin, I. 1994. Antimicrobial susceptibility of Campylobacter jejuni and Campylobacter coli to ciprofloxacin, erythromycin and tetracycline from 1982 to 1992. Med. Microbiol. Lett. 3: 300-3 05. Nachamkin, I. 1997. Microbiologic approaches for studying Campylobacter in patients with Guillain-BarrC syndrome. J. Infect. Dis. 176(Suppl. 2):S106-S114. Nachamkin, I., and S. Barbagallo. 1990. Culture confirmation of Campylobacter spp. by latex agglutination. J. Clin. Microbiol. 28: 817-8 18. Nachamkin, I., C. Stowell, D. Skalina, A. M. Jones, R. M. Roop, and R. M. Smibert. 1984. Campylobacter laridis causing bacteremia in an immunosuppressed patient. Ann. Intern. Med. 101: 55-57. National Antimicrobial Resistance Monitoring System. 2006. Enteric Bacteria. NARMS 2003 Executive Report. U S . Department of Health and Human Services, U.S. Food and Drug Administration, Bethesda, Md. Nelson, J. M., T. M. Chiller, J. H. Powers, and F. J. Angulo. 2007. Fluoroquinolone-resistant Campylobacter species and the withdrawal of fluoroquinolones from use in poultry: a public health success story. Clin. Infect. Dis. 44:977-980. Newell, D. G., J. A. Frost, B. Duim, J. A. Wagenaar, R. H. Madden, J. van der Plas, and s. L. W. On. 2000. New developments in the subtyping of Campylobacter species, p. 27-44. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter. ASM Press, Washington, DC. Ng, L.-K., D. E. Taylor, and M. E. Stiles. 1988. Characterization of freshly isolated Campylobacter coli strains and suitability of selective media for their growth. J. Clin. Microbiol. 2 6 5 18-523. Nolan, C. M., K. E. Johnson, M. B. Coyle, and K. Faler. 1983. Campylobacter jejuni enteritis: efficacy of antimicrobial and antimotility drugs. Am. J. Gastroenterol. 78:621-626. Olsen, S. J., G. R. Hansen, L. Bartlett, C. Fitzgerald, A. Sonder, R. Manjrekar, T. Riggs, J. Kim, R Flahart, G. Pezzino, and D. L. Swerdlow. 2001. An outbreak of Campylobacter jejuni infections associated with food handler contamination: the use of pulsed-field gel electrophoresis. J. Infect. Dis. 183:164-167. On, S. L. 1996. Identification methods for campylobacters, helicobacters, and related organisms. Clin. Microbiol. Rev. 9:405422. On, S. L., and P. J. Jordan. 2003. Evaluation of 11 PCR assays for species-level identification of Campylobacter jejuni and Campylobacter coli. Clin. Microbiol. 41:330-336. On, S . L. W. 2005. Taxonomy, phylogeny and methods for the identification of Campylobacter species, p. 13-42. In J. M. K. Konkel (ed.), Campylobacter Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom.
].
CHAPTER 12
Paisley, J. W., S. Mirrett, b. A. Lauer, M. Roe, and L. B. Reller. 1982. Dark-field microscopy of human feces for presumptive diagnosis of Campylobacter fetus subsp. jejuni enteritis. J. Clin. Microbiol. 1561-63. Papavasileiou, E., A. Voyatzi, K. Papavasileiou, A. Makri, I. Andrianopoulou, and S. Chatzipanagiotou. 2007. Antimicrobial susceptibilities of Campylobacter jejuni isolates from hospitalized children in Athens, Greece, collected during. 2004-2005. Eur. J. Epidemiol. 22:77-78. Park, C. H., D. L. Hixon, A. S. Polhemus, C. B. Ferguson, S. L. Hall, C. C. Risheim, and C. B. Cook. 1983. A rapid diagnosis of campylobacter enteritis by direct smear examination. Am. J. Clin. Pathol. 80:388-390. Patton, C. M., and I. K. Wachsmuth. 1992. Typing schemes: are current methods useful? p. 10-128. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Penner, J. L., and J. N. Hennessy. 1980. Passive hemagglutination technique for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heat-stable antigens. J. Clin. Microbiol. 12: 732-737. Perlman, D. M., N. M. Ampel, R B. Schifman, D. L. Cohn, C. M. Patton, M. L. Aguirre, W.-L. L. Wang, and M. J. Blaser. 1988. Persistent Campylobacter jejuni infections in patients infected with human immunodeficiency virus (HIV). Ann. Intern. Med. 108:540-546. Persson, S., and K. E. Olsen. 2005. Multiplex PCR for identification of Campylobacter coli and Campylobacter jejuni from pure cultures and directly on stool samples. J. Med. Microbiol. 54: 1043-1047. Piddock, L. J. 1997. Quinolone resistance and Campylobacter, p. 191-199. In WHO (ed.), The Medical Impact of the Use ofAntimicrobial Use in Food Animals. Report and proceedings of a WHO meeting, 13 to 17 October 1997, Berlin, Germany. Division of Emerging and Other Communicable Diseases Surveillance and Control, WHO, Geneva. Piddock, L. J. 1995. Quinolone resistance and Campylobacter spp. J. Antimicrob. Chemother. 36:89 1-898. Pigrau, C., R. Bartolome, B. Almirante, A. M. Planes, J. Gavalda, and A. Pahissa. 1997. Bacteremia due to Campylobacter species: clinical findings and antimicrobial susceptibility patterns. Clin. Infect. Dis. 25:1414-1420. Pitkanen, T., T. Pettersson, A. Pomka, and T. U. Kosunen. 1982. Effect of erythromycin on the fecal excretion of Campylobacter fetus subspecies jejuni. J. Infect. Dis. 145:128. Popovic-Uroic, T., C. M. Patton, I. K. Wachsmuth, and P. Roeder. 1991. Evaluation of an oligonucleotide probe for identification of Campylobacter species. Lab. Med. 22533-539. Preston, M. A., A. E. Simor, S. L. Walmsley, S. A. Fuller, A. J. Lastovica, K. Sandstedt, and J. L. Penner. 1990. In vitro susceptibility of “Campylobacter upsaliensis” to twenty-four antimicrobial agents. Eur. J. Clin. Microbiol. Infect. Dis. 9:822-824. Quinones, B., C. T. Parker, J. M. Janda, Jr., W. G. Miller, and R. E. Mandrell. 2007. Detection and genotyping of Arcobacter and Campylobacter isolates from retail chicken samples by use of DNA oligonucleotide arrays. Appl. Environ. Microbiol. 73: 3 645-3 655. Rautelin, H., 0. V. Renkonen, and T. U. Kosunen. 1991. Emergence of fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli in subjects from Finland. Antimicrob. Agents Chemother. 35:2065-2069. Reller, L. B., S. Mirrett, and L. G. Reimer. 1983. Controlled evaluation of an improved selective medium for isolation of Campylobacter jejuni, abstr. C-274. Abst. Ann. Meet. Am. SOC. Microbiol.
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
241
Reina, J., N. Borrell, and A. Serra. 1992. Emergence of resistance to erythromycin and fluoroqninolones in thermotolerant Campylobacter strains isolated from feces 1987-1991. Eur. J. Clin. Microbiol. Infect. Dis. 11:1163-1166. Reina, J., M. J. Ros, and A. Munoz-Lopez. 1996. Isolation of a strain of Campylobacter jejuni resistant to fluoroquinolones and sensitive to nalidixic acid. Enferm. Infecc. Microbiol. Clin. 14: 124-125. Reina, J., M. J. Ros, and A. Serra. 1994. Susceptib antimicrobial agents of 1,220 Campylobacter strains isolated from. 1987 to. 1993 from feces of pediatric patients. Antimicrob. Agents Chernother. 38:2917-2920. Ribot, E. M., C. Fitzgerald, K. Kubota, B. Swaminathan, and T. J. Barrett. 2001. Rapid pulsed-field gel electrophoresis protocol for subtyping Campylobacter jejuni. J. Clin. Microbiol. 39:18891894. Rodriguez-Avial, I., C. Rodriguez-Avail, 0. Lopez, E. Culebras, and J. J. Picazo. 2006. In vitro activity of tigecycline (GAR-936) and other antimicrobials against tetracycline- and ciprofloxacinresistant Campylobacter clinical isolates. Int. J. Antimicrob. Agents 27:303-306. Rubin, S. J., and M. Woodard. 1983. Enhanced isolation of Campylobacter jejuni by cold enrichment in Campy-thio broth. J. Clin. Microbiol. 18:1008-1 0 10. Ruiz, J., P. Goni, F. Marco, F. Gallardo, B. Mirelis, D. A. Jimenez, and J. Vila. 1998. Increased resistance to quinolones in Campylobacter jejuni: a genetic analysis of gyrA gene mutations in quinolone-resistant clinical isolates. Microbiol. Immunol. 42: 223-226. Sails, A. D., B. Swaminathan, and P. I. Fields. 2003. Utility of multilocus sequence typing as an epidemiological tool for investigation of outbreaks of gastroenteritis caused by Campylobacter jejuni. J. Clin. Microbiol. 41:4733-4739. Sanchez, R., B. Fernandez, M. D. Diaz, P. Munoz, C. Rodriguez, and E. Bouza. 1994. Evolution of susceptib bacter spp. to quinolones and macrolides. Antimicrob. Agents Chemother. 38:1879-1882. Sazie, E. S. M., and A. E. Titus. 1982. Rapid diagnosis of Campylobacter enteritis. Ann. Intern. Med. 96:62-63. Schwartz, D., H. Goossens, J. Levy, J. P. Butzler, and J. Goldhar. 1993. Plasmid profiles and antimicrobial susceptibility of Campylobacter jejuni isolated from Israeli children with diarrhea. lnt. J. Med. Microbiol. Virol. Parasitol. Infect. Dis. 279:368-376. Segreti, J., T. D. Gootz, L. J. Goodman, G. W. Parkhurst, J. P. Quinn, B. A. Martin, and G . M. Trenholme. 1992. High-level quinolone resistance in clinical isolates of Campylobacter jejuni. J. Infect. Dis. 165:667-670. Siemer, B. L., E. M. Nielsen, and S. L. On. 2005. Identification and molecular epidemiology of Campylobacter coli isolates from human gastroenteritis, food, and animal sources by amplified fragment length polymorphism analysis and Penner serotyping. Appl. Environ. Microbiol. 71:1953-1958. Simor, A. E., and L. Wilcox. 1987. Enteritis associated with Campylobacter laridis. J. Clin. Microbiol. 25:lO-12. Siqueira, J. F., Jr., and I. N. Rocas. 2003. Campylobacter gracilis and Campylobacter rectus in primary endodontic infections. lnt. Endodon. J. 36:174-180. Sjogren, E., G. B. Lindblom, and B. Kaijser. 1997. Norfloxacin resistance in Campylobacter jejuni and Campylobacter coli isolates from Swedish patients. J. Antimicrob. Chemother. 40:257261. Skirrow, M. B. 1977. Campylobacter enteritis: a “new” disease. Br. Med. J. ii:9-11. Skirrow, M. B., and M. J. Blaser. 1995. Campylobacter jejuni, p. 825-848. In M. J. Blaser, P. D. Smith, J. I. Ravdin, H. B. Green-
242
FITZGERALD ET AL.
berg, and R. L. Guerrant (ed.), Infections o f Gastrointestinal Tract. New York: Raven Press. Skirrow, M. B., D. M. Jones, E. Sutcliffe, and J. Benjamin. 1993. Campylobacter bacteremia in England and Wales, 1981-1991. Epidemiol. Infect. 110567-573. Snijders, F., E. J. Kuijper, B. de Wever, L. van der Hoek, S. A. Danner, and J. Dankert. 1997. Prevalence of Campylobacterassociated diarrhea among patients infected with human immunodeficiency virus. Clin. Infect. Dis. 24:1107-1113. Steinbruckner, B., J. Fehrenbach, G. Philippczik, M. Kist, and T. M. Bauer. 1999. Clinical significance of pure or predominant growth of Pseudomonas aeruginosa in faecal specimens of medical patients. J. Hosp. Infect. 43:164-165. Stucki, U., J. Frey, J. Nicolet, and A. P. Burnens. 1995. Identification of Campylobacter jejuni on the basis of a species-specific gene that encodes a membrane protein. J. Clin. Microbiol. 33: 855-859. Tajada, P., G. Gomez, J. I. Alos, D. Balas, and R. Cogollos. 1996. Antimicrobial susceptibilities of Campylobacter jejuni and Campylobacter coli to 12 beta-lactam agents and combinations with beta-lactamase inhibitors. Antimicrob. Agents Chemother. 40: 1924-1 925. Taylor, B. V., J. Williamson, J. Luck, D. Coleman, D. Jones, and A. McGregor. 2004. Sensitivity and specificity of serology in determining recent acute Campylobacter infection. Internal. Med. J. 34:250-258. Taylor, D. N. 1992. Campylobacter infections in developing countries, p. 20-30. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Taylor, D. N., M. J. Blaser, P. Echeverria, C. Pitarangsi, L. Bodhidatta, and W. L. Wang. 1987. Erythromycin-resistant Campylobacter infections in Thailand. Antimicrob. Agents Chemother. 3 1:438-442. Taylor, D. N., D. M. Perlman, P. D. Echeverria, U. Lexomboon, and M. J. Blaser. 1993. Campylobacter immunity and quantitative excretion rates in Thai children. J. Infect. Dis. 168:754758. Tee, W., and A. Mijch. 1998. Campylobacter jejuni bacteremia in human immunodeficiency virus (H1V)-infected and non-HIVinfected patients: comparison of clinical features and review. Clin. Infect. Dis. 26:91-96. Tee, W., A. Mijch, E. Wright, and A. Yung. 1995. Emergence of multidrug resistance in Campylobacter jejuni isolates from three patients infected with human immunodeficiency virus. Clin. Infect. Dis. 21:634-63 8 . Tenover, F. C., L. Carlson, S. Barbagallo, and I. Nachamkin. 1990. DNA probe culture confirmation assay for identification of thermophilic Campylobacter species. J. Clin. Microbiol. 28: 1284-1287. Thompson, J. S., D. S. Hodge, D. E. Smith, and Y. A. Yong. 1990. Use of tri-gas incubator for routine culture of Campylobacter species from fecal specimens. J. Clin. Microbiol. 28:2802-2803. Thwaites, R. T., and J. A. Frost. 1999. Drug resistance in Campylobacter jejuni, C. coli, and C. lari isolated from humans in north west England and Wales,. 1997. J. Clin. Pathol. 52:812814. Tissari, P., and H. Rautelin. 2007. Evaluation of an enzyme immunoassay-based stool antigen test to detect Campylobacter jejuni and Campylobacter coli. Diagn. Microbiol. Infect. Dis. 58: 171-175. Tolcin, R., M. M. LaSalvia, B. A. Kirkley, E. A. Vetter, F. Cockerill, and G. W. Procop. 2000. Evaluation of the Alexon-Trend ProSpecT Campylobacter microplate assay. J. Clin. Micriobol. 38:3853-3855.
Totten, P. A., C. M. Patton, F. C. Tenover, T. J. Barrett, W. E. Stamm, A. G. Steigerwalt, J. Y. Lin, K. K. Holmes, and D. J. Brenner. 1987. Prevalence and characterization of hippuratenegative Campylobacter jejuni in King County, Washington. J. Clin. Microbiol. 25:1747-1752. Tremblay, C., and C. Gaudreau. 1998. Antimicrobial susceptibility testing of 59 strains of Campylobacter fetus subsp. fetus. Antimicrob. Agents Chemother. 42:1847-1849. Tremblay, C., C. Gaudreau, and M. Lorange. 2003. Epidemiology and antimicrobial susceptibilities of 111 Campylobacter fetus subsp. fetus strains isolated in Quebec, Canada, from 1983 to 2000. J. Clin. Microbiol. 41:463-466. Unicomb, L. E., J. Ferguson, R. J. Stafford, R. Ashbolt, M. D. Kirk, N. G. Becker, M. S. Patel, G. L. Gilbert, M. Valcanis, and L. Mickan, for the Australian Campylobacter Subtyping Study Group. 2006. Low-level fluoroquinolone resistance among Campylobacter jejuni isolates in Australia. Clin. Infect. Dis. 42: 13681374. Valenstein, P., M. Pfaller, and M. Yungbluth. 1996. The use and abuse of routine stool microbiology: a College of American Pathologists Q-probes study of 601 institutions. Arch. Pathol. Lab. Med. 120:206-211. van Bergen, M. A., G. Simons, van der Graaf-van Bloois, J. P. van Putten, J. Rombout, I. Wesley, and J. A. Wagenaar. 2005. Amplified fragment length polymorphism based identification of genetic markers and novel PCR assay for differentiation of Campylobacter fetus subspecies. ]. Med. Microbiol. 54:1217-1224. Van der Auwera, P., and B. Scorneaux. 1985. In vitro susceptibility of Campylobacter jejuni to 27 antimicrobial agents and various combinations of beta-lactams with clavulanic acid or sulbactam. Antimicrob. Agents Chemother. 28:37-40. van Doorn, L. J., A. V. Van Haperen, A. Burnens, M. Huysmans, P. Vandamme, B. A. J. Giesendorf, M. J. Blaser, and W. G. V. Quint. 1999. Rapid identification of thermotolerant Campylobacter jejuni, Campylobacter coli, Campylobacter lari, and Campylobacter upsaliensis from various geographic regions by a GTPase-based PCR-reverse hybridization assay. J. Clin. Micriobol. 3 7: 1790-1 796. van Hees, B. C., M. J. VeldmawAriesen, B. M. de Jongh, M. Tersmette, and W. van Pelt. 2007. Regional and seasonal differences in incidence and antibiotic resistance of Campylobacter from a nationwide surveillance study in the Netherlands: an overview of 2000-2004. Clin. Microbiol. Infect. 13:305-310. Vandamme, P. 2000. Taxonomy of the Family Campylobacteraceae, p. 3-26. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter. ASM Press, Washington, DC. Vandamme, P., and J. De Ley. 1991. Proposal for a new family, Campylobacteraceae.Int. J. Syst. Bacteriol. 41:451-455. Vandamme, P., E. Falsen, R. Rossau, B. Hoste, P. Segers, R. Tytgat, and J. De Ley. 1991. Revision of Campylobacter, Helicobacter, and Wolinella taxonomy: emendation of generic descriptions and proposal of Arcobacter gen. nov. Int. J. Syst. Bacteriol. 41~81-103. Vandamme, P., L. J. van Doorn, S. T. A1 Rashid, W. G. Quint, J. van der Plas, V. L. Chan, and S. L. On. 1997. Campylobacter hyoilei Alderton et al.. 1995 and Campylobacter coli Veron and Chatelain 1973 are subjective synonyms. Int. J. Syst. Bacteriol. 47:1055-1060. Vandamme, P., M. Vancanneyt, B. Pot, L. Mels, B. Hoste, D. Dewettinck, L. Vlaes, C. van den Borre, R. Higgins, J. Hommez, K. Kersters, J. P. Butzler, and H. Goossens. 1992. Polyphasic taxonomic study of the emended genus Arcobacter with Arcobacter butzleri comb. nov. and Arcobacter skirrowii sp. nov., an aerotolerant bacterium isolated from veterinary specimens. Int. J. Syst. Bacteriol. 42:344-356.
CHAPTER 12
Vandenberg, O., A. Dediste, K. Houf, S. Ibekwem, H. Souayah, S . Cadranel, N. Douat, G. Zissis, J. P. Butzler, and P. Vandamme. 2004. Arcobucter species in humans. Emerg. Infect. Dis. 10~1863-1867. Vandenberg, O., K. Houf, N. Douat, L. Vlaes, P. Retore, J. P. Butzler, and A. Dediste. 2006. Antimicrobial susceptibility of clinical isolates of non-jejuni/coli campylobacters and arcobacters from Belgium. J. Antimicrob. Chernother. 57:908-913. Vanhoof, R., M. P. Vanderlinden, R. Dierickx, S . Lauwers, E. Yourassowsky, and J. P. Butzler. 1978. Susceptibility of Campylobacter fetus subsp. jejuni to twenty-nine antimicrobial agents. Antimicrob. Agents Chemother. 14553-556. Vanniasinkam, T., J. A. Lamer, and M. D. Barton. 1999. PCR for the detection of Cumpylobucter spp. in clinical specimens. Lett. Appl. Microbiol. 2852-56. Varela, N. P., R. Friendship, and C. Dewey. 2007. Prevalence of resistance to 11 antimicrobials among C. coli isolated from pigs on 80 grower-finisher farms in Ontario. Can. J. Vet. Res. 71: 189-194. Volokhov, D., V. Chizhikov, K. Chumakov, and A. Rasooly. 2003. Microarray-based identification of thermophilic Campylobacter
DIAGNOSIS AND ANTIMICROBIAL SUSCEPTIBILITY
243
jejuni, C. coli, C. lari, and C. upsaliensis.1.Clin. Microbiol. 41: 4071-4080. Wang, H., and D. R. Murdoch. 2004. Detection of Cumpylobacter species in faecal samples by direct gram stain microscopy. Pathology 36:343-344. Wang, W.-L. L., and M. J. Blaser. 1986. Detection of pathogenic Cumpylobucter species in blood culture systems. I; Clin. Microbiol. 23:709-714. Wang, W.-L. L., L. B. Reller, B. Smallwood, N. W. Luechtefeld, and M. J. Blaser. 1983. Evaluation of transport media for Campylobacter jejuni in human fecal specimens. ]. Clin. Microbiol. 18~803-807. Wasfy, M., B. Oyofo, A. Elgindy, and A. Churilla. 1995. Comparison of preservation media for storage of stool samples. J. Clin. Microbiol. 3 3 :2 176-21 78. Wells, J. G . N. D. Puhr, C. M. Patton, M. A. Nicholson, M. A. Lambert, and R. Jerris. 1989. Comparison of selective media and filtration for the isolation of Campylobucter from feces, abstr. C231. Abstr. Ann. Meet. Am. Soc. Microbiol. Wistrom, J., and S . R. Norrby. 1995. Fluoroquinolones and bacterial enteritis, when and for whom? J. Antimicrob. Chemother. 36~23-39.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 13
Guillain-Barre Syndrome and Campylobacter Infection BART c. JACOBS,
ALEX VAN
BELKUM,AND HUBERT P. ENDTZ
GBS is presently a descriptive disease entity defined by a set of clinical, laboratory, and electrodiagnostic criteria (Table 1). GBS is a polyradiculoneuropathy characterized by an acute progressive and symmetrical motor weakness of the extremities with loss of tendon reflexes (Asbury and Cornblath, 1990). A specific subform of GBS, the Miller Fisher syndrome (MFS), is characterized by areflexia, ataxia, and ophthalmoplegia (Miller Fisher, 1956). Surveillance studies from widely scattered geographical areas showed that the annual incidence rate of GBS per 100,000 subjects ranges from 0.6 to 1.9 (Schonberger et al., 1981). In some areas of the world, the incidence of GBS may be higher, as has been suggested for endemic forms in rural areas in northern China (Ho et al., 1995; Liu and Jou, 1993), or may fluctuate in time, as was demonstrated in CuraSao (van Koningsveld et al., 2001). On the basis of the abovementioned incidence figures, 40,000 to 120,000 new cases are being diagnosed per year worldwide. GBS may affect persons of all ages but has an increased incidence in the elderly (Dowling et al., 1977; Schonberger et al., 1981; van Koningsveld et al., 2000). GBS occurs slightly more often in male subjects than in female subjects (Hurwitz et al., 1983; van Koningsveld et al., 2000). Multiple cases of GBS within one family have been described, although this appears to be an infrequent event (Geleijns et al., 2004). Two-thirds of patients with GBS report symptoms of a preceding gastrointestinal or respiratory infectious illness from which they usually recover spontaneously (van Koningsveld et al., 2000; Winer et al., 1988a, 1988b). One to 3 weeks later, the patient may notice numbness or tingling in the arms or legs with symmetrical loss of strength in hands or feet. The
Guillain-Barrt syndrome (GBS) is the most common form of acute neuromuscular paralysis in countries where poliomyelitis has been eradicated. The major clinical features of the syndrome were first united by J.-B. 0. Landry in 1859. The syndrome was named after G. Guillain and J. A. Barrt, two French army neurologists, who in 1916 described, together with A. Strohl, the typical findings in the cerebrospinal fluid (Guillain et al., 1916). This dissociation albumino-cytologique, the combination of high protein level and normal cell count in cerebrospinal fluid, was in those days an important diagnostic feature that distinguished the syndrome from other neurological diseases, including poliomyelitis. GBS was early recognized as a typical postinfectious disease, although it took several decades before the types of antecedent infection were identified. In 1982, shortly after the development of adequate culture systems, Rhodes and Tattersfield (1982) reported the first isolation of Campylobacter jejuni from the stools of a patient with GBS. This discovery set the stage for research on the pathogenesis of C. jejuni-related GBS. At present, there is convincing evidence that C. jejuni is the predominant preceding infection in GBS and that this infection triggers the production of crossreactive antibodies to gangliosides that are damaging peripheral nerve tissue. GBS is currently considered to be a true case of molecular mimicry mediated disease, at least in those patients with a preceding jejuni infection. All these discoveries were made in the last 25 years. One of the challenges of the next decennium will be to translate the newly acquired knowledge into interventions in order to prevent disease, or to improve recovery and lower morbidity and mortality in GBS patients.
c.
Bart C. Jacobs Departments of Neurology and Immunology, Erasmus University Medical Center, Rotterdam, 3015 CE The Netherlands. Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Center, Alex van Belkum Rotterdam, 3015 CE The Netherlands. Hubert P. Endtz International Centre for Diarrhoea1 Disease Research, Bangladesh, Dhaka 1212, Bangladesh, and Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Center, Rotterdam, 3015 CE The Netherlands.
-
245
246
IACOBS ET AL.
Table 1. Diagnostic criteria for Guillain-Barrt syndrome" Features required for diagnosis Progressive motor weakness of more than one limb Low or absent reflexes * No other identifiable cause Features strongly supporting the diagnosis Clinical Progression of less than 4 weeks Relative symmetry Mild sensory symptoms and signs Cranial nerve involvement Recovery beginning 2 weeks to months after progression ceases Autonomic dysfunction Absence of fever at onset Cerebrospinal fluid (CSF) High concentration of CSF protein after the first week Less than 50 mononuclear cells per pl CSF Electrodiagnostic Conduction slowing or block Features casting doubt on the diagnosis Marked, persistent asymmetry of weakness Persistent bladder or bowel dysfunction Bladder or bowel dysfunction at onset More than 50 mononuclear leukocytes per pl CSF Presence of polymorphonuclear leukocytes in CSF Sharp sensory level Onset with respiratory failure disproportionate to limb weakness "Adapted from the criteria supported by the National Institute of Neurological and Communication Disorders and Stroke (Asbury et al. 1990).
weakness spreads during the next few hours to weeks and may finally lead to a paralysis of limb, trunk, extraocular, facial, pharyngeal, and tongue musculature. Because of the involvement of respiratory muscles, up to one-third of the patients need to be ventilated (Winer et al., 1988a, 1988b). Patients may also have severe pain, diminished sensibility, or autonomic dysfunction. Clinical examination of the patients shows poor or absent tendon reflexes. The diagnosis is further supported by a high protein level in combination with normal cell count in cerebrospinal fluid and by features of demyelination and axonal degeneration in electrodiagnostic studies. The progression of weakness ceases by definition within 4 weeks. Recovery usually begins after a plateau phase of 2 to 4 weeks but may be delayed for months (Fig. 1). General medical support is of the utmost importance in treating GBS patients, especially to prevent secondary complications such as infections and decubitus. Additionally, patients will recover more rapidly from their paresis and have less residual deficits after treatment with plasma exchange (French Cooperative Group on Plasma Exchange in GBS, 1987;
GBS Study Group, 1985) or intravenous immunoglobulins (Plasma ExchangelSandoglobulin GBS Trial Group, 1997; van der MechC et al., 1992; van Koningsveld et al., 2004). The mortality is usually less than 5 to 10% and is mostly the result of respiratory or autonomic complications (Ropper et al., 1991). Recovery may be complete, but most patients have some residual handicap. Approximately 20% of patients have a severe functional deficit and remain unable to walk or bedridden, and they even may need continuous artificial ventilation (Ropper et al., 1991; van Koningsveld et al., 2007). A prognostic model was developed that accurately predicts which patients will have a poor outcome (van Koningsveld et al., 2007). GBS usually has a monophasic course, but treatment-related fluctuations occur in 10 to 15% of patients (Kleyweg and van der Mech6, 1991), and about 1 to 5% may experience relapse (Ropper et al., 1991). HETEROGENEITY AND SUBGROUPS GBS is an accepted disease entity, but there is marked patient-to-patient variation regarding disease severity, distribution of neurological deficits, electrophysiological findings, type of antecedent infection, and specificity of autoantibodies (Table 2). The extent of the motor deficits ranges from minimal weakness of the legs to a total paralysis of all limb, trunk, eye, facial, and bulbar muscles, a grave situation in which the patient has full awareness but is unable to communicate. A vivid description of such a situation was given by Joseph Heller (Heller and Vogel, 1986). The distribution of motor deficits varies between patients, as illustrated by the selective weakness of proximal limb muscles in some cases and of distal muscles in others. Sensory involvement may be completely absent, as in the case of a pure motor GBS, or may be severe and long-lasting and may include all sensory modalities. Some patients may rapidly recover without residual deficits, even from severe GBS, whereas others need ventilation for months or remain bed-bound for years. This clinical heterogeneity is further illustrated by the presence of distinct variants of GBS such as the MFS, in which the paresis is restricted to the muscles involved in eye movements (ophthalmoplegia), frequently in combination with bulbar or facial weakness (Fisher, 1956). Other subforms are the pharyngeal-cervical-brachial and the lower bulbar variants of GBS (Ropper et al., 1991). The extent of demyelization and axonal degeneration of peripheral nerves may also differ considerably between individual patients. The acute inflammatory demyelinating polyneuropathy (AIDP) variant, characterized by segmental demyelination
CHAPTJZR 13
SYNDROME AND CAMPYLOBACTER
GUILLAIN-BA&
247
................... i Admission i
1
,
,
~
,
,
,
,
,
,
,
lInfeaionl I
,
,
,
,
,
,
,
,
,
,
,
,
Weeks Serum antibodies to clanclliosides
1-
I
Plateau phase
with no or minor secondary axonal degeneration, is the predominant type of GBS in Western countries (Hadden et al., 2001; van Koningsveld et al., 2004). The acute motor axonal neuropathy (AMAN) and acute motor sensory axonal neuropathy variants, characterized by predominant axonal involvement,
I Recovery phase
Disability
are more frequent in Asia and developing countries (Griffin et al., 1995; Islam et al., 2007; McKhann et al., 1993). The involvement of axonal degeneration is indicated by electrophysiological findings but can only be proven in histological studies (Cros and Triggs, 1994). It is as yet unclear whether the AIDP
Table 2. Heterogeneity in Guillain-BarrC syndrome Variation Neurological deficits
Subforms
......... Motor-sensory
form Pure motor variant Miller Fisher syndrome (MFS) Lower bulbar variant Pure sensory variant Clinical severity .............Mild variant Total paralysis with respiratory insufficiency and autonomic dysfunction Electrophysiology/ histology .. .Acute inflammatory demyelinating polyneuropathy (AIDP) Acute motor axonal neuropathy (AMAN) Acute motor sensory axonal neuropathy (AMSAN) Antecedent infections ........Campylobacter jejuni Cytomegalovirus Epstein-Barr virus Mycoplasma pneumoniae Haemophilus influenzae Anti-ganglioside antibodies. ...GM1, GMlb, (GalNAc-)GDla (pure motor variant, AMAN) GQlb, GTla, GD3, GD2 (MFS, lower bulbar variant) GM2, GM3, LM1 Clinical outcome ............Spontaneous and full recovery Severely disabled Treatment-related fluctuations Relapse
248
TACOBS ET AL.
and axonal forms are absolute distinct entities resulting from exclusive pathogenic mechanisms, or whether they represent extreme forms in a continuous spectrum of a syndromic disease entity.
IMMUNOPATHOLOGY There is convincing evidence from extensive histology, serology, and animal model studies that GBS is caused by an autoimmune response. This parallels the failure of natural immune tolerance in other disorders such as rheumatoid arthritis, systemic lupus erythematosus, and multiple sclerosis, which are generally classified as autoimmune diseases. However, most of these classic autoimmune diseases have a specific age at onset, are more frequent in female subjects, and are associated with family, human leukocyte antigen (HLA) haplotype, and other autoimmune diseases, and usually show a relapsingremitting or chronic course that frequently improves after treatment with corticosteroids. None of these characteristics, however, is evident in GBS (Table 3). Instead, GBS is strongly associated with preceding acute infections, which are usually much less evident in these classic autoimmune diseases. GBS is therefore frequently classified as a typical postinfectious d’isease. In about half of GBS patients, serum antibodies to various gangliosides can be demonstrated in the acute stage of disease (Willison, 2005; Willison and Yuki, 2002; Yuki, 2007). Gangliosides are glycolipids that are highly enriched in neural membranes (Ledeen, 1985). They consist of a ceramide tail inserted in the lipid bilayer and a highly variable oligosaccharide moiety containing sialic acid that protrudes extracellulary (Ledeen and Yu, 1982). Gangliosides are organized in specialized functional microdomains called lipid rafts, where they play a role in mainte-
nance of the cell membrane structure and are implicated in cell growth, differentiation, and cell-cell recognition. A variety of gangliosides have been identified in the nerve myelin sheets and axons of human peripheral nerves, each with a specific tissue distribution (Ledeen, 1985). Antibodies to more than 20 different gangliosides have been identified in various frequencies in serum from GBS patients. The most frequent antibodies in GBS are directed to the gangliosides LM1, GM1, GMlb, GM2, GDla, GalNAc-GDla, GDlb, GD2, GD3, GTla, and GQlb (Willison and Yuki, 2002). These antibodies specifically recognize parts of the oligosaccharide moieties and may bind to various gangliosides with the similar carbohydrate structures. Other antibodies bind to mixtures or complexes of different gangliosides instead of individual gangliosides (Kaida et al., 2006, 2007; Kuijf et al., 2007). The highest titers of serum antiganglioside antibodies are usually found in the acute stage of disease (Fig. 1).These antibodies usually disappear within weeks, although they may persist for more than 6 months, and antibodies may also occasionally persist in patients who have completely recovered. The antiganglioside antibodies are predominantly immunoglobulin (Ig) G antibodies and usually of the IgGl and IgG3 subclass, but IgM and IgA antibodies can also be demonstrated. Typically they are oligoclonal or even monoclonal, polyreactive, and have a moderate affinity (Boffey et al., 2005; Townson et al., 2007). Antibodies to other glycolipids and even antibodies and T cells to peripheral nerve proteins have also been demonstrated in serum from GBS patients. Some antiganglioside antibodies are highly toxic for peripheral nerves. Antibodies to GQlb and other disialylated gangliosides have an a-latrotoxin-like effect in mice characterized by a dramatic release of acetylcholine, resulting in a depletion of this neurotransmitter at the nerve terminals, final blockade of
Table 3. General comparison between classic autoimmune diseases and Guillain-Barri syndrome“ Classic AID
Characteristic Demographic associations Predominant age at onset Predominant sex Segregation within families Immunologic associations Preceding acute infections Association with HLA types Association with other AID Clinical course Progression Response to corticosteroids
GBS
20-40 yr Females Yes
None; frequently at older age Males Usually not
Weak Strong Yes
Strong None or weak None
Relapsing-remitting, chronic Frequently
Monop hasic None
“AID, autoimmune diseases (classic AID such as rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis); GBS, Guillain-BarrC syndrome; HLA, human leucocyte antigen.
CHAPTER 13
GUILLAIN-BARF& SYNDROME AND CAMPYLOBACTER
249
PRECEDING INFECTIONS ASSOCIATED WITH GBS
nerve transmission, and paralysis of the nerve-muscle preparation (Jacobs et al., 2002; Plomp et al., 1999). There is additional destruction of the architecture of these nerve endings (O’Hanlon et al., 2001). Antibodies to GM1 affect the sodium channels in the nodes of Ranvier of rabbit peripheral nerves (Susuki et al., 2007). All these effects are complement dependent. After binding of the antibodies to their targets on peripheral nerves, there are local depositions of complement factors leading to the formation of the membrane attack complex (Halstead et al., 2004). Obduction studies also demonstrated that local complement activation occurs at the site of nerve damage, which is the axolemma in patients with Ah4AN and the Schwann cell membrane in patients with AIDP (Hafer-Macko et al., 1996a, 199613). The neurotoxic effects of serum anti-GQlb antibodies were inhibited by immunoglobulins, which also have a therapeutic effect in patients with GBS (Jacobs et al., 2003). These studies show that at least some of the serum antiganglioside antibodies are highly neurotoxic in ex vivo models, and probably also in GBS patients. After these antibodies have disappeared, have lowered in titer, or are neutralized by the treatment, the peripheral nervous system has the capacity to slowly recover, provided that some of the nerve architecture has remained intact.
After the first isolation of C. jejuni from stool specimens from a GBS patient in 1982, studies in controlled series of patients confirmed the association between Cumpylobacter infections and GBS (Table 4). These studies indicate that a C. jejuni infection precedes GBS in 20 to 50% in Europe, North and South America, Japan, and Australia. In other areas, the frequency of C. jejuni infections preceding GBS may be even higher, such as in northern China and Bangladesh (Ho et al., 1995; Islam et al., 2007). C. jejuni can been cultured from stool specimens from GBS patients, although this appears to necessitate laborious isolation protocols including enrichment media. Still, stool cultures are often negative at the onset of the neurological symptoms, considering the mean period of bacterial excretion after onset of diarrhea is only 16 days (Svedhem and Kaijser, 1980). GBS typically starts 1 to 3 weeks after a Campylobacter infection. Studies based on stool culture alone will therefore underestimate the frequency of C. jejuni infections in GBS. Recent studies, however, have shown that Campylobacter serology is highly sensitive and specific in GBS patients (Ang et al., 2007). The C. jejuni infections preceding GBS are usually
Table 4. Frequency studies on Campylobacterjejuni infections in patients with Guillain-Barrt syndrome“ Reference
Frequencyb
Axonal‘
Kaldor and Speed (1984) Speed et al. (1987) Winer et al. (1988a) Walsh et al. (1991) Gruenewald et al. (1991) Boucquey et al. (1991) Nobile-Orazio et al. (1992) Gregson et al. (1993) Kuroki et al. (1993) Enders et al. (1993) Vriesendorp et al. (1993) Mishu and Blaser (1993) Von Wulffen et al. (1994) Rees et al. (1995b)
21/56 (38%)” 22/45 (49%)” 14/99 (14%)’ 14/94 (15%)’ 3/17 (18%)’.f 6/42 (14%)’ 3/16 (19%)’ 15/42 (36%)’ 14/46 (30%)f 15/38 (39%)e 10158 (17%)e 431118 (36%)’ 11142 (26%)c 25/96 (26%)e,f 217 (29%)‘ 25/38 (66%)’ 461154 (32%)’ 531229 (23%)’ 11311049 (ll%)f 19/80 (26%)’ 32/78 (41%)”
NT NT NT
Ho et al. (1995) Jacobs et al. (1996) Hadden et al. (2001) Takahashi et al. (2005) Sinha et al. (2006) Nachamkin et al. (2007)
+
NT NT NT NT NT -
+
NT NT
+ + NT + NT + NT
-
Clinical association
Poor outcomee
Severe weakness NT NT NT NT NT NT NT Severe weakness NT NT Pure motor MFS
NT
-
-
Severe, pure motor Pure motor NT NT NT
+
NT NT NT NT NT -
+
NT NT
+
NT
+ after PE; - after IVIg + NT NT NT
a NT, not tested; +, present; -, absent; MFS, Miller Fisher syndrome; PE, plasma exchange; IVIg, intravenous immunoglobulins. “Number of patients in case reports, or number of C. jejuni-infected patients per number of tested patients in frequency studies. ‘Association with severe axonal degeneration. dDetermined at least 3 months after diagnosis of GBS. eDetermined by serology. fDetermined by stool culture.
250
IACOBS ET AL.
self-limiting, and treatment with erythromycin does not prevent the development of weakness (Constant et al., 1983; Kohler and Goldblatt, 1987; Pryor et al., 1984; Rhodes et al., 1982; Ropper, 1988; Sovilla et al., 1988). A remarkable diversity of antecedent infections other than those caused by C. jejuni has been reported in relation with GBS. Most of these microorganisms were reported in case reports only and may represent coincidental findings. In a large comparative and prospective study that tested for 16 different infectious causes, C. jejuni infection was found to be the predominant antecedent event (Jacobs et al., 1996). Other infections associated with GBS in these case-control studies were cytomegalovirus, EpsteinBarr virus, and Mycoplasma pneumoniae. Infections with Huemophilus influenzae (Houliston et al., 2007; Ju et al., 2004; Koga et al., 2001; Mori et al., 2000), human immunodeficiency virus, and vaccinations have also been reported in association with GBS. Other types of infection may also be related to GBS, but their frequency is probably too low to produce a significant association in case-control studies. This wide spectrum of preceding infections in GBS may partly explain the patient-to-patient variation in clinical and electrophysiological phenotype. Patients with a C. jejuni infection may represent a clinical and pathological subgroup distinct from other patients with GBS (Table 4). Some studies in larger series found an association between C. jejuni infection and a more severe form of GBS (Jacobs et al., 1996; Kaldor and Speed, 1984; Visser et al., 1995), with less sensory deficits (Hadden et al., 2001; Jacobs et al., 1996; Rees and Hughes, 1994; Visser et al., 1995), more frequent or extensive axonal degeneration (Hadden et al., 2001; Ho et al., 1995; Rees et al., 1995a; Sinha et al., 2006; Vriesendorp et al., 1993; Walsh et al., 1991), or poorer outcome (Hadden et al., 2001; Jacobs et al., 1996; Rees and Hughes, 1994; Rees et al., 1995; Vriensendorp et al., 1993). In a more recent study including two large independent cohorts of patients, preceding diarrhea or positive C. jejuni serology was identified as a strong prognostic factor, independent of other prognostic factors, which can be used to predict the chance of recovery toward independent walking after 6 months (van Koningsveld et al., 2007). In Northern China, C. jejuni infections are more frequent in patients with AMAN than AIDP (Ho et al., 1995). Also in Japan, C. jejuni infections are strongly associated with the axonal variants of GBS, and some studies concluded that this type of infection is exclusively found in patients with axonal forms of GBS (Kuwabara et al., 2004). Geographical differences in the incidence of C. jejuni infections may therefore also
influence the predominant clinical and electrophysiological phenotype. Whether this exclusive association between C. jejuni and axonal variants is also present in other geographical areas remains to be determined. Preceding infections with C. jejuni are also the predominant type of preceding infection in patients with the MFS (Koga et al., 2005a). Preceding infections with cytomegalovirus, on the other hand, are related to a different phenotype. These patients have severe sensory symptoms and cranial nerve deficits (Jacobs et al., 1997b; Yuki and Tagawa, 1998).
WHOLE-GENOME POLYMORPHISMS IN GBS-RELATED C. JEJUNI C. jejuni is an organism with a relatively wellconserved genome size and composition (Parkhill et al., 2000). Despite this apparent conservatism, many variable target regions suited for bacterialepidemiologic and functional typing have been identified. Such tools are of importance in defining the determinant of GBS-inducing capacities in GBSassociated bacterial isolates. Targets harboring variable DNA sequences range from distinct regions such as the flagellin gene locus or the clustered, regularly interspaced short palindromic repeat domain to variable restriction sites more suited for pan-genomic restriction fragment length polymorphism assessment or pulsed-field gel electrophoresis analyses. The current trend toward direct DNA sequence-based typing has also penetrated the Campylobacter research field. A variety of more or less conserved genes has been identified and used in large-scale genetic typing studies of clinical, veterinary, or specific diseaseassociated collections of Carnpylobacter strains. We will here briefly discuss the international literature on epidemiological and high-throughput functional typing of C. jejuni strains associated with postinfectious neurological syndromes including GBS and MFS (Allos, 1998; Tam et al., 2006). Initial studies pointed in the direction of putative clonality of the possibly neuropathogenic GBS- and MFS-associated strains. Studies from different geographical locations revealed that most of the clinical isolates belonged to a restricted set of serotypes, e.g., HS:19 and HS:44 (Allos et al., 1998; Lastovica et al., 1997; Nishimura et al., 1997; Saida et al., 1997; Yuki et al., 1992). These data strongly proposed clonal dissemination of neuropathogenic strains across restricted areas. In turn, this suggested the existence of specific traits, not widely disseminated among all Campylobacter strains, which were causally related to the postinfectious neurological complications associated with Campylobacter enteritis. Unfortunately, these data were in a sense misleading. Subsequent re-
CHAPTER 13
GUILLAIN-BARF& SYNDROME AND CAMPYLOBACTER
search revealed that C. jejuni strains collected within restricted time frames and geographic locales could also display extensive diversity or polyclonality (Endtz et al., 2000; Engberg et al., 2001). This latter observation suggested that the capacity of strains to cause neuropathological syndromes was more widely disseminated than previously anticipated on the basis of the clonality hypothesis. Because GBS- and MFSassociated strains were now considered to derive from a variety of different genetic backgrounds, focused experimental searches for non-serotypedependent factors important in the induction of such diseases were facilitated.
GENE-SPECIFIC VARIATION IN GBS-RELATED C . JEJUNI Obviously, hunting for genetic traits by epidemiological comparisons was initially focused toward the genes encoding the enzymes responsible for the biosynthesis of the lipooligosaccharides (LOS), the core element in the molecular mimicry hypothesis (Ang et al., 2002a; Aspinall et al., 1994b; Yuki et al., 2004). Further screening for the presence of LOSassociated genes that were identified through genome sequencing (Parker et al., 2005) revealed that several genes were selectively present or absent, depending on the composition of the LOS gene cluster. A diversity of gene clusters has now been identified (ranging from type A to type R), all of which have a characteristic set of core LOS biosynthesis genes present. These genes may be polymorphic with respect to their primary structure, but absence or presence appeared to be strongly correlated with the propensity to cause GBS or MFS. The Campylobacter sialic acid transferase (cstll) gene turned out to be a particularly interesting one. It was discovered, again by comparative genomics technologies, that the presence of the gene and the presence of G Q l b epitopes on the Campylobacter outer LOS surface were intimately linked to the development of MFS, thereby for the first time linking bacterial gene variation to the capacity to cause neurological damage through autoimmune antibody induction (van Belkum et al., 2001). This for the first time supported the molecular mimicry hypothesis from the bacterial genomics perspective. After these initial explorations, several additional large-scale comparative genetic screenings have been performed for C. jejuni, but none of these yielded other molecular markers for GBS and MFS besides the LOS biosynthesis genes. Multilocus sequence typing and flagellin gene sequencing confirmed the polyclonality previously assessed among GBS- and MFS-related Campylobacter strains (Dingle
25 1
et al., 2001). Amplified fragment-length polymorphism assessment also did not contribute substantially. Although its appropriateness for identifying clonal expansion among clinical, veterinary, and environmental isolates was convincingly demonstrated, no specific markers could be identified on the basis of the DNA fingerprints developed for the neuropathogenic isolates (Duim et al., 2000, 2003; Endtz et al., 2003). Even the application of high-throughput amplified fragment-length polymorphism rendering thousands of markers for individual strains failed to add novel markers to the existing ones (Godschalk et al., 2006a). The use of whole-genome DNA arraysthe typing tool used after whole-genome sequencing-did not reveal any GBS-specific markers when comparing gene content for strains derived from noncomplicated enteritis cases versus GBS (Leonard et al., 2004). More recently, Taboada et al. (2007) determined the gene conservation profile of 1,712 virulence and other genes in 56 GBS-associated strains and 102 control strains. They confirmed that genes involved in the sialylation of LOS are significantly associated with neuropathogenic strains. However, some enteritis control strains still bear these genes and share remarkable levels of genomic similarity with their neuropathogenic counterparts. In addition, they described two genes involved in the capsular biosynthesis (Cj1421c and Cj1428c) that also were significantly associated with GBS. Any potential involvement of these genes in the pathogenesis of GBS is yet to be determined.
BACTERIAL GENE CHARACTERISTICS, CLINICAL SYMPTOMS, AND GBS VARIANTS There appears to be an association between the specific clinical symptoms, GBS variants, and particular characteristics of C. jejuni strains that primarily resides in the LOS biosynthesis genes. This has been demonstrated by gene-specific knockout mutagenesis in the locus itself (Godschalk et al., 2005; Perera et al., 2007; Xiang et al., 2006). Typing studies also revealed that different types of C. jejuni are capable of inducing different disease presentations (Kuwabara et al., 2004; Takahashi et al., 2005). The presence of certain genes could be associated with a specific epitope at the bacterial cell surface, even corroborated with sophisticated mass spectrometry techniques, which could in turn be coupled to various classes of autoimmune antibody reactions. In this respect, it is important to note that C. jejuni can acquire large portions and even entire LOS gene clusters from external sources through horizontal transfer (Gilbert et al.,
252
TACOBS ET AL.
2004; Phongsisay et al., 2006). Interestingly, specific point mutations within specific genes could also be associated with LOS variation and hence different clinical presentations of the diseases induced (Kimoto et al., 2006). The LOS locus classification and some highly specific amino acid changes in, for instance, the Cstll gene ultimately determine the LOS biosynthetic capacity. The Thr51 as opposed to the Am51 variant of cstll is responsible for the capacity to synthesize GM1 and GDla ganglioside mimics, respectively, and as such constitutes an important GBS risk factor (Hye and Nachamkin, 2007; Koga et al., 2 0 0 . 5 ~2006). ~ It has to be noted, however, that both these studies were biased because specific, clonally related C. jejuni strains were overrepresented in the stain collections.
EVIDENCE FOR A CAUSAL RELATIONSHIP BETWEEN C. JEJUNI AND GBS Classically, the fulfillment of Koch's postulates confirms infectious causation, but this has become increasingly difficult during the last century (Fredricks and Relman, 1996). Sequence-based identification of microbial pathogens has led to a reconsideration of Koch's postulates. Reliance on Koch's postulates as critical evidence for causation has thus diminished. In fact, the postulates have not been fulfilled for most diseases for which infectious causation has been accepted, including cervical cancer, liver cancer, T-cell leukemia, and AIDS. Therefore, because of the limitations of these postulates, the emphasis has been on alternative ways to identify causative organisms (Cochan et al., 2000). The discovery of C. jejuni as the predominant infection preceding GBS stimulated the research on the pathogenesis of the disease and its role in causing the postinfectious neuropathy. Early reports speculated that the C. jejuni infection provides the antigenic trigger for a transient immune response to peripheral nerves. The antibody response to C. jejuni generally peaks at 1 to 2 weeks after infection (Blaser and Duncan, 1984), which coincides with the delay between infection and the first neurological symptoms in GBS. After eradication of the C. jejuni, the antigenic drive of this immune response would be lost, possibly explaining the monophasic course of GBS. There is now convincing evidence from previous studies that antibodies to gangliosides in patients with a C. jejuni-related GBS are induced by molecular mimicry (Table 5 ) . In short, these studies described the following: (i) among GBS patients, recent C. jejuni infection are associated with the presence of serum antibodies to distinct gangliosides; (ii) C.
Table 5 . Evidence for a causal role of Cumpylobucter jejuni infections in the pathogenesis of Guillain-Barre syndrome" Association between C. jejuni infections and GBS Molecular mimicry between C. jejuni LOS and peripheral nerve ganglisoides Cross-reactive antibodies to C. jejuni LOS and gangliosides in serum from GBS patients Induction of similar cross-reactive antibodies in animals after immunization with C. jejuni LOS Pathogenicity of these cross-reactive antibodies in animal models Prevention of disease in animal models after blocking pathogenic effects of these antibodies a
GBS, Guillain-Barre syndrome; LOS, lipooligosaccharide.
jejuni isolates from GBS patients frequently express a LOS that mimics gangliosides; (iii) among GBS patients, serum antibodies to C. jejuni LOS usually cross-react with gangliosides; (iv) animals immunized with C. jejuni LOS produce similar cross-reactive antiganglioside antibodies as found in humans; and (v) these immunizations induce neurotoxic antibodies or an axonal neuropathy with limb paresis in certain infection models. Association between C. jejuni Infection and Serum Antiganglioside Antibodies Initial studies reported an association between various types of antecedent infection and the presence of specific serum antiganglioside antibodies (Jacobs et al., 1998; Willison and Yuki, 2002). C. jejuni infections are frequently associated with antibodies to GM1, GMlb, GDla, and GalNAc-GDla in patients with the pure motor or axonal variants of GBS (Ang et al., 1999; Ho et al., 1999; Jacobs et al., 1996; Ogawara et al., 2000, 2003; Yuki et al., 1990). C. jejuni infections are also associated with antibodies to GQlb and GTla and oculomotor weakness in patients with GBS and MFS. In contrast, patients with a recent cytomegalovirus infection more frequently have antibodies to GM2 (Jacobs et al., 1997b; Yuki and Tagawa, 1998), and patients with a recent M. gneumoniue infection frequently have antibodies to galactocerebroside (Ang et al., 2002b; Kusunoki et al., 2001), although this was not confirmed by others (Susuki et al., 2004). Molecular Mimicry between C. jejuni LOS and Peripheral Nerve Gangliosides Mass spectrometry studies of the C. jejuni isolates from several patients with GBS and MFS showed a similar carbohydrate moiety in LOS and peripheral nerve gangliosides (Aspinall et al., 1994a, 1994b; Yuki et al., 1993) (Fig. 2). In a recent study,
GUILLAIN-BARF& SYNDROME AND CAMPYLOBACTER
CHAPTER 13
I
C. jejuni isolate
GB2
I
1 pizq
LOS biosynthesis class genotype
1
V l J I
I
J
253
LOS outer core
Cross-reactive antibodies
Anti-GMl/GDla IgG
1
1
GMI
GDla
Nerve gangliosides
Cer
I Neurological deficits
0Galactose,
Anti-GTla/GQl b IgG
1 Cer
GQ1b Cer
Cer
1
1
l-l--l
Severe, pure motor
N-Acetyl-galactosamide
1
G D3
1 MFS
KAcetylneurarninic acid,
0 Heptose,
Glucose, Cer Ceramide
Figure 2. C. jejuni genotype and LOS phenotype in relation to serum antiganglioside antibodies and clinical features in GBS and MFS. Two illustrative examples of C. jejuni strains isolated from a patient with a severe, pure motor form of GBS (strain GB2) and from a patient with MFS (strain MF6) (Godschalk et al., 2007). Strain GB2 with a class A genotype produces a GM1- and GDla-like LOS, which induced cross-reactive antibodies to the peripheral nerve ganglioside GM1 and GDla. Strain MF6 with a class B genotype produces a GMlb- and GDlc-like LOS, which induced cross-reactive antibodies to the peripheral nerve ganglioside GD3 and GQlb.
21 of 26 GBS- and MFS-related C. jejuni strains expressed a ganglioside-like LOS, with up to five identical glycan structures (Godschalk et al., 2007). The majority of these strains expressed a heterogeneous LOS, in which the combination of GM1- and GDlalike structures was most frequently found. This combination was associated with the presence of serum antibodies to GM1 and GDla and with the occurrence of severe pure motor GBS. Other strains showed mimicry with disialylated gangliosides and were associated with antibodies to GQlb, GTla, and GD3 and with the occurrence of oculomotor weakness and the MFS. The LOS of some strains contained no sialic acids and showed no mimicry with gangliosides. Previous studies on C. jejuni isolates from enteritis controls without GBS already found a large heterogeneity in the carbohydrate structure of LOS. LOS from serostrain PEN HS:3 is not sialylated, and this Penner serotype is not associated with GBS (Aspinall et al., 1995). These biochemical studies indicate that C. jejuni strains isolated from GBS and MFS patients more frequently have a ganglioside-like
LOS than strains isolated from enteritis controls. The association, however, is not absolute: GBS-related strains may not have ganglioside mimicry (Godschalk et al., 2006a), whereas some strains from enteritis controls do (Godschalk et al., 2006a). In addition, on infection with multiple strains of C. jejuni, the isolation of only one strain may easily lead to a wrong conclusion with respect to absence or presence of molecular mimicry involved in the pathogenesis of GBS (Godschalk et al., 2006b). Cross-Reactive Antibodies to C . jejuni LOS and Gangliosides Serological studies have clearly demonstrated that antiganglioside antibodies in serum from patients with GBS and MFS cross-react with the LOS from autologous C. jejuni isolates (Jacobs et al., 1995, 1997a; Yuki et al., 1994). The serum antibodies to LOS in these patients can also be depleted by gangliosides. The variation in antibody specificity observed in GBS patients is related to the variation in gangli-
254
TACOBS ET AL.
oside mimicry found in LOS from C. jejuni isolates. Serum antibodies to GM1 and GDla are specifically adsorbed by C. jejuni strains expressing LOS with GM1 and GDla mimics. On the other hand, antibodies to G Q l b are adsorbed by strains expressing a GD3 or GTla mimic (a GQlb-like structure in C. jejuni LOS has not been found). Even antibodies to ganglioside complexes cross-react with C. jejuni LOS (Kuijf et al., 2007). These studies strongly indicate that in C. jejuni-related cases of GBS, the carbohydrate moiety of the LOS determines the specificity of the antiganglioside antibodies and the related neurological deficits (Fig. 2). Induction of Cross-Reactive Antibodies in Animals Immunized with C. jejuni LOS Two conditions need to be fulfilled to induce a cross-reactive immune response against autologous antigens by molecular mimicry. First, the microbial and autologous antigens need to be sufficiently similar to induce antibodies that can cross-react. C. jejuni LOS shows a high similarity with gangliosides, and rabbits immunized with LOS produce antiganglioside antibodies with a similar specificity and crossreactivity as found in the sera of patients with GBS and MFS (Ang et al., 2000a, 2001; Bowes et al., 2002; Goodyear et al., 1999). Second, the microbial antigen needs to be sufficiently different from the autoantigen to break the tolerance of the immune system and induce a specific immune response. This second condition explains why the production of antiganglioside antibodies in wild-type mice may be low or even absent, and why GalNAc transferase knockout mice, which lack complex gangliosides and only express high levels of GM3 and GD3, produce much higher titers (Bowes et al., 2002). Also, the antiganglioside antibodies induced in those experiments cross-react with C. jejuni LOS. Induction of Neurotoxic Antibodies in Animals Immunized with C. jejuni LOS Monoclonal antibodies to GQlb and other disialylated gangliosides raised in mice after immunization with C. jejuni LOS are highly toxic for mouse perisynaptic Schwann cells, peripheral nerve endings, and neuromuscular junctions (Goodyear et al., 1999; Halstead et al., 2004; O’Hanlon et al., 2001). Antibodies to GDla have similar neurotoxic effects on motor nerve terminals in GD3 synthetase knockout mice that overexpress GDla (Goodfellow et al., 2005). These toxic effects are all fully complement dependent and require the presence of a human com-
plement source to exert their neurotoxic effects. In a rabbit model of GBS, immunization with a GM1-like LOS from a C. jejuni isolated from a GBS patient with anti-GM1 antibodies induced anti-GM1 IgG antibodies, an axonal neuropathy, and limb weakness (Yuki et al., 2004). Studies of biopsy samples showed that at the nodes of Ranvier in peripheral nerves of these affected rabbits (Susuki et al., 2007). Complement activation by anti-GM1 antibodies may have precipitated these effects because depositions of IgG and complement products were found along the nodes of Ranvier. Passive transfer studies with antiganglioside antibodies are needed to confirm this hypothesis.
UNRESOLVED ISSUES IN THE PATHOGENESIS OF C. JEJUNI-RELATED GBS Why Do Infections with Ganglioside-Mimicking C. jejuni So Rarely Induce GBS? If C. jejuni is directly involved in the pathogenesis of GBS, how can such a common infection, with an estimated incidence of 1 per 100 individuals per year, be followed by such a rare disease as GBS, with an incidence of 1 to 2 per 100,000 per year? Assuming that one-third of GBS patients have an antecedent C. jejuni infection, one may calculate that approximately only 1 in 1,000 to 5,000 subjects with C. jejuni infection will subsequently develop GBS (Mishu and Blaser, 1993). This discrepancy is further illustrated by a report of a single case of GBS in a family outbreak of C. jejuni enteritis (Ang et al., 2001). Thus, GBS is a rare complication of a Cumpylobucter enteritis. This infrequent event may be partly explained by the observation that ganglioside-like structures are expressed only by a proportion of C. jejuni strains. Other virulence factors might also be involved in triggering GBS because C. jejuni isolates from patients with uncomplicated enteritis without GBS also express ganglioside-like LOS. Presumably, the immunogenicity of these structures may differ, or alternatively, the expression may depend on other bacterial factors-for instance, the LOS itself or other, yet unknown structures. In addition, host factors may also play an important role in the production of antiganglioside antibodies and the development of GBS. This would also explain the absence of epidemics or outbreaks of C. jejuni-associated GBS. Single nucleotide polymorphisms (SNPs) in immune response genes determine the variation in immune responses to infections in general and may also influence the susceptibility to develop GBS after an infection with C. jejuni. Two studies reported an association between C. jejuni infections and specific
CHAPTER 13
GUILLAIN-BARRE SYNDROME AND CAMPYLOBACTER
class I or I1 HLA types in, respectively, Japanese and European patients with GBS (Koga et al., 1998; Rees at al., 1 9 9 5 ~ )A. study on GBS patients from northern China reported an association between distinct HLADQ and HLA-DR types and the occurrence of AIDP (Magira et al., 2003). In a Dutch study, no association between these HLA class I1 alleles and AIDP or preceding infection with C. jejuni was found (Geleijns et al., 2005). Another Japanese study also found no association between HLA and GBS (subgroups) (Ma et al., 1998). In conclusion, at present there is no consistent evidence that HLA alleles are a general susceptibility factor for GBS. Other SNPs may be related to the development of GBS, such as those in interleukin 10 (Myhr et al., 2003). There is more evidence, however, that SNP may play a role as diseasemodifying factors. An association has been demonstrated between disease severity or outcome and SNP in genes coding for mannose binding lectin, Fc gamma receptor 111, matrix metalloproteinase 9, and tumor necrosis factor alpha (Geleijns et al., 2006, 2007; van Sorge et al., 2005). SNP analysis may provide insight in the role of host factors in the pathogenesis of GBS, although these studies require confirmation in large and unselected groups of patients and the demonstration of a functional effect of these genetic associations. What Cellular Mechanism Drives the Production of Cross-Reactive Antibodies to Gangliosides after Infection with a Ganglioside-Mimicking C. jejuni? There is convincing evidence that an immune response to C. jejuni LOS by molecular mimicry leads to the production of cross-reactive and neurotoxic antibodies to peripheral nerve gangliosides. The immune competent cells and mechanisms involved in antigen presentation and B-cell activation, however, are currently unknown. There is increasing evidence that C. jejuni antigens, including LOS, can induce maturation and cytokine production in human dendritic cells (Hu et al., 2006). Whether C. jejuni antigens are presented to T cells and whether T cells are required to provide help to B cell to produce cross-reactive anti-ganglioside antibodies has not been elucidated. T cells reactive to gangliosides have been identified that are activated by bacterial infections (de Libero et al., 2005). Antigen-presenting cells mediate this activation by increasing the endogenous synthesis of glycosphingolipids, including GM1, which are presented by CD1 and activate antiglycosphingolipid T cells. In a patient with a c. jejuni-related GBS, T cells recognizing a C. jejuni “nonprotein antigen” were identified (Cooper et al., 2002). Oligoclonal expansions of T cells in periph-
255
eral blood are frequent in patients with GBS, although the role of this T-cell activation and proliferation in the pathogenesis remains to be determined (Koga et al., 2003b). The immune response responsible for the production of antiganglioside antibodies in GBS is unusual given the subclass distribution, the polyreactivity, the moderate affinity, the oligoclonality, and the temporary presence of these antibodies. In humans, IgG2 is the predominant subclass of IgG antibodies to carbohydrates, but in GBS, the antibodies to C. jejuni are usually IgGl and/or IgG3. The IgG subclass of these antibodies may also indicate the involvement of T-cell help, although the immune response in GBS does not lead to sustained high titers. If T cells indeed are activated and provide help to B cells, the recognized C. jejuni T-cell epitope is unknown. It has been hypothesized that a yet unknown Campylobacter protein, possibly a toxin, may function as a carrier protein to provide adequate T-cell help (Willison and Kennedy, 1993). However, there is as yet no conclusive evidence in support of this hypothesis. Do C. jejuni Infections Exclusively Trigger the Onset of Axonal Forms of GBS?
Many studies have reported an association between preceding C. jejuni infections and axonal forms of GBS. A study from Japan suggested that C. jejuni infections there are exclusively found in patients with axonal variants, although the first electrophysiological studies may show features compatible with demyelination (Kuwabara et al., 2004). In Western countries, however, axonal variants are rare, although preceding C. jejuni infections are found in 20 to 50% of patients with GBS (Hadden et al., 1998). This finding may suggest that C. jejuni may also trigger AIDP, at least in Western countries. In addition, it remains unclear whether all electrophysiological findings in patients with GBS can be attributed to axonal degeneration or demyelination. Is There a “Safe” C. jejuni Strain that Cannot Trigger GBS? C. jejuni isolates from GBS patients more frequently have a distinct LOS genotype and a related ganglioside-like LOS as compared with isolates from patients with uncomplicated enteritis (Ang et al., 2002a; Godschalk et al., 2006a, 2007; van Belkum et al., 2001). These associations, although statistically significant, are far from absolute, and other bacterial factors may also trigger GBS. Moreover, occasionally, C. jejuni strains that do not express ganglioside mimicry are isolated from the stools of GBS patients
256
TACOBS ET AL.
(Godschalk et al., 2006b, 2007). Extensive genotyping studies have shown that these strains are almost indistinguishable from enteritis control strains (Taboada et al., 2007). Identification of additional and as yet unknown bacterial factors involved in the pathogenesis is highly relevant for future vaccine development and the designation of safe challenge strains in chicken but also in human experiments. Further understanding of the microbe-host interaction, including the involvement of the cellular immune response, remains a challenging objective. Can a Single Animal Model Be Developed T o Demonstrate That Antiganglioside Antibodies Triggered by a C . jejuni Infection Induce a Neuropathy Similar to That of Patients with GBS? Studies in current animal models have come a long way in providing evidence for the molecular mimicry hypothesis for the pathogenesis of C. jejunirelated GBS. All these models, however, have some limitations. In the mouse model, the C. jejuni LOSinduced cross-reactive antibodies to gangliosides have an evident and transferable toxic effect on the neuromuscular junction and perisynaptic Schwann cells (Halstead et al., 2004; Jacobs et al., 2002; O’Hanlon et al., 2001; Plomp et al., 1999). To induce these effects, however, the presence of a human complement source is required, and these mice do not develop a similar polyneuropathy as observed in GBS patients. In the rabbit model, immunization with C. jejuni LOS leads to the production of cross-reactive antibodies to gangliosides that induce a complementmediated disruption of sodium channel clusters in peripheral motor nerve fibers and a neuropathy that shows some similarities with AMAN (Susuki et al., 2007; Yuki et al., 2004). Passive transfer studies, however, have not been reported, and because of that, the role of antibodies in the pathogenesis of this axonal polyneuropathy model remains elusive. How Can These Discoveries Be Used to Improve Outcome in Patients with GBS? The progress made in the understanding of the pathogenesis of C. jejuni-related GBS until now has not influenced the treatment and outcome of these patients. C. jejuni-related GBS patients in particular need better treatment because prognosis in these patients in general is poor. About one-third of these patients will be unable to walk independently 6 months after onset of disease. Considering the crucial role of antiganglioside antibodies in the pathogenesis of these forms of GBS, patients may profit from se-
lective depletion of antibodies. Artificially produced ganglioside carbohydrates were found to be effective in depleting antiganglioside antibodies from the serum of GBS patients (Townson et al., 2007; Willison et al., 2004). Alternatively, treatment with recently developed complement inhibitors may block the pathogenic actions of these antibodies (Halstead et al., 2005). New treatment regimens will be further developed and tested in future clinical trials. The above-mentioned two future strategies appear within our abilities and may have a significant impact on the outcome of autoimmune neuropathies, and GBS in particular.
REFERENCES Allos, B. M. 1998. Association between Cumpylobacter infection and Guillain-Barre syndrome. J. Infect. Dis 176(Suppl. 2):S125S128. Allos, B. M., F. T. Lippy, A. Carlsen, R. G. Washburn, and M. J. Blaser. 1998. Campylobucter jejuni strains from patients with Guillain-BarrC syndrome. Emerg. Infect. Dis. 4:263-268. Ang, C. W., M. A. De Klerk, H. P. Endtz, B. C. Jacobs, J. D. Laman, F. G. van der Meche, and P. A. van Doom. 2001. Gullain-Barrt syndrome- and Miller Fisher syndrome-associated Cumpylobucter jejuni lipopolysaccharides induce anti-GM 1 and anti-GQlb antibodies in rabbits. Infect. Immun. 69:24622469. Ang, C. W., H. P. Endtz, B. C. Jacobs, J. D. Laman, M. A. De Klerk, F. G. A. van der Meche, and P. A. van Doorn. 2000a. Cumpylobacter jejuni lipopolysaccharides from Guillain-Barrt syndrome patients induce IgG anti-GM1 antibodies in rabbits. J. Neuroimmunol. 104: 133-13 8. Ang, C. W., J. D. Laman, H. J. Willison, E. R. Wagner, H. P. Endtz, M. A. De Klerk, A. P. Tio-Gillen, N. van den Braak, B. C. Jacobs, and P. A. van Doorn. 2002a. Structure of Cumpylobucter jejuni lipopolysaccharides determines antiganglioside specificity and clinical features of Guillain-Barre and Miller Fisher patients. Infect. Immun. 70: 1202-1208. Ang, C. W., A. P. Tio-Gillen, J. Groen, P. Herbrink, B. C. Jacobs, R. van Koningsveld, A. D. Osterhaus, F. G. van der MechC, and P. A. van Doorn. 2002b. Cross-reactive anti-galactocerebroside antibodies and Mycoplusma pneumoniue infections in GuillainBarrt syndrome. J. Neuroimmunol. 130:179-183. Ang, C. W., P. A. van Doorn, H. P. Endtz, I. S. J. Merkies, B. C. Jacobs, M. A. de Klerk, R. van Koningsveld, and F. G. A. van der MechC. 2000b. A case of Guillain-BarrC syndrome following a family outbreak of Cumpylobacter jejuni enteritis. J. Neuroimmunol. 111:229-233. Ang, C. W., N. Yuki, B. C. Jacobs, P. A. van Doorn, P. I. M. Schmitz, and F. G . A. van der MechC. 1999. Rapidly progressive, predominantly motor Guillain-BarrC syndrome with antiGalNAc-GD1a antibodies. Neurology 5 3 :2 122-2 127. Asbury, A. K., and D. R. Cornblath. 1990. Assessment of current diagnostic criteria for Guillain-BarrC syndrome. Ann. Neurol. 27(Suppl.):21-24. Aspinall, G. O., C. M. Lynch, H. Pang, R. T. Shaver, and A. P. Moran. 1995. Chemical structures of the core region of Cumpylobuctevjejuni HS:3 lipopolysaccharide and an associated polysaccharide. Eur. J. Biochem. 231570-578. Aspinall, G. O., A. G . McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1994a. Lipopolysaccharides of Cumpylobacter je-
CHAPTER 13
GUILLAIN-BARRI? SYNDROME AND CAMPYLOBACTER
juni serotype HS: 19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barrt syndrome. Biochemistry 33:241-249. Aspinall, G. O., S. Fujimoto, A. G. McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 199410. Lipopolysaccharides from Campylobacter jejuni associated with Guillain-Barrt syndrome patients mimic human gangliosides in structure. Infect. lmmun. 62 :2 122-2 125. Blaser, M. J., and D. J. Duncan. 1984. Human serum antibody response to Cumpylobacter jejuni infection as measured in an enzyme-linked immunosorbent assay. Infect. Immun. 44:292298. Boffey, J., M. Odaka, D. Nicoll, E. R. Wagner, K. Townson, T. Wowes, J. Conner, K. Furakawa, and H. J. Willison. 2005. Characterisation of the immunoglobulin Variable region gene usage encoding the murine anti-ganglioside antibody repertoire. J. Neuroimmunol. 165:92-103. Boucquey, D., C. J. Sindic, M. Lamy, M. Delmee, J. P. Tomasi, and E. C. Laterre. 1991. Clinical and serological studies in a series of 45 patients with Guillain-Barrt syndrome. J. Neurol. Sci. 104:56-63. Bowes, T., E. R. Wagner, J. Boffey, D. Nicholl, L. Cochrane, M. Benboubetra, J. Conner, K. Furukawa, K. Furukawa, and H. J. Willison. 2002. Tolerance to self gangliosides is the major factor restricting the antibody response to lipopolysaccharides core oligosaccharides in Campylobacter jejuni strains associated with Guillain-BarrC syndrome. Infect. Immun. 705008-5009. Cochran, G. M., P. W. Ewald, and K. D. Cochran. 2000. Infectious causation of disease: an evolutionary perspective. Perspect. Biol. Med. 43:406-448. Constant, 0. C., C. C. Bentley, A. M. Denman, J. R. Lehane, and H. E. Larson. 1983. The Guillain-Barrt syndrome following Cumpylobacter enteritis with recovery after plasmapheresis. J. Infect. 6: 89-9 1. Cooper, J. C. S. Hughes S., A. Ben-Smith A., C. 0. Savage, and J. B. Winer. 2002. T cell recognition of a non-protein antigen preparation of Campylobacter jejuni in patients with GBS. 1. Neurol. Neurosurg. Psychiatry 72:413-414. Cros, D., and W. J. Triggs. 1994. There are no neurophysiologic features characteristic of “axonal” Guillain-Barrt syndrome. Muscle Nerve 17:675-677. de Libero, G., A. P. Moran, H. J. Gober, E. Rossy, A. Shamshiev, 0. Chelnokova, 2. Mazorra, S . Vedetti, A. Sacchi, M. M. Prendergast, S . Sansano, A. Tonevitsky, R Landmann, and L. Mori. 2005. Bacterial infections promote T cell recognition of selfglycolipids. Immunity 22:763-772. Dingle, K. E., N. van Den Braak, F. M. Colles, L. J. Price, D. L. Woodward, F. G. Rodgers, H. P. Endtz, A. van Belkum, and M. C. Maiden. 2001. Sequence typing confirms that Campylobucter jejuni strains associated with Guillain-Barrt and MillerFisher syndromes are of diverse genetic lineage, serotype, and flagella type. J. Clin. Microbiol. 39:3346-3349. Dowling, P. J., Menonna, and S . Cook. 1977. Cytomegalovirus complement fixation ntibody in Guillain-Barrt syndrome. Neurology. 27:1153-1156. Duim, B., C. W. Ang, A. van Belkum, A. Rigter, N. W. van Leeuwen, H. P. Endtz, and J. A. Wagenaar. 2000. Amplified fragment length polymorphism analysis of Campylobacter jejuni strains isolated from chickens and from patients with gastroenteritis or Guillain-Barrt or Miller Fisher syndrome. Appl. Environ. Microbiol. 66:39 17-3923. Duim, B., P. C. Godschalk, N. van den Braak, K. E. Dingle, J. R. Dijkstra, E. Leyde, J. van der Plas, F. M. Colles, H. P. Endtz, J. A. Wagenaar, M. C. Maiden, and A. van Belkum. 2003. Molecular evidence for dissemination of unique Cumpylobacter je-
257
juni clones in Curacao, Netherlands Antilles. J. Clin. Microbiol. 41r5.593-5597. Enders, U., H. Karch, K. V. Toyka, M. Michels, J. Zielasek, M. Pette, J. Heesemann, and H. P. Hartung. 1993. The spectrum of immune responses to Cumpylobacter jejuni and glycoconjugates in Guillain-Barrt syndrome and in other neuroimmunological disorders. Ann. Neurol. 34:136-144. Endtz, H. P., C. W. Ang, N. van Den Braak, B. Duim, A. Rigter, L. J. Price, D. L. Woodward, F. G. Rodgers, W. M. Johnson, J. A. Wagenaar, B. C. Jacobs, H. A. Verbrugh, and A. van Belkum. 2000. Molecular characterization of Campylobucter jejuni from patients with Guillain-Barrt and Miller Fisher syndromes. J. Clin. Microbiol. 38:2297-230 1. Endtz, H. P., H. van West, P. C. Godschalk, L. de Haan, Y. Halabi, N. van den Braak, B. I. Kesztyus, E. Leyde, A. Ott, R. Verkooyen, L. J. Price, D. L. Woodward, F. G. Rodgers, C. W. Ang, R. van Koningsveld, A. van Belkum, and I Gerstenbluth. 2003. Risk factors associated with Campylobacter jejuni infections in Curacao, Netherlands Antilles. J. Clin. Microbiol. 41: 5588-5592. Engberg, J., I. Nachamkin, V. Fussing, G. M. McKhann, J. W. Griffin, J. C. Piffaretti, E. M. Nielsen, and P. Gerner-Smidt. 2001. Absence of clonality of Cumpylobacter jejuni in serotypes other than HS: 19 associated with Guillain-Barrt syndrome and gastroenteritis. J. Infect. Dis. 15:215-220. Fisher, M. 1956. An unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia, ataxia and areflexia). N. Engl. J. Med. 25557-65. Fredricks, D. N., and D. A. Relman. 1996. Sequence-based identification of microbial pathogens: a reconsideration of Koch‘s postulates. Clin. Microbiol. Rev. 9:18-33. French Cooperative Group on Plasma Exchange in Guillain-Barrt syndrome. 1987. Efficiency of plasma exchange in GuillainBarri syndrome: role of replacement fluids. Ann. Neurol. 22: 753-761. Geleijns, K., B. A. Brouwer, B. C. Jacobs, J. J. HouwingDuistermaat, C. M. van Duijn, and P. A. van Doorn. 2004. The occurrence of Guillain-BarrC syndrome within families. NeurolO ~ Y63~1747-1750. Geleijns, K., M. Emonts, J. D. Laman, W. van Rijs, P. A. van Doorn, P. W. Hermans, and B. C. Jacobs. 2007. Genetic polymorphisms of macrophage-mediators in Guillain-Barrt syndrome. J. Neuroimmunol. 190:127-130. Geleijns, K., Roos A, Houwing-Duistermaat JJ, van Rijs W, TioGillen AP, Laman JD, van Doorn PA, and Jacobs BC. 2006. Mannose-binding lectin contributes to the severity of GuillainBarrt syndrome. J. Immunol. 177:4211-4217. Geleijns, K., G. M. Th. Schreuder, B. C. Jacobs, K. Sintnicolaas, R. van Koningsveld, J. Meulstee, J. D. Laman, and P. A. van Doorn. 2005. HLA class I1 alleles are not a general susceptibility factor in Guillain-Barrt syndrome. Neurology 64:44-49. Gilbert, M., P. C. Godschalk, M. F. Karwaski, C. W. Ang, A. van Belkum, J. Li, W. W. Wakarchuk, and H. P. Endtz. 2004. Evidence for the acquisition of the lipooligosaccharide biosynthesis locus in Campylobucter jejuni GB11, a strain from a patient with Guillain-Barrt syndrome, by horizontal exchange. Infect. Immun. 72:1162-1165. Godschalk, P. C., M. P. Bergman, R. F. Gorkink, G. Simons, N. van den Braak, A. J. Lastovica, H. P. Endtz, H. A. Verbrugh, and A. van Belkum. 2006a. Identification of DNA sequence variation in Campylobacter jejuni strains associated with the Guillain Barri syndrome by high-throughput AFLP analysis. BMC Microbiol. 6:32. Godschalk, P. C., M. Gilbert, B. C. Jacobs, T. Kramers, A. P. TioGillen, C. W. Ang, N. van den Braak, J. Li, H. A. Verbrugh, A. van Belkum, and H. P. Endtz. 2006b. Co-infection with two
258
IACOBS ET AL.
different Campylobacter jejuni strains in a patient with the Guillain Barrt syndrome. Microbes Infect. 8:248-253. Godschalk, P. C. R., A. E. Heikema, M. Gilbert, T. Kornagamine, C. W. Ang, J. Glerum, D. Brochn, J. Li, N. Yuki, B. C. Jacobs, A. van Belkum, and H. P. Endtz. 2004. The crucial role of Cumpylobucter jejuni genes in antiganglioside antibody induction in Guillain Barre syndrome. J. Clin. Investig. 114:1659-1665. Godschalk, P. C., M. L. Kuijf, J. Li, F. St. Michael, C. W. Ang, B. C. Jacobs, M. Karwaski, D. Brochu, H. P. Endtz, A. van Belkum, and M. Gilbert. 2007. Structural characterization of Cumpylobacter jejuni lipooligosaccharide outer cores associated with Guillain-Barrt and Miller Fisher syndromes. Infect. Immun. 75: 1245-1254. Goodfellow, J. A., J. Bowes, K. Sheick, M. Odaka, S. K. Halstead, P. D. Humphreys, E. R. Wagner, N. Ynki, K. Furukawa, K. Furukawa, and H. J. Willison. 2005. Overexpression of GDla ganglioside sensitizes motor nerve terminals to anti-GDla antibody-mediated injury in a model of acute motor axonal neuropathy. J. Neurosci. 25:1620-1628. Goodyear, C. S., G. M. O’Hanlon, J. J. Plomp, E. R. Wagner, I. Morrison, J. Veitch, L. Cochrane, R. W. Bullens, P. C. Molenaar, J. Conner, and H. J. Willison. 1999. Monoclonal antibodies raised against Guillain-Barre syndrome-associated Cumpylobacter jejuni lipopolysaccharides react with neuronal gangliosides and paralyze muscle-nerve preparations. J. Clin. Invest. 104:697-708. Gregson, N. A., S. Koblar, and R. A. Hughes. 1993. Antibodies to gangliosides in Guillain-Barrt syndrome: specificity and relationship to clinical features. Q. J. Med. 86:111-117. Griffin, J. W., C. Y. Li, T. W. Ho, P. Xue, C. Macko, C. Y. Gao, C. Yang, M. Tian, B. Mishu, D. R. G. M. Cornblath, McKhann, and A. K. Asburry. 1995. Guillain-Barrt syndrome in northern China. The spectrum of neuropathological changes in clinically defined cases. Bruin 118577495. Gruenewald, R., A. H. Ropper, H. Lior, J. Chan, R. Lee, and V. S. Molinaro. 1991. Serologic evidence of Campylobacter jejunil coli enteritis in patients with Guillain-BarrC syndrome. Arch. Neurol. 48:1080-1082. Guillain, G., J. A. Barrt, and A. Strohl. 1916. Sur un syndrome de radiculoneurite avec hyperalbuminose du liquide cephalorachidien sans reaction cellulaire. Bull. Mem. Soc. Med. Hop. Paris. 40: 1462- 1470. The Guillain-Barre Syndrome Study Group. 1985. Plasmapheresis and acute Guillain-Barrt syndrome. Neurology 35:1096-1104. Hadden, R. D., D. R. Cornblath, R. A. C. Hughes, J. Zielasek, H. P. Hartung, K. V. Toyka, and A. V. Swan. 1998. Electrophysiological classificatin of Guillain-Barrt syndrome: clinical associations and outcome. Ann. Neurol. 44:225-230. Hadden, R. D., H. Karch, H. P. Hartung, J. Zielasek, B. Weissbrich, J. Schnbert, A. Weishaupt, D. R. Cornblath, A. V. Swan, R. A. Hughes, and K. V. Toyka. 2001. Preceding infections, immune factors, and outcome in Guillain-Barrt syndrome. Neurology 5 6 ~ 78-765. 5 Hafer-Macko, C., S. T. Hsieh, C. Y. Li, T. W. Ho, K. Sheikh, D. R. Cornblath, G. M. McKhann, A. K. Asbury, and J. W. Griffin. 1996a. Acute motor axonal neuropathy: an antibodymediated attack on axolemma. Ann. Neurol. 40:635-644. Hafer-Macko, C. E., K. Sheikh, C. Y.Li, T. W. Ho, D. R. Cornblath, G. M. McKhann, A. K. Asbury, and J. W. Griffin. 1996b. Immune attack on the Schwann cell surface in acute demyelinating polyneuropathy. Ann. Neurol. 39:625-635. Halstead, S. K., G. M. O’Hanlon, P. D. Humphreys, D. B. Morrison, B. P. Morgan, A. J. Todd, J. J. Plomp, and H. J. Willison. 2004. Anti-disialoside antibody kill perisynaptic Schwann cells and damage motor nerve terminals via membrane attack complex in a murine model of neuropathy. Bruin 127:2109-2123.
Halstaed, S. K., P. D. Humphreys, J. A. Goodfellow, E. R. Wagner, R. Smith, and H. J. Willison. 2005. Complement inhibition abrogates nerve terminal injury in Miller Fisher syndrome. Ann. Neurol. 58:203-210. Heller, J., and Vogel S. 1986. No Laughing Matter. Donald I. Fione, Inc., New York. Ho, T. W., B. Mishu C. Y. Li, C. Y. Gao, D. R. Cornblath, J. W. Griffin, A. K. Asbury, M. J. Blaser, and G. M. McKhann. 1995. Guillain-Barre syndrome in northern China. Relationship to Cumpylobacter jejuni infection and anti-glycolopid antibodies. Bruin 118597-605. Ho, T. W., H. J. Willison, I. Nachamkin, C. Y. Li, J. Veitch, H. Ung, G. R. Wang, R. C. Liu, D. R. Cornblath, A. K. Asbury, J. W. Griffin, and G. M. McKhann. 1999. Anti -GDla antibody is associated with axonal but not demyelinating forms of Guillain-Barrt syndrome. Ann. Neurol. 45:168-173. Houliston, R. S., M. Koga, J. Li, H. C. Jarrell, J. C. Richards, V. Vitiazeva, E. K. H. Schweda, N. Yuki, and M. A. Gilbert. 2007. A Haernophilus influenme strain associated with Fisher syndrome expresses a novel disialylated ganglioside mimic. Biochemistry 46:8 164-8 171. Hu, L., M. D. Bray, M. Osorio, and D. J. Kopecko. 2006. Cumpylobucter jejuni induces maturation and cytokine production in human dendritic cells. Infect. lmmun. 74:2697-2705. Hurwitz, E. S., R. C. Holman, D. B. Nelson, and L. B. Schonberger. 1983. National surveillance for Guillain-Barrt syndrome: January 1978-March 1979. Neurology 33:150-157. Hye, J. B., and I. Nachamkin. 2007. Cumpylobacter jejuni CstlI polymorphisms and association with development of GuillainBarrt syndrome. Neurology 68:1633-1634. Islam, Z., B. C. Jacobs, H. P. Endtz, Q. D. Mohammad, M. Islam, G. Nair, D. A. Sack, and K. A. Talukder. 2007. Guillain-Barrt syndrome in Bangladesh: an epidemiological and clinical study. J. Vet. Med. 54(Suppl. 1):152-153. Jacobs, B. C., P. W. Bullens, G. M. O’Hanlon, C. W. Ang, H. J. Willison, and J. J. Plomp. 2002. Detection and prevalence of alpha-latrotoxin-like effects of serum from patients with Guillain-Barre syndrome. Muscle Nerve 25549-558. Jacobs, B. C., H. P. Endtz, F. G. A. van der Meche, M. P. Hazenberg, H. A. M. Achtereekte, and P. A. van Doorn. 1995. Serum anti-GQlb IgG anti-bodies recognize surface epitopes on Cumpylobucter jejuni from patients with Miller Fisher syndrome. Ann. Neurol. 37:260-264. Jacobs, B. C., M. P. Hazenberg, P. A. van Doorn, H. P. Endtz, and F. G. A. van der Mecht, 1997a. Cross-reactive antibodies against gangliosides and Cumpylobacter jejuni lipopolysaccharides in patients with Guillain-Barre or Miller Fisher syndrome. J. Infect. Dis. 175:729-733. Jacobs, B. C., G. W. O’Hanlon, R. W. Bullens, J. Veitch, J. J. Plomp, and H. J. Willison. 2003. Immunoglobulins inhibit pathophysiological effects of anti-GQlb-positive sera at motor nerve terminals through inhibition of antibody binding. Bruin 126: 2220-2234. Jacobs, B. C., P. H. Rothbarth, F. G. A. van der Mecht, P. Herbrink, P. I. M. Schmitz, M. A. de Klerk, and P. A. van Doorn. 1998. The spectrum of antecedent infections in Guillain-Barre syndrome: a case-control study. Neurology 51:lllO-1115. Jacobs, B. C., P. A. van Doorn, J. H. M. Groeneveld, A. P. TioGillen, and F. G. A. van der Mecht. 199713. Cytomegalovirus infections and anti-GM2 antibodies n Guillain-Barrt syndrome. J. Neurol. Neurosurg. Psychiatry 62:641-643. Jacobs, B. C., P. A. van Doorn, P. I. M. Schmitz, A. P. Tio-Gillen, P. Herbrink, L. H. Visser, H. Hooijkaas, and F. G. A. van der Mecht. 1996. Cumpylobacter jejuni infections and anti-GM1 antibodies in Guillain-Barre syndrome. Ann. Neurol. 40:181-187.
CHAPTER 13
GUILLAIN-BARRI? SYNDROME AND CAMPYLOBACTER
Ju, Y. Y., H. Womersley, J. Pritchard, I. Gray, R. A. Hughes, and N. A. Gregson. 2004. Haemophilus influenzae as a possible cause of Guillain-Barrt syndrome. J. Neuroimmunol. 149:160166. Kaida, K., M. Kanzaki, D. Morita, K. Kamakura, K. Motoyoshi, M. Hirakawa, and S. Kusunoki. 2006. Anti-gangliosidecomplex antibodies in Miller Fisher syndrome. J. Neurol. Neurosurg. Psychiatry 77: 1002-1 003. Kaida, K., D. Morita, M. Kanzaki, K. Kamakura, K. Motoyoshi, M. Hirakawa, and S. Kusunoki. 2007. Anti-gangliosidecomplex antibodies associated with severe disability in GBS. J. Neuroimmunol. 182:2 12-2 18. Kaldor, J., and B. R. Speed. 1984. Guillain-Barrt syndrome and Campylobacter jejuni: a serological study. Br. Med. J. 288:18671870. Kimoto, K., M. Koga, M. Odaka, K. Hirata, M. Takahashi, J. Li, M. Gilbert, and N. Yuki. 2006. Relationship of bacterial strains of campylobacter-associated neuropathies. Neurology 67: 18371843. Kleyweg, R. P., and F. G. A. van der Mecht. 1991. Treatment related fluctuations in Guillain-Barrt syndrome after high-dose immunoglobulins or plasma-exchange. J. Neurol. Neurosurg. Psychiatry 54:957-960. Koga, M., M. Gilbert, J. Li, S. Koike, M. Takahashi, K. Furakawa, K. Hirata, and N. Yuki. 2005a. Antecedent infections in Fisher syndrome: a common pathogenesis of molecular mimicry. Neurology 64: 1605-161 1. Koga, M., M. Gilbert, M. Takahashi, J. Li, S. Koike, K. Hirata, and N. Yuki. 2006. Comprehensive analysis of bacterial risk factors for the development of Guillain-BarrC syndrome after Campylobacter jejuni enteritis. J. Infect. Dis. 193547-555. Koga, M., S. Koike, K. Hirata, and N. Yuki. 2005b. Ambiguous value of aemophilus influenzae isolation in Guillain-Barrt and Fisher syndromes. J. Neurol. Neurosurg. Psychiatry 76: 17361738. Koga, M., M. Takahashi, M. Masuda, K. Hirata, and N. Yuki. 2005c. Campylobacter gene polymorphism as a determinant of clinical features of Guillain-Barrt syndrome. Neurology 65: 1350-1351. Koga, M., N. Yuki, K. Hirata, M. Morimatsu, M. Mori, and S. Kuwabara. 2003a. Anti -GM1 antibody IgG subclass: a clinical recovery predictor in Guillain-Barrt syndrome. Neurology 60: 1514-1518. Koga M., N. Yuki, K. Kashiwase, K. Tadokoro, T. Juji, and K. Hirata. 1998. Guillain-Barrt and Fisher’s syndrome subsequent to Campylobacter jejuni enteritis are associated with HLA-BS4 and Cwl independent of anti-ganglioside antibodies. J. Neuroimmunol. 88:62-66. Koga, M., N. Yuki, T. Tai, and K. Hirata. 2001. Miller Fisher syndrome and Haemophilus influenzae infection. Neurology 57: 686-69 1. Koga M., N. Yuki, Y. Tsukada, K. Hirata, Y. Matsumoto. 2003b. CDR3 sectratyping analysis of the T cell receptor in GuillainBarrt and Fisher syndromes. J. Neuoimmunol. 141:112-117. Kohler, P. C., and D. Goldblatt. 1987. Guillain-Barrt syndrome following Campylobacterjejuni enteritis. Arch. Neurol. 44: 1219. Kuijf, M. L., P. C. Godschalk, M. Gilbert, H. P. Endtz, A. P. TioGillen, C. W. Ang, P. A. cvan Doorn, and B. C. Jacobs. 2007. Origin of ganglioside complex antibodies in Guillain-Barrt syndrome. J. Neuroimmunol. 188:69-73. Kuroki, S., T. Saida, M. Nukina, T. Haruta, M. Yoshioka, Y. Kobayashi, and H. Nakanishi. 1993. Campylobacter jejuni strains from patients with Guillain-Barrt syndrome belong mostly to Penner serogroup 19 and contain beta-N-acetylglucosamine residues. Ann. Neurol. 33:243-247.
259
Kusunoki, S., M. Shiina, and I. Kanazawa. 2001. Anti-Gal-C antibodies in GBS subsequent to mycoplasma infection: evidence of molecular mimicry. Neurology 57:736-738. Kuwabara, S., K. Ogawara, S. Misawa, M. Koga, M. Mori, A. Hiraga, T. Kanesaka, T. Hattori, and N. Yuki. 2004. Does Campylobacter jejuni infection elicit demyelinating Guillain Barrt syndrome? Neurology 63529-533. Landry, J. B. 0. 1859. Note sur la paralysie ascendante aigue. Gaz. Hebd. Med. Chir. 6:472-488. Lastovica. A. J., E. A. Goddard, and A. C. Argent. 1997. GuillainBarrt syndrome in South Africa associated with Campylobacter jejuni HS:41 strains. J. Infect. Dis. 176:S139-S143. Ledeen, R. W. 1985. Gangliosides of the neuron. Trends Neurosci. 10~169-174. Ledeen, R. W., and R K. Yu. 1982. Gangliosides: structure, isolation and analysis. Methods Enzymol. 83:139-191. Leonard, E. E., L. S. Tompkins, S. Falkow, and I. Nachamkin. 2004. Comparison of Campylobacter jejuni isolates implicated in Guillain-Barrt syndrome and strains that cause enteritis by a DNA microaray. Infect. Immun. 72:1199-1203. Liu, Y., and L. P. Jou. 1993. Acute motor axonal neuropathy: a frequent cause of acute flaccid paralysis in China. Ann. Neurol. 33:333-342. Ma, J. J., M. Nishimura, H. Mine, S. Kuroki, M. Nubina, M. Ohta, H. Saji, H. Obyashi, T. Saida, H. Kawakami, and T. Uchiyama. 1998. HLA and T-cell receptor gene polymorphisms in Guillain-Barrt syndrome. Neurology 51:379-384. Magira, E. E., M. Papaioakim, I. Nachamkin, A. K. Asburry, C. Y. Li, T. W. Ho, J. W. Grifiin, G. M. MaKhann, and D. S. Monos. 2003. Differential distribution of HLA-DQ beta/DR beta epitopes in the two forms of Guillain-Barrt syndrome, acute motor neuropathy and acute inflammatory demyelinating polyneuropathy (AIDP): identification of DQ beta epitopes associated with susceptibility to and protection from AIDP. J. Immunol. 170: 3 074-3 080. McKhann, G. M., D. R Cornblath, J. W. Griffin, T. W. Ho, C. Y. Li, Z. Jiang, H. S. Wu, G. Zhaori, B. Mishu, A. A. Ilyas, C. L. Koski, F. Vriesendorp, S. D. Cook, F. A. Mithen, and M. J. Blaser. 1993. Serologic evidence of previous Campylobacter jejuni infection in patients with the Guillain-BarrCsyndrome. Ann. Intern. Med. 118:947-953. Mishu, B., and M. J. Blaser. 1993. Role of infection due to Cumpylobacter jejuni in the initiation of Guillain-Barrt syndrome. Clin. Infect. Dis. 17:104-108. Mori, M., S. Kuwabara, M. Miyake, M. Noda, H. Kuroki, H. Kanno, K. Ogawara, and T. Hattori. 2000. Haemophilus influenzae infection and Guillain-Barrt syndrome. Brain. 123: 2171-2178. Myhr, K. M., K. S. Vages, T. H. Maroy, J. H. Aarseth, H. I. Nyland, and C. A. Vedeler. 2003. Interleukin-10 promotor polymorphisms in patients with Guillain-Barrt syndrome. J. Neuroimmunol. 139:81-83. Nachamkin, I., P. A. Barbosa, H. Ung, C. Lobato, A. G. Rivera, P. Rodriguez, A. G. Briseno, L. M. Cordero, L. G. Perea, J. C. Perez, M. Ribera, J. Veitch, C. Fitzgerald, D. Cornblath, M. R. Pinto, J. W. Griffin, H. J. Willison, A. K. Asbury, and G. M. McKhann. 2007. Patterns of Guillain-Barrt syndrome in children: results from a Mexican population. Neurology 69:16651671. Nishimura, M., M. Nukina, S, Kuroki, H. Obayashi, M. Ohta, J. J. Ma, T. Saida, and T. Uchiyama. 1997. Characterization of Cumpylobacter jejuni isolates from patients with Guillain-Barrt syndrome. J. Neurol. Sci. 153:91-99. Nobile-Orazio, E., M. Carpo, N. Meucci, M. P. Grassi, E. Capitani, M. Sciacco, A. Mangoni, and G. Scarlato. 1992. Guillain-
260
TACOBS ET AL.
Barrt syndrome associated with high titers of anti-GM1 antibodies. J. Neurol. Sci. 109:200-206. Ogawara, K., S. Kuwabara, M. Koga, M. Mori, N. Yuki, and T. Hattori. 2003. Anti -GMlb IgG antibody is associated with acute motor axonal neuropathy and Campylobacter jejuni infection. J. Neurol. Sci. 210:41-45. Ogawara, K., S. Kuwabara, M. Mori, T. Hattori, M. Koga, and N. Yuki. 2000. Axonal Guillain-Barrt syndrome: relation to anti-ganglioside antibodies and Campylobacter jejuni infection in Japan. Ann. Neurol. 48:624-631. O’Hanlon, G. M., J. J. Plomp, M. Chakrabarti, I. Morrison, E. R. Wagner, C. S. Goodyear, X. Yin, B. D. Trapp, J. Conner, P. C. Molenaar, S. Stewart, E. G. Rowan, and H. J. Willison. 2001. Anti-GQlb ganglioside antibodies mediate complement-dependent destruction of the motor nerve terminal. Brain 124:893906. Parker, C. T., S. T. Horn, M. Gilbert, W. G. Miller, D. L. Woodward, and R. E. Mandrell. 2005. Comparison of Campylobacter jejuni lipooligosaccharide biosynthesis loci from a variety of sources. J. Clin. Microbiol. 43:2771-2281. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Chucher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quall, M. A. Rajandream, K. M. Rntherford, A. H. M. van Vliet, S. Whitehead, and B. G. Barrtll. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Perera, V. N., I. Nachamkin, H. Ung, J. H. Patterson, M. J. McConville, P. J. Coloe, and B. N. Fry. 2007. Molecular mimicry in Campylobacter jejuni: role of the lipooligosaccharidecore oligosaccharide in inducing anti-ganglioside antibodies. FEMS Immunol. Med. Microbiol. 50:27-36. Phongsisay, V., V. N. Perera, and B. N. Fry. 2006. Exchange of lipooligosaccharide synthesis genes creates potential Guillan Barrt syndrome-inducible strains of Campylobacter jejuni. Infect Immun. 74:1368-1372. Plasma Exchange/Sandoglobulin Guillain-Barrt Syndrome Trial Group. 1997. Randomised trial of plasma exchange, intravenous immunoglobulin, and combined treatments in Guillain-Barrt syndrome. Lancet 349:225-230. Plomp, J. J., P. C. Molenaar, G. M. O’Hanlon, B. C. Jacobs, J. Veitch, M. R. Daha, P. A. van Doorn, F. G. A. van der Mech&, A. Vincent, B. P. Morgan, and H. J. Willison. 1999. Miller Fisher anti-GQlb antibodies: alpha-latrotoxin-like effects on motor endplates. Ann. Neurol. 45:189-199. Pryor, W. M., J. S. Freiman, M. A. Gillies, and R. R. Tuck. 1984. Guillain-Barrt syndrome associated with Cumpylobucter infection. Aust. N.Z. J. Med. 14:687-688. Rees, J. H., and R. A. C. Hughes. 1994. Campylobacter jejuni and Guillan-Barrt syndrome. Ann. Neurol. 35:248-249. Rees, J. H., N. A. Gregson, and R A. C. Hughes. 1995a. Antiganglioside GM 1 antibodies in Guillain-Barrt syndrome and their relationship to Campylobacter jejuni infection. Ann. Neu701. 38:809-816. Rees, J. H., S. E. Soudain, N. A. Gregson, and R. A. C. Hughes. 1995b. Campylobacter jejuni infection and Guillain-Barrt syndrome. N. Engl. J. Med. 333:1374-1379. Rees, J. H., R. W. Vaughan, E. Kondeatis, and R. A. Hughes. 1995c. HLA-class I1 alleles in Guillain-Barrt syndrome and Miller Fisher syndrome and their association with preceding Campylobacter jejuni infection. J. Neuroimmunol. 62:53-57. Rhodes, K. M., and A. E. Tattersfield. 1982. Guillain-Barrt syndrome associated with Campylobacter infection. Br. Med. J. 285: 173-1 74. Ropper, A. H., E. F. M. Wijdicks, and B. T. Truax. 1991. Guillain-Bad Syndrome. F. A. Davis Company, Philadelphia.
Ropper, A. H. 1988. Campylobacter diarrhea and Guillain-Barrt syndrome. Arch. Neurol. 45:655-656. Saida, T., S. Kuroki, Q. Hao, M. Nishimura, M. Nukina, and H. Obayashi. 1997. Campylobacter jejuni isolates from Japanese patients with Guillain-Barrt syndrome. J. Infect. Dis. 1765129S134. Schonberger, L. B., E. S. Hurwitz, P. Katona, R. C. Holman, and D. J. Bregman. 1981. Guillain-Barrt syndrome: its epidemiology and associations with influenza vaccination. Ann. Neurol. 9:3 138. Sinha, S., K. N. Prasad, D. Jain, C. M. Pandey, S. Jha, and S. Pradhan. 2006. Preceding infections and anti-ganglioside antibodies in patients with Guillain-Barrt syndrome: a single centre prospective case-control study. Clin. Microbiol Infect. 13:3 16346. Sovilla, J. Y., R. Regli, and P. B. Francioli. 1988. Guillain-Barrt syndrome following Campylobacter jejuni enteritis. Report of three cases and review of the literature. Arch. Intern. Med.148: 739-74 1. Speed, B. R., J. Kaldor, J. Watson, H. Newton-John, W. Tee, D. Noonan, and B. W. Dwyer. 1987. Campylobacter jejunilcampylobacter coli-associated Guillain-Barre syndrome. Immunoblot confirmation of the serological response. Med. 1.Aust. 147: 13-16. Susuki, K., M. Odaka, M. Mori, K. Hirata, and N. Yuki. 2004. Acute motor axonal neuropathy after Mycoplasma infection. Evidence of molecular mimicry. Neurology 62:949-956. Susuki, K., M. N. Rasband, K. Tohyama, K. Koibuchi, S. Okamoto, K. Funakoshi, K. Hirata, H. Baba, and N. Yuki. 2007. Anti-GM1 antibodies cause complement-mediated disruption of sodium channel clusters in peripheral motor nerve fibers.J. Neurosci. 27:3956-3976. Svedhem, A., and B. Kaijser. 1980. Campylobacter fetus subspecies jejuni: a common sause of diarrhea in Sweden. J. Infect. Dis. 142: 353-3 59. Taboada, E. N., A. F. van Belkum, N. Yuki, R. R. Acedillo, P. C. Godschalk, M. Koga, H. P. Endtz, M. Gilbert, and J. H. Nash. 2007. Comparative genomic analysis of Campylobacter jejuni associated with Guillain-Barrt and Miller Fisher syndromes: neuropathogenic and enteritis-associated isolates can share high levels of genomic similarity. BMC Genomics 8:359. Takahashi, M., M. Koga, K. Yokoyama, and N. Yuki. 2005. Epidemiology of Campylobacter jejuni isolated from patients with Guillain-Barrt and Fisher syndromes in Japan. J. Clin. Microbiol. 43:335-339. Tam, C. C., L. C. Rodrigues, I. Petersen, A. Islam, A. Hayward, and S. J. O’Brien. 2006. Incidence of Guillain Barrt syndrome among patients with Campylobacter infection: a general practice research database study. J. Infect. Dis. 194:95-97. Townson. K., J. Boffey, D. Nicholl, J. Veitch, D. Bundle, P. Zhang, E. Samai, T. Antoine, A. Bernardi, D. Arosio, S. Sonnino, N. Isaacs, and H. Willison. 2007. Solid phase immunoadsorption for therapeutic and analytical studies on neuropathyassociated anti-GM1 antibodies. Glycobiology 17:294-303. Turck, D. C. 1984. The pathogenesis of Haemophilw infection. J. Med. Microbiol. 18:l-16. Winer, J. B., R. A. C. Hughes, M. J. Anderson, D. M. Jones, H. Kangro, and R P. Watkins. 1988a. A prospective study of acute idiopathic neuropathy. 11. Antecedent events. J. Neurol. Neurosurg. Psychiatry 5 1:6 13-6 18. Winer, J. B., R. A. C. Hughes, and C. Osmond. 1988b. A prospective study of acute idiopathic neuropathy. I. Clinical features and their prognostic value. J. Neurol. Neurosurg. Psychiatry 51: 605-612. van Belkum, A., N. van den Braak, P. Godschalk, W. Ang, B. Jacobs, M. Gilbert, W. Wakarchuk, H. Verbrugh, and E. Endtz.
CHAPTER 13
GUILLAIN-BARRI? SYNDROME AND CAMPYLOBACTER
2001. A Campylobacter jejuni gene associated with immunemediated neuropathy. Nut. Med. 7:752-753. van der Mecht, F. G. A., P. I. M. Schmitz, and the Dutch GuillainBarrt Study Group. 1992. A randomized trial comparing intravenous immune globulin and plasma exchange in Guillain-Barrt syndrome. N. Engl.]. Med. 326:1123-1129. van Koningsveld, R., R. Rico, I. Gerstenbluth, P. I. Schmitz, C. W. ang, I. S. Merkies, B. C. Jacobs, Y. Halabi, H. P. Endtz, F. G. van der Mecht, and P. A. van Doorn. 2001. Gastroenteritisassociated Guillain-Barrt syndrome on the Caribbean island Curafao. Neurology 56:1467-1472. van Koningsveld, R., P. I. M. Schmitz, F. G. A. van der MechC, L. H. Visser, J. Meulstee, and P. A. van Doorn, for the Dutch GBS Study Group. 2004. Effect of methylprednisolone when added to standard treatment with intravenous immunoglobulin for Guillain-Barrt syndrome: randomised trial. Lancet 363:192196. van Koningsveld, R., E. W. Steyerberg, R. A. C. Hughes, A. V. Swan, P. A. van Doorn, and B. C. Jacobs. 2007. A clinical prognostic scoring system for Guillain-Bard syndrome. Lancet Neurol. 6:5 89-594. van Koningsveld, R., P. van Doorn, P. I. Schmitz, C. W. Ang, and F. G. S. van der Mecht. 2000. Mild forms of Guillain-Barrt syndrome in an epidemioloc survey in The Netherlands. Neurology 54:620-625. van Sorge, N. M., W. L. van der Pol, M. D. Jansen, K. P. Geleijns, S. Kalmijn, R. A. Hughes, J. H. Rees, J. Pritchard, C. A. Vedeler, K. M. Myhr, C. Shaw, I. N. van Schaik, J. H. Wokke, P. A. van Doorn, B. C. Jacobs, J. G. van de Winkel, and L. H. van den Berg. 2005. Severity of Guillain-Barrt syndrome with Fc gamma Receptor 111 polymorphisms. J. Neuroimmunol. 162:157-164. Visser, L. H., F. G. A. van der Mecht, P. A. van Doorn, J. Meulstee, B. C. Jacobs, P. G. Oomes, and R. P. Kleyweg for the Dutch Guillain-Barrt Study Group. 1995. Guillain-Barrt syndrome without sensory loss (acute motor neuropathy). Brain 118: 84 1-847. Von Wulffen, H., C. Hartard, and E. Scharein. 1994. Seroreactivity to Campylobacter jejuni and gangliosides in patients with Guillain-Barrt syndrome. 1.Infect. Dis.170:828-833. Vriesendorp, F. J., B. Mishu, M. J. Blaser, and C. L. Koski. 1993. Serum antibodies to GM1, GDlb, peripheral nerve myelin, and Campylobacter jejuni in patients with Guillain-BarrC syndrome and controls: correlation and prognosis. Ann. Neurol. 34:130135.
261
Walsh, F. S., M. Cronin, S. Koblar, P. Doherty, J. Winer, A, Leon, and R. A. C. Hughes. 1991. Association between glycoconjugate antibodies and Campylobacter infection in patients with Guillain-Barrt syndrome. J. Neuroimmunol. 34:43-5 1. Willison, H. J., and P. G. Kennedy. 1993. Gangliosides and bacterial toxins in Guillain-Barrt syndrome. J. Neuroimmunol. 46: 105-112. Willison, H. J., K. Towson, J. Veitch, J. Boffey, N. Isaacs, S. M. Andersen, P. Zhang, C. Ling, and D. R. Brundle. 2004. Synthetic disialylgalactose immunoadsorbents deplete anit-GQlb antibodies from autoimmune neuropathy sera. Bruin 127:680691. Willison, H. J., and N. Yuki. 2002. Peripheral neuropathies and anti-glycolipid antibodies. Brain 125 :2591-2625. Willison, H. J. 2005. The immunobiology of Guillain-Barrt syndromes. J. Peripher. New. Syst. 10:94-112. Xiang, S. L., M. Zhong, F. C. Cai, B. Deng, and X. P. Zhang. 2006. The sialic acid residue is a crucial component of Campylobucter jejuni ganglioside mimicry in the induction of the Guillain Barrt syndrome.]. Neuroimmunol. 174:126-132. Yuki, N. 2007. Ganglioside mimicry and peripheral nerve disease. Muscle Nerve 35:691-711. Yuki, N., S. Sato, S. Fujimoto, S. Yamada, Y. Tsujino, A. Kinoshita, and T. Itoh. 1992. Serotype of Carnpylobacter jejuni, HLA, and the Guillain-BarrC syndrome. Muscle Nerve 15:968969. Yuki, N., K. Susuki, M. Koga, Y. Nishimoto, M. Odaka, K. Hirata, K. Taguchi, T. Miyatake, K. Furukawa, T. Kobata, and M. Yamada. 2004. Carbohydrate mimicry between human ganglioside GMI and Campylobacter jejuni lipooligosaccharide causes Guillain-Barrt syndrome. Proc. Natl. Acad. Sci. USA 101:1140411409. Yuki, N., and Y. Tagawa. 1998. Acute cytomegalovirus infection and IgM anti-GM2 antibody. J. Neurol. Sci. 154:14-17. Yuki, N., T. Taki, F. Inagaki, T. Kasama, M. Takahashi, K. Saito, S. Handa, and T. Miyatake. 1993. A bacterium lipopolysaccharide that elicits Guillain-Barrt syndrome has a GMI ganglioside-like structure. 1.Exp. Med. 178:1771-1775. Yuki, N., T. Taki, M. Takahashi, K. Saito, H. Yoshino, T. Tai, S. Handa, and T. Miyatake. 1994. Molecular mimicry between G Q l b ganglioside and lipopolysaccharides of Campylobacter jejuni isolated from patients with Fisher’s syndrome. Ann. Neurol. 36:79 1-793. Yuki, N., H. Yoshino, S. Sato, and T. Miyatake. 1990. Acute axonal polyneuropathy associated with anti-GM1 antibodies following Campylobacter enteritis. Neurology 40: 1900-1902.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser Q 2008 ASM Press, Washington, DC
Chapter 14
Mechanisms of Antibiotic Resistance in Campylobacter QIJINGZHANG
AND PAUL J. PLUMMER
organisms (Appelbaum and Hunter, 2000; Hooper, 1998). They are among the most valuable antimicrobials available to treat bacterial infections because of their broad spectrum of activity, pharmacodynamics, safety, and ease of administration (Appelbaum and Hunter, 2000; Chu, 1999b). Once inside bacterial cells, FQ antimicrobials interact with two target enzymes, DNA gyrase and topoisomerase IV (Drlica and Zhao, 1997; Hooper, 2001), forming stable complexes with these enzymes and trapping the enzymes on DNA, which results in double-stranded breaks in DNA and bacterial death (Drlica and Malik, 2003; Shea and Hiasa, 1999; Willmott et al., 1994). In gram-negative bacteria, DNA gyrase is the primary target of FQ antibiotics, while topoisomerase IV is the main target of FQs in gram-positive bacteria (Hooper, 2001). FQ antimicrobials are important in the treatment of enteric infections including campylobacteriosis (Adachi et al., 2000). The mechanisms involved in FQ resistance include modification of DNA gyrase and/ or topoisomerase IV, active efflux, altered membrane permeability, and reduced target enzyme expression (Hooper, 2001; Levy, 1992; Poole, 2000). All of these resistance mechanisms are encoded by chromosomal elements. However, plasmid-mediated quinolone resistance has been recently reported in Klebsiella pneumoniae and Escherichia coli, in which the plasmid-carried quinolone-resistance gene encodes a protein (Qnr) that protects DNA gyrase from inhibition by the drug (Tran et al., 2005; Wang et al., 2003, 2004). Alterations in target enzymes (DNA gyrase and topoisomerase IV) have been studied extensively in bacterial pathogens and are mainly due to spontaneous mutations in the quinoloneresistance-determining region of the genes encoding GyrA or ParC (Khodursky et al., 1995; Yoshida et
Campylobacter spp., especially Campylobacter jejuni and Campylobacter coli, are increasingly resistant to antimicrobials, which has become a significant concern for public health. To counteract the selection pressure from antimicrobial usage, the Campylobacter pathogens have evolved multiple mechanisms for antimicrobial resistance, including (i) synthesis of antibiotic-inactivating/modifying enzymes (e.g., plactamase), (ii) alteration or protection of antibiotic targets (e.g., mutations in gyrA or 23s rRNA genes), (iii) active extrusion of drugs out of bacterial cells through drug efflux transporters (e.g., CmeABC), and (iv) reduced permeability to antibiotics due to unique membrane structures. Some of the resistanceassociated traits are endogenous to Campylobacter, whereas others are acquired by mutation or genetic transfer. In this chapter, we will discuss the mechanisms involved in Campylobacter resistance to various antibiotics, with a particular emphasis on new information obtained during the last few years. It should be pointed out that our understanding of the resistance mechanisms is limited to C. jejuni and C. coli, and little information is available on antimicrobial resistance in other Campylobacter spp. In addition, the epidemiology of antibiotic resistance and the transfer of antibiotic-resistant Campylobacter from animals to humans through the food chain are discussed elsewhere (chapter 36) and will not be covered in this chapter.
FLUOROQUINOLONE (FQ) RESISTANCE FQ antimicrobials, such as ciprofloxacin, enrofloxacin, and levofloxacin, are a class of structurally related synthetic compounds with potent bactericidal activity against both gram-positive and gram-negative
Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, 1116 Veterinary Medicine, Qijing Zhang Department of Veterinary Microbiology and Preventive Medicine, Iowa State University, 2180 Ames, IA 50011. Paul J. Plummer Veterinary Medicine, Ames, LA 50011.
263
264
ZHANG AND PLUMMER
al., 1990). Mutations in the quinolone-resistancedetermining region drastically reduce the affinity of gyrases to quinolone antimicrobials, resulting in resistance to the bactericidal action. Resistance-associated mutations can also occur in GyrB and ParE (Hooper, 2001). Campylobacter has become increasingly resistant to FQ antimicrobials (Aarestrup and Wegener, 1999; Bachoual et al., 2001; Ge et al., 2003; Gupta et al., 2004; Nachamkin et al., 2002; Piddock, 1995; Rautelin et al., 1991; Ruiz et al., 1998; Saenz et al., 2000; Sanchez et al., 1994; Smith et al., 2000; Van Looveren et al., 2001). The resistance is mediated by point mutations in the gyrA gene in conjunction with the function of the multidrug efflux pump CmeABC (Ge et al., 2005; Luo et al., 2003; Zhang et al., 2003). Specific mutations at positions Thr-86, Asp90 and Ala-70 in GyrA have been linked to FQ resistance in C. jejuni (Engberg et al., 2001; Luo et al., 2003; Wang et al., 1993). Specifically, the Thr-86-Ile change (mediated by the C257T mutation) is the most commonly observed mutation in FQ-resistant Campylobacter isolates and confers high-level (ciprofloxacin MIC 2 1 6 pg/ml) resistance to FQs, while the Asp-90-Asn and Thr-86-Lys mutations are less common and are associated with intermediate-level FQ resistance (Gootz and Martin, 1991; Luo et al., 2003; Ruiz et al., 1998; Wang et al., 1993). No mutations in gyrB have been associated with FQ resistance in Campylobacter (Bachoual et al., 2001; Payot et al., 2002; Piddock et al., 2003). Although a single report documented the involvement of ParC mutations in the resistance to FQs (Gibreel et al., 1998), subsequent work revealed that Campylobacter lack the parC and parE genes (Bachoual et al., 2001; Cooper et al., 2002; Luo et al., 2003; Parkhill et al., 2000; Payot et al., 2002; Piddock et al., 2003). Thus, it can be concluded that parC and parE mutations are not involved in Campylobacter resistance to FQ antimicrobials. The contribution of an efflux mechanism to FQ resistance in Campylobacter was reported in an early study (Charvalos et al., 1995), but the identity of the efflux pump was not determined. It has been found that the CmeABC efflux pump contributes significantly to both intrinsic and acquired resistance of C. jejuni to FQ antimicrobials (Ge et al., 2005; Lin et al., 2002; Luo et al., 2003; Pumbwe and Piddock, 2002). This efflux pump reduces the accumulation of FQs in Campylobacter cells and functions synergistically with the gyrA mutations in conferring and maintaining high-level FQ resistance in clinical isolates (Ge et al., 2005; Luo et al., 2003). Overexpression of CmeABC reduces the accumulation of ciprofloxacin in C. jejuni and increases the resistance
conferred by various types of GyrA mutations (Yan et al., 2006). This efflux pump also plays an important role in the emergence of FQ-resistant Campylobacter under selection pressure because many of the spontaneous gyrA mutants cannot survive the selection by ciprofloxacin in the absence of CmeABC (Yan et al., 2006). On the other hand, overexpression of CmeABC increases the frequency of emergence of FQ resistant mutants that are highly resistant to ciprofloxacin (Yan et al., 2006). One unique feature of FQ resistance in Campylobacter is that acquisition of high-level resistance does not require stepwise accumulation of point mutations in gyrA. Instead, a single point mutation in gyrA can lead to clinically relevant levels of resistance to FQ antimicrobials (Ge et al., 2005; Gootz and Martin, 1991; Luo et al., 2003; Ruiz et al., 1998; Wang et al., 1993; Zhang et al., 2003). In vitro, the frequencies of emergence of FQ-resistant GyrA mutants range from approximately to lo-', depending on the concentrations of ciprofloxacin used on the enumerating plates (Yan et al., 2006). The higher the selection pressure, the lower the frequencies of emergence. The reason for the variable frequencies is that multiple types of spontaneous point mutations occur in the gyrA genes, but different point mutations confer varied levels of FQ resistance and thus survive different levels of selection pressure (Yan et al., 2006). When 4 pg/ml of ciprofloxacin was used in the plating media, the frequency of mutant emergence was consistently measured in the order of lo-', and the mutants predominantly harbor the C257T mutation in gyrA (Gootz and Martin, 1991; Yan et al., 2006). In vivo, FQ-resistant Campylobacter mutants occur rapidly in chickens treated with enrofloxacin and reached to as high as lo7 CFU/g feces within 24 to 48 h after the initiation of treatment (Farnell et al., 2005; Griggs et al., 2005; Luo et al., 2003; McDermott et al., 2002; van Boven et al., 2003). FQ-resistant C. jejuni could also rapidly develop in Campylobacter-infected patients treated with ciprofloxacin (Ellis-Pegler et al., 1995; Segreti et al., 1992; Wretlind et al., 1992). These findings suggest that Campylobacter is highly mutable to FQ treatment and that FQ antimicrobials may not be the ideal antibiotics for clinical therapy of campylobacteriosis. The GyrA mutations not only confer FQ resistance, but also affect the fitness of C. jejuni in the chicken host. Luo et al. (2005) examined the fitness of FQ-resistant Carnpylobacter in the absence of selection pressure by using clonally related and genetically isogenic mutants. When monoinoculated into the chickens, FQ-resistant Campylobacter was able to colonize and persist in chickens as efficiently as the
CHAPTER 14
FQ-resistant strains in the absence of FQ antimicrobials. The prolonged colonization in vivo did not result in the reversion or loss of the specific resistance-conferring mutation (C257T) in gyrA. When coinoculated into chickens as pairwise competition, the FQ-resistant Campylobacter outcompeted the FQ susceptible in the majority of pairs. The fitness change was directly linked to the C257T mutation in gyrA. These findings suggest that the resistanceassociated gyrA mutation does not incur a fitness cost and even enhances the fitness of Campylobacter in the absence of selection pressure. This notion is supported by two surveys conducted in the United States (Price et al., 2005, 2007), which revealed that FQresistant Campylobacter was still prevalent in poultry from farms that had not used FQ antimicrobials for several years. The physiological and molecular basis for the enhanced fitness in FQ-resistant Campylobacter is unknown and remains to be determined in future work.
MACROLIDE RESISTANCE Macrolide antibiotics, such as erythromycin, clarithromycin, azithromycin, and telithromycin, are important drugs for the treatment of respiratory tract infections and gastric diseases caused by Helicobacter pylori and Campylobacter in humans (Chu, 1999a; Kasbekar and Acharya, 2005). This class of antibiotics including erythromycin, tylosin, spiramycin, tilmicosin, and roxithromycin are also approved for growth promotion and therapeutic purposes in animal agriculture (McEwen and Fedorka-Cray , 2002; Prescott, 2000). Macrolide antibiotics target the 50s subunit of bacterial ribosome and produce antibacterial effects by inhibiting protein synthesis (Poehlsgaard and Douthwaite, 2005). Bacterial resistance to macrolides is mediated by three mechanisms, including antibiotic modifications, target site modification or alteration, and drug efflux (Leclercq, 2002). Some bacteria produce enzymes that inactivate macrolide antibiotics by phosphorylation or glycosylation (Kuo et al., 1989; Noguchi et al., 1995). Target modification/alteration is via either rRNA methylases encoded by the erm genes or mutations in 23s rRNA or ribosomal proteins L4 and L22 (Jacobs and Johnson, 2003; Maravic, 2004). These changes interfere with the binding of macrolides to the 50s ribosomal subunit and confer macrolide resistance. Effluxmediated macrolide resistance is through various efflux pumps that reduce accumulation of the drugs in bacterial cells Uacobs and Johnson, 2003). Macrolides such as erythromycin are considered the drug of choice for treating human campylobac-
MECHANISMS OF ANTIBIOTIC RESISTANCE
265
teriosis, and there is a trend of rising resistance in Campylobacter to erythromycin and related macrolides (Engberg et al., 2001; Gibreel and Taylor, 2006). Two mechanisms, target modification and efflux, are involved in macrolide resistance in Campylobacter, but there is no evidence that Campylobacter produces macrolide-modifying enzymes (Payot et al., 2006). Target modification involves chromosomal mutations in the 23s rRNA gene or the genes encoding ribosomal proteins L4 and L22. Several point mutations in 23s rRNA have been associated with Campylobacter resistance to macrolides (Engberg et al., 2001; Gibreel et al., 2005; Harrow et al., 2004; Jensen and Aarestrup, 2001; Lin et al., 2007; Mamelli et al., 2005), which occur at the base position 2074 (A2074C, A2074G, or A2074T) or 2075 (A2075G, A2075T, or A2075C) or both. These positions correspond to the base positions 2058 and 2059, respectively, in E. coli. Both the A2074G and A2075G changes can confer high-level resistance to macrolide antibiotics, but the A2075G mutation is more frequently observed than the A2074G mutation (Gibreel and Taylor, 2006). Although a previous study indicated that the A2074G mutation was unstable (Gibreel et al., 2005), recent work on the development of macrolide-resistant Campylobacter revealed the predominance of the A2074G mutation in chickens fed with tylosin, and this mutation was stably maintained in antibiotic-free media (Lin et al., 2007). There are three copies of the 23s rRNA gene in the chromosome of C. jejuni and C. coli (Fouts et al., 2005; Parkhill et al., 2000). In most cases, all three copies of the 23s rRNA gene carry the resistanceassociated mutations, but both the wild-type and mutated alleles can coexist in a single macrolide-resistant mutant (Gibreel et al., 2005; Jensen and Aarestrup, 2001; Lin et al., 2007). These point mutations in the 23s rRNA likely block the interaction of macrolide antibiotics with the 50s ribosome and thus confer resistance to this class of antibiotics. In addition to the 23s rRNA mutations, mutations in the ribosomal protein L4 (G74D) and L22 (insertions at position 86 or 98) were also shown to confer macrolide resistance in Campylobacter (Cagliero et al., 2006). The contribution of efflux systems to macrolide resistance in Campylobacter has been described in several recent studies (Cagliero et al., 2005; Gibreel et al., 2007; Lin et al., 2007; Mamelli et al., 2005; Payot et al., 2004). Inhibition by efflux pump inhibitors decreased the MICs of macrolides in various Campylobacter isolates and also significantly decreased the frequency of emergence of erythromycinresistant C. jejuni mutants (Cagliero et al., 2005; Martinez and Lin, 2006; Payot et al., 2004), indicating the involvement of efflux pumps in macrolide re-
266
ZHANG AND PLUMMER
sistance. By use of insertional mutagenesis, CmeABC was found to be a contributor to the intrinsic resistance to macrolides in Campylobacter (Lin et al., 2002; Pumbwe and Piddock, 2002). Subsequently, it was found that CmeABC also contributes significantly to the acquired resistance to macrolide antibiotics (Cagliero et al., 2005, 2006; Lin et al., 2007). In the macrolide-resistant mutants that lack target mutations and show low to intermediate resistance to macrolide, inactivation of CmeABC completely reversed the susceptibility (Cagliero et al., 2005; Lin et al., 2007), while in those mutants that harbor resistanceassociated mutation in L4, L22, or 23s rRNA, CmeABC functions synergistically with the target mutations in conferring the acquired resistance to macrolides (Cagliero et al., 2005, 2006; Lin et al., 2007). These findings clearly indicate that CmeABC is a significant player in conferring the acquired resistance to macrolides. However, there is also some evidence that a CmeABC-independent efflux system also contributes to macrolide resistance in Campylobacter (Mamelli et al., 2005), but the identity of this efflux pump has not been defined. Similar to the situation in other bacteria, the target mutations and efflux pumps in Campylobacter confer cross-resistance to different members of the macrolide family (e.g., erythromycin, clarithromycin, azithromycin, and tylosin), including ketolides (e.g., telithromycin) of this class (Cagliero et al., 2005; Mamelli et al., 2005). It is also reported that crossresistance between macrolides and lincosamides occurs in Campylobacter (Luangtongkum et al., 2006; Taylor and Courvalin, 1988), suggesting that use of lincosamides may potentially affect the development of macrolide resistance in Campylobacter. This crossresistance is probably due to the fact that both classes of antibiotics interact with the same site on the bacteria ribosomal subunit (Poehlsgaard and Douthwaite, 2003). Different from FQ resistance conferred by gyrA mutations, the spontaneous mutation rates for macrolide resistance appear to be low in C. jejuni and C. coli when measured by both in vitro and in vivo studies. As shown by experimental study, treatment of Campylobacter-infected chickens with a therapeutic dose of tylosin (a macrolide) did not select for erythromycin-resistant Campylobacter even after three courses of treatment (Lin et al., 2007). This is in clear contrast to the development of FQ-resistant Campylobacter in chickens treated with enrofloxacin, in which FQ-resistant Campylobacter developed rapidly-in some cases within 24 h (Luo et al., 2003). However, when Campylobacter-infected chickens were fed tylosin at a growth-promoting dose, macrolide-resistant Campylobacter occurred in the
birds after several weeks of exposure, suggesting that emergence of macrolide-resistant Campylobacter requires a prolonged selection process (Lin et al., 2007). The low spontaneous mutation rate and the slow process of resistance development may explain why macrolide-resistant Campylobacter isolates are generally less prevalent than FQ-resistant Cumpylobacter.
TETRACYCLINE RESISTANCE Tetracyclines are an important class of antibiotics widely used in both human and animal medicine. The lipophilic nature of tetracyclines allows them to freely diffuse through the cell membrane of grampositive bacteria and accumulate in the cytoplasm (Chopra and Roberts, 2001). In contrast, transfer across the outer membrane of gram-negative bacteria appears to involve binding of the drug to Mg2+cations and passage through outer membrane porins into the periplasmic space, where the drug dissociates from the magnesium and moves passively through the inner membrane, as is the case in gram-positive bacteria (Chopra and Roberts, 2001). Once in the cytoplasm, the drug binds reversibly with the ribosome, inhibiting peptide elongation by preventing the incoming aminoacyl-tRNA from binding to the A site of the ribosome (Connell et al., 2003a, 2003b). A number of tetracycline resistance genes have been identified in various bacteria to date (Chopra and Roberts, 2001). General mechanisms of tetracycline resistance can be classified into four major categories: efflux pumps, ribosomal protection proteins, chemical modification of the drug, and mutations of the ribosome that alter the ability of the antibiotic to bind. The ribosomal protection protein class of molecules represents a common mechanism of resistance and often confers a broader range of resistance to the tetracycline class of antibiotics (Chopra and Roberts, 2001). To date, two mechanisms have been described for Campylobacter resistance to tetracyclines. The first mechanism involves a ribosomal protection protein termed Tet(0) (Manavathu et al., 1988; Sougakoff et al., 1987; Taylor et al., 1987). The molecular mechanisms underlying the resistance conferred by the Tet(0) protein has been documented recently (Connell et al., 2003a, 2003b). The current model suggests that Tet(0) recognizes the presence of an unoccupied A site on tetracycline blocked ribosomal complexes. Upon binding of the Tet(0) protein, a conformational change occurs in the region of the decoding site and results in the sequential release of the tetracycline molecule and the GTPase-mediated
CHAPTER 14
release of the Tet(0) protein. Interestingly, the conformational state that is induced by the binding of Tet(0) is able to persist after the release of the protein and thus allows the A site to function in protein elongation. The second mechanism of tetracycline resistance in Campylobacter involves the efflux system. The multidrug efflux pump CmeABC has been shown to contribute to both intrinsic and acquired resistance to tetracycline (Gibreel et al., 2007; Lin et al., 2002; Pumbwe and Piddock, 2002). Different from the tetracycline-specific efflux pumps, such as Tet(A) and Tet(B), CmeABC is not tetracycline specific and has a broad substrate spectrum. CmeABC functions synergistically with Tet(0) in conferring high-level resistance to tetracyclines (Gibreel et al., 2007; Lin et al., 2002). The t e t ( 0 ) gene is mainly encoded on transferable plasmids of various sizes, but the gene can be chromosomally encoded in some strains (Akhtar, 1988; Dasti et al., 2007; Manavathu et al., 1988; Ng et al. 1987; Pratt and Korolik, 2005; Sougakoff et al., 1987; Taylor et al., 1987). The pTet plasmid carrying t e t ( 0 ) was fully sequenced, revealing the colocalization of t e t ( 0 )with conjugation-associated genes on the same plasmid (Batchelor et al., 2004). te t (0 )carrying plasmids can be transferred from donor strains to recipient strains via conjugation in laboratory systems (Batchelor et al., 2004; Gibreel et al., 2004b; Nirdnoy et al., 2005; Pratt and Korolik, 2005). In addition, in vivo transfer studies have demonstrated rapid transmission of the t e t ( 0 ) gene between Campylobacter strains via horizontal gene transfer in the intestinal environment of chickens, a natural host for this organism (Avrain et al., 2004). Evolutionarily, the t e t ( 0 )gene likely originated from gram-positive bacteria (Sougakoff et al., 1987; Zilhao et al., 1988). In support of this theory is the fact that the nucleotide sequence of the Campylobacter t e t ( 0 ) gene shares 99.4% homology (98.9% amino acid identity) with that of the t e t ( 0 ) gene in Streptococcus mutans DL5; the G+C content (40%) of the t e t ( 0 ) gene is substantially higher than the remainder sequences of Campylobacter plasmids or genomes (approximately 30%); the codon usage of t e t ( 0 ) is more similar to that of Streptococcus spp. than to that of Campylobacter spp.; and the ribosomal binding site of t e t ( 0 ) matches complementarily to the 3' end of the 16s rRNA of Bacillus subtilis (Batchelor et al., 2004; Khachatryan et al., 2006; Sougakoff et al., 1987; Taylor et al., 1983). In addition, Southern blot experiments have demonstrated the hybridization of the Campylobacter t e t ( 0 )gene to genomic DNA obtained from isolates of tetracycline-resistant Streptococcus and Enterococcus species (Taylor et al., 1983; Zilhao et al., 1988). Collectively, these observations
MECHANISMS OF ANTIBIOTIC RESISTANCE
267
strongly suggest Campylobacter obtained the t e t ( 0 ) gene from gram-positive bacteria by horizontal gene transfer.
AMINOGLYCOSIDE RESISTANCE Aminoglycosides are structurally characterized by an aminocyclitol ring bound to one or more amino sugars by pseudoglycosidic bonds Uana and Deb, 2006; Magnet and Blanchard, 2005; Smith and Baker, 2002). This class of antimicrobials is widely used in human and animal medicine and is generally considered to have broad-spectrum bacteriocidal activity. The molecular properties of aminoglycosides tend to make them polycationic, which facilitates the passive binding of the drug to negatively charged molecules such as the gram-negative outer membrane and bacterial DNA (Jana and Deb, 2006). Transfer of the compounds across the bacterial cytoplasmic membrane is energy dependent, requiring an intact electron transport chain and oxygen (Jana and Deb, 2006). For this reason, aminogycosides are considered to have limited activity in anaerobic environments. On the basis of the in vitro susceptibility of many Campylobacter isolates to aminoglycosides (Gibreel et al., 2004a; Tenover and Elvrum, 1988), it appears that the oxygen levels present in the microaerophilic environments preferred by Campylobacter are sufficient to allow for transport of the compounds into the intracellular environment. The primary mechanism of action of aminogycosides involves disruption of elongation of nascent proteins by both interfering with the normal proofreading activity of the ribosome and by preventing the translocation of the growing peptide from the A site to the P site of the ribosome Uana and Deb, 2006; Magnet and Blanchard, 2005). Consequently, proteins are either not formed or are inappropriately translated because of the lack of proofreading. Four general mechanisms of aminoglycoside resistance have been described in bacteria (Jana and Deb, 2006; Magnet and Blanchard, 2005; Smith and Baker, 2002). The first mechanism involves reduced accumulation of the drug in the intracellular environment. This decrease in intracellular concentration is mediated by either the activity of a nonspecific multidrug efflux pump working to transport the drug back into the extracellular environment or by reducing the permeability of bacterial cellular membrane to the drug Uana and Deb, 2006). The second mechanism of resistance involves the methylation of 16s rRNA in sites that interfere with efficient binding of the drug. Methyltransferase genes are most commonly found in bacterial species that produce endog-
268
ZHANG AND PLUMMER
enous aminoglycosides; however, their presence in other species of bacteria, including Pseudomonas aeruginosa, Serratia marcescens, Klebsiella pneumoniae, and E. coli, has also been reported (Magnet and Blanchard, 2005; Perichon et al., 2007). The third mechanism of resistance is of minor importance and involves mutations of the binding sites of the ribosomal RNA. Because most bacterial species have multiple copies of the rRNA sequence, this mechanism is thought to be effectively confined to Mycobacterium spp. that have a single copy of the ribosomal operon (Magnet and Blanchard, 2005). The fourth, and perhaps the most important, mechanism of aminoglycoside resistance is enzymatic modification of the drug (Jana and Deb, 2006). Enzymes responsible for aminoglycoside modification are roughly grouped into three categories on the basis of the type of reaction that they catalyze and subgrouped by denoting the molecular position to which that modification occurs and the range of drugs affected by that enzyme (Smith and Baker, 2002). For instance, a 3’aminoglycoside phosphotransferase type I11 would phosphorylate the 3’ hydroxyl site of the drugs in the type I11 group of aminoglycosides, which includes kanamycin, neomycin, lividomycin, and butirosin (Smith and Baker, 2002). Aminoglycoside adenyltransferases (also sometimes called aminoglycoside nucelotidyltransferases or ANTS) catalyze the formation of an 0-adenylated aminoglycoside from a free Mg-ATP molecule. This class of enzymes can be either chromosomally or plasmid encoded (Magnet and Blanchard, 2005). The second group of enzymes catalyzes the transfer of a phosphate group to a hydroxyl substituent of the aminoglycoside and is descriptively termed aminoglycoside phosphotransferases (APHs). Genes for these enzymes are most commonly found on multidrug-resistant plasmids, integrons, and transposons of gram-positive bacteria (Jana and Deb, 2006; Magnet and Blanchard, 2005). Acetylation of an amino group located on the aminoglycosides is catalyzed by the third set of enzymes known as aminoglycoside acetyltransferases. Such a modification is quite effective in conferring resistance and has been demonstrated to reduce the drug binding to the ribosomes by four orders of magnitude in some cases (Jana and Deb, 2006). 3 ‘-Aminoglycoside phosphotyransferases (APH(3‘)) account for the majority of aminoglycoside modifying enzymes reported in Campylobacter spp. at this time. By definition, this class of enzymes is responsible for the phosphorylation of the 3‘ hydroxyl group of aminoglycosides, and their activity is restricted to drugs having a hydroxyl group at this location. In general, resistance to kanamycin and neomycin is conferred by all APH(3’) enzymes, and re-
sistance to additional specific aminoglycosides is used to subclassify these enzymes into eight additional groups symbolized by roman numerals (I to VIII) (Smith and Baker, 2002). Only types I, 111, IVYand VII APH(3’) enzymes have been described in Campylobacter. A 3 ‘-aminoglycoside phosphotransferase type 1 gene (APH(3’)I) was identified in an isolate obtained from a human enteritis patient in France (Ouellette et al., 1987). This gene, named aphA-I, appeared to be located on the genomic DNA and was hybridized with a probe specific for the type I class of enzymes. Type I APH(3’) genes are generally considered to be specific to gram-negative bacteria, and sequence analysis of the last 5 13 bp of the open reading frame from the Campylobacter strain demonstrated nearly identical homology to the kanamycin resistance determinant of the Tn903 transposon in Escherichia coli (Ouellette et al., 1987). Furthermore, the resistance gene is linked to the IS15-A insertion sequence commonly found in the Enterobacteriaceae family of organisms (Ouellette et al., 1987). These findings would suggest that the aphA-2 gene identified in this clinical Campylobacter isolate may have originated from a member of Enterobacteriaceae (Ouellette et al., 1987). In contrast, the identification of a 3 ‘-aminoglycoside phosphotransferase type I11 gene in Campylobacter suggests that the transfer of aminoglycoside resistance determinants may also occur from gram-positive bacteria. This gene has been identified in several different clinical isolates of Campylobacter and is carried on either a transmissible plasmid or the chromosome. (Gibreel et al., 2004a; Lambert et al., 1985; Taylor et al., 1988). In 1985, Lambert and colleagues first made the observation that this gene was similar to the type I11 genes that had previously been associated only with gram-positive bacteria. Subsequently it was demonstrated that the gene sequence was identical to that of an aphA-3 gene in Streptococcus and had quasi-identity to a similar gene in Staphylococcus with only two nucleotide substitutions and a single codon deletion (Trieu-Cuot et al., 1985). Recently, Gibreel et al. (2004a) described eight C. jejuni plasmids carrying the aphA-3 gene. In two of the eight plasmids, the gene was located as part of a resistance cluster containing two other aminoglycoside resistance determinants (a 6‘-adenylyltransferase and a streptothricin acetyltransferase) and appeared to originate from a gram-positive source. In contrast, in six of the plasmids the aphA-3 gene was associated with an insertion sequence that is similar to IS607 of Helicobacter pylori. Furthermore, the G+C content of the aphA-3 gene identified in these six plasmids was approximately 30.8%, which is similar to the remainder of the Campylobacter genome.
CHAPTER 14
These findings would suggest that in addition to transfer from gram-positive bacteria, the aphA-3 gene may in some cases be indigenous to the Campylobacter species. Another class of aminoglycoside phosphotransferase enzymes recognized in Campylobacter spp. are the 3 '-aminoglycoside phosphotransferase type VII (aphA-7) (Tenover and Elvrum, 1988; Tenover et al., 1989, 1992). This plasmid-encoded resistance gene is similar to some of the type I11 genes discussed above in that it may be an indigenous Campylobacter gene. Support for this hypothesis includes the G+C content at 32.8%, which is similar to that of Campylobacter genome (Tenover et al., 1989). A single report also documented the presence of type IV APH(3') in C. coli (Rivera et al., 1986). Apart from the 3'-aminoglycoside phosphotransferases, several other aminoglycoside modifying enzymes have been described in Campylobacter spp. All belong to the aminoglycoside adenyltransferase class of enzymes responsible for the adenylation of specific amino groups of susceptible aminoglycosides. The 3rr,9-aminoglycoside adenyltransferase (aadA) gene was identified in a strain that also carried the type I phosphotransferase described above (Ouellette et al., 1987; Pinto-Alphandary et al., 1990). This class of adenyltransferases is commonly found in gramnegative bacteria and is believed to confer resistance to streptomycin and spectinomycin. The other gene, a 6-aminoglycoside adenyltransferase (aadE) was identified in five human and animal clinical isolates (3 C. coli and 2 C. jejuni) (Pinto-Alphandary et al., 1990). Unlike the aadA gene, the aadE gene in this strain only confers resistance to streptomycin and not to spectinomycin. The presence of this gene has previously been associated with only gram-positive bacteria and thus again suggest that natural transfer of antimicrobial resistance determinants occurs between Campylobacter spp. and gram-positive bacteria (Pinto-Alphandary et al., 1990). Recently, an integron-associated aacA4 gene, encoding an 6'aminoglycoside adenyltransferase, was identified in C. jejuni (Lee et al., 2002). Campylobacter isolates harboring aacA4 are highly resistant to tobramycin and gentamicin (Lee et al., 2002).
RESISTANCE TO p-LACTAMS The p-lactam class of antibiotics represents a diverse range of antimicrobial compounds including penicillin. These drugs are recognized by the presence of the p-lactam ring in association with a variety of side groups. This class of antibiotics disrupts peptidoglycan cross-linking during formation of the bac-
MECHANISMS OF ANTIBIOTIC RESISTANCE
269
terial cell wall. Specifically, the drugs bind to the penicillin binding proteins, which are actually D-alanylD-alanyl transpeptidases, and prevent the crosslinking of the peptidoglycan network (Martin and Kaye, 2004; Siu, 2002). Loss of a functionally crosslinked cell wall results in cellular swelling and subsequent death and lysis of bacterial cells. Although access to the site of action in gram-positive bacteria poses little problem for the drug, access to the cell wall of gram-negative bacteria is hindered by the outer membrane. In these bacteria, movement of the drug commonly occurs through the protein porins formed in the outer membrane (Siu, 2002). Four general mechanisms of p-lactam resistance have been described in the literature (Li et al., 2007). The predominant mechanism of resistance for gramnegative bacteria is the production of chromosomally or plasmid-encoded enzymes capable of hydrolyzing the lactam ring. These enzymes are commonly referred to as /?-lactamases, and their presence is widely documented in a diverse group of bacterial species including Campylobacter spp. (Li et al., 2007). The second general mechanism of /?-lactam resistance is the reduced uptake of the p-lactam drug from the environment as a result of alterations in membrane structures or porin proteins (Siu, 2002). The alteration of the drug target, penicillin binding proteins, has been described as a third mechanism of resistance in some bacteria but has not been reported in Campylobacter. The fourth and final mechanism of resistance involves the active transport of the drug out of the cell by the action of drug efflux pumps. Evaluation of the porin proteins of C. jejuni and C. coli suggest that access of penicillins to the periplasmic space may be hindered in comparison to other gram-negative species such as E. coli (Huyer et al., 1986; Page et al., 1989). Work by Page et al. (1989) evaluated the ability of substrates with known molecular weights to passively diffuse through the purified porin proteins of C. jejuni and C. coli. Their findings demonstrated that the pores formed by Campylobacter porins were smaller than the ones formed by E. coli porins and only allowed the passage of solutes with a molecular weight of 342 or less. It also appears that the Campylobacter porins, especially C. jejuni porins, are cation selective and are particularly capable of discriminating against divalent anions (Page et al., 1989). Because most p-lactams have molecular weights greater then 360, Campylobacter is generally less susceptible to this class of antibiotics and is particularly resistant to most of the anionic plactams (Page et al., 1989). Several studies reported that the majority of C. jejuni and C. coli isolates produce P-lactamases (Lachance et al., 1991; Li et al., 2007; Tajada et al.,
270
ZHANG AND PLUMMER
1996; Wright and Knowles, 1980). The genomic sequence of C. jejuni NCTC l l 168 revealed the presence of a putative class D P-lactamase (Cj0299) (Parkhill et al., 2000). More recently, an ortholog of Cj0299 was functionally characterized in clinical C. jejuni isolate GCO15, and this P-lactamase was named OXA-61 (Alfredson and Korolik, 2005). This product conferred 232-fold increase in the resistance to ampicillin, piperacillin, and carbenicillin but did not affect the susceptibility to cefotaxime and imipenem in C. jejuni (Alfredson and Korolik, 2005). There are also studies suggesting that Cumpylobucter carries a second class of p-lactamase enzymes different from the class D serine p-lactamases, but their identities remain to be determined (Gaudreau et al., 1987; Lachance et al., 1991).
RESISTANCE TO OTHER ANTIBIOTICS Cumpylobacter resistance to chloramphenicol rarely occurs, but a chloramphenicol resistance gene carried on a C. coli plasmid has been documented (Wang and Taylor, 1990). This gene encodes a acetyltransferase that modifies chloramphenicol and confers resistance in Cumpylobucter to this antibiotic (Tayler and Tracz, 2005; Wang and Taylor, 1990). Cumpylobucter spp. are intrinsically resistant to multiple antimicrobials, such as polymixin, bacitracin, novobiocin, vancomycin, cephalothin, cefoperazone, rifampin, streptogramin By trimethoprim, and cycloheximide (Corry et al., 1995; Taylor and Courvalin, 1988). Some of these antibiotics have been used in selective media to isolate Cumpylobucter spp. from clinical samples because they inhibit other bacteria but allow the growth of Cumpylobucter when used at appropriate concentrations (Corry et al., 1995). How Cumpylobucter is intrinsically resistant to these antibiotics is not totally clear, but advances in understanding antibiotic resistance have provided some insights into the intrinsic resistance mechanisms. One possible mechanism is that Cumpylobucter has unique membrane structural features (e.g., small pores), which reduce the permeability of certain antibiotics across the outer membrane (Huyer et al., 1986; Page et al., 1989). Another mechanism involves the multidrug efflux pump CmeABC, which plays an important role in the resistance to rifampin, cephalothin, cefoperazone, and novobiocin (Lin et al., 2002; B. Guo and Q. Zhang, unpublished data). In C. jejuni, resistance to trimethoprim has been linked to the dfrl and dfr9 genes, which encode for variant dihydrofolate reductases that are resistant to the action of trimethoprim (Gibreel and Skold, 1998, 2000). The dfrl and dfr9 genes are located on the Cumpylobacter
chromosome and are linked to an integron and transposon, respectively, suggesting that Cumpylobacter acquired the genes from a foreign source via horizontal gene transfer (Gibreel and Skold, 1998,2000). Thus, although the resistance to trimethoprim is widely distributed in C. jejuni and C. coli, it may no longer be appropriate to regard the resistance as a trait intrinsic to Cumpylobucter.
MULTIDRUG EFFLUX PUMPS Different from the specific resistance mechanisms conferred by target modification/alteration and antibiotic inaction, bacterial multidrug efflux pumps confer a broad spectrum of resistance to structurally diverse antimicrobials (Li and Nikaido, 2004; Neyfakh, 2002; Poole, 2005; Putman et al., 2000; Van Bambeke et al., 2000). On the basis of the differences in their structures and energy sources, bacterial efflux transporters are divided into distinct families including ABC (ATP-binding cassette), RND (resistancenodulation-division), MATE (multidrug and toxic compound extrusion), MF (major facilitator), and SMR (small multidrug resistance) (Li and Nikaido, 2004; Poole, 2005). The genomic sequences of Cumpylobucter revealed the presence of multiple putative multidrug efflux transporters of different families (Fouts et al., 2005; Hofreuter et al., 2006; Parkhill et al., 2000; Poly et al., 2007). For example, C. jejuni 11168 harbors 14 putative drug efflux pumps, including 1ABC (Cj1187c), 3 RND (CjO366c, Cj1033, and Cj1373), 2 MATE (Cj0560 and Cj0619), 4 MF (CjOO35c, Cj1257C, Cj1375, and Cj1687), and 4 SMR (Cj0309c, CjO31Oc, Cj1173, and Cj1174) transporters (Lin et al. 2 0 0 5 ~ )At . present, CmeABC and CmeDEF are the only functionally characterized antibiotic efflux transporters in Cumpylobucter (Akiba et al., 2006; Lin et al., 2002; Pumbwe and Piddock, 2002). CmeABC consists of three components, including an outer membrane protein (CmeC), an inner membrane drug transporter (CmeB), and a periplasmic fusion protein (CmeA). CmeABC contributes significantly to the intrinsic and acquired resistance of Cumpylobucter to structurally diverse antimicrobials (Cagliero et al., 2005; Ge et al., 2005; Gibreel et al., 2007; Lin et al., 2002; Luo et al., 2003; Mamelli et al., 2005; Pumbwe and Piddock, 2002). CmeDEF is also a RND-type efflux pump encoded by a three-gene operon (Cj1031, Cj1032, and Cj1033). It also contributes to the intrinsic resistance to antimicrobials and toxic compounds, but the contribution is not at the same scale as that of CmeABC and is normally masked by the function of CmeABC (Akiba et al., 2006). Both immunoblotting and pro-
CHAPTER 14
moter fusion demonstrated that the expression of cmeDEF is at a low level in laboratory media and is at least 10-fold lower than that of cmeABC, which may explain why CmeDEF only plays a moderate role in the intrinsic resistance to antimicrobials. Interestingly, double mutations in CmeABC and CmeDEF appeared to be lethal to strain 11168 and affected the growth of strain 81-176 in MuellerHinton broth (Akiba et al., 2006), suggesting that the interplay of the two efflux pumps plays an important role in maintaining optimal cell viability in Cumpylobacter, possibly by extruding endogenous toxic metabolites. In contrast to CmeABC and CmeDEF, the functions of other drug efflux transporters in Cumpylobucter remain unknown. Ge et al. (2005) inactivated multiple efflux transporter genes in C. jejuni 81-176 and evaluated the susceptibility of the individual mutants to 4 antimicrobials (chloramphenicol, ciprofloxacin, erythromycin, and tetracycline). Except for the cmeB mutant, other transporter mutants did not show obvious changes in the susceptibility to the 4 antimicrobials. However, the list of antimicrobials tested in the study was limited, and it was unclear whether the lack of change with the single mutants was due to low-level expression or complementary functions of the transporters, or due to potential masking by CmeABC. In addition to conferring resistance to antibiotics, CmeABC is also a key player in the resistance to bile compounds and is essential for Cumpylobucter colonization in the intestinal tract (Lin et al., 2003). The CmeB mutant grows normally in antibiotic-free media but shows a severe growth defect in bilecontaining media or in chicken intestinal extracts. When inoculated into chickens, the CmeB mutant failed to colonize the inoculated birds. Complementation of the mutant with a wild-type cmeABC allele in trans fully restored the growth of the CmeB mutant in bile-containing media and its ability to colonize chicken cecum (Lin et al., 2003). These findings provide compelling evidence that bile resistance is a natural function of CmeABC. Because C. jejuni and C. coli live in the animal intestinal tract, where bile compounds are normally present in high concentrations, bile resistance conferred by CmeABC may have been evolutionally selected and maintained, which may also explain the ubiquitous presence of this efflux pump in different strains of C. jejuni and C. coli (Fouts et al., 2005; Hofreuter et al., 2006; Lin et al., 2002; Parkhill et al., 2000; Poly et al., 2007). Expression of bacteria efflux pumps is often regulated by transcriptional factors (Li and Nikaido, 2004; Poole, 2005). In Campylobacter, CmeABC is regulated by CmeR, a transcriptional regulator en-
MECHANISMS OF ANTIBIOTIC RESISTANCE
271
coded by a gene located immediately upstream of the cmeABC operon (Lin et al., 2005a). CmeR belongs to the TetR family of transcriptional regulators (Grkovic et al., 2002; Ramos et al., 2005). Like other members of the TetR family, CmeR contains a predicted N-terminal DNA binding motif (a-helix-turna-helix) and a C-terminal domain potentially involved in the interaction with inducing ligands (Lin et al., 2005a). As a transcriptional factor, CmeR directly binds to the promoter region of cmeABC and represses the transcription of this efflux operon (Lin et al., 2005a). Deletion of cmeR or mutations in the CmeR binding site releases the repression and results in overexpression of CmeABC, leading to enhanced resistance to multiple antibiotics (Cagliero et al., 2007; Lin et al., 2005a). CmeR does not regulate cmeDEF (Akiba et al., 2006), suggesting that the two RND efflux pumps have different regulatory mechanisms. Consistent with its important role in bile resistance, cmeABC is inducible by bile salts (Lin et al., 2005b). The induction occurs with both conjugated and unconjugated bile salts and in a dose- and timedependent manner. The mechanism responsible for the induction was investigated by using surface plasmon resonance, which revealed that bile salts inhibit the binding of CmeR to the promoter DNA of cmeABC and promote the disassociation of CmeR from the promoter DNA (Lin et al., 2005b). This finding explains why bile salts induce the expression of cmeABC and strongly suggests that bile compounds are inducing ligands of CmeR. Several members in the TetR family of transcriptional factors, such as TetR and QacR, are subject to induction by various ligands, which bind to the C-terminal regions of the regulatory proteins and induce the conformational changes in the N-terminal DNA binding domains (Hillen and Berens, 1994; Hinrichs et al., 1994; Orth et al., 2000; Schumacher et al., 2001, 2002). Thus, it is likely that bile salts bind to the Cterminal region of CmeR and the interaction subsequently triggers conformational changes in the Nterminal region, prohibiting the binding of CmeR to its target promoter. Indeed, the most recent crystallization work determined the three-dimensional structures of dimeric CmeR and revealed a DNAbinding motif formed by the N termini of the dimer and a large flexible ligand-binding pocket in the Cterminal domain of each monomer (Gu et al., 2007). The binding pocket is predicted to be able to accommodate the binding by bile salts and other potential ligands. Cocrystallization of CmeR with bile salts are being performed, which will reveal whether bile salts indeed interact with the ligand-binding pocket and if such interaction cause a conformational change in the
272
ZHANG AND PLUMMER
DNA-binding domain of CmeR. Notably, the expression of cmeABC was also greatly increased in rabbit ileal loops (Stintzi et al., 2005) and chicken ceca (Y.W. Barton and Q. Zhang, unpublished data) when compared with the Campylobacter cells grown in culture media. This enhanced expression of cmeABC in vivo is likely the result of induction by bile compounds, which are present in the intestinal tract of animals. Together, these findings strongly indicate that the CmeR-regulated expression of CmeABC is important for Campylobacter adaptation to the intestinal environment. From the information discussed above, it is clear that the antibiotic efflux machinery in Campylobacter plays a significant role not only in antimicrobial resistance, but also in facilitating Campylobacter adaptation to various environments, including the inhost conditions. Despite these recent advances, the majority of the efflux transporters in Campylobacter have not been functionally characterized, and their regulatory mechanisms remain unknown. Further efforts are required to understand the natural functions of these efflux transporters and their interplay with other physiological processes in facilitating Campylobacter adaptation to various environments.
CONCLUSIONS It is clear that Campylobacter has evolved various mechanisms for antimicrobial resistance. The unique surface structures of Campylobacter render its outer membrane less permeable to certain hydrophilic antibiotics, conferring intrinsic resistance to these antimicrobials (Page et al., 1989). Campylobacter also has the ability to mutate under antibiotic selection pressure, and the relatively high mutation rate may be because Campylobacter lacks many genes involved in the DNA repair systems (Payot et al., 2006). However, the mutation rates for chromosomally encoded resistance vary with different antimicrobials and the associated resistance mechanisms. In addition, Campylobacter may acquire foreign resistance determinants via horizontal gene transfer, possibly mediated by natural transformation and/ or conjugation. Finally, the contribution of efflux pumps to antimicrobial resistance in Campylobacter is increasingly recognized. The multidrug efflux transporters facilitate the emergence of mutation-based resistance and function together with other resistance determinants in conferring high-level resistance to various antimicrobials. The efflux pumps (e.g., CmeABC) also have important physiological functions and play important roles in facilitating Campylobacter adaptation to various niches. Inhibition of the efflux sys-
tems may not only control antibiotic resistance but also prevent Campylobacter colonization in vivo. Future efforts should be directed to understanding the interplay of various mechanisms in conferring antimicrobial resistance and the molecular basis underlying the emergence and transmission of antibioticresistant Campylobacter. These efforts will ultimately facilitate the development of strategies to control antimicrobial-resistant Campylobacter in animal reservoirs and human clinical therapy. Acknowledgments. This work is supported by National Institute of Health grant ROlDK063008 and National Research Initiative Competitive grants 2003-35212-133 16 and 2005-51 110-03273 from the U.S. Department of Agriculture Cooperative State Research, Education, and Extension Service. REFERENCES Aarestrup, F. M., and H. C. Wegener. 1999. The effects of antibiotic usage in food animals on the development of antimicrobial resistance of importance for humans in Campylobacter and Escherichia coli. Microbes Infect. 1~639-644. Adachi, J. A., L. 0. Zeichner, H. L. Dupont, and C. D. Ericsson. 2000. Empirical antimicrobial therapy for traveler’s diarrhea. Clin. Infect. Dis. 31:1079-1083. Akhtar, S . Q. 1988. Antimicrobial sensitivity and plasmidmediated tetracycline resistance in Campylobacter jejuni isolated in Bangladesh. Chemotherapy 34:326-33 1. Akiba, M., J. Lin, Y. W. Barton, and Q. J. Zhang. 2006. Interaction of CmeABC and CmeDEF in conferring antimicrobial resistance and maintaining cell viability in Campylobacter jejuni. J. Antimicrob. Chemother. 5752-60. Alfredson, D. A., and V. Korolik. 2005. Isolation and expression of a novel molecular class D beta-lactamase, OXA-61, from Campylobacter jejuni. Antimicrob. Agents Chemother. 49:25 152518. Appelbaum, P. C., and P. A. Hunter. 2000. The fluoroquinolone antibacterials: past, present and future perspectives. Int. J. Antimicrob. Agents 165-15. Avrain, L., C. Vernozy-Rozand, and I. Kempf. 2004. Evidence for natural horizontal transfer of tetO gene between Campylobacter jejuni strains in chickens. J. Appl. Microbiol. 97:134-140. Bachoual, R., S . Ouabdesselam, F. Mory, C. Lascols, C. J. Soussy, and J. Tankovic. 2001. Single or double mutational alterations of gyrA associated with fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli. Microb. Drug Resist. 7: 257-26 1. Batchelor, R. A., B. M. Pearson, L. M. Friis, P. Guerry, and J. M. Wells. 2004. Nucleotide sequences and comparison of two large conjugative plasmids from different Campylobacter species. Microbiology 150:3507-35 17. Cagliero, C., M. C. Maurel, A. Cloeckaert, and S . Payot. 2007. Regulation of the expression of the CmeABC efflux pump in Campylobacter jejuni: identification of a point mutation abolishing the binding of the CmeR repressor in an in vitro-selected multidrug-resistant mutant. FEMS Microbiol. Lett. 267:89-94. Cagliero, C., C. Mouline, A. Cloeckaert, and S . Payot. 2006. Synergy between efflux pump CmeABC and modifications in ribosomal proteins L4 and L22 in conferring macrolide resistance in Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 50:3893-3896. Cagliero, C., C. Mouline, S. Payot, and A. Cloeckaert. 2005. Involvement of the CmeABC efflux pump in the macrolide resist-
CHAPTER 14
ance of Campylobacter coli. J. Antimicrob. Chemother. 56:948950. Charvalos, E., Y. Tselentis, M. M. Hamzehpour, T. Kohler, and J. C. Pechere. 1995. Evidence for an efflux pump in multidrugresistant Campylobacter jejuni. Antimicrob. Agents Chemother. 39~2019-2022. Chopra, I., and M. Roberts. 2001. Tetracycline antibiotics: mode of action, applications, molecular biology, and epidemiology of bacterial resistance. Microbiol. Mol. Biol. Rev. 65:232-260. Chu, D. T. 1999a. Recent developments in macrolides and ketolides. Curr. Opin. Microbiol. 2:467-474. Chu, D. T. 1999b. Recent progress in novel macrolides, quinolones, and 2-pyridones to overcome bacterial resistance. Med. Res. Rev. 19:497-520. Connell, S . R., D. M. Tracz, K. H. Nierhaus, and D. E. Taylor. 2003a. Ribosomal protection proteins and their mechanism of tetracycline resistance. Antimicrob. Agents Chemother. 47:36753681. Connell, S. R., C. A. Trieber, G. P. Dinos, E. Einfeldt, D. E. Taylor, and K. H. Nierhaus. 2003b. Mechanism of Tet(0)-mediated tetracycline resistance. EMBO J. 22:945-953. Cooper, R., H. Segal, A. J. Lastovica, and B. G. Elisha. 2002. Genetic basis of quinolone resistance and epidemiology of resistant and susceptible isolates of porcine Campylobacter coli strains. J. Appl. Microbiol. 93:241-249. Corry, J. E., D. E. Post, P. Colin, and M. J. Laisney. 1995. Culture media for the isolation of campylobacters. Int. J. FoodMicrobiol. 26~43-76. Dasti, J. I., U. Gross, S. Pohl, R. Lugert, M. Weig, and R. SchmidtOtt. 2007. Role of the plasmid-encoded t e t ( 0 ) gene in tetracycline-resistant clinical isolates of Campylobacter jejuni and Campylobacter coli. 1.Med. Microbiol. 56:833-837. Drlica, K., and M. Malik. 2003. Fluoroquinolones: action and resistance. Curr. Top. Med. Chem. 3:249-282. Drlica, K., and X. Zhao. 1997. DNA gyrase, topoisomerase IV, and the 4-quinolones. Microbiol. Mol. Biol. Rev. 61:377-392. Ellis-Pegler, R. B., L. K. Hyman, R. J. Ingram, and M. McCarthy. 1995. A placebo controlled evaluation of lomefloxacin in the treatment of bacterial diarrhoea in the community. J. Antimicrob. Chemother. 36:259-263. Engberg, J., F. M. Aarestrup, D. E. Taylor, P. Gerner-Smidt, and I. Nachamkin. 2001. Quinolone and macrolide resistance in Campylobacter jejuni and C. coli: resistance mechanisms and trends in human isolates. Emerg. Infect. Dis. 7:24-34. Farnell, M. B., A. M. Donoghue, K. Cole, I. Reyes-Herrera, P. J. Blore, and D. J. Donoghue. 2005. Campylobacter susceptibility to ciprofloxacin and corresponding fluoroquinolone concentrations within the gastrointestinal tracts of chickens. J. Appl. Microbiol. 99:1043-1050. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. Deboy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Gaudreau, C. L., L. A. Lariviere, J. C. Lauzer, and F. F. Turgeon. 1987. Effect of clavulanic acid on susceptibility of Campylobacter jejuni and Campylobacter coli to eight beta-lactam antibiotics. Antimicrob. Agents Chemother. 31:940-942. Ge, B., P. F. McDermott, D. G. White, and J. Meng. 2005. Role of efflux pumps and topoisomerase mutations in fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 49:3347-3354.
MECHANISMS OF ANTIBIOTIC RESISTANCE
273
Ge, B., D. G. White, P. F. McDermott, W. Girard, S. Zhao, S. Hubert, and J. Meng. 2003. Antimicrobial-resistant Campylobacter species from retail raw meats. Appl. Environ. Microbiol. 69:3005-3007. Gibreel, A., V. N. Kos, M. Keelan, C. A. Trieber, S. Levesque, S. Michaud, and D. E. Taylor. 2005. Macrolide resistance in Campylobacter jejuni and Campylobacter coli: molecular mechanism and stability of the resistance phenotype. Antimicrob. Agents Chemother. 49:2753-2759. Gibreel, A., E. Sjogren, B. Kaijser, B. Wretlind, and 0. Skold. 1998. Rapid emergence of high-level resistance to quinolones in Campylobacter jejuni associated with mutational changes ingyrA and parC. Antimicrob. Agents Chemother. 42:3276-3278. Gibreel, A., and 0. Skold. 1998. High-level resistance to trimethoprim in clinical isolates of Campylobacter jejuni by acquisition of foreign genes (dfrl and dfr9) expressing drug-insensitive dihydrofolate reductases. Antimicrob. Agents Chemother. 42: 3059-3064. Gibreel, A., and 0. Skold. 2000. An integron cassette carrying dfrl with 90-bp repeat sequences located on the chromosome of trimethoprim-resistant isolates of Campylobacter jejuni. Microb. Drug Resist. 6:91-98. Gibreel, A., 0. Skold, and D. E. Taylor. 2004a. Characterization of plasmid-mediated aphA-3 kanamycin resistance in Campylobacter jejuni. Microb. Drug Resist. 10:98-105. Gibreel, A., and D. E. Taylor. 2006. Macrolide resistance in Campylobacter jejuni and Campylobacter coli. J. Antimicrob. Chemother. 58:243-255. Gibreel, A., D. M. Tracz, L. Nonaka, T. M. Ngo, S. R. Connell, and D. E. Taylor. 2004b. Incidence of antibiotic resistance in Campylobacter jejuni isolated in Alberta, Canada, from 1999 to 2002, with special reference to tet(0)-mediated tetracycline resistance. Antimicrob. Agents Chemother. 48:3442-3450. Gibreel, A., N. M. Wetsch, and D. E. Taylor. 2007. Contribution of the CmeABC efflux pump to macrolide and tetracycline resistance in Campylobacter jejuni. Antimicrob. Agents Chemother. Gootz, T. D., and B. A. Martin. 1991. Characterization of highlevel quinolone resistance in Campylobacter jejuni. Antimicrob. Agents Chemother. 35:840-845. Griggs, D. J., M. M. Johnson, J. A. Frost, T. Humphrey, F. Jorgensen, and L. J. Piddock. 2005. Incidence and mechanism of ciprofloxacin resistance in Campylobacter spp. isolated from commercial poultry flocks in the United Kingdom before, during, and after fluoroquinolone treatment. Antimicrob. Agents Chemother. 49:699-707. Grkovic, S., M. H. Brown, and R. A. Sknrray. 2002. Regulation of bacterial drug export systems. Microbiol. Mol. Biol. Rev. 66: 671-701. Gu, R., C. C. Su, F. Shi, M. Li, G. McDermott, Q. Zhang, and E. W. Yu. 2007. Crystal structure of the transcriptional regulator CmeR from Campylobacter jejuni. J. Mol. Biol. 372583593. Gupta, A., J. M. Nelson, T. J. Barrett, R. V. Tauxe, S. P. Rossiter, C. R. Friedman, K. W. Joyce, K. E. Smith, T. F. Jones, M. A. Hawkins, B. Shiferaw, J. L. Beebe, D. J. Vugia, T. RabatskyEhr, J. A. Benson, T. P. Root, and F. J. Angulo. 2004. Antimicrobial resistance among Campylobacter strains, United States, 1997-2001. Emerg. Infect. Dis. 10:1102-1109. Harrow, S . A., B. J. Gilpin, and J. D. Klena. 2004. Characterization of erythromycin resistance in Campylobacter coli and Campylobacter jejuni isolated from pig offal in New Zealand. J. Appl. Microbiol. 97:141-148. Hillen, W., and C. Berens. 1994. Mechanisms underlying expression of TnlO encoded tetracycline resistance. Annu. Rev. Microbiol. 48:345-369.
274
ZHANG AND PLUMMER
Hinrichs, W., C. Kisker, M. Duvel, A. Muller, K. Tovar, W. Hillen, and W. Saenger. 1994. Structure of the Tet repressortetracycline complex and regulation of antibiotic resistance. Science 264:418-420. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Hooper, D. C. 1998. Clinical applications of quinolones. Biochim. Biophys. Acta 1400:45-61. Hooper, D. C. 2001. Emerging mechanisms of fluoroquinolone resistance. Emerg. Infect. Dis. 7:337-341. Huyer, M., T. R. Parr, R. E. W. Hancock, and W. J. Page. 1986. Outer membrane porin protein of Campylobacter jejuni. FEMS Microbiol. Lett. 37:247-250. Jacobs, M. R., and C. E. Johnson. 2003. Macrolide resistance: an increasing concern for treatment failure in children. Emerg. lnfect. Dis. 22:S131-S138. Jana, S., and J. K. Deb. 2006. Molecular understanding of aminoglycoside action and resistance. Appl. Microbiol. Biotechnol. 70:140-150. Jensen, L. B., and F. M. Aarestrup. 2001. Macrolide resistance in Campylobacter coli of animal origin in Denmark. Antimicrob. Agents Chemother. 45:371-372. Kasbekar, N., and P. S . Acharya. 2005. Telithromycin: the first ketolide for the treatment of respiratory infections. Am. J. Health Syst. Pharm. 62:905-916. Khachatryan, A. R., D. D. Hancock, T. E. Besser, and D. R. Call. 2006. Antimicrobial drug resistance genes do not convey a secondary fitness advantage to calf-adapted Escherichia coli. Appl. Environ. Microbiol. 72:443-448. Khodursky, A. B., E. L. Zechiedrich, and N. R. Cozzarelli. 1995. Topoisomerase IV is a target of quinolones in Escherichia coli. Proc. Natl. Acad. Sci. USA 92:11801-11805. Kuo, M. S., D. G. Chirby, A. D. Argoudelis, J. I. Cialdella, J. H. Coats, and V. P. Marshall. 1989. Microbial glycosylation of erythromycin A. Antimicrob. Agents Chemother. 33:2089-2091. Lachance, N., C. Gaudreau, F. Lamothe, and L. A. Lariviere. 1991. Role of the beta-lactamase of Campylobacter jejuni in resistance to beta-lactam agents. Antimicrob. Agents Cbemother. 35:813-818. Lambert, T., G. Gerbaud, P. Trieu-Cuot, and P. Courvalin. 1985. Structural relationship between the genes encoding 3'-aminoglycoside phosphotransferases in Campylobacter and in grampositive cocci. Ann. Inst. Pasteur Microbiol. 136B:135-150. Leclercq, R. 2002. Mechanisms of resistance to macrolides and lincosamides: nature of the resistance elements and their clinical implications. Clin. Infect. Dis. 34:482-492. Lee, M. D., S. Sanchez, M. Zimmer, U. Idris, M. E. Berrang, and P. F. McDermott. 2002. Class 1integron-associated tobramycingentamicin resistance in Campylobacter jejuni isolated from the broiler chicken house environment. Antimicrob. Agents Chemother. 46:3660-3664. Levy, S. B. 1992. Active efflux mechanisms for antimicrobial resistance. Antimicrob. Agents Chemotber. 36:695-703. Li, X. Z., M. Mehrotra, S. Ghimire, and L. Adewoye. 2007. betaLactam resistance and beta-lactamases in bacteria of animal origin. Vet. Microbiol. 121:197-214. Li, X. Z., and H. Nikaido. 2004. Efflux-mediated drug resistance in bacteria. Drugs 64:159-204. Lin, J., M. Akiba, 0. Sahin, and Q. Zhang. 2005a. CmeR functions as a transcriptional repressor for the multidrug efflux pump CmeABC in Campylobacter jejuni. Antimicrob. Agents Chemother. 49: 1067-1075. Lin, J., C. Cagliero, B. Guo, Y. W. Barton, M. C. Maurel, S. Payot, and Q. Zhang. 2005b. Bile salts modulate expression of the
CmeABC multidrug efflux pump in Campylobacter jejuni. J. Bacteriol. 187:74 17-7424. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Chemother. 46:2124-2131. Lin, J., 0. Sahin, L. 0. Michel, and Q. Zhang. 2003. Critical role of multidrug efflux pump CmeABC in bile resistance and in vivo colonization of Campylobacter jejuni. Infect. lmmun. 71:42504259. Lin, J., M. Yan, 0. Sahin, S . Pereira, Y. J. Chang, and Q. Zhang. 2007. Effect of macrolide usage on emergence of erythromycinresistant Campylobacter isolates in chickens. Antimicrob. Agents Chemother. 5 1:1678-1 686. Lin, J., M. Akiba, and Q. Zhang. 2005c. Multidrug efflux systems in Campylobacter, p. 205-218. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Luangtongkum, T., T. Y. Morishita, A. J. Ison, S. Huang, P. F. McDermott, and Q. Zhang. 2006. Effect of conventional and organic production practices on the prevalence and antimicrobial resistance of Campylobacter spp. in poultry. Appl. Environ. Microbiol. 72:3 600-3 607. Luo, N., S. Pereira, 0. Sahin, J. Lin, S. Huang, L. Michel, and Q. Zhang. 2005. Enhanced in vivo fitness of fluoroquinoloneresistant Campylobacter jejuni in the absence of antibiotic selection pressure. Proc. Natl. Acad. Sci. USA 102541-546. Luo, N., 0. Sahin, J. Lin, L. 0. Michel, and Q. Zhang. 2003. In vivo selection of Campylobacter isolates with high levels of fluoroquinolone resistance associated with gyrA mutations and the function of the CmeABC efflux pump. Antimicrob. Agents Chemother. 47:390-394. Magnet, S., and J. S . Blanchard. 2005. Molecular insights into aminoglycoside action and resistance. Chem. Rev. 105:477-498. Mamelli, L., V. Prouzet-Mauleon, J. M. Pages, F. Megraud, and J. M. Bolla. 2005. Molecular basis of macrolide resistance in Campylobacter: role of efflux pumps and target mutations. J. Antimicrob. Chemother. 56:491-497. Manavathu, E. K., K. Hiratsuka, and D. E. Taylor. 1988. Nucleotide sequence analysis and expression of a tetracyclineresistance gene from Campylobacter jejuni. Gene 62: 17-26. Maravic, G. 2004. Macrolide resistance based on the Ermmediated rRNA methylation. Curr. Drug Targets Infect. Disord. 4~193-202. Martin, S. I., and K. M. Kaye. 2004. Beta-lactam antibiotics: newer formulations and newer agents. Infect. Dis. Clin. North Am. 18:603-619. Martinez, A., and J. Lin. 2006. Effect of an efflux pump inhibitor on the function of the multidrug efflux pump CmeABC and antimicrobial resistance in Campylobacter. Foodborne Pathog. Dis. 3~393-402. McDermott, P. F., S . M. Bodeis, L. L. English, D. G. White, R. D. Walker, S . Zhao, S . Simjee, and D. D. Wagner. 2002. Ciprofloxacin resistance in Campylobacter jejuni evolves rapidly in chickens treated with fluoroquinolones. J. Infect. Dis. 185:837840. McEwen, S. A., and P. J. Fedorka-Cray. 2002. Antimicrobial use and resistance in animals. Clin. Infect. Dis. 34 (Suppl 3):S93S106. Nachamkin, I., H. Ung, and M. Li. 2002. Increasing fluoroquinolone resistance in Campylobacter jejuni, Pennsylvania, USA, 1982-2001. Emerg. Infect. Dis. 8:1501-1503. Neyfakh, A. A. 2002. Mystery of multidrug transporters: the answer can be simple. Mol. Microbiol. 44:1123-1130. Ng, L. K., M. E. Stiles, and D. E. Taylor. 1987. DNA probes for identification of tetracycline resistance genes in Campylobacter
CHAPTER 14
species isolated from swine and cattle. Antimicrob. Agents Chemother. 31:1669-1674. Nirdnoy, W., C. J. Mason, and P. Guerry. 2005. Mosaic structure of a multiple-drug-resistant, conjugative plasmid from Campylobacter jejuni. Antimicrob. Agents Chemother. 49:2454-2459. Noguchi, N., A. Emura, H. Matsuyama, K. OHara, M. Sasatsu, and M. Kono. 1995. Nucleotide sequence and characterization of erythromycin resistance determinant that encodes macrolide 2'-phosphotransferase I in Escherichia coli. Antimicrob. Agents Chemother. 39:2359-2363. Orth, P., D. Schnappinger, W. Hillen, W. Saenger, and W. Hinrichs. 2000. Structural basis of gene regulation by the tetracycline inducible Tet repressor-operator system. Nut. Struct. Biol. 7:215-2 19. Ouellette, M., G. Gerbaud, T. Lambert, and P. Courvalin. 1987. Acquisition by a Cumpylobacter-like strain of aphA-2, a kanamycin resistance determinant from members of the family Enterobacteriaceae. Antimicrob. Agents Chemother. 31:10211026. Page, W. J., G. Huyer, M. Huyer, and E. A. Worobec. 1989. Characterization of the porins of Campylobacterjejuni and Campylobacter coli and implications for antibiotic susceptibility. Antimicrob. Agents Chemother. 33:297-303. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Mode, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the foodborne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Payot, S., L. Avrain, C. Magras, K. Praud, A. Cloeckaert, and E. Chaslus-Dancla. 2004. Relative contribution of target gene mutation and efflux to fluoroquinolone and erythromycin resistance, in French poultry and pig isolates of Campylobactercoli. Int. 1.Antimicrob. Agents 23:468-472. Payot, S., J. M. Bolla, D. Corcoran, S. Fanning, F. Megraud, and Q. Zhang. 2006. Mechanisms of fluoroquinolone and macrolide resistance in Campylobacter spp. Microbes Infect. 8:1967-1971. Payot, S., A. Cloeckaert, and E. Chaslus-Dancla. 2002. Selection and characterization of fluoroquinolone-resistant mutants of Campylobacter jejuni using enrofloxacin. Microb. Drug Resist. 8: 335-343. Perichon, B., P. Courvalin, and M. Galimand. 2007. Transferable resistance to aminoglycosides by methylation of G1405 in 16s rRNA and to hydrophilic fluoroquinolones by QepA-mediated efflux in Escherichia coli. Antimicrob. Agents Chemother. 51: 2464-2469. Piddock, L. J. 1995. Quinolone resistance and Campylobacterspp. J. Antimicrob. Chemother. 36:891-898. Piddock, L. J., V. Ricci, L. Pumbwe, M. J. Everett, and D. J. Griggs. 2003. Fluoroquinolone resistance in Campylobacter species from man and animals: detection of mutations in topoisomerase genes. 1.Antimicrob. Chemother. 51:19-26. Pinto-Alphandary, H., C. Mabilat, and P. Courvalin. 1990. Emergence of aminoglycoside resistance genes aadA and aadE in the genus Campylobacter. Antimicrob. Agents Chemother. 34:12941296. Poehlsgaard, J., and S. Douthwaite. 2003. Macrolide antibiotic interaction and resistance on the bacterial ribosome. Curr. Opin. Investig. Drugs 4:140-148. Poehlsgaard, J., and S. Douthwaite. 2005. The bacterial ribosome as a target for antibiotics. Nut. Rev. Microbiol. 3:870-881. Poly, F., T. Read, D. R. Tribble, S. Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. Immun. 75:3425-3433.
MECHANISMS OF ANTIBIOTIC RESISTANCE
275
Poole, K. 2000. Efflux-mediated resistance to fluoroquinolones in gram-negative bacteria. Antimicrob. Agents Chemother. 44: 2233-2241, Poole, K. 2005. Efflux-mediated antimicrobial resistance. J. Antimicrob. Chemother. 56:20-51. Pratt, A., and V. Korolik. 2005. Tetracycline resistance of Australian Campylobacter jejuni and Campylobactercoli isolates.J. Antimicrob. Chemother. 55:452-460. Prescott, J. F. 2000. Lincosamides, macrolides, and pleuromutilins. p. 229-263. In J. F. Prescott, J. D. Baggot, and R. D. Walker (ed.), Antimicrobial Therapy in Veterinary Medicine, 3rd ed. Iowa State Press, Ames. Price, L. B., E. Johnson, R. Vailes, and E. Silbergeld. 2005. Fluoroquinolone-resistant Campylobacter isolates from conventional and antibiotic-free chicken products. Environ. Health Perspect. 113557-5 60. Price, L. B., L. G. Lackey, R Vailes, and E. Silbergeld. 2007. The persistence of fluoroquinolone-resistant Campylobacter in poultry production. Environ. Health Perspect. 115:1035-1039. Pumbwe, L., and L. J. Piddock. 2002. Identification and molecular characterisation of CmeB, a Campylobacter jejuni multidrug efflux pump. FEMS Microbiol. Lett. 206:185-189. Putman, M., H. W. van Veen, and W. N. Konings. 2000. Molecular properties of bacterial multidrug transporters. Microbiol. Mol. Biol. Rev. 64:672-693. Ramos, J. L., M. Martinez-Bueno, A. J. Molina-Henares, W. Teran, K. Watanabe, X. D. Zhang, M. T. Gallegos, R Brennan, and R. Tobes. 2005. The TetR family of transcriptional repressors. Microbiol. Mol. Biol. Rev. 69:326-356. Rautelin, H., 0. V. Renkonen, and T. U. Kosunen. 1991. Emergence of fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli in subjects from Finland. Antimicrob. Agents Chemother. 35:2065-2069. Rivera, M. J., J. Castillo, C. Martin, M. Navarro, and R. GomezLus. 1986. Aminoglycoside-phosphotransferases APH(3')-IV and APH(3") synthesized by a strain of Campylobacter coli. 1. Antimicrob. Chemother. 18:153-15 8. Ruiz, J., P. Goni, F. Marco, F. Gallardo, B. Mirelis, d. A. Jimenez, and J. Vila. 1998. Increased resistance to quinolones in Campylobacter jejuni: a genetic analysis of gyrA gene mutations in quinolone-resistant clinical isolates. Microbiol. Immunol. 42: 223-226. Saenz, Y., M. Zarazaga, M. Lantero, M. J. Gastanares, F. Baquero, and C. Torres. 2000. Antibiotic resistance in Campylobacter strains isolated from animals, foods, and humans in Spain in 1997-1998. Antimicrob. Agents Chemother. 44:267-271. Sanchez, R., V. Fernandez-Baca, M. D. Diaz, P. Munoz, M. Rodriguez-Creixems, and E. Bouza. 1994. Evolution of susceptibilities of Campylobacter spp. to quinolones and macrolides. Antimicrob. Agents Chemother. 38: 1879-1882. Schumacher, M. A., M. C. Miller, S. Grkovic, M. H. Brown, R. A. Skurray, and R. G. Brennan. 2001. Structural mechanisms of QacR induction and multidrug recognition. Science 294:21582163. Schumacher, M. A., M. C. Miller, S. Grkovic, M. H. Brown, R. A. Skurray, and R. G . Brennan. 2002. Structural basis for cooperative DNA binding by two dimers of the multidrug-binding protein QacR. EMBOJ. 21:1210-1218. Segreti, J., T. D. Gootz, L. J. Goodman, G. W. Parkhurst, J. P. Quinn, B. A. Martin, and G. M. Trenholme. 1992. High-level quinolone resistance in clinical isolates of Campylobacter jejuni. J. Infect. Dis. 165:667-670. Shea, M. E., and H. Hiasa. 1999. Interactions between DNA helicases and frozen topoisomerase IV-quinolone-DNA ternary complexes. 1.Biol. Chem. 274:22747-22754.
276
ZHANG AND PLUMMER
Siu, L. K. 2002. Antibiotics: action and resistance in gram-negative bacteria. 1.Microbiol. Immunol. Infect. 35:1-11. Smith, C. A., and E. N. Baker. 2002. Aminoglycoside antibiotic resistance by enzymatic deactivation. Cuw. Drug Targets Infect. Disord. 2:143-160. Smith, K. E., J. B. Bender, and M. T. Osterholm. 2000. Antimicrobial resistance in animals and relevance to human infections, p. 483-495. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. American Society for Microbiology, Washington, DC. Sougakoff, W., B. Papadopoulou, P. Nordmann, and P. Courvalin. 1987. Nucleotide-sequence and distribution of gene tet(0) encoding tetracycline resistance in Campylobacter coli. FEMS Microbiol. Lett. 44:153-159. Stintzi, A., D. Marlow, K. Palyada, H. Naikare, R. Panciera, L. Whitworth, and C. Clarke. 2005. Use of genome-wide expression profiling and mutagenesis to study the intestinal lifestyle of Campylobacter jeiuni. Infect. Immun. 73:1797-1810. Tajada, P., J. L. Gomez-Graces, J. I. Alos, D. Balas, and R. Cogollos. 1996. Antimicrobial susceptibilities of Campylobacter jejuni and Campylobacter coli to 12 beta-lactam agents and combinations with beta-lactamase inhibitors. Antimicrob. Agents Chemother. 40: 1924-1 925. Taylor, D. E., and P. Courvalin. 1988. Mechanisms of antibiotic resistance in Campylobacter species. Antimicrob. Agents Chemother. 32:1107-1112. Taylor, D. E., and D. M. Tracz. 2005. Mechanisms of antimicrobial resistance in Campylobacter, p. 193-204. In J. M.Ketley and M. E. Konkel (ed.), Campylobacter: Molecular Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Taylor, D. E., R. S. Garner, and B. J. Allan. 1983. Characterization of tetracycline resistance plasmids from Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 24: 930-935. Taylor, D. E., K. Hiratsuka, H. Ray, and E. K. Manavathu. 1987. Characterization and expression of a cloned tetracycline resistance determinant from Campylobacter jejuni plasmid pUA466. 1.Bacteriol. 169:2984-2989. Taylor, D. E., W. Yan, L. K. Ng, E. K. Manavathu, and P. Courvalin. 1988. Genetic characterization of kanamycin resistance in Campylobacter coli. Ann. Inst. Pasteur Microbiol. 139:665-676. Tenover, F. C., and P. M. Elvrum. 1988. Detection of two different kanamycin resistance genes in naturally occurring isolates of Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 32:1170-1173. Tenover, F. C.,C. L. Fennell, L. Lee, and D. J. LeBlanc. 1992. Characterization of two plasmids from Campylobacter jejuni isolates that carry the aphA-7 kanamycin resistance determinant. Antimicrob. Agents Chemother. 36:7 12-7 16. Tenover, F. C., T. Gilbert, and P. O'Hara. 1989. Nucleotide sequence of a novel kanamycin resistance gene, aphA-7, from Campylobacter jejuni and comparison to other kanamycin phosphotransferase genes. Plasmid 2252-58. Tran, J. H., G. A. Jacoby, and D. C. Hooper. 2005. Interaction of the plasmid-encoded quinolone resistance protein Qnr with
Escherichia coli DNA gyrase. Antimicrob. Agents Chemother. 49: 118-125. Trieu-Cuot, P., G. Gerbaud, T. Lambert, and P. Courvalin. 1985. In vivo transfer of genetic information between gram-positive and gram-negative bacteria. EMBO J. 4:3583-3587. Van Bambeke, F., E. Balzi, and P. M.Tulkens. 2000. Antibiotic efflux pumps. Biochem. Pharmacol. 60:457-470. van Boven, M., K. T. Veldman, M. C. de Jong, and D. J. Mevius. 2003. Rapid selection of quinolone resistance in Campylobacter jejuni but not in Escherichia coli in individually housed broilers. J. Antimicrob. Chemother. 52:719-723. Van Looveren, M., G . Daube, L. De Zutter, J. M. Dumont, C. Lammens, M. Wijdooghe, P. Vandamme, M. Jouret, M. Cornelis, and H. Goossens. 2001. Antimicrobial susceptibilities of Campylobacter strains isolated from food animals in Belgium.]. Antimicrob. Chemother. 48:235-240. Wang, M., D. F. Sahm, G. A. Jacoby, Y. Zhang, and D. C. Hooper. 2004. Activities of newer quinolones against Escherichia coli and Klebsiella pneumoniae containing the plasmidmediated quinolone resistance determinant qnr. Antimicrob. Agents Chemother. 48:1400-1401. Wang, M., J. H. Tran, G. A. Jacoby, Y. Zhang, F. Wang, and D. C. Hooper. 2003. Plasmid-mediated quinolone resistance in clinical isolates of Escherichia coli from Shanghai, China. Antimicrob. Agents Chemother. 47:2242-2248. Wang, Y.,W. M. Huang, and D. E. Taylor. 1993. Cloning and nucleotide sequence of the Campylobacter jejuni gyrA gene and characterization of quinolone resistance mutations. Antimicrob. Agents Chemother. 37:45 7-463. Wang, Y., and D. E. Taylor. 1990. Chloramphenicol resistance in Campylobacter coli: nucleotide sequence, expression, and cloning vector construction. Gene 94:23-28. Willmott, C. J., S. E. Critchlow, I. C. Eperon, and A. Maxwell. 1994. The complex of DNA gyrase and quinolone drugs with DNA forms a barrier to transcription by RNA polymerase. 1. Mol. Biol. 242:351-363. Wretlind, B., A. Stromberg, L. Ostlund, E. Sjogren, and B. Kaijser. 1992. Rapid emergence of quinolone resistance in Campylobacter jejuni in patients treated with norfloxacin. Scand. I. Infect. Dis. 24:685-686. Wright, E. P., and M. A. Knowles. 1980. Beta-lactamase production by Campylobacter jejuni. J. Clin. Pathol. 33:904-905. Yan, M., 0. Sahin, J. Lin, and Q. Zhang. 2006. Role of the CmeABC efflux pump in the emergence of fluoroquinoloneresistant Campylobacter under selection pressure. 1.Antimicrob. Chemother. 58:1154-1159. Yoshida, H., M. Bogaki, M. Nakamura, and S. Nakamura. 1990. Quinolone resistance-determining region in the DNA gyrase gyrA gene of Escherichia coli. Antimicrob. Agents Chemother. 34: 1271-1272. Zhang, Q., J. Lin, and S. Pereira. 2003. Fluoroquinolone-resistant Campylobacter in animal reservoirs: dynamics of development, resistance mechanisms and ecological fitness. Anim. Health Res. Rev. 4:63-71. Zilhao, R., B. Papadopoulou, and P. Courvalin. 1988. Occurrence of the Campylobacter resistance gene tetO in Enterococcus and Streptococcus spp. Antimicrob. Agents Chemother. 32: 17931796.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blnser 0 2008 ASM Press, Washington, DC
Chapter 15
National Molecular Subtyping Network for Food-Borne Bacterial Disease Surveillance in the United States PETERGEFWER-SMIDT, STEVEN G. STROIKA, AND COLLETTE FITZGERALD
Molecular subtyping has evolved during the past 30 years as an important tool in the study of the taxonomy, phylogeny, and epidemiology of microbial pathogens (Struelens, 1996). In the latter context, restriction fragment length polymorphism (RFLP) methods have been instrumental in detection of clusters and investigation of outbreaks of food-borne infections during the last 10 to 15 years (Wiedmann, 2002). Among these RFLP methods, macrorestriction endonuclease analysis of genomic DNA by pulsedfield gel electrophoresis (PFGE) has become the gold standard (Olive and Bean, 1999). Typically, an electrophoresis-based method like PFGE is comparative by nature, i.e., comparison can only be performed between isolates subtyped in the same experiment. However, through strict standardization of the subtyping procedure and by the use of sophisticated image analysis software, it is possible to compare PFGE patterns generated in different laboratories at different times. Through the 1980s and the early 1990s, PFGE proved useful as a high discriminatory tool to delineate outbreaks of food-borne pathogens all over the world as a supplement to the less discriminatory phenotypic methods like serotyping and biotyping and the first-generation DNA fingerprinting methods, e.g., plasmid profiling and ribotyping (Olive and Bean, 1999; Struelens, 1996; Wiedmann, 2002). The former methods were rarely sufficiently discriminatory to differentiate outbreak-related cases from sporadic cases, and the latter were either not very stable or were labor intensive. The first PFGE protocols were slow, taking 3 to 4 days to complete, thereby hampering their utility in a real-time event. However, more important factors slowed down the microbiological investigations of food-borne outbreaks; only a few laboratories had the capacity to perform molec-
ular subtyping, and the protocols they used differed from each other. Therefore, the isolates had to be transported to the laboratories, which were often in a different state from where the outbreak was occurring, and if isolates from the same outbreak were sent to different laboratories, the results could not be easily compared. In the early 1 9 9 0 ~the ~ food-borne pathogen laboratory at CDC became increasingly involved in the investigation of outbreaks caused by Shiga toxin (Vero cytotoxin)-producing Escherichia coli (STEC) 0157:H7, and the laboratory could not keep pace with the demand for outbreak-related PFGE analyses. At the same time, the first software packages to analyze DNA fingerprints became available, enabling the analysis of PFGE profiles that had been run on different gels, and the first rapid 24-h PFGE was described (Gautom, 1997). This led to a paradigm shift in the microbiological outbreak investigations of food-borne infections and to the birth of the molecular surveillance network PulseNet (Swaminathan et al., 2001). PULSENET USA PulseNet USA is a network of public health laboratories from all 50 states, federal food regulatory agency laboratories, and a few state agricultural laboratories in the United States dedicated to molecular surveillance of food-borne bacterial infections (Fig. 1).The network is administered by the Association of Public Health Laboratories and coordinated by CDC, and it receives its funding from both federal sources and state funds. In PulseNet, PFGE is performed locally, where and when the outbreaks occur, by using protocols that are highly standardized with regard to methodology, reagents, equipment, data ac-
Peter Gerner-Smidt, Steven G . Stroika, and Collette Fitzgerald Enteric Diseases Laboratory Branch, Centers for Disease Control and Prevention, 1600 Clifton Road, Mail Stop C03, Atlanta, GA 30333.
277
278
GERNER-SMIDT ET AL.
West Mountain South Central North Central ~
i
~ M~d~Atlantic ~ ~ s Southeast t Northeast
Figure 1. Organizational map of PulseNet USA.
quisition, and analysis. In the beginning, images of all PFGE gels were sent by e-mail to the database managers at CDC, where they were analyzed and compared with all patterns in the database. The PulseNet system rapidly proved its utility not just for the outbreak investigation but also for detection of clusters of infections caused by isolates with indistinguishable profiles representing common source outbreaks (CDC, 2005a, 2005b; Gerner-Smidt et al., 2005; Swaminathan e t al., 2001). Because all STEC 0 1 5 7 that are submitted to PulseNet participating laboratories are subtyped and the patterns submitted to the database in real time, PulseNet often facilitates early detection of outbreaks, thus making it possible to investigate them and control them at an early stage (CDC, 2007; Gerner-Smidt et al., 2005). The PulseNet system is equally efficient in detecting single-state and multistate outbreaks. Since the inception of the network in 1996, rapid standardized PulseNet PFGE protocols have been developed for STEC 0157, Salmonella, Shigellu (Ribot et al., 2006), Listeria monocytogenes (Graves and Swaminathan, 2001), Clostridium perfringens (Maslanka et al., 1999), thermophilic Campylobacter (Ribot et al., 2001), Vibrio cholerae (Cooper et al., 2006), Vibrio
parahaemolyticus (Parsons et al., 2007), and the biothreat agent Yersinia pestis (http: //www.cdc.gov/ pulsenet/protocols.htm). Protocols for Clostridium botulinum, Yersinia enterocolitica, non-0157 STEC, and Francisella tularensis are under development. Today, almost all PulseNet participants have a local database containing the PFGE patterns generated in their laboratory. The pattern analysis software used is a highly customized version of the BioNumerics serverlclient software (Applied Maths, Sint-Martens-Latem, Belgium). The participants have client versions of the databases installed that allow them to connect via the Internet to the central national server databases at CDC. After performing PFGE on isolates received in their respective laboratories, a digital image of the gel is captured and transferred to their client database. The profiles are analyzed and compared with the local database, then uploaded to the national database via the Internet. The participants can also compare any profile in their database against the whole national database through the Internet. If the participants detect a cluster of isolates with indistinguishable patterns, they will notify their local food-borne epidemiologists and, as appropriate, begin an outbreak investigation. Concurrently,
CHAPTER 15
they will alert the other PulseNet participants about the cluster investigation by posting a message to the PulseNet listserv. The other participants can then compare their own profiles against the cluster patterns and notify each other if they believe that they have isolates that might be part of the cluster. All clusters posted on the PulseNet listserv are confirmed by the database team at CDC, and if multiple states are involved, the food-borne epidemiologists at CDC will usually be involved in the investigation. At least once a week, the PulseNet database team at CDC performs comparison of the patterns that have been uploaded to the national server within a 60- to 120day window (a cluster search). If a cluster is identified, the participants in the states involved are notified immediately, an alert is posted on the mailing list, and the cluster findings are discussed with foodborne epidemiologists at CDC to determine whether an outbreak investigation should be initiated. The following parameters of a cluster are considered when deciding to initiate a formal outbreak investigation: the number of cases, the commonality and history of the pattern involved, geographic diversity of the involved cases, and recent uploads of nonhuman isolates with the cluster pattern.
-
NATIONAL MOLECULAR SUBTYPING NETWORK
279
site on the chromosome, and therefore, isolates with similar patterns may not be as related as they seem. For these reasons, PulseNet has initiated the development and implementation of next-generation subtyping methods, multilocus variable number tandem repeat analysis, and single nucleotide polymorphism analysis. The first of these methods, multilocus variable number tandem repeat analysis for STEC 0157, has been validated and is under implementation in the PulseNet network (Hyytia-Trees et al., 2006). Both of these methods provide phylogenic information, and the data are easier to analyze because the data produced are sets of characters rather than a banding pattern. However, the differentiation criteria or answering the question “How different is different?” needs careful evaluation before these methods can be routinely implemented in PulseNet; the new methods are not as universal as PFGE-that is, the same PFGE method can be used to study all Salmonella serotypes, but the same variable number of tandem repeat sites or single nucleotide polymorphisms are not informative or present in all organisms. For this reason, it seems likely that PFGE will still be used in PulseNet for years to come.
PULSENET PARTNERSHIPS NEXT-GENERATION SUBTYPING METHOD IMPLEMENTATION IN THE PULSENET NETWORK Even though PFGE as used by PulseNet is a fairly rapid subtyping method, comparison of the patterns becomes increasingly tedious as the databases grow because it is based on electronic images. With the current size of the databases (Table l),comparisons are fairly slow, and the chance of overlooking clusters is a real threat. Additionally, PFGE analysis does not provide phylogenetic information, i.e., restriction fragments of the same size from different isolates do not necessarily originate from the same
Table 1. Number of entries in the PulseNet national PFGE databases as of the end of 2006
No. of entries with: Database
Campylobacter E . coli Listeria Salmonella Shigella V. cholerae Y. pestis
No. of entries 4,493 25,748 7,642 140,281 24,s 11 187 1,503
1st enzyme
2nd enzyme
4,464 23,884 7,443 138,246 24,473 180 1,501
1,501 11,309 6,455 19,924 1,507 173 15
In cluster evaluations and outbreak investigations, PulseNet works with many partners in the United States: United States Department of Agriculture (USDA) VetNet, OutbreakNet, FoodNet, the National Antimicrobial Resistance Monitoring System (NARMS), and PulseNet International. The USDA VetNet performs PFGE on isolates of nontyphoid Salmonella and Campylobacter received in the animal arm of NARMS, including isolates from diagnostic specimens, healthy farm animals, and carcasses of food animals at slaughter (Jackson et al., 2007). The database managers in USDA VetNet and PulseNet can search each other’s databases, and this functionality is used when any cluster of human infections are detected to get a hint about its possible source. VetNet also alerts PulseNet about clusters of nonhuman patterns that may later show up in human infections. OutbreakNet, FoodNet, and NARMS are, together with PulseNet, the key surveillance systems for food-borne infections in the United States coordinated by CDC. OutbreakNet is a network of food-borne epidemiologists in the United States coordinated by CDC. This network is conducting the investigation of foodborne outbreaks and is for that reason an essential partner of PulseNet (http://www.cdc.gov/about/
280
GERNER-SMIDT ET AL.
stateofcdc/everyday/pulseNet.htm) (CDC, 2006, 2007). FoodNet is the Foodborne Diseases Active Surveillance Network coordinated by CDC (http: // www.cdc.gov/foodnet/). It collects data from 10 U.S. states regarding diseases caused by enteric pathogens transmitted commonly through food. FoodNet quantifies and monitors the incidence of these infections by conducting active, population-based surveillance for laboratory-confirmed illness. It also quantifies food consumption habits in the United States and in that way provides invaluable information for outbreak investigations. NARMS is a collaboration of CDC, USDA, and the U.S. Food and Drug Administration gathering data about antimicrobial resistance and their trends in food-borne bacterial pathogens in humans, food, and food animals in the United States (http://www. cdc.gov/narms/). This information may also be useful in investigation of outbreaks of food-borne illnesses. With the increasing international trade of food and travel, outbreaks occurring in one part of the world may have its origin in a different country or even on a different continent. Regional and national PulseNet networks inspired by PulseNet USA have been established in different parts of the world dedicated to molecular surveillance of food-borne infections. These networks collaborate under the umbrella of PulseNet International (Swaminathan et al., 2006). PulseNets are currently present in Canada, Europe,
the Asia-Pacific region, Latin America, and the Middle East (Fig. 2). They work together in investigations of international outbreaks (CDC, 2005a) in building capacity for molecular surveillance all over the world and in the development and validation of new PulseNet methods to ensure that data generated in all participating networks are comparable (Cooper et al., 2006). UTILITY OF ADDING CAMPYLOBACTER TO PULSENET PFGE was first used to subtype C. jejuni and C. coli in 1991 (Yan et al., 1991). Since then, the method quickly emerged as one of the most commonly used molecular subtyping techniques to characterize C. jejuni in the 1990s because of its broad applicability, high discriminatory power, and epidemiologic concordance (Fitzgerald et al., 2001; Gibson et al., 1995; On et al., 1998). PFGE was first used to assist with a C. jejuni outbreak in the United States in 1998 (Olsen et al., 2001); the outbreak was associated with a food handler at a school luncheon in Kansas, and PFGE using SmaI proved to be a valuable tool to link the food handler to lunch attendees and to show that epidemiologically unrelated cases in the community were not linked to the outbreak. At that time, no standardized PFGE protocol was available; multiple PFGE protocols had been described, each using different procedures for plug
Figure 2. PulseNet International networks.
CHAPTER 15
preparation, restriction enzymes, electrophoretic conditions, and data acquisition and analysis, and protocols typically took 3 to 4 days to complete. In addition, each laboratory had its own custom nomenclature system for designating subtypes that were meaningless outside that particular laboratory. Thus, many of the advantages offered by the highly discriminatory PFGE were negated by the lack of standardization and the inability to compare or exchange the PFGE data. The immediate need for a standardized, rapid, and validated Campylobacter PFGE protocol to assist with outbreak investigations of C. jejuni was apparent after the Kansas outbreak in 1998. This led to the development of a rapid PFGE protocol that used SmaI for subtyping of C. jejuni on the basis of the standardized protocols used by PulseNet laboratories for subtyping of other food-borne pathogens in 2001 (Ribot et al., 2001). With the addition of a second enzyme with increased discriminatory power (KpnI), this protocol became the standardized PFGE protocol for C. jejuni to be used by PulseNet laboratories (http://www.cdc.gov/pulsenet/protocols.htm). At present, PulseNet recommends that participants use KpnI on isolates that are indistinguishable by SmaI before inferring strain relatedness between isolates. Establishment of a PulseNet Campylobacter database later in 2001 and subsequent customization and refinements to the Campylobacter server database and development of Campylobacter-specific client software led to the PulseNet USA Campylobacter database going live in 2004. Since then, PulseNet participants may compare PFGE patterns within their own laboratories and online with patterns in the National database at CDC. PulseNet was designed to have two central functions; the first is to use routine subtyping of isolates for the detection of outbreaks that would otherwise be missed, and the second is to provide helpful information to epidemiologists during an outbreak investigation. Although routine subtyping has proved to be important for outbreaks of E. coli 0157:H7, Listeria monocytogenes, and a variety of Salmonella serotypes (as described earlier), routine subtyping of Campylobacter has been shown to be of limited value because of the high genetic diversity (Hedberg et al., 2001), the weakly clonal population structure of C. jejuni, and genetic instability that can lead to changes in PFGE profile, thereby complicating interpretation of results (Wassenaar and Newell, 2000). Unlike other pathogens covered in the network, PulseNet does not recommend routine subtyping of Campylobacter isolates; instead, we suggest only confirmatory subtyping of the strains when outbreaks are detected by other means. Because Campylobacter
NATIONAL MOLECULAR SUBTYPING NETWORK
281
species are a common cause of diarrheal illness, active case finding during an outbreak will inevitably turn up additional Campylobacter infections that may or may not be related to the outbreak. In this setting, as was the case for the Kansas outbreak, PFGE can be useful for separating the outbreak-associated cases from the unrelated background cases. To date, public health laboratories from 25 states, one international partner (PulseNet Canada), NARMS, and the U.S. Food and Drug Administration’s Center for Veterinary Medicine have contributed to the PulseNet USA Campylobacter database. At the end of 2006, the Campylobacter database had grown to 4,493 entries. Of the entries, 4,464 contained first enzyme (SmaI) patterns, and 1,501 had second enzyme (KpnI) patterns available (Table 1). The database included 453 entries that were associated with 46 different outbreaks. These outbreaks occurred between 1980 and 2006 and represented selected outbreaks for which strains were available for PFGE subtyping. The sources of these outbreaks were identified in most cases: 18 were milk borne, 12 were associated with a particular food vehicle, 6 were waterborne, 3 were linked to food handlers, 2 were associated with animal contact, and 5 were unidentified. Sixty-one different SmaI, 68 different KpnI, and 72 SmaI and KpnI combinations of PFGE patterns were observed among the outbreak strains. A single SmaI and KpnI pattern combination was observed for each of 32 outbreaks, multiple SmaI and KpnI pattern combinations were associated with each of 10 outbreaks, and strains from the remaining 4 outbreaks had only SmaI patterns available in the database. KpnI has been reported to be more discriminatory than SmaI and to discriminate among isolates with indistinguishable SmaI profiles (Michaud et al., 200 1; On et al., 1998). PulseNet Campylobacter PFGE data also support this finding. Eleven SmaI PFGE patterns were associated with more than one outbreak; KpnI analysis on isolates from 9 of these 11 SmaI pattern groups differentiated 5 of these SmaI patterns by outbreak, and isolates from the 4 remaining SmaI pattern groups were all indistinguishable by second enzyme analysis, suggesting these four strains were responsible for multiple outbreaks (Fig. 3). For example, the strain with the PulseNet designated patterns DBRS16.0084 and DBRK02.0111 was responsible for the raw milk outbreaks in Kansas in 1991, Oklahoma in 1988, and Las Vegas in 1993. In the context of public health, the strength of PFGE analysis of Campylobacter strains in PulseNet lies in the availability of a 1-day standardized rapid and robust protocol that should be used in combination with epidemiologic information to facilitate the timely investigation of outbreaks. Similar at-
282
GERNER-SMIDTET AL.
Figure 3. PFGE cluster analysis of C. jejuni outbreak strains from different outbreaks with indistinguishable SmaI and KpnI patterns.
tempts were made by the Campynet network to harmonize and standardize molecular subtyping methods, including PFGE for C. jejuni and C. coli in Europe (CampyNet I , 2002). The PulseNet network has the added advantage of providing a platform in the form of the Web-based PulseNet listserv for rapid exchange of data so participating laboratories can share information in real time, a central national database at CDC, and a uniform nomenclature for assigning unique Campylobacter PFGE patterns. To ensure comparability of data entered into the PulseNet Campylobacter database, an extensive quality assurance system must be followed by participating laboratories. Given the enormous amount of PFGE data generated for C. jejuni and published in the literature since 1991, for the first time, we are able to compare PFGE data from different laboratories to start addressing the question that has been asked for many years now: what is the point of performing PFGE subtyping of C. jejuni? Studies from the United Kingdom and Denmark have shown that concurrent Campylobacter illness in the home and/or in the community occurs more frequently than might be expected (Ethelberg et al., 2004; Gillespie et al., 2003). Although prospective subtyping has aided the identification of clusters of infection possibly representing general outbreaks that would otherwise have been missed by routine public health surveillance (Fussing et al., 2007), the identification of epidemiological links between cases shown to be part of PFGE clusters have been unsuccessful (Gilpin et al., 2006; Hedberg et al., 2001; Michaud
et al., 2005). Clusters of apparently sporadic isolates that were indistinguishable by two enzymes has also been observed in PulseNet. At the end of 2006, the PulseNet USA Campylobacter database contained PFGE profiles of 4,067 sporadic isolates in addition to outbreak PFGE data. Second enzyme data are currently available on only 1,223 of the 4,067 isolates; the limited amount of second enzyme data on isolates that are indistinguishable by SmaI currently restricts our ability to determine strain relatedness of these isolates. However, preliminary analysis of 399 isolates from 16 state public health laboratories for which two enzyme PFGE data were available clustered 75% (301 of 399) of the isolates into groups of 2 to 26 isolates. The six most common SmaI and KpnI pattern combinations accounted for 42% of the total number of isolates, and these isolates were from between two and five states, respectively. Because PulseNet does not perform outbreak detection for Campylobacter, the significance of apparently sporadic isolates indistinguishable by two enzymes is unclear; these data do not represent real-time prospective surveillance of Campylobacter infections, but rather originate from specific epidemiological projects conducted independently by the state public health laboratories for which PFGE was generated. Further work is needed to investigate the epidemiological relevance of finding stable clones from sporadic cases of infection that are separated temporally and geographically. As a result of the well-documented limitations of PFGE (de Boer et al., 2002; Wassenaar et al., 1998), the investigation of the next generation of subtyping
CHAPTER 15
methods is also well underway for subtyping of campylobacters and will be available for implementation within the network in the future. Amplified fragmentlength polymorphism analysis has been shown to be a promising tool for national and global epidemiological studies (Duim et al., 1999; On and Harrington, 2000; Schouls et al., 2003) and appears to be insensitive to the genetic instability that complicates interpretation of Campylobacter PFGE data (Wassenaar and Newell, 2000). However, PulseNet does not consider this method further because interlaboratory reproducibility of Campylobacter amplified fragmentlength polymorphism data have been shown to be poor (CampyNet I, 2002). Multilocus sequencing typing (MLST) has recently emerged as the current method of choice for subtyping of campylobacters. This method has been found to be useful for investigating the genetic diversity of C. jejuni (Dingle et al., 2001) and for estimating the role of recombination in shaping its population structure (Fearnhead et al., 2005), and it has been successfully used to identify strains from outbreaks (Schouls et al., 2003), although recognition of outbreaks caused by strains with common sequence types (STs) may be problematic (Clark et al., 2005; Sails et al., 2003). Characterization of strains by MLST from 12 of the 46 outbreaks in PulseNet showed MLST to be as discriminatory as PFGE for distinguishing temporally related isolates from the epidemic strain in all of the outbreaks. However, MLST was not as discriminatory as PFGE for strain discrimination between outbreaks. Sequencing of an additional locus, flaA, was needed to provide sufficient discrimination within the more common sequence types (Sails et al., 2003), a finding also noted during the investigation of a waterborne outbreak in Canada (Clark et al., 2005). Strains from 6 of the 12 outbreaks belonged to the ST-21 complex, which is also the most common among isolates associated with human gastroenteritis (Dingle et al., 2002; Sopwith et al., 2006). This clonal complex has also been reported in strains from a wide variety of sources, including chicken, cattle, sheep, wild birds, milk, and water (Colles et al., 2003; Dingle et al., 2001; French et al., 2005). Therefore, our current data suggest that before MLST can be used to replace PFGE to assist with outbreak investigations, further examination of additional loci may be required to improve the discriminatory power of the technique and our understanding of the large clonal complexes currently described. A new priority area for PulseNet USA is to identify the sources of sporadic food-borne infections (Gerner-Smidt et al., 2006). PulseNet is collaborating with FoodNet, NARMS, and USDA VetNet on this
NATIONAL MOLECULAR SUBTYPING NETWORK
283
project, and studies are currently underway to determine the fraction of human Salmonella infections attributable to different food sources by using a similar approach to the Danish attribution model (Hald et al., 2004). This task is more challenging for Campylobacter, given the many possible sources of infection, the enormous reservoir of extremely polymorphic Campylobacter strains in animals (Dingle et al., 2001; Duim et al., 1999; Schouls et al., 2003), the carriage of multiple strains in individual food animals (Schouls et al., 2003), the diversity of subtypes contaminating a food product (Hedberg et al., 2001), the known ability of campylobacters to undergo frequent genetic recombination resulting in the reported instability of some C. jejuni subtypes (de Boer et al., 2002; Wassenaar et al., 1998), and the lack of representative population samples. Genetic markers predictive of infection source were reported when the phylogeny of C. jejuni was modeled by genomotyping combined with Basyesian-based algorithms (Champion et al., 2005). However, no unambiguous associations were found with these livestock-associated markers when Finnish livestock C. jejuni strains were tested (Karenlampi et al., 2007), suggesting the need for further characterization of additional data sets representing geographically diverse strains from a wide source range. MLST has been used to show an association between particular genotypes of C. jejuni strains and host source (Colles et al., 2003; Dingle et al., 2001, 2002; Manning et al., 2003), as suggested by PFGE (Fitzgerald et al., 2001; Karenlampi et al., 2007) and serotyping (Wareing et al., 2002). More recent analysis of MLST data that uses the full allelic profile (instead of the more commonly used sequence type and clonal complex), in combination with STRUCTURE (Pritchard et al., 2000), a model-based clustering method, showed 80% accuracy in distinguishing isolates from chickens and bovids (cattle and sheep) (McCarthy et al., 2007). As these sophisticated genetic models advance, PulseNet would like to collaborate with national and international partners to analyze by MLST representative Campylobacter population samples, so that more accurate host assignment of strains of unknown source can be determined and better estimates of statistical uncertainty achieved. In conclusion, PFGE is a good tool to confirm relatedness between Campylobacter strains during outbreaks, and in PulseNet, this method will be used in conjunction with MLST and fZaA sequencing of selected strains in outbreak investigations and for attribution analysis. The use and the content of the PulseNet Campylobacter database is therefore different from the other PulseNet databases, which primarily aimed at detection of disease clusters by
284
GERNER-SMIDT ET AL.
PFGE. However, the introduction of multilocus variable number tandem repeat analysis and single nucleotide polymorphism analysis will open the door to reliable phylogenetic analysis and attribution analysis for the non-Cumpylobacter organisms in PulseNet too. Acknowledgments. Use of trade names is for identification only and does not imply endorsement by the Public Health Service or by the U.S. Department of Health and Human Services. The findings and conclusions in this chapter have not been formally disseminated by CDC and should not be construed to represent any agency determination or policy.
REFERENCES CampyNet I. 2002. Final report: the standardization and harmonization of molecular subtyping techniques for C. jejuni (Campynet). www.medvetnet.org. Centers for Disease Control and Prevention. 2005a. Escherichia coli 0157:H7 infections associated with ground beef from a U.S. military installation-Okinawa, Japan, February 2004. Morb. Mortal Wkly Rep. 54:40-42. Centers for Disease Control and Prevention. 2007. Multistate outbreak of Salmonella serotype Tennessee infections associated with peanut butter-United States, 2006-2007. Morb. Mortal Wkly Rep. 5 6 5 2 1-524. Centers for Disease Control and Prevention. 2006. Ongoing multistate outbreak of Escherichia coli serotype 0157:H7 infections associated with consumption of fresh spinach-United States, September 2006. Morb. Mortal Wkly Rep. 55:1045-1046. Centers for Disease Control and Prevention. 2005b. Outbreak of multidrug-resistant Salmonella typhimurium associated with rodents purchased at retail pet stores-United States, December 2003-October 2004. Morb. Mortal Wkly Rep. 54:429-433. Champion, 0. L., M. W. Gaunt, 0. Gundogdu, A. Elmi, A. A. Witney, J. Hinds, N. Dorrell, and B. W. Wren. 2005. Comparative phylogenomics of the food-borne pathogen Campylobacter jejuni reveals genetic markers predictive of infection source. Proc. Natl. Acad. Sci. USA 102:16043-16048. lark, C. G., L. Bryden, W. R. Cuff, P. L. Johnson, F. Jamieson, B. Ciebin, and G. Wang. 2005. Use of the oxford multilocus sequence typing protocol and sequencing of the flagellin short variable region to characterize isolates from a large outbreak of waterborne Campylobacter sp. strains in Walkerton, Ontario, Canada. ]. Clin. Microbiol. 43:2080-2091. :olles, F. M., K. Jones, R. M. Harding, and M. C. Maiden. 2003. Genetic diversity of Campylobacter jejuni isolates from farm animals and the farm environment. Appl. Environ. Microbiol. 69: 7409-7413. Cooper, K. L., C. K. Luey, Mv Bird, J. Terajima, G. B. Nair, K. M. Kam, E. Arakawa, A. Safa, D. T. Chenng, C. P. Law, H. Watanabe, K. Kubota, B. Swaminathan, and E. M. Ribot. 2006. Development and validation of a PulseNet standardized pulsedfield gel electrophoresis protocol for subtyping of Vibrio cholerae. Foodborne Pathog. Dis. 3 5 - 5 8 . de Boer, P., J. A. Wagenaar, R P. Achterberg, J. P. van Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:351-359. Dingle, K. E., F. M. Colles, R. Ure, J. A. Wagenaar, B. Duim, F. J. Bolton, A. J. Fox, D. R. Wareing, and M. C. Maiden. 2002. Molecular characterization of Campylobacter jejuni clones: a basis for epidemiologic investigation. Emerg. Infect. Dis. 8:949955.
Dingle, K. E., F. M. Colles, D. R. Wareing, R. Ure, A. J. Fox, F. E. Bolton, H. J. Bootsma, R. J. Willems, R. Urwin, and M. C. Maiden. 2001. Multilocus sequence typing system for Campylobacter jejuni. J. Clin. Microbiol. 39:14-23. Duim, B., T. M. Wassenaar, A. Rigter, and J. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with amplified fragment length polymorphism fingerprinting. Appl. Environ. Microbiol. 65:236923 75. Ethelberg, S., K. E. Olsen, P. Gerner-Smidt, and K. Molbak. 2004. Household outbreaks among culture-confirmed cases of bacterial gastrointestinal disease. Am. J. Epidemiol. 159:406-412. Fearnhead, P., N. G. Smith, M. Barrigas, A. Fox, and N. French. 2005. Analysis of recombination in Campylobacter jejuni from MLST population data. J. Mol. Evol. 61:333-340. Fitzgerald, C., K. Stanley, S. Andrew, and K. Jones. 2001. Use of pulsed-field gel electrophoresis and flagellin gene typing in identifying clonal groups of Campylobacter jejuni and Campylobacter coli in farm and clinical environments. Appl. Environ. Microbiol. 67:1429-1436. French, N., M. Barrigas, P. Brown, P. Ribiero, N. Williams, H. Leatherbarrow, R Birtles, E. Bolton, P. Fearnhead, and A. Fox. 2005. Spatial epidemiology and natural population structure of Campylobacter jejuni colonizing a farmland ecosystem. Environ. Microbiol. 7:1116-1126. Fussing, V., E. Moller Nielsen, J. Neimann, and J. Engberg. 2007. Systematic serotyping and riboprinting of Campylobacter spp. improves surveillance: experiences from two Danish counties. Clin. Microbiol. Infect. 13:635-642. Gautom, R. K. 1997. Rapid pulsed-field gel electrophoresis protocol for typing of Escherichia coli 0157:H7 and other gramnegative organisms in 1 day. ]. Clin. Microbiol. 35:2977-2980. Gerner-Smidt, P., K. Hise, J. Kincaid, S. Hunter, S. Rolando, E. Hyytia-Trees, E. M. Ribot, and B. Swaminathan for the Pulsenet Taskforce. 2006. PulseNet USA: a five-year update. Foodborne Pathog. Dis. 3:9-19. Gerner-Smidt, P., J. Kincaid, K. Kubota, K. Hise, S. B. Hunter, M. A. Fair, D. Norton, A. Woo-Ming, T. Kurzynski, M. J. Sotir, M. Head, K. Holt, and B. Swaminathan. 2005. Molecular surveillance of shiga toxigenic Escherichia coli 0 1 5 7 by PulseNet USA. J. Food Prot. 68:1926-1931. Gibson, J. R., C . Fitzgerald, and R. J. Owen. 1995. Comparison of PFGE, ribotyping and phage-typing in the epidemiological analysis of Campylobacter jejuni serotype HS2 infections. Epidemiol. Infect. 115215-225. Gillespie, I. A., S . J. O’Brien, G. K. Adak, C. C. Tam, J. A. Frost, F. J. Bolton, and D. S . Tompkins. 2003. Point source outbreaks of Campylobacter jejuni infection-are they more common than we think and what might cause them? Epidemiol. Infect. 130: 367-375. Gilpin, B., A. Cornelius, B. Robson, N. Boxall, A. Ferguson, C. Nicol, and T. Henderson. 2006. Application of pulsed-field gel electrophoresis to identify potential outbreaks of campylobacteriosis in New Zealand. J. Clin. Microbiol. 44:406-412. Graves, L. M, and B. Swaminathan. 2001. PulseNet standardized protocol for subtyping Listeria monocytogenes by macrorestriction and pulsed-field gel electrophoresis. Int. J. Food Microbiol. 65:55-62. Hald, T., D. Vose, H. C . Wegener, and T. Koupeev. 2004. A Bayesian approach t o quantify the contribution of animal-food sources to human salmonellosis. Risk Anal. 24:255-269. Hedberg, C. W., K. E. Smith, J. M. Besser, D. J. Boxrud, T. W. Hennessy, J. B. Bender, F. A. Anderson, and M. T. Osterholm. 2001. Limitations of pulsed-field gel electrophoresis for the routine surveillance of Campylobacter infections. J. Infect. Dis. 184: 242-244.
CHAPTER 15
Hyytia-Trees, E., S. C. Smole, P. I. Fields, B. Swaminathan, and E. M. Ribot. 2006. Second generation subtyping: a proposed PulseNet protocol for multiple-locus variable-number tandem repeat analysis of Shiga toxin-producing Escherichia coli 0 1 5 7 (STEC 0157). Foodborne Pathog. Dis. 3:118-131. Jackson, C. R., P. J. Fedorka-Cray, N. Wineland, J. D. Tankson, J. B. Barrett, A. Douris, C. P. Gresham, C. Jackson-Hall, B. M. McGlinchey, and M. V. Price. 2007. Introduction to United States Department of Agriculture VetNet: status of Salmonella and Campylobacter databases from 2004 through 2005. Foodborne Pathog. Dis. 4:241-248. Karenlampi, R., H. Rautelin, and M. L. Hanninen. 2007. Evaluation of genetic markers and molecular typing methods for prediction of sources of Campylobacter jejuni and C. coli infections. Appl. Environ. Microbiol. 73 :1683-1685. Manning, G., C. G. Dowson, M. C. Bagnall, I. H. Ahmed, M. West, and D. G. Newell. 2003. Multilocus sequence typing for comparison of veterinary and human isolates of Campylobacter jejuni. Appl. Environ. Microbiol. 69:6370-6379. Maslanka, S. E., J. G. Kerr, G. Williams, J. M. Barbaree, L. A. Carson, J. M. Miller, and B. Swaminathan. 1999. Molecular subtyping of Clostridium perfringens by pulsed-field gel electrophoresis to facilitate food-borne-disease outbreak investigations. J. Clin. Microbiol. 37:2209-2214. McCarthy, N. D., F. M. Colles, K. E. Dingle, M. C. Bagnall, G. Manning, M. C. Maiden, and D. Falush. 2007. Host-associated genetic import in Campylobacter jejuni. Emerg. Infect. Dis. 13: 267-272. Michand, S., S. Menard, and R. D. Arbeit. 2005. Role of real-time molecular typing in the surveillance of Campylobacter enteritis and comparison of pulsed-field gel electrophoresis profiles from chicken and human isolates. J. Clin. Microbiol. 43:1105-1111. Michaud, S., S. Menard, C. Gaudreau, and R. D. Arbeit. 2001. Comparison of SmaI-defined genotypes of Campylobacter jejuni examined by KpnI: a population-based study. J. Med. Microbiol. 50: 1075-108 1. Olive, D. M., and I?. Bean. 1999. Principles and applications of methods for DNA-based typing of microbial organisms. J. Clin. Microbiol. 37:1661-1669. Olsen, S. J., G. R. Hansen, L. Bartlett, C. Fitzgerald, A. Sonder, R. Manjrekar, T. Riggs, J. Kim, R Flahart, G. Pezzino, and D. L. Swerdlow. 2001. An outbreak of Campylobacter jejuni infections associated with food handler contamination: the use of pulsed-field gel electrophoresis. J. Infect. Dis. 183:164-167. On, S. L., and C. S. Harrington. 2000. Identification of taxonomic and epidemiological relationships among Campylobacter species by numerical analysis of AFLP profiles. FEMS Microbiol. Lett. 193:161-169. On, S. L., E. M. Nielsen, J. Engberg, and M. Madsen. 1998. Validity of SmaI-defined genotypes of Campylobacter jejuni examined by SalI, KpnI, and BamHI polymorphisms: evidence of identical clones infecting humans, poultry, and cattle. Epidemiol. Infect. 120:231-237. Parsons, M. B., K. L. F. Cooper, K. Kubota, N. F’uhr, S. Siminington, P. S. Calimilim, D. Schoonmaker-Bopp, C. Bopp, B. Swaminathan, P. Gerner-Smidt, and E. Ribot. 2007. PulseNet USA standardized pulsed-field gel electrophoresis protocol for
-
NATIONAL MOLECULAR SUBTYPING NETWORK
285
subtyping Vibrio parahaemolyticus. Foodborne Pathog. Dis. 4: 285-292. Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945-959. Ribot, E. M., M. A. Fair, R. Gautom, D. N. Cameron, S. B. Hunter, B. Swaminathan, and T. J. Barrett. 2006. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli 0157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 359-67. Ribot, E. M., C. Fitzgerald, K.Kubota, B. Swaminathan, and T. J. Barrett. 2001. Rapid pulsed-field gel electrophoresis protocol for subtyping of Campylobacter jejuni. J. Clin. Microbiol. 39: 1889-1894. Sails, A. D., B. Swaminathan, and P. I. Fields. 2003. Utility of multilocus sequence typing as an epidemiological tool for investigation of outbreaks of gastroenteritis caused by Campylobacter jejuni. J. Clin. Microbiol. 41:4733-4739. Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. J. Clin. Microbiol. 41:15-26. Sopwith, W., A. Birtles, M. Matthews, A. Fox, S. Gee, M. Painter, M. Regan, Q. Syed, and E. Bolton. 2006. Carnpylobacterjejuni multilocus sequence types in humans, northwest England, 20032004. Emerg. Infect. Dis. 12:1500-1507. Struelens, M. J. 1996. Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clin. Microbiol. Infect. 2:2-11. Swaminathan, B., T. J. Barrett, S. B. Hunter, and R. V. Tauxe. 2001. PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg. Infect. Dis. 7~382-389. Swaminathan, B., P. Gerner-Smidt, L. K. Ng, S. Lukinmaa, K. M. Kam, S. Rolando, E. P. GutiCrrez, and N. Binsztein, on behalf of all participants in the five PulseNet networks. 2006. Building PulseNet International: an interconnected system of laboratory networks to facilitate timely public health recognition and response to foodborne disease outbreaks and emerging foodborne diseases. Foodborne Pathog. Dis. 3:3 6-50. Wareing, D. R., F. J. Bolton, A. J. Fox, P. A. Wright, and D. L. Greenway. 2002. Phenotypic diversity of Campylobacter isolates from sporadic cases of human enteritis in the UK. J. Appl. Microbiol. 92502-509. Wassenaar, T. M., B. Geilhansen, and D. G. Newell. 1998. Evidence of genomic instability in Campylobacter jejuni isolated from poultry. Appl. Environ. Microbiol. 64:1816-1821. Wassenaar, T. M., and D. G. Newell. 2000. Genotyping of Campylobacter spp. Appl. Environ. Microbiol. 66:l-9. Wiedmann, M. 2002. Subtyping of bacterial foodborne pathogens. Nutr. Rev. 60:20 1-208. Yan, W., N. Chang, and D. E. Taylor. 1991. Pulsed-field gel electrophoresis of Campylobacter jejuni and Campylobacter coli genomic DNA and its epidemiologic application. J. Infect. Dis. 163:1068-1072.
111. PATHOGENESIS AND IMMUNITY
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 16
Interaction of Campylobacter jejuni with Host Cells ROBERT0. WATSONAND JORGEE. GALAN
C . JEJUNI INTERNALIZATION INTO NONPHAGOCYTIC INTESTINAL EPITHELIAL CELLS
Microbial pathogens that have maintained longstanding associations with their hosts have evolved complex functional interfaces to secure their own replication and survival. Often, this functional interface involves bacterial determinants that interact with host molecules to modulate cellular processes for the pathogen’s benefit. Therefore, the understanding of these complex interactions requires not only the identification of the bacterial determinants involved in shaping this functional interface, but also the knowledge of the cell biological processes that they modulate. Campylobacter jejuni is an example of a pathogen that has established a close association with a variety of vertebrate hosts and that consequently has evolved specific adaptations to modulate cellular functions. However, the details of its functional interface with host cells are poorly understood because the studies of the cell biology of C. jejuni-host cell interactions have lagged far behind those of other better-understood pathogens such as Salmonella enterica and Listeria monocytogenes. Particularly limiting has been the lack of knowledge of bacterial determinants that directly modulate these cellular processes. Despite this paucity of knowledge, it is already clear that understanding the interaction of C. jejuni with host cells will be essential for understanding its mechanisms of pathogenesis. In this chapter, we will discuss three critical aspects of the interaction of C. jejuni with host cells: (i) its ability to mediate its own uptake into nonphagocytic cells, (ii) its ability to modulate vesicular trafficking pathways to avoid delivery into lysosomes, and (iii) its ability to reprogram host cell gene expression to stimulate the production of proinflammatory cytokines. We will not attempt to provide a comprehensive review of the literature. Rather, we intend to provide our view of these events based on our own interpretation of the information available.
Examination of intestinal biopsy samples of patients infected with C. jejuni (van Spreeuwel et al., 1985), as well as experiments conducted with primates (Russell et al., 1993) and other animals (Babakhani and Joens, 1993; Newel1 and Pearson, 1984; Yao et al., 1997), together with in vitro experiments that used cultured human intestinal epithelial cells (De Melo et al., 1989; Ketley, 1997; Konkel and Joens, 1989; Oelschlaeger et al., 1993), have clearly established that C. jejuni can gain intracellular access to nonphagocytic intestinal epithelial cells. However, despite the demonstrated importance of this phenomenon, little is known about the cellular mechanisms that lead to C. jejuni internalization or the bacterial determinants that trigger the internalization event. Although several bacterial determinants have been proposed to be involved in the internalization of C. jejuni on the basis of the phenotype of specific mutants, with the exception of motility, the internalization defect of these mutants is relatively minor, suggesting an indirect role in bacterial entry. In fact, studies in other much better understood systems such as the mechanisms of Salmonella enterica serovar Typhimurium entry into cells have demonstrated that the invasion phenotype can be indirectly affected by dozens of mutations. To further complicate the interpretation and comparison of results from different laboratories, it is now apparent that different C. jejuni strains exhibit rather different ability to enter into nonphagocytic cells. It is therefore likely that the contribution of most if not all the C. jejuni determinants so far implicated in the internalization event is indirect.
~~
Robert 0. Watson and Jorge E. GalAn 06536.
Section of Microbial Pathogenesis, Yale University, School of Medicine, New Haven, CT
289
290
WATSON AND G A L h
Many studies have demonstrated the role of motility in invasion of C. jejuni (Carrillo et al., 2004; Golden and Acheson, 2002; Grant et al., 1993; Naess et al., 1988; Newel1 et al., 1985; Wassenaar et al., 1991; Yao et al., 1994, 1997). Indeed, unlike other mutants that affect bacterial entry into host cells, mutations that eliminate motility almost completely eliminate bacterial internalization. However, there is no evidence supporting a direct role for any component of the flagellar apparatus in directly triggering bacterial entry. Studies have also proposed that nonflagellar proteins secreted through the flagellar apparatus may trigger internalization (Konkel et al., 2004; Song et al., 2004). However, the evidence to support this conclusion is scant at best. In fact, mutations that affect motility (i.e., mot mutants) but not the structure of the flagellar apparatus are completely defective for entry (Yao et al., 1994). Because these mutants have a fully functional flagellar secretion machinery, they should be fully competent to secrete the putative invasive determinants and therefore should be invasive if the hypothesis was correct. The available evidence therefore indicates that motility per se is essential for entry, perhaps to facilitate the intimate contact of the bacterial with host cells, thereby allowing other yet-to-be-identified bacterial determinants to trigger the internalization process. Studies have also proposed that a C. jejuni protein, CiaB, may mediate bacterial internalization through its translocation by a type I11 secretion system (Rivera-Amill et al., 2001). This ascertainment was made on the basis of a postulated amino acid sequence similarity of CiaB and components of the proteins translocation machinery of type I11 secretion systems. However, this hypothesis is clearly incorrect because first, there are no type I11 secretion systems encoded in C. jejuni (Fouts et al., 2005; Hofreuter et al., 2006; Parkhill et al., 2000; Poly et al., 2007), and second, the amino acid sequence similarity of CiaB with type I11 secretion components is too limited to be of any significance. In addition to motility, mutations affecting protein glycosylation or capsular synthesis also exhibit a measurable defect in bacterial entry (Bachtiar et al., 2007; Bacon et al., 2001; Guerry et al., 2006; Kakuda and DiRita, 2006; Karlyshev et al., 2004; Szymanski et al., 2002; Vijayakumar et al., 2006). Some of these mutations have pleiotropic effects because, for example, glycosylation is essential for flagellar assembly and thus motility (Vijayakumar et al., 2006). Therefore, the phenotype of glycosylation mutants that affect motility is difficult to interpret given the essential role of motility in bacterial entry. The polysaccharide capsule is also important for bacterial colonization and virulence (Bacon et al.,
2001; Jones et al., 2004; Watson et al., 2008). However, the strong phenotype associated with capsular mutants in different animal models is likely to be due to reasons other than their rather minor effect in bacterial internalization. Therefore, it is likely that the capsular polysaccharide may contribute to internalization indirectly, perhaps by promoting bacterial attachment to host cells. Knowledge of the cell biology of the internalization process is also limited. However, the information available suggests that C. jejuni is internalized by mechanisms that differ significantly from those utilized by other pathogens to enter into cells. For example, the actin cytoskeleton is known to be essential for the internalization of many invasive pathogenic bacteria such as S. enterica serovar Typhimurium (Finlay et al., 1991; Galan et al., 1992), Listeria monocytogenes (Gaillard et al., 1987; Kuhn et al., 1990), or Shigellu flexneri (Clerc and Sansonetti, 1987; Dramsi and Cossart, 1998). However, disruption of the actin cytoskeleton does not prevent C. jejuni internalization (Oelschlaeger et al., 1993). Instead, it appears that C. jejuni internalization requires microtubules because drugs that inhibit microtubule dynamics block bacterial internalization (Oelschlaeger et al., 1993). There are reports implicating the actin cytoskeleton and Rho-family GTPases in the internalization of some strains of C. jejuni (Biswas et al., 2003; De Melo et al., 1989; Krause-Gruszczynska et al., 2007; Monteville et al., 2003). However, those studies were carried out with C. jejuni strains that have a rather modest invasive ability, or with cell lines that do not support robust invasion, which may easily lead to artifacts. It is, however, clear that in the case of invasive strains such as C. jejuni 81-176, the invasion process does not require the actin cytoskeleton, but rather requires intact microtubules. However, specifically how microtubules contribute to bacterial entry is unknown. A number of inhibitor studies have also implicated other cellular processes in C. jejuni internalization. For example, broad specificity tyrosine kinase inhibitors have been reported to inhibit bacterial internalization into cultured intestinal epithelial cells (Biswas et al., 2004; Wooldridge et al., 1996). In addition, increased tyrosine phosphorylation of host cellular proteins upon C. jejuni infection has also been reported (Biswas et al., 2004). These results suggest the involvement of a tyrosine kinase in C. jejuni entry, although the identity of the tyrosine kinase or kinases required for bacterial entry or the specific role that they may play in the entry process is unknown. A role for phosphoinositide 3 kinase or protein kinase C in C. jejuni entry has also been suggested on the basis of the observation that addition of inhibitors of these kinases results in a modest reduction in bac-
CHAPTER 16
terial entry (Biswas et al., 2000; Hu et al., 2005; Wooldridge et al., 1996). However, because these experiments relied only on inhibitors, the significance of these observations is unclear. Lipid rafts or caveolae are microdomains within plasma membranes enriched for cholesterol, glycolipids, sphingolipids, and signaling molecules such as receptor tyrosine kinases. Many pathogens have been shown to exploit these domains in order to gain intracellular access. consistent with lipid rafts or caveolae playing a role in C. jejuni internalization, sequestration of host cell cholesterol inhibits C. jejuni entry (Hu et al., 2006; Wooldridge et al., 1996). Further experiments revealed that the C. jejunicontaining vacuole (CCV) acquires lipid raft or caveolae markers at early time points during infection and that the function of caveolin-1 is required for efficient C. jejuni entry (Watson and Galan, 2008). Taken together, these data indicate that C. jejuni internalization into intestinal epithelial cells requires caveolae. However, as discussed below, bacterial internalization does not require dynamin, an essential component of the endocytic machinery associated with cavaealoe. It is therefore possible that caveolae may be required for the proper assembly of signaling molecules important for the signal transduction pathways that lead to C. jejuni entry rather than to mediate bacterial entry through its associated endocytic machinery. In summary, the available information, largely relying on the use of inhibitors, indicates that C. jejuni internalization may require signaling to yet unidentified kinases to trigger a microtubuledependent entry event. Clearly, more studies will be required to understand the cellular events that lead to internalization.
INTRACELLULAR SURVIVAL AND TRAFFICKING OF C. JEJUNI Intracellular pathogens utilize a variety of strategies to survive and replicate within host cells. For example, some pathogens such as Trypanosoma cruzi (Brener, 1973), Listeria monocytogenes (Goebel and Kuhn, 2000), and Shigella flexneri (Cossart and Sansonetti, 2004; Ogawa and Sasakawa, 2006) break out of phagocytic vacuoles, gaining access to the cytosol where they can replicate. Other pathogens, such as Leishmania, have evolved an array of adaptations to survive in the hostile environment of the phagolysosome (Alexander et al., 1999). Yet another group of intracellular pathogens, including S. enterica serovar Typhimurium (Knodler and Steele-Mortimer, 2003) and Mycobacterium tuberculosis (Deretic et al., 2006), alter the biogenesis and dynamics of their vac-
INTERACTION OF C. TETUNI WITH HOST CELLS
291
uolar compartment in order to prevent fusion with lysosomes. Like other intracellular pathogens, C. jejuni must have evolved specific adaptations to survive within host cells. In general, most studies have found that intracellular C. jejuni loses viability within intestinal epithelial cells over the course of 24 h, with no evidence of intracellular replication (Candon et al., 2007; Day et al., 2000; De Melo et al., 1989; Konkel et al., 1992; Naikare et al., 2006). Consistent with these observations, a recent study showed a similar decrease in intracellular viability over 24 h when C. jejuni is cultured under standard laboratory conditions (Watson and Galan, 2008). However, both LIVE/DEAD staining coupled with fluorescenceactivated cell sorter analysis and cultivation under oxygen-limiting conditions revealed that there is no significant decrease in the viability of intracellular C. jejuni over 24 h (Watson and Galan, 2008). These results indicate not only that intracellular C. jejuni remains viable for up to 24 h, but also that this bacterium shifts to a physiological state such that it can only be initially cultured under oxygen-limiting conditions. Such a shift in respiration may be a way for the bacteria to adapt to the low oxygen environment within the cell. All evidence to date indicates that after internalization into intestinal epithelial cells, C. jejunz’ resides within a membrane bound compartment (Kiehlbauch et al., 1985; Russell and Blake, 1994). Characterizations of this compartment indicate that it is functionally distinct from lysosomes (Watson and Galan, 2008). The CCV is not accessible to endocytic tracers, demonstrating that it is functionally separated from the canonical endocytic pathway. However, such adaptations must not operate in macrophages because in these cells, C. jejuni colocalizes with endocytic tracers and quickly loses viability (Banfi et al., 1986; Day et al., 2000; Kiehlbauch et al., 1985; Myszewski and Stern, 1991; Wassenaar et al., 1997). Furthermore, when C. jejuni is targeted to lysosomes via the Fc receptor, the bacterium’s ability to survive intracellularly significantly decreases, suggesting that C. jejuni has not evolved to survive within lysosomes (Watson and Galan, 2008). Upon entry, the CCV appears to interact with early endosomes because it associates with early endosomal markers such as EEA-1, Rab5, Rab4, and PX-GFP (which labels PUP) (Watson and Galan, 2008). However, this interaction is transient as this population matures and acquires the marker lamp- 1. In spite of lamp-1 acquisition, the CCV is devoid of the lysosomal protein cathepsin By further providing evidence that C. jejuni does not reside within a lysosomal compartment in intestinal epithelial cells. In
292
WATSON AND GALAN
fact, the acquisition of lamp-1, which takes place early in the CCV maturation, occurs by an unusual mechanism that does not require the GTPases Rab5 or Rab7. S. enterica serovar Typhimurium also resides within a vacuole that is apparently segregated from the canonical endocytic pathway (Garcia-del Portillo and Finlay, 1995; Knodler and SteeleMortimer, 2003) and also harbors lamp-1, although in this case acquisition of this marker appears to require Rab7 (Meresse et al., 1999). Figure 1 provides a diagram illustrating C. jejuni intracellular trafficking. Collectively, available data suggest that C. jejuni’s unusual entry mechanism may be central to its ability to avoid delivery into lysosomes. Indeed, when internalized via the Fc receptor, C. jejuni cannot avoid delivery into lysosomes, suggesting that the diversion from the canonical endocytic pathway must be dictated upon entry. Although it appears that C. jejuni interacts with the early endocytic pathway, the
CCV also acquires markers of lipid-associated rafts and caveolae at early time points during infection (Watson and Galan, 2007). Thus, this bacterium may reside in a compartment that is functionally distinct from early endosomes. Efficient C. jejuni internalization requires caveolae and caveolin-1 (Hu et al., 2006; Watson and Galan, 2007; Wooldridge et al., 1996); however, the entry process is independent of dynamin, the function of which is essential for clathrin- and caveolae-mediated endocytosis (Henley et al., 1998; Oh et al., 1998; Takei et al., 1995). Therefore, caveolin-1 may be required for proper signaling through tyrosine kinases, which are also required for C. jejuni internalization. Indeed, it has been shown that efficient signaling through receptor tyrosine kinases requires lipid rafts or caveolae (Simons and Toomre, 2000). Another unique property of the CCV is its close association with the Golgi apparatus, which is dependent on microtubules and the motor protein dynein
C.jejuni
Rab7
Rab7
0
&
Late Endosome
Lysosome
Figure 1. Model for C. jejuni internalization and trafficking within epithelial cells. C. jejuni enters intestinal epithelial cell via a microtubule and caveolae-dependent process. After internalization, the CCV transiently acquires different markers of the endocytic pathway and ultimately survives within a compartment that is functionally separated from the canonical endocytic pathway. Adapted from Watson and Galan (2008).
CHAPTER 16
(Hu and Kopecko, 1999; Watson and Galan, 2008). More studies will be required to better define the nature of this compartment and the significance of this association. Identification of the signaling pathway involved in C. jejuni entry may better elucidate the mechanism by which this bacterium is able to avoid delivery to lysosomes.
C. JEJUNI STIMULATION OF INNATE IMMUNE RESPONSES I N INTESTINAL EPITHELIAL CELLS
Intestinal epithelial cells are equipped to mount innate immune responses upon detection of microbial pathogens (Dann and Eckmann, 2007). However, the mechanisms and sensing systems involved in triggering these responses varied depending on the nature of the pathogen. In general, however, conserved bacterial products (erroneously but widely referred to as pathogen-associated molecular patterns, or PAMPs) serve as elicitors of responses sense by Toll-like receptors (TLRs) or a family of intracellular receptors belonging to the nucleotide-binding oligomerization domain (Nod) family of proteins (Fritz et al., 2006; Meylan et al., 2006; Medzhitov, 2001). One of the canonical innate immune responses of intestinal epithelial cells to the presence of bacteria is a significant reprogramming of gene expression leading to the production of proinflammatory cytokines. This reprogramming most often requires the activation of the NF-KBtranscription factor as well as the mitogenactivated protein kinase signaling pathways. C. jejuni interaction with intestinal epithelial cells results in the production of proinflammatory cytokines, which is thought to be important for its ability to induce intestinal inflammation (Bakhiet et al., 2004; Chen et al., 2006; Hickey et al., 2000; Hu and Hickey, 2005; Johanesen and Dwinell, 2006; Jones et al., 2003; Rinella et al., 2006; Stephen, 2001; Watson and Galan, 2005). The mechanisms by which C. jejuni reprograms gene expression in intestinal cells is poorly understood but is known to be the consequence of activation of NF-KB and the mitogen-activated protein kinase signaling pathways (Mellits et al., 2002; Watson and Galan, 2005). It is currently unclear whether these responses are elicited by C. jejuni-associated PAMPs. Indeed, the C. jejuni lipopolysaccharide is capable of stimulating TLR4-dependent innate immune responses, but it is uncertain whether lipopolysaccharide can stimulate signaling from the apical side of a polarized epithelial cell. On the other hand, TLR-5, which recognizes bacterial flagellin, is thought to be important in the bacterial sensing
INTERACTION OF C . 7E7UNI WITH HOST CELLS
293
mechanisms of intestinal epithelial cells. In contrast to other pathogenic bacteria, C. jejuni flagellin is unable to stimulate TLR-5-dependent responses as a result of amino acid substitutions in a region of flagellin that is known to be critical for TLR-5 stimulation (Andersen-Nissen et al., 2005; Watson and Galan, 2005), indicating that C. jejuni’s flagellin may have evolved to avoid detection by TLR-5-dependent signaling mechanisms. Data also suggest that C. jejuni may have evolved specific adaptations that allow it to stimulate nuclear responses independently of the canonical innate immune sensors (i.e., TLRs or Nod receptors). For example, it has been proposed that C. jejuni may stimulate the production of interleukin-8 through the activity of its cytolethal distending toxin (Hickey et al., 2000). Cytolethal distending toxin is an A/B toxin that induces DNA damage through the nuclease activity of its a subunit (Lara-Tejero and Galan, 2000; Whitehouse et al., 1998). Thus, it is possible that DNA damage itself may result in stress responses leading to cytokine production. Another piece of evidence suggesting the requirement of specific adaptation for C. jejuni to stimulate innate immune responses comes from the observation that infection of intestinal epithelial cells in the presence of chloramphenicol, a bacterial protein synthesis inhibitor, does not lead to the stimulation of proinflammatory cytokine production (Watson and Galan, 2005). Because the addition of protein synthesis inhibitors under these conditions does not affect the ability of C. jejuni to display potential agonists of innate immune receptors such as lipopolysaccharide, these results suggest that gene products that are synthesized upon contact with host cells are required for the stimulation of these responses. Because under these same experimental conditions addition of chloramphenicol also blocks the ability of C. jejuni to enter into cells (Oelschlaeger et al., 1993), it is possible that stimulation of proinflammatory cytokine production may require bacterial internalization, perhaps, as previously proposed, to stimulate Nod-protein family members (Zilbauer et al., 2007). However, more experiments will be required to demonstrate this possibility. Alternatively, signal transduction pathways specifically triggered by newly synthesized C. jejuni determinants at the cell surface may ultimately be responsible for the reprogramming of gene expression.
CONCLUDING REMARKS Studies have clearly shown that like many other pathogens, C. jejuni has evolved specific adaptations to modulate the activities of the host cell cytoskele-
294
WATSON AND G A L m
ton, the vesicular trafficking machinery, and the innate immune system. The last few years have seen advances in the understanding of the cell biology of infection, although more studies will be required to gain an understanding of these issues on par with that of other pathogenic bacteria. However, the main shortcoming is the fact that no bacterial determinants specifically and directly involved in modulating these responses have been identified. Although many C. jejuni mutants apparently defective in some of these process have been identified (e.g., bacterial entry), the direct involvement of these determinants in C. jejunihost cell interactions has not been demonstrated. This is clearly the main challenge in the field, and it is expected that the availability of powerful genetic tools, coupled with a better understanding of the cell biology of infection, will help to identify those bacterial determinants.
REFERENCES Alexander, J., A. R. Satoskar, and D. G. Russell. 1999. Leishmania species: models of intracellular parasitism. 1. Cell. Sci. 112(Pt. 18):2993-3002. Andersen-Nissen, E., K. D. Smith, K. L. Strobe, S. L. Barrett, B. T. Cookson, S. M. Logan, and A. Aderem. 2005. Evasion of Tolllike receptor 5 by flagellated bacteria. Proc. Natl. Acad. Sci. USA 102~9247-9252. Babakhani, F. K., and L. A. Joens. 1993. Primary swine intestinal cells as a model for studying Campylobacter jejuni invasiveness. Infect. lmmun. 61:2723-2726. Bachtiar, B. M., P. J. Coloe, and B. N. Fry. 2007. Knockout mutagenesis of the kpsE gene of Campylobacter jejuni 81116 and its involvement in bacterium-host interactions. FEMS lmmunol. Med. Microbiol. 49:149-154. Bacon, D. J., C. M. Szymanski, D. H. Burr, R. P. Silver, R. A. A h , and P. Guerry. 2001. A phase-variable capsule is involved in virulence of Campylobacter jejuni 8 1-176. Mol. Microbiol. 40~769-777. Bakhiet, M., F. S. Al-Salloom, A. Qareiballa, K. Bindayna, I. Farid, and G. A. Botta. 2004. Induction of alpha and beta chemokines by intestinal epithelial cells stimulated with Campylobacter jejuni. J. Infect. 48:236-244. Banfi, E., M. Cinco, and G. Zabucchi. 1986. Phagocytosis of Campylobacter jejuni and C. coli by peritoneal macrophages.]. Gen. Microbiol. 132:2409-2412. Biswas, D., K. Itoh, and C. Sasakawa. 2000. Uptake pathways of clinical and healthy animal isolates of Campylobacter jejuni into "-407 cells. FEMS Immunol. Med. Microbiol. 29:203-211. Biswas, D., K. Itoh, and C. Sasakawa. 2003. Role of microfilaments and microtubules in the invasion of I N - 4 0 7 cells by Campylobacter jejuni. Microbiol. Immunol. 47A69-473. Biswas, D., H. Niwa, and K. Itoh. 2004. Infection with Campylobacter jejuni induces tyrosine-phosphorylated proteins into INT-407 cells. Microbiol. lmmunol. 48:221-228. Brener, Z. 1973. Biology of Trypanosoma cruzi. Annu. Rev. Microbiol. 27:347-382. Candon, H. L., B. J. Allan, C. D. Fraley, and E. C. Gaynor. 2007. Polyphosphate kinase 1 (PPK1) is a pathogenesis determinant in Campylobacter jejuni. 1. Bacteriol. 189:8099-8108.
Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K. Amoako, S. Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Genome-wide expression analyses of Campylobacter jejuni N C T C l l l 6 8 reveals coordinate regulation of motility and virulence by flhA. 1.Biol. Chem. 279:20327-203 3 8. Chen, M. L., Z. Ge, J. G. Fox, and D. B. Schauer. 2006. Disruption of tight junctions and induction of proinflammatory cytokine responses in colonic epithelial cells by Campylobacter jejuni. Infect. lmmun. 746581-6589. Clerc, P., and P. J. Sansonetti. 1987. Entry of Shigella flexneri into HeLa cells: evidence for directed phagocytosis involving actin polymerization and myosin accumulation. Infect. Immun. 55: 2681-2688. Cossart, P., and P. J. Sansonetti. 2004. Bacterial invasion: the paradigms of enteroinvasive pathogens. Science 304:242-248. Dann, S. M., and L. Eckmann. 2007. Innate immune defenses in the intestinal tract. Curr. Opin. Gastroenterol. 23:115-120. Day, W. A., Jr., J. L. Sajecki, T. M. Pitts, and L. A. Joens. 2000. Role of catalase in Campylobacter jejuni intracellular survival. Infect. lmmun. 68:6337-6345. De Melo, M. A., G. Gabbiani, and J. C. Pechere. 1989. Cellular events and intracellular survival of Campylobacter jejuni during infection of HEp-2 cells. Infect. lmmun. 57:2214-2222. Deretic, V., S. Singh, S. Master, J. Harris, E. Roberts, G. Kyei, A. Davis, S . de Haro, J. Naylor, H. H. Lee, and I. Vergne. 2006. Mycobacterium tuberculosis inhibition of phagolysosome biogenesis and autophagy as a host defence mechanism. Cell. Microbiol. 8:719-727. Dramsi, S., and P. Cossart. 1998. Intracellular pathogens and the actin cytoskeleton. Annu. Rev. Cell. Dev. Biol. 14:137-166. Finlay, B. B., S. Ruschkowski, and S . Dedhar. 1991. Cytoskeletal rearrangements accompanying salmonella entry into epithelial cells. J. Cell. Sci. 99(Pt. 2):283-296. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn,M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Fritz, J. H., R. L. Ferrero, D. J. Philpott, and S. E. Girardin. 2006. Nod-like proteins in immunity, inflammation and disease. Nut. lmmunol. 7:1250-1257. Gaillard, J. L., P. Berche, J. Mounier, S . Richard, and P. Sansonetti. 1987. In vitro model of penetration and intracellular growth of Listeria monocytogenes in the human enterocyte-like cell line Caco-2. Infect. lmmun. 55:2822-2829. Galan, J. E., C. Ginocchio, and P. Costeas. 1992. Molecular and functional characterization of the Salmonella invasion gene invA: homology of InvA to members of a new protein family. 1.Bacteriol. 174:4338-4349. Garcia-del Portillo, F., and B. B. Finlay. 1995. Targeting of Salmonella typhimurium to vesicles containing lysosomal membrane glycoproteins bypasses compartments with mannose 6-phosphate receptors.]. Cell Biol. 129:81-97. Goebel, W., and M. Kuhn. 2000. Bacterial replication in the host cell cytosol. Cur. Opin. Microbiol. 3:49-53. Golden, N. J., and D. W. Acheson. 2002. Identification of motility and autoagglutination Campylobacter jejuni mutants by random transposon mutagenesis. Infect. Immun. 70:1761-1771. Grant, C. C., M. E. Konkel, W. Cieplak, Jr., and L. S. Tompkins. 1993. Role of flagella in adherence, internalization, and trans-
CHAPTER 16
location of Campylobacter jejuni in nonpolarized and polarized epithelial cell cultures. Infect. lmmun. 61: 1764-1771. Guerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Pattarini, G. Majam, P. Thibault, and S. Logan. 2006. Changes in flagellin glycosylation affect Campylobacter autoagglutination and virulence. Mol. Microbiol. 60:299-3 11. Henley, J. R., E. W. Krueger, B. J. Oswald, and M. A. McNiven. 1998. Dynamin-mediated internalization of caveolae. J. Cell Biol. 14153-99. Hickey, T. E., A. L. McVeigh, D. A. Scott, R. E. Michielutti, A. Bixby, S. A. Carroll, A. L. Bourgeois, and P. Guerry. 2000. Campylobacter jejuni cytolethal distending toxin mediates release of interleukin-8 from intestinal epithelial cells. Infect. Immun. 68:6535-6541. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. lmmun. 74:4694-4707. Hu, L., and D. J. Kopecko. 1999. Campylobacter jejuni 81-176 associates with microtubules and dynein during invasion of human intestinal cells. Infect. Immun. 67:4171-4182. Hu, L., and T. E. Hickey. 2005. Carnpylobacter jejuni induces secretion of proinflammatory chemokines from human intestinal epithelial cells. Infect. lmmun. 73:4437-4440. Hu, L., J. P. McDaniel, and D. J. Kopecko. 2006. Signal transduction events involved in human epithelial cell invasion by Campylobacter jejuni 81-176. Microb. Pathog. 40:91-100. Hu, L., R. B. Raybourne, and D. J. Kopecko. 2005. Caz+ release from host intracellular stores and related signal transduction during Campylobacter jejuni 8 1-176 internalization into human intestinal cells. Microbiology 151:3097-3 105. Johanesen, P. A., and M. B. Dwinell. 2006. Flagellin-independent regulation of chemokine host defense in Campylobacter jejuniinfected intestinal epithelium. Infect. Immun. 74:3437-3447. Jones, M. A., K. L. Marston, C. A. Woodall, D. J. Maskell, D. Linton, A. V. Karlyshev, N. Dorrell, B. W. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni NCTClll68 to high-level colonization of the avian gastrointestinal tract. Infect. lmmun. 72:3769-3776. Jones, M. A., S. Totemeyer, D. J. Maskell, C. E. Bryant, and P. A. Barrow. 2003. Induction of proinflammatory responses in the human monocytic cell line THP-1 by Campylobacter jejuni. lnfect. lmmun. 71:2626-2633. Kakuda, T., and V. J. DiRita. 2006. Cj1496c encodes a Campylobacter jejuni glycoprotein that influences invasion of human epithelial cells and colonization of the chick gastrointestinal tract. Infect. lmmun. 74:4715-4723. Karlyshev, A. V., P. Everest, D. Linton, S. Cawthraw, D. G. Newell, and B. W. Wren. 2004. The Campylobacter jejuni general glycosylation system is important for attachment to human epithelial cells and in the colonization of chicks. Microbiology 150: 1957-1964. Ketley, J. M. 1997. Pathogenesis of enteric infection by Campylobacter. Microbiology 143(Pt. 1):s-21. Kiehlbauch, J. A., R. A. Albach, L. L. Baum, and K. P. Chang. 1985. Phagocytosis of Campylobacter jejuni and its intracellular survival in mononuclear phagocytes. Infect. lmmun. 48:446451. Knodler, L. A., and 0. Steele-Mortimer. 2003. Taking possession: biogenesis of the Salmonella-containing vacuole. Traffic 4587599. Konkel, M. E., S. F. Hayes, L. A. Joens, and W. Cieplak, Jr. 1992. Characteristics of the internalization and intracellular survival of Campylobacter jejuni in human epithelial cell cultures. Microb. Pathog. 13:357-370.
INTERACTION OF C. JEJUnrI WITH HOST CELLS
295
Konkel, M. E., and L. A. Joens. 1989. Adhesion to and invasion of HEp-2 cells by Campylobacter spp. Infect. lmmun. 57:29842990. Konkel, M. E., J. D. Klena, V. Rivera-Amill, M. R. Monteville, D. Biswas, B. Raphael, and J. Mickelson. 2004. Secretion of virulence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. 1. Bacteriol. 186:32963303. Krause-Gruszczynska, M., M. Rohde, R. Hartig, H. Genth, G. Schmidt, T. Keo, W. Konig, W. G. Miller, M. E. Konkel, and S. Backert. 2007. Role of the small Rho GTPases Racl and Cdc42 in host cell invasion of Campylobacter jejuni. Cell. Microbiol. 9:243 1-2444. Kuhn, M., M. C. Prevost, J. Mounier, and P. J. Sansonetti. 1990. A nonvirulent mutant of Listeria monocytogenes does not move intracellularly but still induces polymerization of actin. Infect. Immun. 58:3477-3486. Lara-Tejero, M., and J. E. Galan. 2000. A bacterial toxin that controls cell cycle progression as a deoxyribonuclease I-like protein. Science 290:354-357. Medzhitov, R. 2001. Toll-like receptors and innate immunity. Nut. Rev. Immunol. 1:135-145. Mellits, K. H., J. Mullen, M. Wand, G. Armbruster, A. Patel, P. L. Connerton, M. Skelly, and I. F. Connerton. 2002. Activation of the transcription factor NF-kappaB by Campylobacterjejuni. Microbiology 148:2753-2763. Meresse, S., 0. Steele-Mortimer, B. B. Finlay, and J. P. Gorvel. 1999. The rab7 GTPase controls the maturation of Salmonella typhimurium-containing vacuoles in HeLa cells. EMBO 1. 18: 43 94-4403. Meylan, E., J. Tschopp, and M. Karin. 2006. Intracellular pattern recognition receptors in the host response. Nature 442:39-44. Monteville, M. R., J. E. Yoon, and M. E. Konkel. 2003. Maximal adherence and invasion of INT 407 cells by Campylobacter jejuni requires the CadF outer-membrane protein and microfilament reorganization. Microbiology 149: 153-1 65. Myszewski, M. A., and N. J. Stern. 1991. Phagocytosis and intracellular killing of Campylobacter jejuni by elicited chicken peritoneal macrophages. Avian Dis. 35:750-755. Naess, V., C. Johannessen, and T. Hofstad. 1988. Adherence of Campylobacter jejuni and Campylobacter coli to porcine intestinal brush border membranes. APMlS 96:681-687. Naikare, H., K. Palyada, R. Panciera, D. Marlow, and A. Stintzi. 2006. Major role for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infect. Immun. 745433-5444. Newell, D. G., H. McBride, and J. M. Dolby. 1985. Investigations on the role of flagella in the colonization of infant mice with Campylobacter jejuni and attachment of Campylobacter jejuni to human epithelial cell lines. J. Hyg. (Lond.) 95:217-227. Newell, D. G., and A. Pearson. 1984. The invasion of epithelial cell lines and the intestinal epithelium of infant mice by Campylobacter jejunilcoli. ]. Diarrhoea1 Dis. Res. 2: 19-26. Oelschlaeger, T. A., P. Guerry, and D. J. Kopecko. 1993. Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proc. Natl. Acad. Sci. USA 90:6884-6888. Ogawa, M., and C. Sasakawa. 2006. Intracellular survival of Shigella. Cell. Microbiol. 8:177-184. Oh, P., D. P. McIntosh, and J. E. Schnitzer. 1998. Dynamin at the neck of caveolae mediates their budding to form transport vesicles by GTP-driven fission from the plasma membrane of endothelium.]. Cell. Biol. 141:lOl-114. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W.
296
WATSON AND G A L m
Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Poly, F., T. Read, D. R. Tribble, S. Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. Immun. 75:3425-3433. Rinella, E. S., C. D. Eversley, I. M. Carroll, J. M. Andrns, D. W. Threadgill, and D. S. Threadgill. 2006. Human epithelialspecific response to pathogenic Campylobacter jejuni. FEMS Microbiol. Lett. 262:23 6-243. Rivera-Amill, V., B. J. Kim, J. Seshu, and M. E. Konkel. 2001. Secretion of the virulence-associated Campylobacter invasion antigens from Campylobacter jejuni requires a stimulatory signal. J. Infect. Dis. 183:1607-1616. Russell, R., and D. J. Blake. 1994. Cell association and invasion of Caco-2 cells by Campylobacter jejuni. Infect. Immun. 62: 3773-3779. Russell, R. G., M. O’Donnoghue, D. C. Blake, Jr., J. Zulty, and L. J. DeTolla. 1993. Early colonic damage and invasion of Campylobacter jejuni in experimentally challenged infant Macaca mulatta. J. Infect. Dis. 168:210-215. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nut. Rev. Mol. Cell. Biol. 1:31-39. Song, Y. C., S. Jin, H. Louie, D. Ng, R. Lau, Y. Zhang, R. Weerasekera, S. Al Rashid, L. A. Ward, S. D. Der, and V. L. Chan. 2004. FlaC, a protein of Campylobacter jejuni TGH9011 (ATCC43431) secreted through the flagellar apparatus, binds epithelial cells and influences cell invasion. Mol. Microbiol. 53: 541-553. Stephen, J. 2001. Pathogenesis of infectious diarrhea. Can. J. Gastroenterol. 15 :669-683. Szymanski, C . M., D. H. Burr, and P. Guerry. 2002. Campylobucter protein glycosylation affects host cell interactions. Infect. Immun. 70:2242-2244. Takei, K., P. S. McPherson, S. L. Schmid, and P. De Camilli. 1995. Tubular membrane invaginations coated by dynamin rings are induced by GTP-gamma S in nerve terminals. Nature 374:186190. van Spreeuwel, J. P., G . C. Duursma, C. J. Meijer, R. Bax, P. C. Rosekrans, and J. Lindeman. 1985. Campylobacter colitis: histological immunohistochemical and ultrastructural findings. Gut 26~945-951.
Vijayakumar, S., A. Merkx-Jacques, D. B. Ratnayake, I. Gryski, R. K. Obhi, S. Houle, C. M. Dozois, and C. Creuzenet. 2006. CjllZlc, a novel UDP-4-keto-6-deoxy-GlcNAc C-4 aminotransferase essential for protein glycosylation and virulence in Campylobacter jejuni. J. Biol. Chem. 281:27733-27743. Wassenaar, T. M., N. M. Bleumink-Pluym, and B. A. van der Zeijst. 1991. Inactivation of Campylobacter jejuni flagellin genes by homologous recombination demonstrates that flaA but not flaB is required for invasion. EMBO J. 10:2055-2061. Wassenaar, T. M., M. Engelskirchen, S. Park, and A. Lastovica. 1997. Differential uptake and killing potential of Campylobacter jejuni by human peripheral monocytesImacrophages. Med. Microbiol. Immunol. 186:139-144. Watson, R. O., and J. E. Galan. 2005. Signal transduction in Campylobacter jejuni-induced cytokine production. Cell. Microbiol. 7:655-665. Watson, R. O., and J. E. Galan. 2008. C. jejuni survives within epithelial cells by avoiding delivery to lysosomes. PloS Pathog. 4~69-83. Watson, R. O., V. Novik, D. Hofreuter, M. Lara-Tejero, and J. E. Galan. 2007. A MyD88-deficient mouse model reveals a role for Nrampl in Campylobacter jejuni infection. Infect. Immun. 75:1994-2003. Whitehouse, C. A., P. B. Balbo, E. C. Pesci, D. L. Cottle, P. M. Mirabito, and C . L. Pickett. 1998. Campylobacter jejuni cytolethal distending toxin causes a G2-phase cell cycle block. Infect. Immun. 66:1934-1940. Wooldridge, K. G., P. H. Williams, and J. M. Ketley. 1996. Host signal transduction and endocytosis of Campylobacter jejuni. Microb. Pathog. 21:299-305. Yao, R, D. H. Burr, P. Doig, T. J. Trust, H. Niu, and P. Guerry. 1994. Isolation of motile and nonmotile insertional mutants of Campylobacter jejuni: the role of motility in adherence and invasion of eukaryotic cells. Mol. Microbiol. 14:883-893. Yao, R, D. H. Burr, and P. Guerry. 1997. CheY-mediated modulation of Campylobacter jejuni virulence. Mol. Microbiol. 23: 1021-103 1. Zilbauer, M., N. Dorrell, A. Elmi, K. J. Lindley, S. Schuller, H. E. Jones, N. J. Klein, G. Nunez, B. W. Wren, and M. Bajaj-Elliott. 2007. A major role for intestinal epithelial nucleotide oligomerization domain 1 (NOD1) in eliciting host bactericidal immune responses to Campylobacter jejuni. Cell. Microbiol. 9:2541.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 17
Cell Biology of Human Host Cell Entry by Campylobacter jejuni LAN Hu AND DENNISJ. KOPECKO
Campylobacter jejuni, a zoonotic pathogen, is part of the normal gut flora in many animals, including birds, rodents, and dogs. In mice, C. jejuni colonizes the mucous layer and crypts of the intestinal mucosa, mainly in the colon and cecum. In other animals, different parts of the gastrointestinal (GI) tract may be preferentially colonized, depending on the ecological microenvironment (Lee et al., 1983). Humans typically acquire campylobacters from contaminated food or water. After passage through the stomach, these organisms colonize the ileum and colon, where they can interfere with the normal secretory or absorptive capacity of the intestine. As a consequence, this bacterium causes a GI disease characterized by watery or bloody diarrhea, abdominal pain, fever, headache, and nausea. C. jejuni also causes postdiarrheal irritable bowel syndrome, Guillain-Barrt syndrome, and Reiter’s reactive arthritis. Bacterial diseases of the GI tract typically result from a complex set of interactions between the offending bacteria and the host. In the process of evolution, humans have developed a variety of innate mechanisms that, among other purposes, help protect them from pathogenic organisms (e.g., gastric acidity, gut peristalsis, mucous layer, epithelial barrier, innate immunity). At the same time, bacteria have evolved pathogenic attributes to circumvent these innate host defenses. Many bacterial pathogens are known to develop specific interactions (i.e., attachment, invasion, or both) with host mucosal surfaces that result in disease initiation (e.g., Salmonella spp., Shigella spp., enterotoxigenic Escherichia coli, enteropathogenic E. coli [EPEC], and C. jejuni). These pathogens typically exploit host cell machinery, which contributes to their pathogenic ability. For example, on establishing a unique association with human intestinal epithelial cells, Salmonella, Shigella, and EPEC secrete effector proteins via type I11 secretion pathways into the eu-
karyotic host cell, which initiates host signal transduction events that lead to cytoskeletal rearrangements and eventual intimate colonization (i.e., for EPEC) or internalization (i.e., for Salmonella and Shigella) of the adjacent pathogen. The results of analysis of intestinal biopsy samples of patients, infected primates, and other experimentally infected model animals, together with experimental infection of cultured human intestinal epithelial cells, have clearly demonstrated that C. jejuni can adhere to and invade the intestine. Further, these collective data emphasize the importance of bacterial invasiveness as a virulence factor for Campylobacter pathogenesis. After passage through the upper GI tract, Campylobacter adheres to colonic epithelial cells and triggers signal transduction events that induce host cytoskeletal rearrangements, bacterial internalization, and translocation across the mucosa. During mucosal penetration, other specific bacteria-host cell interactions induce interleukin and chemokine production, the recruitment of inflammatory cells, and host cell death and cause diarrhea, colitis, or both. Genomic sequencing data from several strains reveals that C. jejuni does not contain homologues of classical bacterial enterotoxins, adhesins, invasins, type I11 protein secretion systems, or pathogenicity islands (Eppinger et al., 2004; Fouts et al., 2005; Hofreuter et al., 2006; Parkhill et al., 2000). The virulence factors of the related Helicobacter pylori are absent in C. jejuni, with the only revealed homologues being housekeeping genes. It seems that Campylobacter may have specific pathogenic mechanisms that differ from many other enteric bacterial pathogens. In a limited number of Campylobacter strains, type IV secretion homologues have been identified on a large plasmid (Bacon et al., 2000, 2002), but a functional type IV system has not been established for C. jejuni.
Lan Hu and Dennis J. Kopeck0 Laboratory of Enteric and Sexually Transmitted Diseases, Center for Biologics Evaluation and Research, U.S. Food and Drug Administration, Bethesda, MD 20892.
297
298
HU AND KOPECK0
Many studies have revealed that both C. jejuni adherence and invasion are multifactorial processes (e.g., Hu and Kopecko, 2000; Ketley, 1997; Konkel et al., 1999; Wassenaar and Blaser, 1999). To gain a full understanding of Campylobacter invasion mechanisms, readers are directed to earlier reviews of Campylobacter bacteria-host cell interactions (e.g., Hu and Kopecko, 2000; Kopecko et al., 2001; Wooldridge and Ketley, 1997). Rather than summarizing all previous studies, our aim here is to give an update of the current cell biological understanding of C. jejuni invasion of the gut epithelium and to describe how these events may impact disease pathogenesis. This chapter presents an overview of the interaction of C. jejuni with intestinal host cells and focuses on bacterial adherence and invasion into the intestinal epithelium, transcytosis across the epithelial mucosa, and ensuing damage to host cells.
C. JEJUNI ADHERE TO INTESTINAL EPITHELIAL CELLS After ingestion of contaminated food or water, enteric bacterial pathogens pass through stomach acid and the highly alkaline secretion from the bile duct in the upper small intestine, yet unexplored stresses that may be important in pathogen gene modulation. Early visible interactions between a pathogenic enteric organism and its host usually entail colonization of the mucous barrier and then specific attachment to the mucosal cell surface. Bacterial pathogens have numerous ways of attaching themselves to host cells, and most pathogens encode multiple different adherence functions. Bacterial adherence is typically due to a specific interaction between molecules on the bacterial surface (i.e., adhesins) and molecules on the host surface (i.e., receptors). As is the case for other intestinal pathogens, the ability of C. jejuni to colonize the GI tract by binding to epithelial cells has been proposed to be essential for disease production. Fauchere et al. (1986) initially reported that C. jejuni isolated from individuals with fever and diarrhea exhibited much greater binding to epithelial cells than strains isolated from asymptomatic individuals, but they did not examine specific adherence mechanisms. Most studies on C. jejuni adherence have focused on binding to immortalized epithelial cells and not to the mucous layer coating the gut epithelium. Numerous studies have been performed to identify and characterize potential Campylobacter adhesins that mediate the organism’s attachment to host cells. Reported adhesins include flagella, outer membrane proteins (Omps), and surface polysaccharide moieties (Fauchere et al., 1986; McSweegan and
Walker, 1986; McSweegan et al., 1987). Functional flagella are involved in epithelial cell adherence and are essential for the internalization of C. jejuni into cultured epithelial cells and for colonization of the mouse intestine (Alm et al., 1993; Grant et al., 1993; Konkel et al., 1992a, 1992c; Wassenaar et al., 1991; Yanagawa et al., 1994; Yao et al., 1994; L. Hu et al., unpublished data). In addition to motility, chemotaxis functions appear to be important for Campylobacter adherence (Hendrixson and DiRita, 2004). A 28-kDa Campylobacter protein, termed PEB1, has been identified as a conserved antigen in C. jejuni and C. coli strains and proposed to be an adhesin (Pei and Blaser, 1993; Pei et al., 1991, 1998). PEBl shares homology with a periplasmic binding protein involved in nutrient acquisition (Garvis et al., 1996; Pei and Blaser, 1993). Many pathogenic microorganisms are capable of binding to components of the extracellular matrix, such as fibronectin, laminin, vitronectin, and collagen. Scanning electron microscopy of infected INT407 cells first indicated that C. jejuni binds to fibronectin. This suspected binding was confirmed to be mediated by a 37-kDa OMP (CadF), which is conserved among C. jejuni isolates (Konkel et al., 1999) and also appears to be essential for uptake preferentially at the basolateral host cell surface (Monteville and Konkel, 2002). Anti-CadF antibody reduced the binding of two C. jejuni clinical isolates to immobilized Fn by >50% (Monteville et al., 2003). A surface-exposed lipoprotein, JlpA, has also been reported as a C. jejuni adhesin (Jin et al., 2001). JlpA, a 42.3-kDa protein, confers C. jejuni adherence by binding to Hsp9O-alpha on the surface of Hep-2 epithelial cells (Jin et al., 2003). Lipooligosaccharides and polysaccharide capsules are the major surface antigens of campylobacters and play important roles in the interaction of these bacteria with human or animal hosts and/or the environment. These surface polysaccharide molecules in C. jejuni also serve as host mucosal adherence factors. The high-molecularweight glycan capsule and lipooligosaccharide produced by some C. jejuni strains is required for cell invasion in vitro and full virulence in the ferret animal model (Bacon et al., 2001; Kanipes et al., 2004; Karlyshev et al., 2000). The N-linked general protein glycosylation pathway (Pgl) modifies many bacterial surface proteins. C. jejuni pgl mutants are reduced in their ability to adhere to and invade Caco-2 cells. Also, the pglH mutant of strain 81116 is severely affected in its ability to colonize chicks (Karlyshev et al., 2004). Thus, Campylobacter glycosylated proteins appear to be important in intestinal colonization. Colonization with C. jejuni in poultry is normally asymptomatic. It is commonly assumed that hu-
CHAPTER 17
man handling of contaminated poultry carcasses and the consumption of undercooked poultry meat are major sources of campylobacters. Reducing or eradicating colonization of Campylobacter at the poultry flock level may lead to control or prevention of human campylobacteriosis (Rosenquist et al., 2003). Understanding how C. jejuni colonize poultry is therefore important. There have been numerous studies of C. jejuni short-term colonization in young chickens (Ahmed et al., 2002; Cawthraw et al., 1996; Ringoir and Korolik, 2003; Stern et al., 1988), but few factors associated with long-term persistence have been defined. Several C. jejuni mutations that eliminate motility (maf5, PaA) or capsule expression (kpsM) eliminated colonization in chickens, whereas pglH reduced but did not eliminate intestinal colonization (Jones et al., 2004). In contrast, Byrd et al. (2007) reported no significant role of capsular polysaccharides in colonization of the chicken gut. No association has been found between differing FlaA types and invasion or colonization ability in the chick gut (Hanel et al., 2004). Several two-component gene regulation systems of Campylobacter are involved in colonization (Bras and Ketley, 1999; MacKichan et al., 2004). For example, the DccR-DccS system has been reported to be important for in vivo colonization but dispensable for in vitro growth (MacKichan et al., 2004). It appears that these two-component signal transduction systems allow bacteria to respond to environmental cues by controlling the expression of many genes (e.g., pathogenicity factors). The yglutamyl transpeptidase of C. jejuni is not required for in vitro growth or initial colonization of l-dayold chicks but is somehow required for persistent colonization of the avian gut (Barnes et al., 2007). In brief, C. jejuni exhibit numerous host cell adhesins, but the relative importance of each is not well defined. The mechanism of binding and the nature of the host receptors involved in such attachment remain largely uncharacterized, as are the specific involvements of two-component gene regulatory systems and other surface factors. Nevertheless, C. jejuni attachment to host cells appears to be a prerequisite for subsequent invasion.
C. JEJUM INVADE HUMAN INTESTINAL EPITHELIAL CELLS
Campylobacter invasion into the epithelial mucosa appears to be an essential process leading to colitis (Allos, 1997; Russell and Blake, 1994; Van Spreeuwel et al., 1985). As with other intestinal bacterial pathogens, Campylobacter interacts with the intestinal mucosal surface, triggering special host signal
HUMAN HOST CELL ENTRY
299
transduction events that induce host cytoskeletal rearrangements, resulting in bacterial uptake. Although many researchers would agree with this general summary of Campylobacter invasion events, there still remains considerable confusion regarding how Campylobacter enter and cross the intestinal mucosa. The following sections are intended to present an update of current data and an attempt at rational interpretation of these results. Host Cytoskeletal Requirements Microfilaments (MFs) and microtubules (MTs), which are composed of actin and tubulin, respectively, are involved in both cellular and subcellular movements, and together with intermediate filaments, they determine host cell shape. Bacterial internalization into host epithelial cells has typically been observed to involve pathogen-induced rearrangement of host cytoskeletal structures, resulting in endocytosis of the pathogen. Most invasive enteric organisms (e.g., Salmonella, Sbigella, Listeria, Yersinia; Elsinghorst et al., 1989; Finlay and Falkow, 1988, 1989, 1990; Finlay et al., 1991; Isberg et al., 1987; Kihlstrom and Nilsson, 1977; Miller et al., 1988) have been found to trigger largely MFdependent entry pathways. More recent studies of these pathogens, however, have revealed the involvement of MTs in various aspects of their intracellular bacterial replication and survival (Abrahams and Hensal, 2006; Henry et al., 2006; Yoshida and Sasakawa, 2003). Campylobacter internalization has been variously reported to require mainly MFs (de Melo et al., 1989; Konkel and Joens, 1989; Konkel et al., 1992b), mainly MTs (Hu and Kopecko, 1999; Oelschlaeger et al., 1993), both MFs and MTs (Biswas et al., 2003; Monteville et al., 2003; Oelschlaeger et al., 1993), or neither (Russell and Blake, 1994), depending on the methods used (i.e., different and sometimes lacking appropriate experimental controls) and the C. jejuni strain or strains studied. Further contributing to the confusion, only a few C. jejuni strains have been studied in both molecular and microscopic detail for invasion mechanism or mechanisms. The available data, although inconclusive, suggest that Campylobacter may encode separate MF-dependent (Konkel and Joens, 1989; Konkel et al., 1992b) or MT-dependent (Hu and Kopecko, 1999; Oelschlaeger et al., 1993) pathways for host invasion. Many strains express a dependence on both MTs and MFs during invasion ( e g , Biswas et al., 2003; Monteville et al., 2003; Oelschlaeger et al., 1993 [e.g., C. jejuni VC841) that may reflect the involvement of one or more different mechanisms.
300
HU AND KOPECK0
What are possible explanations for the differences observed in host cell cytoskeletal requirements for C. jejuni invasion? These investigations typically involve culturing human epithelial (e.g., HEp-2, INT407, Caco-2, or T-84) cells as a semiconfluent or confluent monolayer, and treating host cells with or without agents that depolymerize MTs or MFs before infecting them with C. jejuni at the apical surface for 2 to 3 h. After this period, the infected monolayer is washed to remove unattached bacteria and incubated for 2 to 3 h in fresh media containing gentamicin to kill all extracellular bacteria before host cell lysis and enumeration of internalized bacteria. Appropriate experimental controls ensure that inhibitors actually function as expected and do not affect bacterial or host cell survival. Although the level of monolayer confluence or ineffective gentamicin treatment could affect outcomes, experimental controls in the various studies noted below suggest that these factors had no meaningful effect on reported results. However, there is an important difference in host cell culture conditions utilized in different studies of C. jejuni invasion. Under assay conditions in which host cells are grown and maintained throughout the assay in 10% (i.e., for INT407 cells) or 20% (i.e., for Caco-2 cells) fetal bovine serum, C. jejuni 81-176 and M129 exhibit a predominant requirement for polymerized MTs (Hu and Kopecko, 1999,2000; Kopecko et al., 2001; Oelschlaeger et al., 1993; L. Hu and D. J. Kopecko, unpublished data for strain M129). Other investigations have utilized similar host cells cultured in 10% fetal bovine serum, which are then washed and maintained in 1% serum during the 3-h invasion period and subsequent 3-h gentamicin kill period. Under the latter conditions, strains 81176 and M129 as well as many other C. jejuni strains display a strong dependence on both polymerized MFs and MTs for invasion (e.g., Biswas et al., 2000, 2003; Konkel and Joens, 1989; Konkel et al., 199210; Monteville et al., 2003). We hypothesize that maintenance of host cells in 1% serum is nutrient limiting to many cultured cell lines, and we propose that it may affect the host cell membrane, decrease intercellular adhesion, and alter host cell receptor availability and distribution. In fact, Monteville and Konkel (2002) concluded, after treating T-84 cells with EGTA, that CadF-dependent invasion of epithelial cells occurs preferentially at the basolateral surface, which normally interacts with fibronectin. In this study, the number of internalized C. jejuni F38011 increased approximately threefold after EGTA treatment and was then reduced -80% by treating with anti-fibronectin antibody. We suggest that host cell maintenance in 1% serum relaxes intercellular adhesion, exposing some basolateral surface and allowing
C. jejuni to utilize a CadF-dependent entry mechanism. CadF is conserved among C. jejuni and mediates binding to fibronectin, which leads to internalization that clearly involves MFs (Konkel and Joens, 1989; Konkel et al., 1992b, 1999; Monteville and Konkel, 2002). In an attempt to explain the differences noted in the above referenced studies, we propose that C. jejuni can display two apparently different invasion mechanisms: the first is predominantly MT dependent and observed at the apical host cell surface in nutrient-enriched host cells and the second, MFdependent invasion, occurs preferentially at the basolateral surface (which presumably is partially exposed in cells maintained in 1% serum). Although apparently differing in host cytoskeletal requirements, these two uptake mechanisms may share some properties. It is unclear whether CadF-dependent entry predominantly requires MFs; the reported involvement of MTs in these studies may simply reflect simultaneous invasion by the alternative MTdependent pathway. Thus, studies conducted on host cells maintained in 1% serum may reflect uptake by multiple Campylobacter invasion pathways. Alternatively, relaxed host intercellular adhesion and availability of fibronectin receptors may somehow generate (through host membrane receptor aggregation events) an invasion pathway with mixed host cytoskeletal involvement. Oelschlaeger et al. (1993) demonstrated that the entry of C. jejuni 81-176 into cultured human intestinal INT407 cells requires intact MTs and was actually enhanced (150 to 250%) by the specific depolymerization of actin filaments. Evidence for the MT dependence of this process was strengthened by the consistently strong inhibitory behavior of a series of mechanistically different MT depolymerizing agents. Thus, each MT depolymerizing agent (colchicine, demecolcine, nocodazole, vincristine, or vinblastine) caused >90% reduction in 81-176 invasion over a 2-h period. In these studies, concurrent Salmonella enterica serovar Typhi Ty2 uptake was not reduced by MT depolymerization but, by contrast, was totally blocked by cytochalasin D, which depolymerizes actin filaments (Hu and Kopecko, 1999; Oelschlaeger et al., 1993). Further studies have verified these cytoskeletal requirements and have extended to Caco-2 cells this strong MT dependence of the C. jejuni 81-176 invasion mechanism (Hu and Kopecko, 1999). Thus, strain 8 1-176 enters epithelial cells that are cultured and maintained in 10 to 20% fetal bovine serum by a mechanism that requires predominantly polymerized host cell MTs and is not inhibited by compounds that depolymerize MFs. Invasion by other Campylobacter strains, such as C.
CHAPTER 17
jejuni M129, also show strong MT dependence under these nutrient-enriched cell culture maintenance conditions (Oelschlaeger et al., 1993; L. Hu and D. J. Kopecko, unpublished data). Because CadFdependent invasion apparently requires polymerized MFs (and maybe MTs) but MFs do not appear to be involved in Campylobacter invasion under nutrientenriched host cell maintenance conditions, we suggest that mainly the MT-dependent pathway is functional under these conditions. For this reason, the remainder of this section will focus on characterization of this MT-dependent mechanism. Further evidence of MT involvement has been obtained with immunofluorescence (IF) confocal microscopy. IF confocal microscopic study of differentially labeled, time-course-infected INT407 cells has revealed a close association of C. jejuni 81-176 with MTs during the entire invasion process (Hu and Kopecko, 1999). Early in infection, C. jejuni are observed interacting with the host cell at the tips of fingerlike protrusions of the host cell membrane, apparently being extended by one or a few bundled MTs (Hu and Kopecko, 1999; Kopecko et al., 2001). These early membrane protrusions, initially detected by IF microscopy, have now been observed by electron microscopy (EM) (L. Hu et al., unpublished data). The appearance of these structures suggests that initiation of invasion involves localized reorganization of the host cytoskeleton (i.e., depolymerization of cortical actin filaments and the formation of a MT-based membrane projection) in apparent response to a signal or signals from bound C. jejuni; similar membrane protrusions are not triggered by either the noninvasive RY213 mutant of 81-176 or by S. enterica serovar Typhi in control studies (Hu and Kopecko, 1999). During epithelial cell invasion, C. jejuni 8 1-176 cells have been observed by confocal microscopy to colocalize with MTs and to move, over time, from the periphery of the host cell to the perinuclear region (i.e., the MT organizing center; Hu and Kopecko, 1999). The perinuclear migration of internalized C. jejuni was first reported after IF microscopy of several different strains (Konkel et al., 1992b). EM studies have shown for several different strains that C. jejuni remains within a membrane-bound endosome during passage through the host cell (Konkel et al., 1992c; Oelschlaeger et al., 1993; L. Hu et al., unpublished data). Movement of the endosome containing C. jejuni may occur along MTs via the MTassociated motor proteins dynein and kinesin, which are involved in a multiplicity of biological trafficking. IF confocal microscopic studies have confirmed that C. jejuni 81-176 are colocalized with both MTs and dynein during the invasion process (Hu and Ko-
HUMAN HOST CELL ENTRY
301
pecko, 1999). Further, orthovanadate, an inhibitor of dynein activity, was observed to reduce C. jejuni invasion into epithelial cells significantly, indicating a role for this MT-minus-end-directed molecular motor in the C. jejuni invasion process. MTs and MT-associated motor proteins have more recently been observed to play major or minor roles in the invasion and intracellular survival mechanisms of other bacterial pathogens (Yoshida and Sasakawa, 2003). For example, Chlamydia trachomatis has been shown to move via MTs within host cells toward the MT organizing center (Clausen et al., 1997; Grieshaber et al., 2003). Disruption of MTs was initially reported to effect intracellular Salmonella replication (Garcia-del Portillo et al., 1993). Subsequently, MTs have been shown to serve as a scaffold along which Salmonella-induced filaments are formed (Brumell et al., 2002). MTs accumulate around Salmonella-containing vacuoles. Salmonellacontaining vacuoles migrate from the cell periphery to a perinuclear location, a process involving the recruitment of active Rab7, its effector Rab7-interacting lysosomal protein, and an intact MT network (Guignot et al., 2004; Harrison et al., 2004; Kuhle et al., 2004). Intracellular Salmonella redirect exocytic transport processes in a SPI2-dependent manner. The effector proteins SifA, SseF, and SseG are involved in the recruitment of motor proteins to the Salmonella-containing vacuole (Boucrot et al., 2005; Kuhle et al., 2006). Effector protein EspG (EPEC and Citrobacter rodentium) is found to interact with host cell tubulin. Binding by EspG to tubulin caused localized MT depolymerization, resulting in actin stress fiber formation (Hardwidge et al., 2005). Shigella has been reported to destroy surrounding MTs by secreting VirA via the type 3 secretion system. Degradation of MTs by VirA was dependent on a tubulin-specific, cysteine protease-like activity (Yoshida et al., 2006). It is important to explore further the interaction of Campylobacter virulence factors with host cytoskeletal elements. Because the host cell cytoskeleton is highly interdependent on MTs, MFs, and intermediate filaments, it seems likely that all play a role in Campylobacter internalization, subsequent survival, and passage through the host cell. Additional study is needed to determine the specific bacterial and host molecules interacting with MTs (or MFs in the CadF-dependent pathway) during Campylobacter internalization, intracellular survival, and transcytosis across the epithelial cell barrier. Signal Transduction A common theme among pathogenic invasive microorganisms is their ability to usurp the eukar-
302
HU AND KOPECK0
yotic cell signaling systems both to allow for invasion and to trigger disease pathogenesis. The divalent calcium cation (Ca2+)plays a pivotal role in host signal transduction and other cellular processes, and its cytosolic concentration is of necessity tightly regulated. Increased free intracellular Ca2+ has been demonstrated to link cell surface receptor stimulation, via signaling pathways, with intracellular effectors, and to modulate cytoskeletal structure, chemotaxis, membrane fluidity, chromosome segregation, cell cycle transition, enzyme activity, transmembrane ion fluxes, proteolysis, and other key cell functions (Clapham, 1995; Jacob, 1990; Marks and Maxfield, 1990a, 1990b; Meldolesi et al., 1991; Clapham, 1995). The host cell has two potential sources of Ca2+: entry from the external medium and release from internal stores. Eukaryotic cells maintain an internal free calcium ion concentration ( [Ca2+Ii)that is typically far below that of the extracellular environment and modulate [Ca2+Ii in response to specific signaling events. Thus, eukaryotic cytosolic Ca2+levels are mainly controlled by the action of specific molecular pumps and channels in the plasma membrane and in subcellular Ca2+ storage organelles. Certain enteric bacterial pathogens have been found to cause a specific increase in host [Ca2+Ii(Bierne et al., 2000; Norris et al., 1996; Ruschkowski et al., 1992; Tran Van Nhieu et al., 2004) that may play a role in ensuing pathogenesis (Fig. 1). When extracellular Ca2+ chelators EGTA or 1,2-bis- (2)-ethane-N, N,N',N' - tetra-acetic acid (BAPTA) were added to INT407 cells in Ca2+-free, serum-free culture medium, no reduction in internalization of added C. jejuni 81-176 or control S. enterica serovar Typhi Ty2W was observed (Fig. 1A; Hu et al., ZOOS), suggesting that extracellular Ca2+is
Figure 1. Effects of calcium chelators, stimulator, or inhibitors on invasion of INT407 cells by C. jejuni 81-176 or S. enterica serovar Typhi Ty2w. INT407 cells were grown in complete culture medium containing 10% fetal bovine serum to -80% confluence in wells of a 24-well plate coated with poly-L-lysine. The host cells were washed and incubated in Caz+-free, serum-free minimal essential medium (S-MEM) before pretreatment for 30 min with the various compounds before a 2-h invasion period with C. jejuni at a multiplicity of infection of 20. After invasion, the monolayer was washed and incubated with 100 pg/rnl gentamicin in fresh S-MEM for 2 h before enumeration of intracellular bacteria. (A) Effect of extracellular cationic chelators on host cell invasion. (B) Effect of intracellular CaZ+chelator, BAPTA/AM, or BayK8644, a plasma membrane Caz+ channel agonist, on bacterial invasion. (C) Effect of dantrolene or U73122, which act in different ways to inhibit release of CaZ+from intracellular stores. Bacterial invasion over a 2-h period in the presence of inhibitors or stimulator was compared with invasion without stimulator or inhibitor compounds and is presented as percentage of invasion relative to the no inhibitor/stimulator control (from Hu et al., 2005).
i No inhibitor
EGTA
2oo
5
'5 Q L L .-
BAPTA
I
150
.-%
c
rd -
!! 100a,
0
c C
a,
!!
a"
50-
0
I
No inhibitor
BAPTA, AM
BAY K 8644
T T
li Dantrolene
No inhibitor
U73122
0C.jeiuni S. Typhi
CHAPTER 17
not typically needed for C. jejuni internalization. In contrast to EGTA or BAPTA, which are host cell impermeable, the lipophilic BAPTA/acetoxymethyl ester (AM) can cross into host cells where it is hydrolyzed by a cellular esterase to BAPTA. INT407 monolayer pretreatment with the intracellular chelator BAPTAIAM was observed to markedly reduce invasion of host cells by C. jejuni or a control S. enterica serovar Typhi (Fig. 1B). Further, pretreatment of epithelial cell monolayers with the inhibitors dantrolene or U73122, each of which blocks the mobilization of stored Ca2+ via different receptors in the endoplasmic reticulum, also reduced C. jejuni invasion (Fig. 1C). At the concentrations utilized, these compounds showed no adverse effect on bacterial or host cell viability over a 2.5-h period. Thus, an increase in [Ca2+],, typically induced by Ca2+ release from host intracellular stores, appears to be necessary for efficient C. jejuni internalization. Interestingly, this requirement for increased [Ca2+],to trigger efficient Carnpylobacter invasion could be augmented in the presence of the Ca2+ agonist Bay K8644, which activates the plasma membrane slow Ca2+pumps that allow for influx of Ca2+ from the extracellular environment (Fig. 1B). Measurement of intracellular free Ca2+ by fluorescence dye probes of infected INT407 cells further showed that -10% of total host cells exhibited a marked increase in cytosolic free Ca2+ at any time point measured during 60 min of invasion, suggesting (i) that only a fraction of the host cell population (i.e., -10% of cells) are undergoing active invasion at any time point (i.e., would exhibit the highest levels of cytosolic free Ca2+),and (ii) that infected host cells lower the [Ca2+],within minutes of bacterial entry. However, because continued invasion of INT407 cells by C. jejuni occurs during a 2-h invasion period under the conditions used, [Ca2+],averaged over all cells is raised above resting levels throughout this 2-h period (Hu et al., 2005). Further characterization of these dynamic events awaits the measurement of [Ca2+],fluxes over time in individually infected versus uninfected host cells. Calmodulin (CaM), a low-molecular-mass, highaffinity Ca2+binding protein activated by Ca2+,is responsible for mediating signals in a multitude of systems (e.g., activation of phosphodiesterase or myosin light chain kinase). W7 is a widely used selective CaM antagonist at lower concentrations, but it can also inhibit protein kinase C (PKC) at a 10-fold higher concentration (about 40 p M )than required to inhibit CaM (Tanaka et al., 1982). W7 at a 5 pmol concentration markedly reduced the invasion ability of C. jejuni 81-176, suggesting that a CaMdependent process or processes appears to play a prominent role in host cell invasion by C. jejuni. In addition, a marked reduction in C. jejuni 81-176 in-
HUMAN HOST CELL ENTRY
303
vasion observed in the presence of the PKC-specific inhibitor calphostin C, indicating a role for PKC in the C. jejuni internalization process (Hu et al., 2005). Calphostin C has been shown to reduce invasion of many different C. jejuni strains (Biswas et al., 2000). Figure 2 schematically summarizes the processes involved in regulating intracellular free Ca2+ levels in the host cell and the steps affected by the abovementioned inhibitory or stimulatory compounds (Fig. 2). Most host cell-surface receptor proteins belong to one of three large families: ion-channel-linked receptors, G protein-coupled receptors, and enzymelinked receptors. These host cell-surface receptor proteins, located within discrete membrane domains, convert an extracellular ligand-binding event into intracellular signals that alter the behavior (e.g., the cytoskeleton) of the target cell. Outer domains of the host cell membrane serve as scaffolding for surface receptor proteins and also interact cytoplasmically with signaling molecules such as G proteins and monomeric GTPases. Lipid rafts are dynamic assemblies of cholesterol and sphingolipids in the exoplasmic leaflet of the plasma membrane. Proteins specifically associated with lipid rafts include glycosylphosphatidylinositol-anchored proteins, double-acylated proteins such as Src-family kinases, the a subunits of heterotrimeric G proteins, and palmitoylated transmembrane proteins (Ilangumaran and Hoessli, 1998; Simons and Toomre, 2000). Lipid rafts represent membrane patches of -50 nm in diameter and carry 10 to 30 proteins per patch. Caveolae, a lipidraft subset, are cell surface, flask-shaped membrane invaginations formed by polymerization of caveolins. The antibiotic filipin sequesters cholesterol from lipid rafts, causing their disruption. Filipin I11 pretreatment of cultured epithelial cells reduced subsequent C. jejuni invasion by 99.9% (Hu et al., 2006b; Wooldridge et al., 1996) but had no inhibitory effect on invasion by control S. enterica serovar Typhi Ty2, which utilizes a mechanistically different, MFdependent uptake pathway (Huang et al., 1998). Filipin I11 treatment of host cells also reverses the deposition of caveohn in the plasma membrane, which was observed by Hu and coworkers (2006b). Other inhibitors that interfere with maturation of membrane domains (e.g., g-strophanthin, monodansylcadaverine) have also been found to reduce host cell entry of several C. jejuni strains, including 8 1-176 (Konkel et al., 1992b; Biswas et al., 2000; Oelschlaeger et al., 1993). At present, these data suggest that C. jejuni interact with host membrane receptors located within discrete membrane microdomains (i.e., most likely, lipid rafts). G protein-coupled receptors comprise the largest family of cell-surface receptors and regulate the
304
HU AND KOPECK0
PIPg
DAG
Bay U8644 agonia Ca2+
channel
I'
Figure 2. Schematic diagram depicting the processes involved in regulating intracellular free Ca2+ levels in the host cell and the steps affected by inhibitory/stimulatory compounds. Typically, interaction of a specific bacterial ligand (e.g., C. jejuni protein) with a tyrosine kinase receptor (TKR) located in the host plasma membrane can activate an associated messenger G protein, which in turn can activate phospholipase C (PLC). PLC cleaves membrane-bound phosphatidylinositol 4,5bisphosphate (PIP 2) to diffusible inositol triphosphate (IP 3) and diacylglycerol (DAG). IP 3 interacts with cognate receptors (i.e., IP 3R) in the endoplasmic reticulum (ER) and causes the release of Ca2+ from intracellular stores. The released Ca2+ in turn activates ryanodine receptors (RyR) on the ER, resulting in augmented Ca2+ release. Membrane-bound DAG causes activation of PKC, which ultimately modulates diverse cellular responses. External free Ca2+can be internalized through Ca2+ channels in the plasma membrane, a process that is stimulated by the slow channel agonist Bay K8644. U73122 blocks the activity of PLC, and dantrolene blocks the activation of RyRs, both of which result in reduced release of Ca2+from intracellular stores. BAPTAiAM enters host cells, is cleaved to BAPTA, and then can effectively chelate all intracellular free Ca2+.W7 inhibits the activation of calmodulin, and calphostin C blocks the activation of PKC.
activity of a separate plasma-membrane-bound target. The interaction between the receptor and this target protein is mediated by a trimeric G protein. When used to pretreat INT407 cells, the G-protein inhibitors cholera toxin and pertussis toxin were observed to increase host CAMP levels and to cause a marked reduction in C. jejuni 81-176 invasion ability, indicating the involvement of G proteins in Cumpylobucter uptake (Hu et al., 2006b; Wooldridge et al., 1996). Many membrane receptors are protein kinases or are linked to protein kinases. Protein kinase receptor-ligand interaction leads to the phosphorylation (i.e., activation) of specific sets of proteins in the target cell. Both the general kinase inhibitor staurosporine and the specific tyrosine kinase inhibitor genistein markedly reduced C. jejuni 8 1-176 invasion (Hu et al., 2006b; Kopeck0 et al., 2001), as was also
reported for other C. jejuni strains (Biswas et al., 2004; Wooldridge et al., 1996). These findings suggest that specific tyrosine protein kinase activity is essential for C. jejuni internalization. The use of two different, specific phosphoinositol (PI)-3 kinase inhibitors, supported by simultaneous PI-3 kinase immunopreciptation analyses, demonstrated that PI-3 kinase activation is also essential for C. jejuni 81-176 invasion ability (Hu et al., 2006b). Phosphorylation of PI-3 kinase increased over the first 60 min of invasion; this increase was blocked in the presence of PI-3 kinase inhibitor wortmannin, which also markedly reduced C. jejuni invasion. Examination of total protein tyrosine phosphorylation patterns during C. jejuni 81-176 invasion revealed that at least nine host proteins varying from 15 to 65 kDa, in addition to PI-3 kinase, increased in phosphorylation over a 2-h invasion period (Hu et al., 2006b); these proteins re-
CHAPTER 17
main uncharacterized. Biswas et al. (2004) also reported finding tyrosine-phosphorylated proteins during C. jejuni invasion. Mitogen-activated protein kinase (MAPK) pathways ultimately regulate nuclear transcription events, the products of which may sequentially control cytoskeletal rearrangements that result in bacterial uptake. C. jejuni 8 1-176 invasion was markedly reduced by selective inhibition of extracellular-signal-regulated kinase (ERK) and p3 8 MAPK in INT407 cells (Hu and Kopecko, 2006b), suggesting the direct involvement of these kinases in invasion. In an earlier study, MacCallum and coworkers (2005) showed activation of ERK 1/2, Jun N-terminal protein kinase, and p3 8 MAPK during C. jejuni infection of Caco-2 cells and human colonic tissue explants; these studies did not specifically assess inhibition of MAPK activation on invasion. The small Rho GTPases of host cells act as molecular switches that control a wide variety of signal transduction pathways, including those affecting cellular actin dynamics. In a recent study of C. jejuni 81-176 invasion, activated Racl and Cdc42 (but not RhoA) were reported to be involved. Activation of these GTPases was only partially due to CadF (Krause-Gruszczynska et al., 2007). Although the data from these studies is compelling, the long 6-h invasion period (also conducted in host cell culture media lacking serum) does not allow for clear differentiation of effect on invasion versus intracellular survival. Cortical actin depolymerization would be expected to occur to facilitate a largely MT-dependent uptake mechanism and would eventually require restoration of the terminal polymerized actin web. Teasing apart these factors over shorter versus longer invasion periods will be necessary to gain a full understanding of these processes. How do C. jejuni trigger the above-noted host signal transduction events? Reversible adherence factors may create initial bacteria-epithelial cell membrane contacts that allow for subsequent invasion-specific bacterial ligand-host receptor interactions. These latter interactions, by themselves, could stimulate specific host signaling events that lead to cytoskeletal rearrangements and bacterial uptake. Konkel and coworkers (1999, 2004) have presented data to suggest that a limited subset of C. jejuni proteins are secreted through the flagella, an event that correlates with maximal invasion levels. Further studies to reveal how these secreted proteins may effect invasion are necessary. It is intriguing to note that recent EM studies of C. jejuni 81-176 invasion of Caco-2 cells has revealed contacts in which the bacteria interact perpendicularly with the host cell apical surface, possibly by inserting one polar flagellum into the host cell membrane (L. Hu et al., unpublished
-
HUMAN HOST CELL ENTRY
305
data). These observations of Konkel et al. (2004) and L. Hu and coworkers (unpublished data) await further verification and molecular description. The current data on signal transduction events involved in C. jejuni invasion suggest that host cell invasion receptors” reside in filipin 111-sensitive membrane microdomains (i.e., lipid rafts). Bacterial ligand-host receptor interaction activates one or more G protein-coupled receptors, leading to sustained increased levels of CAMP (Hu et al., 2006b). We speculate that a host tyrosine kinase-linked membrane receptor may also be involved, and ligand-receptor contact leads to activation of PI-3 kinase and the MAPKs. Increased free intracellular Ca2+ levels are essential for invasion and result in activation of PKC and calmodulin, which ultimately effect nuclear transcription events that cause cytoskeletal rearrangements and bacterial internalization. These events are summarized schematically in Fig. 3. Although many host signaling events have been found to affect C. jejuni invasion, there are still significant gaps in our understanding of the signal transduction pathways involved in bacterial internalization. For example, other than the reported involvement of paxillin, a focal adhesin signaling molecule, in the CadF-dependent invasion process (Monteville et al., 2003), there is little understanding of involved signal transduction pathways. Not much is yet understood about the bacterial factors involved in inducing entry. Konkel and coworkers (2004) have reported the requirement for C. jejuni CiaB in invasion, but its precise role remains unclear. Crucial to a better understanding will be deciphering those events essential to invasion versus intracellular survival, and isolating each discrete invasion process (assuming that there is more than one) with appropriate experimental conditions to be able to ascertain essential cell biological interactions.
Campylobacter Translocation across the Mucosa Bacterial translocation across the intestinal mucosa encompasses a series of events through which some enteric pathogens pass across the normally impermeable epithelial barrier to reach the lamina propria, from which inflammation may result or further dissemination may occur. The pathways used by bacteria to traverse the intestinal barrier are transcellular (bacteria pass through the absorptive enterocytes or M cells) or paracellular (bacteria pass between adjacent epithelial cells). Although animal models are useful for studying intestinal translocation, polarized epithelial cell lines provide a simple and controlled experimental alternative to animal models. Polarized human colonic carcinoma (Caco-2) cells, with differ-
306
HU AND KOPECK0
1’
I
PKA
AKT
I
-
I
ERKlR
I
I
cytoskeletal rearrangements, nuclear transcription, gene regulation Figure 3. Schematic representation of host receptors in membrane caveolae/lipid rafts and signal transduction events initiated during C. jejuni invasion and showing the steps blocked by inhibitors. Abbreviations: AKT, anti-protein kinase; CT, cholera toxin; ERK-1/2, extracellular signal-regulated kinase 1 and 2; GDP, guanosine 5’-diphosphate; MAPK, mitogen-activated protein kinase; ERK and p38, MAPK; MEK-1, MAPK/ERK kinase; MKKS, MAPK 3; PDK, 3-phosphoinositide-dependent protein kinase; PI-3 kinase, phosphoinositide 3-kinase; PKA, protein kinase A; PTX, pertussis toxin; RTK, receptor tyrosine kinase. Ras, Rac, and Rafl are family members of small, monometric GTP-binding proteins; filipin I11 disrupts lipid rafts; wortmannin and LY294002 are specific inhibitors of PI-3 kinase; staurosporine and genistein are general protein phosphorylation inhibitors. C. jejuni apparently interact with receptors located in lipid rafts/caveolae, which results in a cascade of host signaling pathways as shown.
entiated apical and basolateral surfaces separated by tight junctions (TJs), express several markers characteristic of normal small intestinal cells and have a well-defined brush border (Everest et al., 1992; Finlay and Falkow, 1990). TJs are the most apical structures of intercellular junctions and are specialized regions of the plasma membrane that form a tight adhesive barrier with adjacent epithelial cells. TJs completely encircle a polarized cell and are composed of host transmembrane proteins (occludins, claudins, and junction adhesion molecules) and scaffolding proteins (ZO-1, -2, -3). Bacterial translocation was first demonstrated with S. enterica serovar Typhimurium by apical addition to polarized cells. This process involves first bacterial endocytosis, then passage through the host cell, and
finally release from the basolateral domain. This transcellular transcytosis process occurred in the presence of tight cell junctions during the first few hours but caused a significant decrease in monolayer electrical resistance by 6 h after infection (Finlay and Falkow, 1990). Campylobacters have also been observed to translocate across tight polarized epithelial cell monolayers (Bras and Ketley, 1999; Everest et al., 1992; Grant et al., 1993; Harvey et al., 1999; Konkel et al., 1 9 9 2 ~ )In . aggregate, these time-course studies show that C. jejuni strains utilize a transcellular route of translocation and appear at the basolateral surface as early as 15 min after infection, a process that peaks in efficiency after 3 to 4 h. C. jejuni infection also did not disrupt TJ integrity at early times after infection. Several of these reports hinted
CHAPTER 17
that paracellular C. jejuni translocation might occur but provided no convincing experimental support. Once internalized, C. jejuni do not appear to encode functions that allow escape from the endosome, and these bacteria undergo limited intracellular replication during the first 8 to 10 h after infection. EM studies have revealed that C. jejuni move intracellularly within endosomes to the basolateral membrane before exocytosis (Konkel et al., 1992c; L. Hu et al., unpublished data). Recent EM studies have revealed the exocytosis event in which C. jejuni endosomes fuse with the basolateral membrane, a process that releases the bacterium basolaterally and completes the transcytosis pathway (L. Hu et al., unpublished data). After translocation across the intestinal mucosa, continued infection of the epithelium may involve basolateral invasion of epithelial cells. A recent report described a putative paracellular pathway of C. jejuni mucosal translocation, termed subvasion (van Alphen et al., 2008). During infection of nonpolarized Chang epithelial cells (grown and maintained without any serum), C. jejuni were observed to pass between cells and proceed basally before internalization. Repeated passage uncovered a chemotaxis mutant that was enhanced in this subvasion mechanism. Translocation by this process was independent of JlpA, PEB1, or CadF adhesins. The findings that subvasion does not occur in polarized epithelial cells and is only observed in cultured epithelial cells maintained under nutrient-limiting conditions (i.e., without serum) raise questions about how relevant this process is for C. jejuni pathogenesis. Nevertheless, further studies are needed to explore this unusual process. Infection of polarized Caco-2 cells with C. jejuni at high multiplicities (Le., 10,000) was found to cause loss of transepithelial electrical resistance by 24 h after infection (MacCallum et al., 2005). This loss of transepithelial electrical resistance accompanied a rearrangement of TJ proteins such as occludin. Similar finding of redistribution of TJ proteins and loss of transepithelial electrical resistance was observed during infection of polarized T84 cells with C. jejuni (Chen et al., 2006). These studies suggest that C. jejuni translocation may cause breaks in the epithelial barrier over 1 to 2 days of infection. Key questions about C. jejuni translocation that need addressing include the following. (i) Do significant numbers of C. jejuni pass through M cells to reach the lamina propria during human disease? (ii) Does translocation across the epithelium proceed mainly by a transcellular invasion of and passage through absorptive enterocytes? (iii) Does paracellular entry (or subvasion) contribute to C. jejuni pathogenesis? (iv) Are pathogenic mechanisms and ulti-
HUMAN HOST CELL ENTRY
307
mate symptoms dependent on initial infectious dose? (v) Do C. jejuni colonize the gut luminal surface, subsequently forming surface microcolonies, which then cause mucosal barrier disruption from without?
HOST DAMAGE CAUSED BY CAMPYLOBACTER While adhering to and/or invading host cells (e.g., epithelial, macrophage, or dendritic cells [DCs]), Campylobacter produce secreted proteins, toxins, and other molecules that can adversely affect the host. These effectors work by different mechanisms and can stimulate or inhibit host signal transduction pathways, promote cytokine production, trigger inflammation, change host cell mitogenesis (causing hyperplasia), cause host cell death, or cause water malabsorption/secretion. These host cell alterations are effectively responsible for the disease symptoms (e.g., abdominal cramps, diarrhea, dysentery). Toxins and Diarrhea/Dysentery
C. jejuni has previously been reported to produce a thermolabile-protein cytotoxin (Misawa et al., 1994), an enterotoxin with antigenic similarity to heat-labile enterotoxin (Klipstein and Engert, 1984a, 1984b; Ruiz-Palacios et al., 1983), and a cytolethal distending toxin (CDT) (Whitehouse et al., 1998). Recent genome analyses show no evidence of homologues of heat-labile enterotoxin in C. jejuni, and the role of cytotoxin or enterotoxin in initiating inflammation or diarrhea remains unclear (Florin and Antillon, 1992; Ketley, 1997; Wassenaar and Blaser, 1999). How then does Campylobacter trigger diarrhea? Kanwar et al. (1995) provided compelling data to show that increased intracellular Ca2+ levels and activated PKC are involved in C. jejuni-induced fluid secretion from the rat ileum. Hu et al. (2005) reported a sustained increase in intracellular Ca2+ and activated PKC in INT407 cells infected with either invasive or a noninvasive mutant of 81-176, but did not assess their involvement in fluid secretion. MacCallum and coworkers (2005) reported that C. jejuni infection inhibited the absorptive function of Caco-2 cells, which may contribute to the cause of diarrhea. These limited results suggest that C. jejuni diarrhea may be due to net fluid secretion in the small intestine and/or malabsorption in the large intestine. CDT, first discovered in E. coli and now known to be produced by strains of Shigella flexneri and S. enterica serovar Typhi, is encoded by most C. jejuni strains (Johnson and Lior, 1988a, 1988b; Lara-Tejero
308
HU AND KOPECK0
and Galan, 2000). CDT causes cell cycle arrest at the G2/M phase, progressive cellular distension, chromatin fragmentation, and apoptotic cell death (Frisan et al., 2002; Lara-Tejero and Galan, 2001; Whitehouse et al., 1998). The active subunit of this toxin, CdtB, is a type I deoxyribonuclease and is responsible for cell cycle arrest and chromosome degradation (Lara-Tejero and Galan, 2000). Pooled human sera from patients previously infected with C. jejuni was found to neutralize CDT activity, indicating that this protein is immunogenic during human infection (AbuOun et al., 2005). CDT production by C. jejuni has been reported to induce secretion of the proinflammatory cytokine interleukin-8 from intestinal epithelial cells (Hickey et al., 2000). Further, C. jejuni cdtB mutants exhibited impaired ability to cause disease in immunodeficient mice (Fox et al., 2004; Purdy et al., 2000). In contrast, CDT does not seem to play a role in the colonization of chickens. C. jejuni cdt mutants were found to be normal in adherence to HeLa and HD-11 cells but were decreased 10-fold in invasion ability (Biswas et al., 2006). Thus far, evidence of CDT involvement in human pathogenesis is indirect. However, CDT may facilitate epithelial (or other) cell death before normal epithelial cell sloughing after 4 to 6 days at the luminal surface and promote further Campylobacter translocation across the mucosa. Dysenteric symptoms may result from C. jejuni-induced epithelial cell death or translocation across the epithelium, which has been associated with a relaxation of TJs leading to compromised mucosal barrier function (Chen et al., 2006) Proinflammatory Factors Inflammation is an innate host response to a physical or infectious trauma to tissue. An infectious insult could entail epithelial cell invasion, C. jejuni translocation to the lamina propria, or bacterial uptake by inflammatory cells. The ensuing infected cells then may, for example, trigger production of defensins, stimulate intestinal motility, affect water transport, or produce proinflammatory chemokines or cytokines, among many possible cellular responses. Although these responses are essential to controlling the infection, immune hyperresponsiveness to invasive diseases not only contributes to disease symptoms, but also exacerbates tissue damage. C. jejuni has been reported to induce secretion of interleukin (1L)-8 in cultured epithelial cells (Hickey et al., 1999, 2000; MacCallum et al., 2006; Mellits et al., 2002). The activation of the ERK and p38 MAPK pathways by C. jejuni infection is essential for IL-8 production. Unlike salmonellae, C. jejuni flagellin does not play a significant role in the stimulation of IL-8 produc-
tion (Watson and Galan, 2005). Other cytokines induced in human epithelial cells after Campylobacter infection include IL-la, IL-4, IL-10, tumor necrosis factor (TNF)-a, and interferon gamma (Al-Salloom et al., 2003). During infection, Cumpylobacter can traverse the epithelial barrier and interact with numerous leukocytes (Walker et al., 1986; Wallis, 1994). Monocytes/macrophages, considered early host responders to infection, are also an important source of proinflammatory cytokines. C. jejuni infection of human monocytes has been shown to induce proinflammatory cytokines IL-la, IL-lp, IL-6, IL-8, and TNF-a (Hickey et al., 2005; Jones et al., 2003). DCs play important roles in both the innate and adaptive immune responses to microbial pathogens. These major antigen-presenting cells are widely distributed in tissues including the intestinal mucosa, and are also an important source of proinflammatory cytokines. DCs have been shown to produce, via activation of NF-KB, significant amounts of IL-la, IL-6, IL-8, IL10, IL-12, interferon gamma, and TNF-a after infection with C. jejuni (Hu et al., 2006a). Among the cytokines produced in the earliest phases of infection are members of a family of chemoattractant cytokines known as chemokines. Chemokines function mainly as chemoattractants for leukocytes recruiting monocytes, neutrophils, and other effector cells from blood to the site of infection. Chemokines can be released by many different cell types in response to bacterial products or other agents. C. jejuni 81-176 infection has been shown to upregulate the growth-related oncogene a (GROa), GROy, macrophage inflammatory protein 1, monocyte chemoattractant protein 1 (MCP-l), and gamma interferon-inducible protein 10 (IP-10). Infection with viable campylobacters was necessary for sustained chemokine transcription, which was NF-KB dependent (Bakhiet et al., 2004; Hu and Hickey, 2005). T84 cell infection with C. jejuni was reported to upregulate macrophage inflammatory protein 3 a, IP10, GROa, and ENA-78 transcription and secretion. These chemokines both activated immature DCs and initiated a T-cell-mediated immune response. The upregulation of theses chemokines by C. jejuni was flagellin independent (Johanesen and Dwinell, 2006). The secretion of proinflammatory chemokines and cytokines likely plays a role in disease clearance, in disease pathology, and in postinfection immune sequelae by enlisting polymorphonuclear neutrophils, macrophage, DCs, and T cells to sites of infection. Apoptosis After interaction with or internalization into epithelial (or other) cells, the offending bacterium can
CHAPTER 1 7
presumably cause host cell death by apoptosis or necrosis. Overwhelming intracellular bacterial multiplication could presumably result in host cell necrosis. In a recent study, C. jejuni strains were reported to produce one of two variants, FspAl or FspA2, of a small acidic protein that is secreted extracellularly through the flagella. Addition of FspA2, but not FspAl, to INT407 epithelial cells resulted in rapid induction of apoptosis, suggesting alternate virulence potential among different C. jejuni strains (Poly et al., 2007). C. jejuni, like Shigella and Salmonella, can trigger programmed cell death in macrophages. Shigella causes apoptosis in macrophages by injecting invasion plasmid antigen B (IpaB) into the cytosol via the type I11 secretion system, and activate procaspase 1, initiating the apoptotic process. Salmonella SipB, a homologue of IpaB, also causes macrophages to undergo apoptosis (Uchiya et al., 1999; Vazquez-Torres et al., 2000). Siegesmund et al. (2004) reported that C. jejuni induce apoptosis in 63% of infected THP-1 monocytes; the Campylobacter invasion antigen CiaB was required for this THP-1 cell apoptosis, which was caspase 1 and caspase 9 independent. In a separate study by Hickey and coworkers (2005), C. jejuni CDT was reported to activate caspases 8 and 9, leading to apoptosis of human SC28 monocytes. In this latter report, C. jejuni 81-176 mutants that were defective in ciaB and flgB induced similar levels of apoptosis to that of wild-type 81-176, leaving the role of CiaB unclear. Although C. jejuni CDT was previously shown to stimulate IL-8 production in INT407 cells, CDT did not induce either IL-6 or IL-8 in infected monocytes (Hickey et al., 2005).
SUMMARY AND FUTURE DIRECTIONS Clinical infections, experimental infections in humans and animals, and in vitro analyses in cultured human cells have now clearly demonstrated that cell adherence and invasiveness are necessary steps in Campylobacter-induced inflammatory diarrhea. Much progress has been made in the past 10 years in our understanding of the cell biology of these events. However, despite these advances, there remain many unanswered questions. Do C. jejuni strains differ in their pathogenic abilities? Are some strains invasive while others have the ability to cause noninflammatory diarrhea? Specific C. jejuni surface proteins and polysaccharides have been shown to be necessary for adherence to host cells. However, definition of the host receptors and the relative importance of each bacterial ligand await future study. C. jejuni invasion-
HUMAN HOST CELL ENTRY
309
specific host interaction appears to involve host receptors located in filipin 111-sensitive membrane microdomains, most likely lipid rafts, or to occur via CadF protein binding to host fibronectin receptors. Cell culture assays for invasiveness have been variable in methodology and have led to different host cytoskeletal requirements for the same strain in different studies. Although the available data are limited, we have hypothesized that there may be at least two different C. jejuni invasion pathways with different host receptors and likely different cytoskeletal requirements. If true, these systems need to be isolated by mutation and/ or by pathway-specific cell culture assays to identify their cognate structural and biochemical requirements. At least one of the proposed uptake pathways (i.e., strong MT dependence) triggers host membrane receptors that result in activation of G proteins, release of Ca2+ from intracellular stores, and activation of calmodulin, PKC, PI-3 kinase, and MAPKs, which somehow lead to an entry process that requires polymerized MTs. What are the relevant cell biological interconnections involved in these multiple signaling events? Despite this current base of molecular understanding, the bacterial invasion ligand and host invasion receptor remain undefined. An important technical issue is the lack of a small animal model of typical Campylobacter disease, making it difficult to assess the pathogenic potential of targeted bacterial mutants. Nevertheless, the selected use of cultured human cell lines or explant tissues with targeted bacterial mutants can be further used to create a more complete cell biological understanding of epithelial cell invasion, transcytosis, and upregulation of proinflammatory and other diseaserelated processes. It is likely that this improved knowledge of C. jejuni invasion and mucosal translocation ability will aid the development of new diagnostic, chemotherapeutic, and prophylactic approaches and may help in our dissection of postdiarrhea immune-related sequelae. REFERENCES Abrahams, G. L., and M. Hensel. 2006. Manipulating cellular transport and immune responses: dynamic interactions between intracellular Salmonella enterica and its host cells. Cell. Microbiol. 8:728-737. AbuOun, M., G. Manning, S. A. Cawthraw, A. Ridley, I. H. Ahmed, T. M. Wassenaar, and D. G. Newell. 2005. Cytolethal distending toxin (CDT)-negative Campylobacter jejuni strains and anti-CDT neutralizing antibodies are induced during human infection but not during colonization in chickens. Infect. Immun. 73:3053-3062. Ahmed, I. H., G. Manning, T. M. Wassenaar, S. Cawthraw, and D. G . Newell. 2002. Identification of genetic differences between two Campylobacter jejuni strains with different colonization potentials. Microbiology 148:1203-1212.
310
HU AND KOPECK0
A h , R. A., P. Guerry, and T. J. Trust. 1993. The Campylobacter sigma 54 flaB flagellin promoter is subject to environmental regulation. J. Bacteriol. 175:4448-4455. Allos, B. M. 1997. Association between Campylobacter infection and Guillain-BarrCsyndrome. J. Infect. Dis. 176(Suppl. 2):S125S128. Al-Salloom, F. S., A. A1 Mahmeed, A. Ismaeel, G. A. Botta, and M. Bakhiet. 2003. Campylobacter-stimulated INT407 cells produce dissociated cytokine profiles. J. Infect. 47:217-224. Bacon, D. J., R. A. Alm, D. H. Burr, L. Hu, D. J. Kopecko, C. P. Ewing, T. J. Trust, and P. Guerry. 2000. Involvement of a plasmid in virulence of Campylobacter jejuni 8 1-176. Infect. Immun. 68~4384-4390. Bacon, D. J., R. A. Alm, L. Hu, T. E. Hickey, C. P. Ewing, R. A. Batchelor, T. J. Trust, and P. Guerry. 2002. DNA sequence and mutational analyses of the pVir plasmid of Campylobacter jejuni 81-176. Infect. lmmun. 70:6242-6250. Bacon, D. J., C. M. Szymanski, D. H. Burr, R. P. Silver, R. A. Alm, and P. Guerry. 2001. A phase-variable capsule is involved in virulence of Campylobacter jejuni 8 1-176. Mol. Microbiol. 40~769-777. Bakhiet, M., F. S. Al-Salloom, A. Qareiballa, K. Bindayna, I. Farid, and G. A. Botta. 2004. Induction of alpha and beta chemokines by intestinal epithelial cells stimulated with Campylobacter jejuni. J. Infect. 48:236-244. Barnes, I. H., M. C. Bagnall, D. D. Browning, S. A. Thompson, G. Manning, and D. G. Newell. 2007. Gamma-glutamyl transpeptidase has a role in the persistent colonization of the avian gut by Campylobacter jejuni. Microb. Pathog. 43:198-207. Bierne, H., S. Dramsi, M. P. Gratacap, C. Randriamampita, G. Carpenter, B. Payrastre, and P. Cossart. 2000. The invasionprotein InlB from Listeria monocytogenes activates PLC-gamma1 downstream from PI 3-kinase. Cell. Microbiol. 2:465-476. Biswas, D., U. Fernando, C. Reiman, P. Willson, A. Potter, and B. Allan. 2006. Effect of cytolethal distending toxin of Campylobacter jejuni on adhesion and internalization in cultured cells and in colonization of the chicken gut. Avian Dis. 50586593. Biswas, D., K. Itoh, and C. Sasakawa. 2000. Uptake pathways of clinical and healthy animal isolates of Campylobacter jejuni into INT407 cells. FEMS Immunol. Med. Microbiol. 29:203-211. Biswas, D., K. Itoh, and C. Sasakawa. 2003. Role of Microfilaments and microtubules in the Invasion of INT407 Cells by Campylobacter jejuni. Microbiol. Immunol. 47:469-473. Biswas D., H. Niwa, and K. Itoh. 2004. Infection with Campylobacter jejuni induces tyrosine-phosphorylated proteins into INT-407 cells. Microbiol. Immunol. 48:221-228. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 198 8. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Boucrot, E., T. Henry, J. P. Borg, J. P. Gorvel, and S. Meresse. 2005. The intracellular fate of Salmonella depends on the recruitment of kinesin. Science 308:1174-1178. Bras, A. M., and J. M. Ketley. 1999. Transcellular translocation of Campylobacter jejuni across human polarised epithelial monolayers. FEMS Microbiol. Lett. 179:209-215. Brumell, J. H., D. L. Goosney, and B. B. Finlay. 2002. SifA, a type 111 secreted effector of Salmonella typhimurium, directs Salmonella-induced filament (Sif) formation along microtubules. Trafic 3:407-415. Byrd, J., R. H. Bailey, R Wills, and D. Nisbet. 2007. Recovery of Campylobacter from commercial broiler hatchery trayliners. Poult. Sci. 86:26-29. Cawthraw, S. A., Wassenaar, T. M., Ayling, R., and D. G. Newell. 1996. Increased colonization potential of Campylobacter jejuni strain 81116 after passage through chickens and its implication
on the rate of transmission within flocks. Epidemiol. Infect. 117: 213-215. Chen, M. L., 2. Ge, J. G. Fox, and D. B. Schauer. 2006. Disruption of tight junctions and induction of proinflammatory cytokine responses in colonic epithelial cells by Campylobacter jejuni. Infect. Immun. 74:6581-6589. Clapham, D. E. 1995. Calcium signaling. Cell 80:259-268. Clausen, J. D., G. Christiansen, H. U. Holst, and S. Birkelund. 1997. Chlamydia trachomatis utilizes the host cell microtubule network during early events of infection. Mol. Microbiol. 25: 44 1-449. de Melo, M. A., G. Gabbiani, and J. C. Pechere. 1989. Cellular events and intracellular survival of Campylobacter jejuni during infection of HEp-2 cells. Infect. Immun. 57:2214-2222. Elsinghorst, E. A., L. S. Baron, and D. J. Kopecko. 1989. Penetration of human intestinal epithelial cells by Salmonella: molecular cloning and expression of Salmonella typhi invasion determinants in Escherichia coli. Proc. Natl. Acad. Sci. USA 86: 5 173-5 177. Eppinger, M., C. Baar, G. Raddatz, D. H. Huson, and S. C. Schuster. 2004. Comparative analysis of four campylobacterales.Nut. Rev. Microbiol. 2: 872-8 8 5. Everest, P. H., H. Goossens, J. P. Butzler, D. Lloyd, S. Knutton, J. M. Ketley, and P. H. Williams. 1992. Differentiated Caco-2 cells as a model for enteric invasion by Campylobacter jejuni and C. coli. J. Med. Microbiol. 37:319-325. Fauchere, J. L., A. Rosenau, M. Veron, E. N. Moyen, S. Richard, and A. Pfister. 1986. Association with HeLa cells of Campylobacter jejuni and Campylobacter coli isolated from human feces. Infect. Immun. 54:283-287. Finlay, B. B., and S. Falkow. 1989. Common themes in microbial pathogenicity. Microbiol. Rev. 53:210-230. Finlay, B. B., and S. Falkow. 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. Finlay, B. B., and S. Falkow. 1990. Salmonella interactions with polarized human intestinal Caco-2 epithelial cells. J. Infect. Dis. 162:1096-1106. Finlay, B. B., S. Ruschkowski, and S. Dedhar. 1991. Cytoskeletal rearrangements accompanying salmonella entry into epithelial cells. J. Cell. Sci. 99:283-296. Florin, I., and F. Antillon. 1992. Production of enterotoxin and cytotoxin in Campylobacter jejuni strains isolated in Costa Rica. J. Med. Microbiol. 37:22-29. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PloS Biol. 3:e15. Fox, J. G., A. B. Rogers, M. T. Whary, 2. Ge, N. S. Taylor, S. Xu, B. H. Honvitz, and S. E. Erdman. 2004. Gastroenteritis in NF-kappaB-deficient mice is produced with wild-type Camplyobacter jejuni but not with C. jejuni lacking cytolethal distending toxin despite persistent colonization with both strains. Infect. lmmun. 72:1116-1125. Frisan, T., X. Cortes-Bratti, and M. Thelestam. 2002. Cytolethal distending toxins and activation of DNA damage-dependent checkpoint responses. Int. J. Med. Microbiol. 291:495-499. Garcia-del Portillo, F., M. B. Zwick, K. Y. Leung, and B. B. Finlay. 1993, Intracellular replication of Salmonella within epithelial
CHAPTER 17
cells is associated with filamentous structures containing lysosoma1 membrane glycoproteins. Infect. Agents Dis. 2:227-23 1. Garvis, S . G., G. J. Puzon, and M. E. Konkel. 1996. Molecular characterization of a Campylobacter jejuni 29-kilodalton periplasmic binding protein. Infect. Immun. 64:3537-3543. Grant, C. C., M. E. Konkel, W. Cieplak, Jr., and L. S. Tompkins. 1993. Role of flagella in adherence, internalization, and translocation of Cumpylobucter jejuni in nonpolarized and polarized epithelial cell cultures. Infect. Immun. 61:1764-1771. Grieshaber, S . S., N. A. Grieshaber, and T. Hackstadt. 2003. Chlamydia trachomatis uses host cell dynein to traffic to the microtubule-organizing center in a p50 dynamitin-independent process. J. Cell Sci. 116:3793-3802. Guignot, J., E. Caron, C. Beuzon, C. Bucci, J. Kagan, C. Roy, and D. W. Holden. 2004. Microtubule motors control membrane dynamics of Salmonella-containing vacuoles. J. Cell Sci. 117: 1033-1045. Hanel, I., J. Muller, W. Muller, and F. Schulze. 2004. Correlation between invasion of Caco-2 eukaryotic cells and colonization ability in the chick gut in Campylobacter jejuni. Vet. Microbiol. 101~75-82. Hardwidge, P. R., W. Deng, B. A. Vallance, I. RodriguezEscudero, V. J. Cid, M. Molina, and B. B. Finlay. 2005. Modulation of host cytoskeleton function by the enteropathogenic Escherichia coli and Citrobacter rodentium effector protein EspG. Infect. lmmun. 73:2586-2594. Harrison, R. E., J. H. Brumell, A. Khandani, C. Bucci, C. C. Scott, X. Jiang, B. B. Finlay, and S. Grinstein. 2004. Salmonella impairs RILP recruitment to Rab7 during maturation of invasion vacuoles. Mol. Biol. Cell. 15:3146-3 154. Harvey, P., Battle, T., and S. Leach. 1999. Different invasion phenotypes of Cumpylobacter isolates in Caco-2 cell monolayers. J. Med. Microbiol. 48:461-69. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Cumpylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Henry, T., J. P. Gorvel, and S . Meresse. 2006. Molecular motors hijacking by intracellular pathogens. Cell. Microbiol. 8:23-32. Hickey, T. E., S . Baqar, A. L. Bourgeois, C. P. Ewing, and P. Guerry. 1999. Campylobacter jejuni-stimulated secretion of interleukin-8 by INT407 cells. Infect. lmmun. 67:88-93. Hickey, T. E., G. Majam, and P. Guerry. 2005. Intracellular survival of Campylobacter jejuni in human monocytic cells and induction of apoptotic death by cytholethal distending toxin. Infect. Immun. 735194-5197. Hickey, T. E., A. L. McVeigh, D. A. Scott, R E. Michielutti, A. Bixby, S . A. Carroll, A. L. Bourgeois, and P. Guerry. 2000. Campylobacter jejuni cytolethal distending toxin mediates release of interleukin-8 from intestinal epithelial cells. Infect. lmmun. 68:6535-6541. Hofrenter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 744694-4707. Hu, L., M. D. Bray, M. Osorio, and D. J. Kopecko. 2006a. Cumpylobacter jejuni induces maturation and cytokine production in human dendritic cells. Infect. Immun. 74:2697-2705. Hu, L., and T. E. Hickey. 2005. Campylobacter jejuni induces secretion of proinflammatory chemokines from human intestinal epithelial cells. Infect. Immun. 73:4437-4440. Hu, L., and D. J. Kopecko. 1999. Campylobacter jejuni 81-176 associates with microtubules and dynein during invasion of human intestinal cells. Infect. Immun. 67:4171-4182. Hu, L., and D. J. Kopecko. 2000. Interactions of Campylobacter with eukaryotic cells: gut luminal colonization and mucosal invasion mechanisms, p. 191-215. In I. Nachamkin and M. J.
HUMAN HOST CELL ENTRY
311
Blaser (ed.), Campylobacter, 2nd ed. American Society for Microbiology, Washington, DC. Hu, L., J. P. McDaniel, and D. J. Kopecko. 2006b. Signal transduction events involved in human epithelial cell invasion by Campylobacter jejuni 81-176. Microb. Pathog. 40:91-100. Hu, L, R. B. Raybourne, and D. J. Kopecko. 2005. Ca2+ release from host intracellular stores and related signal transduction during Campylobacter jejuni 8 1-176 internalization into human intestinal cells. Microbiology 151:3097-3 105. Huang, X. Z., B. Tall, W. R. Schwan, and D. J. Kopecko. 1998. Physical limitations on Salmonella typhi entry into cultured human intestinal epithelial cells. Infect. Immun. 66:2928-2937. Ilangumaran, S., and D. C. Hoessli. 1998. Effects of cholesterol depletion by cyclodextrin on the sphingolipid microdomains of the plasma membrane. Biochem. J. 335(Pt. 2):433-440. Isberg, R. R., D. L. Voorhis, and S. Falkow. 1987. Identification of invasin: a protein that allows enteric bacteria to penetrate cultured mammalian cells. Cell 50:769-778. Jacob, R. 1990. Calcium oscillations in electrically non-excitable cells erratum appears. Biochim. Biopbys. Acta 1052:427-438. Jin, S., A. Joe, J. Lynett, E. K. Hani, P. Sherman, and V. L. Chan. 2001. JlpA, a novel surface-exposed lipoprotein specific to Campylobacter jejuni, mediates adherence to host epithelial cells. Mol. Microbiol. 39:1225-1236. Jin, S., Song, Y. C., Emil, A., Sherman, P. A., and V. L. Chan. 2003. JplA of Campylobacter jejuni interacts with surfaceexposed heat shoch protein 90alpha and triggers signaling pathways leading to the activation of NF-kappaB and p38 MAP kinase in epithelial cells. Cell. Microbiol. 5:165-174. Johanesen, P. A., and M. B. Dwinell. 2006. Flagellin-independent regulation of chemokine host defense in Campylobacter jejuniinfected intestinal epithelium. Infect. lmmun. 74:3437-3447. Johnson, W. M., and H. Lior. 1988a. A new heat-labile cytolethal distending toxin (CLDT) produced by Campylobacter spp. Microb. Pathog. 4:115-126. Johnson, W. M., and H. Lior. 1988b. A new heat-labile cytolethal distending toxin (CLDT) produced by Escherichia coli isolates from clinical material. Microb. Patbog. 4:103-113. Jones, M. A., K. L. Marston, C. A. Woodall, D. J. Maskell, D. Linton, A. V. Karlyshev, N. Dorrell, B. W. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni N C T C l l l 6 8 to high-level colonization of the avian gastrointestinal tract. Infect. Immun. 72:3769-3776. Jones, M. A., S . Totemeyer, D. J. Maskell, C. E. Bryant, and P. A. Barrow. 2003. Induction of proinflammatory responses in the human monocytic cell line THP-1 by Cumpylobacter jejuni. Infect. Immun. 71:2626-2633. Kanipes, M. I., L. C. Holder, A. T. Corcoran, A. P. Moran, and P. Guerry. 2004. A deep-rough mutant of Campylobacter jejuni 81-176 is noninvasive for intestinal epithelial cells. Infect. Immun. 72:2452-4255. Kanwar, R. K., N. K. Ganguly, L. Kumar, J. Rakesh, D. Panigrahi, and B. N. S. Walia. 1995. Calcium and protein kinase C play an important role in Cumpylobacter jejuni-induced changes in Na+ and C1, transport in rat ileum in vitro. Biochim. Biophys. Acts 1270~179-192. Karlyshev, A. V., P. Everest, D. Linton, S . Cawthraw, D. G. Newell, and B. W. Wren. 2004. The Campylobacter jejuni general glycosylation system is important for attachment to human epithelial cells and in the colonization of chicks. Microbiology 150: 1957-1964. Karlyshev, A. V., D. Linton, N. A. Gregson, A. J. Lastovica, and B. W. Wren. 2000. Genetic and biochemical evidence of a Campylobacter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 35529-541.
312
HU AND KOPECK0
Ketley, J. M. 1997. Pathogenesis of enteric infection by Campylobacter. Microbiology 143(Pt. 1):5-21. Kihlstrom, E., and L. Nilsson. 1977. Endocytosis of Salmonella typhimurium 395 MS and MRIO by HeLa cells. Acta Pathol. Microbiol. Scand. B 85B:322-328. Klipstein, F. A., and R. F. Engert. 1984a. Properties of crude Campylobacter jejuni heat-labile enterotoxin. Infect. Immun. 45: 3 14-3 19. Klipstein, F. A., and R. F. Engert. 1984b. Purification of Campylobacter jejuni enterotoxin. Lancet i:ll23-1124. Konkel, M. E., M. D. Convin, L. A. Joens, and W. Cieplak. 1992a. Factors that influence the interaction of Campylobucter jejuni with cultured mammalian cells. J. Med. Microbiol. 37:30-37. Konkel, M. E., S . A. Gray, B. J. Kim, S. G. Garvis, and J. Yoon. 1999. Identification of the enteropathogens Campylobacter jejuni and Campylobacter coli based on the cadF virulence gene and its product.]. Clin. Microbiol. 37510-517. Konkel, M. E., S. F., Hayes, L. A. Joens, and W. Cieplak. Jr. 1992b. Characteristics of the internalization and intracellular survival of Campylobacter jejuni in human epithelial cell cultures. Microb. Pathog. 13:357-370. Konkel, M. E., and L. A. Joens. 1989. Adhesion to and invasion of HEp-2 cells by Campylobacter spp. Infect. Immun. 57:29842990. Konkel, M. E., J. D. Klena, V. Rivera-Amill, M. R. Monteville, D. Biswas, B. Raphael, and J. Mickelson. 2004. Secretion of virulence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. J. Bacteriol. 186:32963303. Konkel, M. E., D. J. Mead, S . F. Hayes, and W. Cieplak, Jr. 1992c. Translocation of Campylobacter jejuni across human polarized epithelial cell monolayer cultures. J. Infect. Dis. 166:308-3 15. Kopecko, D. J., Hu, L., and K. J. M. Zaal. 2001. Campylobacter jejuni-microtubule-dependent invasion. Trends Microbiol. 9: 3 89-3 96. Krause-Gruszczynska, M., M. Rohde, R. Hartig, H. Genth, G. Schmidt, T. Keo, W. Konig, W. G. Miller, M. E. Konkel, and S. Backert. 2007. Role of the small Rho GTPases Racl and Cdc42 in host cell invasion of Carnpylobacter jejuni. Cell. Microbiol. 9:243 1-2444. Kuhle, V., G. L. Abrahams, and M. Hensel. 2006. Intracellular Salmonella enterica redirect exocytic transport processes in a Salmonella pathogenicity island 2-dependent manner. Traffic 7: 716-73 0. Kuhle, V., D. Jackel, and M. Hensel. 2004. Effector proteins encoded by Salmonella pathogenicity island 2 interfere with the microtubule cytoskeleton after translocation into host cells. Traffic 5~356-570. Lara-Tejero, M., and J. E. Galan. 2000. A bacterial toxin that controls cell cycle progression as a deoxyribonuclease I-like protein. Science 290:354-357. Lara-Tejero, M., and J. E. Galan. 2001. CdtA, CdtB, and CdtC form a tripartite complex that is required for cytolethal distending toxin activity. Infect. Immun. 69:4358-4365. Lee, A., J. O’Rourke, M. B. Phillips, and P. Barrington. 1983. Campylobacter jejuni as a Mucosa Associated Organism: An Ecological Study. Public Health Laboratory Service, London, United Kindom. MacCallum, A., S . P. Hardy, and P. H. Everest. 2005. Campylobacter jejuni inhibits the absorptive transport functions of Caco2 cells and disrupts cellular tight junctions. Microbiology 151: 245 1-2458. MacCaIlum, A. J., D. Harris, G. Haddock, and P. H. Everest. 2006. Campylobucter jejuni-infected human epithelial cell lines vary in their ability to secrete interleukin-8 compared to in vi-
tro-infected primary human intestinal tissue. Microbiology 152: 3661-3665. MacKichan, J. K., E. C. Gaynor, C. Chang, S. Cawthraw, D. G. Newell, J. F. Miller, and S . Falkow. 2004. The Campylobacter jejuni dccRS two-component system is required for optimal in vivo colonization but is dispensable for in vitro growth. Mol. Microbiol. 54: 1269-128 6. Marks, P. W., and F. R. Maxfield. 1990a. Local and global changes in cytosolic free calcium in neutrophils during chemotaxis and phagocytosis. Cell Calcium 11:181-190. Marks, P. W., and F. R Maxfield. 1990b. Transient increases in cytosolic free calcium appear to be required for the migration of adherent human neutrophils. J. Cell Biol. 110:43-52. McSweegan, E., D. H. Burr, and R. I. Walker. 1987. Intestinal mucus gel and secretory antibody are barriers to Campylobacter jejuni adherence to INT 407 cells. Infect. Immun. 55:14311435. McSweegan, E., and R. I. Walker. 1986. Identification and characterization of two Campylobacter jejuni adhesins for cellular and mucous substrates. Infect. Immun. 53:141-148. Meldolesi, J., E. Clementi, C. Fasolato, D. Zacchetti, and T. Pozzan. 1991. Caz+ influx following receptor activation. Trends Pharmacol. Sci. 12:289-292. Mellits, K. H., J. Mullen, M. Wand, G. Armbruster, A. Patel, P. L. Connerton, M. Skelly, and I. F. Connerton. 2002. Activation of the transcription factor NF-kappaB by Campylobacter jejuni. Microbiology 148:2753-2763. Miller, V. L., B. B. Finlay, and S . Falkow. 1988. Factors essential for the penetration of mammalian cells by Yersinia. Curr. Top. Microbiol. Immunol. 138:15-39. Misawa, N., T. Ohnishi, K. Itoh, and E. Takahashi. 1994. Development of a tissue culture assay system for Campylobacter jejuni cytotoxin and the influence of culture conditions on cytotoxin production. J. Med. Microbiol. 41:224-230. Monteville, M. R., and M. E. Konkel. 2002. Fibronectin-facilitated invasion of T84 eukaryotic cells by Campylobacter jejuni occurs preferentially at the basolateral cell surface. Infect. Immun. 70: 6665-6671. Monteville, M. R., J. E. Yoon, and M. E. Konkel. 2003. Maximal adherence and invasion of INT 407 cells by Campylobacter jejuni requires the CadF outer-membrane protein and microfilament reorganization. Microbiology 149: 153-165. Norris, V., S . Grant, P. Freestone, J. Canvin, F. N. Sheikh, I. Toth, M. Trinei, K. Modha, and R. I. Norman.1996. Calcium signalling in bacteria. ]. Bacteriol. 178:3677-3682. Oelschlaeger, T. A., P. Guerry, and D. J. Kopecko. 1993. Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proc. Natl. Acad. Sci. USA 90:6884-6888. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R M. Davies, T. Feltwell, S . Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S . Whitehead, and B. G. BarrC11. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pei, Z., and M. J. Blaser. 1993. PEB1, the major cell-binding factor of Campylobacter jejuni, is a homolog of the binding component in gram-negative nutrient transport systems. ]. Biol. Chem. 268: 18717-1 8725. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. 2. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943.
CHAPTER 1 7 ~~~~
~~
HUMAN HOST CELL ENTRY
313
~~
Pei, Z. H., R. T. Ellison, 111, and M. J. Blaser. 1991. Identification, purification, and characterization of major antigenic proteins of Campylobacter jejuni. J. Biol. Chem. 266: 16363-16369. Poly, F., C. Ewing, S. Goon, T. E. Hickey, D. Rockabrand, G. Majam, L. Lee, J. Phan, N. J. Savarino, and P. Guerry. 2007. Heterogeneity of a Campylobacter jejuni protein that is secreted through the flagellar filament. Infect. lmmun. 75:3859-3867. Purdy, D., C. M. Buswell, A. E. Hodgson, K. McAlpine, I. Henderson, and s. A. Leach. 2000. Characterisation of cytolethal distending toxin (CDT) mutants of Campylobacter jejuni. 1. Med. Microbiol. 49:473-479. Ringoir, D. D., and V. Korolik. 2003. Colonisation phenotype and colonization potential differences in Campylobacter jejuni strains in chickens before and after passage in vivo. Vet. Microbiol. 92:225-235. Rosenquist, H., N. L. Nielsen, H. M. Sommer, B. Norrung, and B. B. Christensen. 2003. Quantitative risk assessment of human campylobacteriosis associated with thermophilic Campylobacter species in chickens. Int. J. Food Microbiol. 83537-103. Ruiz-Palacios, G. M., J. Torres, N. I. Torres, E. Escamilla, B. R. Ruiz-Palacios, and J. Tamayo. 1983. Cholera-like enterotoxin produced by Campylobacter jejuni. Characterisation and clinical significance. Lancet ik2.50-253. Ruschkowski, S., I. Rosenshine, and B. B. Finlay. 1992. Salmonella typhimurium induces an inositol phosphate flux in infected epithelial cells. FEMS Microbiol. Lett. 35:121-126. Russell, R. G., and D. C. Blake, Jr. 1994. Cell association and invasion of Caco-2 cells by Campylobacter jejuni. Infect. lmmun. 62~3773-3779. Siegesmund, A. M., M. E. Konkel, J. D. Klena, and P. F. Mixter. 2004. Campylobacter jejuni infection of differentiated THP-1 macrophages results in interleukin 1 beta release and caspase-lindependent apoptosis. Microbiology 150561-569. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nut. Rev. Mol. Cell Biol. 1:31-39. Stern, N. J., J. S. Bailey, L. C. Blankenship, N. A. Cox, and F. McHan. 1988. Colonization characteristics of Campylobacter jejuni in chick ceca. Avian Dis. 32:330-334. Tanaka, T., T. Ohmura, T. Yamakado, and H. Hidaka. 1982. Two types of calcium-dependent protein phosphorylations modulated by calmodulin antagonists. Naphthalenesulfonamide derivatives. Mol. Phamzacol. 22:408-412. Tran Van Nhieu, G., C. Clair, G. Grompone, and P. Sansonetti. 2004. Calcium signalling during cell interactions with bacterial pathogens. Biol. Cell. 96:93-101. Tsien, R. W., and R. Y. Tsien. 1990. Calcium channels, stores, and oscillations. Annu. Rev. Cell Biol. 6:715-760. Uchiya, K., M. A. Barbieri, K. Funato, A. H. Shah, P. D. Stahl, and E. A. Groisman. 1999. A Salmonella virulence protein that inhibits cellular trafficking. EMBO]. 18:3924-3933.
Van Alphen, L. B., Bleumink-Pluym, N. M. C., Rochat, K. D., van Balkom, B. W. M., Wosten, M. M. S. M., and J. P. M. van Putten. 2008. Active migration into the subcellular space precedes Campylobacter jejuni invasion of epithelial cells. Cell. Microbiol. 1053-66. van Spreeuwel, J. P., G. C. Duursma, C. J. Meijer, R. Bax, P. C. Rosekrans, and J. Lindeman. 1985. Campylobacter colitis: histological. immunohistochemical and ultrastructural findings. Gut 26~945-951. Vazquez-Torres, A., Y. Xu, J. Jones-Carson, D. W. Holden, S. M. Lucia, M. C. Dinauer, P. Mastroeni, and F. C. Fang. 2000. Salmonella pathogenicity island 2-dependent evasion of the phagocyte NADPH oxidase. Science 287: 1655-1658. Walker, R. I., M. B. Caldwell, E. C. Lee, P. Guerry, T. J. Trust, and G. M. Ruiz-Palacios. 1986. Pathophysiology of Campylobacter enteritis. Microbiol. Rev. 50:8 1-94. Wallis, M. R. 1994. The pathogenesis of Campylobacter jejuni. Br. J. Biomed. Sci. 5157-64. Wassenaar, T. M., and M. J. Blaser. 1999. Pathophysiology of Campylobacter jejuni infections of humans. Microbes Infect. 1: 1023-1033. Wassenaar, T. M., N. M. Bleumink-Pluym, and B. A. van der Zeijst. 1991. Inactivation of Campylobacter jejuni flagellin genes by homologous recombination demonstrates that flaA but not flaB is required for invasion. EMBO J. 10:2055-2061. Watson, R. O., and J. E. Galan. 2005. Signal transduction in Campylobacter jejuni-induced cytokine production. Cell. Microbiol. 7~6.55-665. Whitehouse, C. A., P. B. Balbo, E. C. Pesci, D. L. Cottle, P. M. Mirabito, and C. L. Pickett. 1998. Campylobacter jejuni cytolethal distending toxin causes a G2-phase cell cycle block. Infect. lmmun. 66:1934-1940. Wooldridge, K. G., and J. M. Ketley. 1997. Campylobacter-host cell interactions. Trends Microbiol. 5:96-102. Wooldridge, K. G., P. H. Williams, and J. M. Ketley. 1996. Host signal transduction and endocytosis of Campylobacter jejuni. Microb. Pathog. 21:299-305. Yanagawa, Y., M. Takahashi, and T. Itoh. 1994. The role of flagella of Campylobacter jejuni in colonization in the intestinal tract in mice and the cultured-cell infectivity. Nippon Saikingaku Zasshi 49:395-403. Yao, R., D. H. Burr, P. Doig, T. J. Trust, H. Niu, and P. Guerry. 1994. Isolation of motile and non-motile insertional mutants of Campylobacter jejuni: the role of motility in adherence and invasion of eukaryotic cells. Mol. Microbiol. 14:883-893. Yoshida, S., Y. Handa, T. Suzuki, M. Ogawa, M. Suzuki, A. Tamai, A. Abe, E. Katayama, and C. Sasakawa. 2006. Microtubule-severing activity of Shigella is pivotal for intercellular spreading. Science 314:985-989. Yoshida, S., and C. Sasakawa. 2003. Exploiting host microtubule dynamics: a new aspect of bacterial invasion. Trends Microbiol. 11~139-143.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 18
Campylobacter jejuni Secretes Proteins via the Flagellar Type I11 Secretion System That Contribute to Host Cell Invasion and Gastroenteritis CHARLESL. LARSON,JEFFREY E. CHRISTENSEN, SOPHIAA. PACHECO,SCOTTA. MINNICH, AND MICHAEL E. KONKEL
juni protein export via the flagellar type I11 secretion system (T3SS), the development of an assay to identify C. jejuni secreted proteins, the evolutionary relatedness of the flagellum and virulence T3SS, and the putative roles of C. jejuni secreted proteins in disease. Although much remains unknown regarding the identity and functional characteristics of the proteins exported via the flagellar apparatus, we will highlight evidence supporting the proposal that these proteins contribute to C. jejuni-mediated enteritis.
Campylobacter jejuni is a leading bacterial cause of gastroenteritis worldwide. The clinical presentation of C. jejuni-mediated disease varies in symptoms, severity, and duration. The spectrum of disease observed in infected individuals likely results from differences in C. jejuni strain virulence and host immunity. Although research indicates that C. jejuni strains have differences in gene content and expression, C. jejuni virulence requires motility, host (target) cell adherence, host cell invasion, alteration of host cell signaling pathways, induction of host cell death, evasion of host immune defenses, iron acquisition, and drug/detergent resistance. This list is not comprehensive but rather illustrates our belief that C. jejuni disease occurs in a susceptible host from a combination of virulence attributes working in concert. We propose that the most severe form of disease, which is characterized by fever, severe abdominal cramps, and diarrhea containing blood and leukocytes, involves C. jejuni invasion of the intestinal epithelium. In the context of bacteria-host cell invasion, we have found that C. jejuni secrete proteins and that the secreted proteins contribute to the organism’s ability to maximally invade epithelial cells. The body of this chapter is divided into three major sections. In the first section, we present a model of C. jejuni-mediated enteritis. The second section presents a general overview of the organism’s pathogenic mechanisms and virulence determinants. Finally, in the third section, we discuss various aspects of C. jejuni-host cell invasion and protein secretion. Specifically, in this section we discuss C. je-
MODEL OF CAMPYLOBACTER JEJUNIMEDIATED ENTERITIS Figure 1 illustrates our model of C. jejuni pathogenesis. An accumulation of events during the C. jejuni-host interaction produce the clinical manifestation of C. jejuni infection. Although the exact sequence of events that occur between colonization of the host intestine and C. jejuni-mediated enteritis is unclear, we have incorporated knowledge from different sources to produce a disease model. The model presented herein describes severe cases of C. jejuni infection, in which individuals exhibit fever, abdominal cramps and diarrhea containing blood and leukocytes (Fig. 1).Clearly, this is not the only disease course resulting from C. jejuni infections. Individuals infected with C. jejuni may experience mild diarrhea symptoms, or they may develop postinfection sequelae such as Guillain-BarrC syndrome or irritable bowel syndrome (Schwerer,
Charles L. Larson, Jeffrey E. Christensen, Sophia A. Pacheco, and Michael E. Konkel School of Molecular Biosciences, Washington State University, Pullman, WA 99164-4234. Scott A. Minnich Department of Microbiology, Molecular Biology and Biochemistry, Life Science South 146, P.O. Box 443052, University of Idaho, Moscow, ID 83844-3052.
315
316
LARSON ET AL.
Figure 1. Model of C. jejuni pathogenesis. The clinical manifestation of C. jejuni infection is a result of bacterial activities (white box) and the host immune response (gray box). Dotted lines indicate processes that can potentiate infection.
2002; Thornley et al., 2001). Clinical manifestations of C. jejuni infection, other than the acute disease state described herein, require separate models. In essence, our model is a platform to analyze the specific C. jejuni virulence factors involved during various stages of acute infection. Early Infection: Key Roles for Virulence Factors Ingestion of viable C. jejuni is the first step of the infectious process. Individuals most often acquire C. jejuni infection from the consumption of foods cross-contaminated with raw or undercooked poultry products; however, milk, eggs, untreated water, and contact with animals colonized with C. jejuni have also been implicated as sources of infection (Friedman et al., 2004; Gillespie et al., 2006). The number of C. jejuni required to infect an individual varies significantly. One study revealed that 800 organisms
were sufficient to cause disease in 50% of the individuals inoculated, while approximately 10' bacteria were required to infect 100% of a population. Moreover, doses of >800 organisms did not produce more severe disease or increase its duration (Black et al., 1988). The factors that determine the infective dose of C. jejuni are not well defined; however, the pathogenicity of the C. jejuni strain and the host immune response are both important determinants. C. jejuni undergoes an adaptive response in which global changes in gene expression occur that facilitate survival in and infection of the host (Andersen et al., 2005; Hendrixson, 2006; Lin et al., 2005; Malik-Kale et al., 2007). Passage of the organism through the gastrointestinal tract triggers the expression of genes encoding virulence factors that help establish infection, including an efflux pump that confers resistance to bile salts and antibiotics (Lin et al., 2002, 2005; Raphael et al., 2005). Once through
C. J E T W I PROTEIN SECRETION
CHAPTER 18
the stomach and duodenum, C. jejuni localizes to the jejunum and/or ileum of the small intestine and occupies specific niches within the host intestinal epithelium in order to resist the peristaltic forces and flushing of the small intestine. C. jejuni responds to the environment, and in a directed manner, the flagellum drives the bacterium through the viscous mucosal layer covering the epithelium where it adheres to the host cell surface. Debate remains regarding the specific cell types within the ileum and/or jejunum that C. jejuni is localized, but studies have indicated targeting of the glandular crypts and M cells (Everest et al., 1993a, 199310, Walker et al., 1988). Regardless, C. jejuni is able to survive, replicate, and interact with host epithelial cells once it traverses the mucosal layer. Motility is pivotal throughout C. jejuni infection. Flagellar locomotion is controlled by chemotaxis and quorum-sensing mechanisms (Jeon et al., 2003; Yao et al., 1997). During the initial stages of infection, bacterial motility promotes access to the apical and basolateral surfaces of host cells. Upon
bacteria-host cell contact, flagellar activity may be modulated to maintain adherence to host cell surfaces and secrete proteins that contribute to infection. The binding of C. jejuni adhesins to specific host cell ligands, which are concentrated on the basolateral surface of the epithelium, facilitates maximal bacteriahost cell invasion (Chen et al., 2006; Monteville and Konkel, 2002). Establishment of adherence facilitates the invasive and cytotoxic activities of C. jejuni responsible for the manifestation of enteric disease. The initiation and establishment of C. jejuni infection involves adherence, protein secretion, and invasion, which can stimulate the host cell inflammatory response and in turn promote additional bacteria-host cell interactions (Fig. 2). C. jejuni appears to have properties that diminish its recognition by immune system and aid in establishing a niche in the host. Although the C. jejuni FlaA and FlaB filament proteins are highly immunogenic, the monomers comprising the filament lack the recognized Toll-like receptor 5 (TLRS) consensus domain of other enteric pathogens. As such, the c.
Consumption of C. jejuni contaminated products
1
n
Survival in the host environment (Adaptive response)
I
c
c
Persistence in mucosal layer
+ Replication
c
Paracellular transepithelial
M cell adsorption
migration
_ _ - - - -Protein secretion
-_I--
4 Invasion (Intracellular survival)
A/
\
\
317
Host cells release proinflammatory cytokines
4
I
Induction of the inflammatory response
Figure 2. Adherence, protein secretion, and invasion are a few of the C. jejuni virulence attributes that contribute to acute infection. As depicted, bacterial colonization of the intestinal tract can occur by different routes. Several virulence attributes may stimulate the host inflammatory response and in turn promote additional bacteria-host cell interactions. Other factors (not listed) are also capable of triggering the host inflammatory response. The dotted line represents the possibility that secreted proteins may enhance the cytokine response.
318
LARSON ET AL.
jejuni flagellin proteins do not trigger interleukin-8 secretion (Ramos et al., 2004). In addition, the C. jejuni flagellar filament is heavily glycosylated with pseudaminic acid, containing more glycan modifications than any other known bacterial proteins (chapter 26) (Guerry et al., 2006). Glycosylation of outer surface proteins by mucosal pathogens contributes to protection against proteolytic cleavage, provides antigenic variation, and aids in immune evasion (Szymanski and Wren, 2005). It is also interesting that the genes involved in flagellar biosynthesis and glycosylation contain numerous polymorphisms among C. jejuni strains, resulting in antigenic diversity (Szymanski et al., 2003). Late Infection During late infection, the disruption of tight junctions, alteration of net water flow, and induction of premature apoptosis (or necrosis) contribute to the severe disease state. These consequences are likely due to a combination of C. jejuni-host cell interactions that include adherence, invasion, and secretion of bacterial effector proteins. The disruption of tight junctions during bacterial infections may also result from host immune processes (Chen et al., 2006; MacCallum et al., 2005; Perdomo et al., 1994b; Sansonetti et al., 1999). Although the precise virulence attributes that contribute to this stage of disease remain to be defined, the net result is severe gastroenteritis. As mentioned previously, C. jejuni binds and invades more efficiently from the basolateral surface of cells, suggesting localization to the lamina propria may be an important step in infection. M-cell adsorption, demonstrated with the rabbit ileal loop model, is one route that C. jejuni can pass through the epithelium into the lamina propria (Walker et al., 1988). In vitro observations indicate transepithelial migration is accompanied by tight junction alteration (Chen et al., 2006; MacCallum et al., 2005; Monteville and Konkel, 2002; Walker et al., 1988). Tight junctions connect adjacent epithelial cells separating the apical and basolateral surfaces, thus creating a barrier that impedes bacteria in the lumen from entering the lamina propria. In addition, tight junctions provide a fencing function, which serve to limit receptors (integrins, focal adhesions, etc.) to the basolateral surface of the epithelium (Blikslager et al., 2007). Tight junctions are an important facet of the gut innate immune defense, and their disruption would result in two processes that could potentiate C. jejuni infection. First, their disruption would allow a bacterium increased access to the lamina propria and to bind the basolateral surfaces of host cells,
which might further increase host cell invasion. Second, the loss of fencing function provided by tight junctions would allow host membrane receptors to migrate to the apical surfaces of cells, thereby facilitating additional bacterial adherence and invasion directly from the lumen. The influx of fluid and professional phagocytes during C. jejuni infection could result in the disruption of tight junctions (Black et al., 1988; Blaser et al., 1983; Gillespie et al., 2006). During the inflammatory response, neutrophils recruited by the release of cytokines migrate across the intestinal epithelium and disrupt tight junctions. In this process, host ligands like fibronectin and integrin receptors become available at the apical surface of the cell, thereby facilitating adherence and invasion. In addition to loss of fencing function, temporary lesions created by neutrophil transepithelial migration may provide a path for C. jejuni to enter the lamina propria. Evidence of this phenomenon is observed during Shigella infection (Perdomo et al., 1994a, 1994b). Before the manifestation of diarrhea, individuals infected with C. jejuni are often febrile. The increased interaction of C. jejuni with cells in the lamina propria likely coincides with the intestinal inflammation characteristic of later stages of infection. Specialized cells located within the lamina propria have an abundance of pathogen recognition receptors that recognize specific pathogen-associated molecular patterns. Recognition of C. jejuni specific pathogenassociated molecular patterns by host pathogen recognition receptors stimulates the release of proinflammatory cytokines, causing an influx of fluid and professional phagocytes (Magalhaes et al., 2007; O’Hara and Shanahan, 2006). In addition, resident macrophages and dendritic cells phagocytize bacteria and present antigens to cells involved in acquired humoral immunity. C. jejuni are rapidly killed by complement mediated lysis by both classical and alternative pathways (Blaser et al., 1985). In vitro observations indicate C. jejuni can persist and replicate in both epithelial cells and macrophages, which possibly represents an adaptation by C. jejuni to avoid complement-mediated destruction (De Melo et al., 1989; Kiehlbauch et al., 1985; Konkel et al., 1992; Myszewski and Stern, 1991; Naikare et al., 2006; Wassenaar et al., 1997). Although the breakdown of some innate defenses during the inflammatory response facilitates infection, the self-limiting nature of C. jejuni-mediated disease within several days of the onset of symptoms indicates the pathogen can be effectively cleared by innate processes before full activation of humoral responses (Blaser et al., 1983; Perez-Perez et al., 1989).
CHAPTER 18
CAMPYLOBACTER JEJUNI FACTORS THAT CONTRIBUTE TO DISEASE Diverse Virulence Phenotypes of C. jejuni Strains A simple, cost-effective animal model that accurately mimics Campylobacter infection of humans is not yet widely available to researchers. Therefore, virulence determinants and their phenotypes have been studied primarily by performing in vitro assays to assess bacterial adherence, invasion, protein secretion, intracellular survival, and toxin production. On the basis of these assays, it is clear that C. jejuni strains exhibit significant genotypic and phenotypic variations (Dorrell et al., 2001; Konkel and Joens, 1989; Newel1 et al., 1985b). Comparisons of C. jejuni genomes by sequencing and subtractive hybridization analysis have revealed variation in genomic content among strains, providing a basis for some of the phenotypic variation observed. C. jejuni are naturally competent for DNA transformation, and horizontal gene transfer between strains within a host has been documented (Hepworth et al., 2007). The exchange of DNA between strains within a host organism is widely supported as the mechanism responsible for C. jejuni genome diversity (Manning et al., 2003; McCarthy et al., 2007). However, differences in genomic content do not account for all the phenotypic diversity observed. Different adherence, invasion, and colonization phenotypes have been recorded with genetically matched strains of C. jejuni and subsequently attributed to point mutations in a sensor kinase or response regulator gene (Hendrixson, 2006; Malik-Kale et al., 2007). These findings suggest the ability of C. jejuni to sense the environment and regulate expression of the genome varies between strains and contributes to their pathogenicity. Thus, it is a combination of genomic content and gene regulation that ultimately determines the pathogenicity of C. jejuni. Pathogenic Mechanisms of C. jejuni Motility, adherence, invasion, protein secretion, intracellular survival, and toxin production may contribute to the pathogenicity of a given C. jejuni strain. As these topics and others dealing with C. jejuni virulence factors are covered in detail elsewhere in this book, we present a general overview of motility, adherence, and invasion to set a foundation for an indepth discussion of C. jejuni protein secretion and the possible roles of secreted proteins in the development of gastrointestinal disease. C. jejuni motility is provided by either monotrichous or amphitrichous flagella, and C. jejuni must be flagellated to colonize chickens or cause disease in
C. TETUNI PROTEIN SECRETION
319
humans (Black et al., 1988; Hendrixson, 2006; Newell et al., 1985a). Flagellar locomotion facilitates penetration of the mucosal boundary lining the host intestine. Motility also allows C. jejuni to colonize specific niches within the host where the bacterium can avoid the peristaltic motion and flushing of the intestine. Disrupting flagellar function impairs motility and reduces C. jejuni binding and invasion of host cells (Carrillo et al., 2004; Konkel and Joens, 1989; Konkel et al., 2004, Malik-Kale et al., 2007; Wassenaar et al., 1991). Adherence to intestinal epithelial cells is proposed to be fundamental to C. jejuni colonization and pathogenesis. Proteins proposed to act as C. jejuni adhesins include CadF, CapA, PorA (major outer membrane protein [MOMP]), PEB1, and JlpA. CadF (Campylobacter adhesion to fibronectin) is a 37-kDa outer membraneprotein (Koikel et al., 1997). A C. jejuni cadF mutant shows a reduction in binding and invasion of INT 407 cells when compared with a wild-type isolate (Monteville and Konkel, 2002; Monteville et al., 2003) and is incapable of colonizing chickens (Ziprin et al., 1999). CapA is a putative member of the autotransporter family of exported proteins. Similar to a mutation in cadF, a capA mutant shows reduced adherence to Caco-2 cells and does not colonize chickens (Ashgar et al., 2007). The porA gene of C. jejuni encodes a 43-kDa MOMP, which facilitates transport of hydrophilic molecules across the bacterium’s outer membrane barrier and provides structural stability to the outer membrane (Bolla et al., 1995; De et al., 2000). MOMP purified from outer membrane preparations was shown to bind to INT 407 cells via ligand immunoblot assays and microadhesion enzyme-linked immunosorbent assays (Moser et al., 1997; Schroder and Moser, 1997). C. jejuni porA mutants have yet to be characterized because mutations in porA are lethal to C. jejuni as a result of the proteins’ critical structural and transport functions. PEBl is a 28-kDa protein. Disruption of pebZA reduces c. jejuni adherence to HeLa cells by 50- to 100-fold (Pei and Blaser, 1993; Pei et al., 1998). In addition, a C. jejuni pebZA null mutant exhibits a reduction in the duration of mouse intestinal colonization when compared with the C. jejuni wild-type isolate (Pei et al., 1998). JlpA (jejuni lipoprotein A) is a C. jejuni 43.2-kDa protein: Dis;upcon of j l j A reduces c. jejuni adherence to HEp2 cells by 18 to 19.4% relative to the wild-type strain (Jin et al., 2001). These studies support the hypothesis that adherence mediated by specific adhesins is necessary for colonization of a host. The most extensively characterized adhesin is the outer membrane protein CadF, which mediates binding to the fibronectin component of the host cell extracellular matrix via four amino acid residues-
320
LARSON ET AL.
phenylalanine, arginine, leucine, and serine (FRLS, residues 133 to 137 of the full-length CadF protein) (Konkel et al., 2005). Fibronectin is a ligand for the aspl integrin receptor. Binding of fibronectin to the extracellular domain triggers lateral migration of multiple integrins across the plasma membrane. This integrin clustering results in increased signaling activity and subsequent formation of focal adhesion complexes, which provide a physical link and transmit signals between the extracellular matrix and the actin cytoskeleton. Numerous scaffolding and signaling molecules are associated with focal adhesions that regulate actin polymerization in response to external stimuli (Gilcrease, 2007). CadF binding to fibronectin has been shown to specifically induce phosphorylation of the focal adhesion component paxillin, indicating C. jejuni attachment to fibronectin triggers host cell signal transduction from the extracellular matrix through the aspl integrin receptors to focal adhesion complexes (Hu et al., 2006; Monteville et al., 2003). The ability of pathogens to influence focal adhesion signaling cascades has several putative implications for invasion (Bruce-Staskal et al., 2002; Eto et al., 2007; Kierbel et al., 2007; McCormick et al., 1997; Shi and Casanova, 2006). The binding of CadF may trigger receptor clustering that recruits focal adhesion complexes to the site of bacterial adherence. During this process, host cell signaling molecules would be brought into close proximity to bacteria, increasing the probability that other C. jejuni virulence proteins would bind to host cell factors and promote bacterial internalization. Noteworthy is that GTPase-dependent cell-signaling events, which are necessary for C. jejuni invasion of human cells, were decreased with a cadF mutant relative to the wildtype strain (Krause-Gruszczynska et al., 2007). Although the binding of C. jejuni to a host cell is sufficient to trigger focal adhesion signaling, host cell invasion is not merely a consequence of the interaction between bacterial adhesins and host cell receptors alone; bacteria must be viable and able to secrete proteins for maximal invasion (Konkel et al., 1992, 1999). Many putative pathogenic mechanisms of C. jejuni are derived from processes known to occur during infections of other enteric pathogens. C. jejuni utilizes a T3SS, the flagellum, to secrete proteins that contribute to host cell invasion (Konkel et al., 2004; Rivera-Amill and Konkel, 1999). We speculate that these secreted proteins modulate host cell signaling pathways. Microfilament reorganization is observed during internalization of C. jejuni (De Melo et al., 1989). Inhibition of either host cell actin dynamics, or C. jejuni protein secretion, reduces the invasive potential of C. jejuni (Konkel et al., 2004; Krause-
Gruszczynska et al., 2007; Rivera-Amill and Konkel, 1999). Studies have shown that microtubules also play a role during the internalization of C. jejuni (Biswas et al., 2003; Kopeck0 et al., 2001). In addition to facilitating invasion, other C. jejuni proteins disrupt host cell processes and trigger apoptosis (Poly et al., 2007; Siegesmund et al., 2004), potentially contributing to the degradation of the host intestinal epithelium (Blaser et al., 1983; Everest et al., 1993b).
INVASION AND PROTEIN SECRETION Proteins Required for Host Cell Invasion Our knowledge of C. jejuni host cell interactions indicates that C. jejuni synthesizes and secretes a set of proteins on host cell contact by using the flagellar type I11 secretion system. It has been demonstrated that at least two of these proteins, CjO914c (CiaB) and CjO859c (FspA), are delivered to epithelial cells, indicating that the flagellum of C. jejuni serves the dual function of cell motility and virulence protein secretion. In this section, the observations leading to this conclusion will be reviewed. Invasion Studies To gain a better understanding of the ability of C. jejuni to enter, survive, and replicate in eukaryotic cells, researchers have used the gentamicin protection assay. This in vitro assay involves inoculation of a monolayer of eukaryotic cells with a known number of bacteria, followed by an incubation period to allow the bacteria to bind to and internalize within the eukaryotic cells. After this incubation period, the cell monolayer is rinsed, and medium containing gentamicin, which does not penetrate eukaryotic cell membranes (Hale and Bonventre, 1979), is added to kill the extracellular bacteria. The number of intracellular C. jejuni is determined by dilution plating after lysis of the host cells with a detergent. In vitro tissue culture assays have provided a method to study bacteria-host cell binding and invasion and to characterize specific C. jejuni mutants. As noted in the previous section, a number of genes have been identified, including cadF, capA, porA, p e b l , and jlpA, that encode putative adhesins. As an alternative to the percentage of the inoculum internalized as a measure of a strain’s invasive potential, we report the percent of adherent bacteria that are internalized, as follows: [(number of internalized bacteria divided by the number of adherent bacteria) X 1001. The reason for this distinction is to normalize the effect of variable adherence on a strain’s invasive capacity, because the ability of a C. jejuni strain to bind to a eukaryotic
CHAPTER 18
cell is a prerequisite for host cell invasion. Moreover, ample evidence exists demonstrating that generating a knockout in a C. jejuni gene encoding an adhesin (i.e., CadF, CapA, and PEBla) results in a reduction in host cell adherence, with a corresponding decrease in host cell invasion (Ashgar et al., 2007; Monteville et al., 2003; Pei et al., 1998). Similarly, if a C. jejuni strain is nonmotile, it shows a reduction in binding to host cells relative to its isogenic, motile counterpart. We consider a C. jejuni strain yielding a percentage I/A of >1 as both invasive and pathogenic. The reason for using the percentage of I/A is because strains yielding a value of >1 cause piglets to develop clinical symptoms that resemble those of human campylobacteriosis, including diarrhea with blood in the stool. It is well documented that some bacteria must be metabolically active for maximal cell invasion, as shown for Haernophilus influenzae (St Geme and Falkow, 1990), Neisseria gonorrhoeae (Richardson and Sadoff, 1988), Rickettsia prowazekii (Walker and Winkler, 1978), Salmonella enterica serovar Typhimurium (Finlay et al., 1989; Lee and Falkow, 1990), and Shigella flexneri (Hale and Bonventre, 1979; Headley and Payne, 1990). Early studies on C. jejuni invasion were consistent with this theme (Konkel and Cieplak, 1992). Internalization of C. jejuni is significantly reduced when protein synthesis is inhibited by exposure to chloramphenicol before coculture with host cells (Konkel and Cieplak, 1992; Konkel et al., 1993; Oelschlaeger et al., 1993). Similar results (loss of invasion capacity) are obtained if C. jejuni is heat or sodium azide killed. Together, these results strongly imply that de novo protein synthesis was required on target cell contact. Indeed, examination of this parameter in more detail demonstrated that C. jejuni responds to culture with epithelial cells by synthesizing a novel set of proteins. One- and twodimensional electrophoretic analyses of metabolically labeled C. jejuni cultured with and without epithelial cells revealed that proteins were synthesized either exclusively or preferentially in the presence of epithelial cells, whereas others were selectively repressed (Konkel and Cieplak, 1992; Konkel et al., 1993). Panigrahi et al. (1992) also reported that in rabbit ileal loops, C. jejuni synthesized a number of proteins that were not synthesized under standard laboratory conditions. Two of the newly synthesized proteins, with apparent molecular masses of 84 and 47 kDa, were detectable by using convalescent sera from C. jejuni-infected individuals. Additional work revealed that the de novo synthesized proteins by C. jejuni upon cocultivation with INT 407 cells were unique from those proteins induced by thermal stress of C. jejuni (Konkel et al., 1998). These findings suggest C.
C. IEJUNI PROTEIN SECRETION
321
jejuni responds in a coordinated fashion to the host epithelial cell microenvironment. As noted below, this response includes induction of genes encoding invasion-promoting proteins (Konkel et al., 1992). Link between Cell Invasion and Protein Secretion In an attempt to identify C. jejuni proteins induced by epithelial cell contact, we screened a C. jejuni genomic DNA phage expression library with two antisera (Konkel et al., 1999). One antiserum was collected from a rabbit injected with a whole-cell lysate of C. jejuni cultured with INT 407 epithelial cells, and the other antiserum was collected from a rabbit injected with a whole-cell lysate of C. jejuni cultured in the absence of epithelial cells (i.e., tissue culture medium alone). Phage clones that reacted positively with the antiserum with epithelial cells were then screened with the antiserum that lacked epithelial cells. From this differential screen, one recombinant phage was identified that reacted with the antiserum with epithelial cells but that did not react with the antiserum that lacked epithelial cells. The C. jejuni genomic DNA fragment within this phage clone contained an open reading frame (ORF) predicted to encode a 73-kDa protein. A C. jejuni mutant strain was constructed and demonstrated an invasion-deficient phenotype compared with the isogenic C. jejuni wild-type strain. This gene was designated ciaB (Campylobacterinvasion antigen B). Confocalmicroscopy examination of C. jejuni infected INT 407 cells with an anti-CiaB antibody revealed staining of the host cell cytoplasm, suggesting that the CiaB protein was secreted from the bacterial cell (Konkel et al., 1999). The induction of specific genes upon host cell contact further suggested that C. jejuni may secrete protein as demonstrated for other enteric pathogens. Therefore, screening of cellfree culture supernatants for candidate secreted proteins was conducted. Cells were grown in medium supplemented with [35S]-methionineunder invasionconducive conditions with subsequent removal of cells by filtration and protein concentration of the culture supernatants. The resultant concentrate was screened for radiolabeled proteins by autoradiography. By this method, at least eight proteins, ranging in size from 12.8 to 108 kDa, were originally identified in culture supernatants from C. jejuni cells in contact with INT 407 epithelial cells (Konkel et al., 1999). Protein secretion was not detected when C. jejuni was incubated in the absence of epithelial cells. Modification of the original secretion assay has allowed for greater sensitivity and improved resolution, revealing additional secreted proteins. A profile of C.
322
LARSON ET AL.
jejuni proteins is shown in Fig. 3. The M , of the secreted proteins is provided in Table 1. Further experiments showed that the ciuB mutant was noninvasive and defective in protein secretion, demonstrating a phenotypic link between invasion and protein secretion (Konkel et al., 1999). Both invasion and protein secretion deficiencies of the ciuB mutant are restored by in trans complementation. Finally, by means of a polyclonal rabbit anti-CiaB antibody, it was determined that CiaB is one of the eight proteins originally detected in culture supernatants under invasion-conducive growth conditions. The “By’ designation reflects that CiaB was the second protein in descending molecular weight order detected in the secretion protein gel profile (Konkel et al., 1999).
kDa
I
Table 1. Relative molecular mass of C. jejuni secreted proteins Protein no.
% of total secreted protein
M,
1
0.2 0.7 2.0 5.4 1.2 19.0 1.5 1.6 1.8 2.4 17.2 1.2 12.0 1.5
86.0 72.0 63.0 60.5 51.5 38.5 36.0 32.0 30.0 29.5 27.0 24.0 22.0 18.5 17.0 15.5 13.0 11.5
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
6.4
14.6 7.0 4.3
2
97.4 66.0 -
45.0 -
31 .O -
21.514.4-
Figure 3. Secretion of the Cia proteins is dependent on an intact flagellar T3SS. Culture medium supplemented with fetal bovine serum (FBS) triggers secretion of the Cia proteins. C. jejuni were suspended in minimal essential medium, either with or without FBS, and radiolabeled with [3’S]-methionine for 3 h. Supernatant fluids were harvested, concentrated fourfold, and solubilized in double-strength sample buffer. Equal volumes of samples were separated in a sodium dodecyl sulfate 12.5% polyacrylamide gel. The gel was dried and analyzed by phosphorimaging (Molecular Dynamics, Inc., ImageQuant, Sunnyvale, CA). Lanes: 1, C. jejuni 81176 (+) FBS; 2, C. jejuni 81-176 (-) FBS.
The correlation between the differential protein synthetic response, protein secretion, and host cell invasion is further supported by the finding that there is a temporal association shared among the three responses. C. jejuni demonstrates an increased rate of radioactive methionine incorporation, demonstrates an altered synthetic response, and secretes the Cia proteins immediately before a rapid increase in C. jejuni-host cell internalization (Konkel et al., 1993; Rivera-Amill and Konkel, 1999). We have also noted that inclusion of serum in the growth medium induces Cia secretion and thus could substitute the requirement of C. jejuni-eukaryotic cell cocultivation. It was also determined that inclusion of bile salts promoted Cia expression even though this stimulus did not promote Cia secretion; hence, expression and secretion signals could be uncoupled by varying the constituents of the medium (Rivera-Amill and Konkel, 1999). In summary, Cia protein secretion is dependent on C. jejuni-host cell contact or another biological stimulus (serum). Furthermore, CiaB lacks a cleavable signal peptide leader, suggesting export is accomplished in lieu of the Sec-dependent general secretory pathway. Although - these characteristics are reminisof-either a virulence-associated T ~ S Sor a type IV secretion system, the genomic sequence of C. jeNCTC 11168 did not reveal candidate genes predicted to encode a nonflagellar T3SS or type IV secretion system. The only T3SS encoded by the c. jejuni genome is the flagellar apparatus (parkhill et 2ooo) (htt~”’www~san~er~ac~uk~Pro~ects’C~ jebni). On the basis of these findings, we hypothesized that the CiaB protein was a T3SS substrate and
cent
CHAPTER 18
was exported by the C. jejuni flagellar T3SS. In part, this prediction was based on the precedence that Yersinia enterocolitica was reported to secrete at least one virulence factor, a phospholipase (YplA), from the flagellum. To test this hypothesis, we have followed two basic strategies. First, we reasoned that if the flagellum is required for secretion of Cia proteins, mutations in both flagellar regulatory and structural genes should be nonmotile and should block both Cia export and C. jejuni host cell invasion. Second, because T3SS are generally indiscriminant in recognition and secretion of substrates (Galan and WolfWatz, 2006; Lloyd et al., 2002; Schlumberger and Hardt, 2006; Sorg et al., 2005), we reasoned that if CiaB is a T3SS substrate, it should be recognized and exported by heterologous T3SS such as those harbored by Y. enterocolitica.
C. jejuni Flagellar Mutants
To test the prediction that the flagellar system was used as the export apparatus for virulence proteins, several flagellar gene mutations were constructed to test whether the Cia proteins require a functional flagellum for secretion (Konkel et al., 2004). From our genetic studies, we determined that the secretion of the Cia proteins required a functional basal body, hook, and at least one of the filament proteins. Mutations that affect either the export of flagellar components (flhB), or the nonfilament structural components (flgB, flgC, and flgE2), likewise result in a Cia secretion-negative phenotype. At least a partial filament is also required for Cia secretion. CiaB is secreted in a fZaA mutant, but CiaB secretion is not detected when both filament genes are deleted @a'm mutation). Finally, we recovered C. jejuni strains from poultry that were nonmotile, as they did not synthesize the flagellum, and found that they were Cia secretion negative and poorly invasive for INT 407 cells (Malik-Kale et al., 2007). Together, the genetic evidence obtained to date is consistent with Cia protein secretion through the flagellar export system (Fig. 4). Two observations suggest that the roles of the Cia proteins are distinct from flagellar proteins. First, a C. jejuni ciaB mutant is motile (Konkel et al., 2004). Second, the expression of the cia genes is regulated in a manner distinct from flagellar genes. More specifically, d4is responsible for directing the expression of the flagellar class I1 components that comprise the basal body, hook proteins, and the FlaB filament protein, while a2*is responsible for the expression of the C. jejuni flagellar class I11 genes, which includes the gene encoding the FlaA filament protein. The expression of the C. jejuni ciaB gene
C. TETUNI PROTEIN SECRETION
323
appears to be independent of both the d4and v2' factors as judged by real-time reverse transcriptionPCR (M. E. Konkel et al., unpublished observations). Studies are currently in progress to dissect the regulation of the C. jejuni Cia-encoding genes in relation to flagellar gene regulation. On the basis of these data, it is evident that the regulation of the genes encoding the flagellar and Cia proteins is separate, but that both sets of proteins require a flagellar T3SS for export. We have designated the proteins secreted from the flagellum as the Campylobacter secreted proteins (Csp). As discussed azove, the subset of the c s p proteins that contributes to host cell invasion has been termed the Campylobacter invasion antigens (Cia). To date, the c s p and Cia mostly remah unidentified. The low level of cia gene expression and Cia protein export in vitro has made it difficult to identify invasion-associated proteins. Detection of these proteins requires addition of serum to the culture medium and radiolabeling of cells. The presence of serum proteins in these preparations, combined with low concentration of secreted proteins, has negated traditional proteomic approaches to identify additional Cia proteins. However, the recognition that the flagellar T3SS is required for Cia secretion and host cell invasion has provided the insight for the development of an assay to screen for and identify the Csp. Recognition and Export of CiaB by a Heterologous T3SS Gram-negative bacteria possess at least six different mechanisms to actively transport proteins across the bacterial membranes, one of which is the T3SS (Kostakioti et al., 2005). Although the flagellum is a T3SS, other T3SS have evolved solely to transport bacterial effector proteins from the bacterial cytoplasm into the host cell cytosol via a specialized conduit comprising a basal body and a translocon. Secretion of effector proteins is triggered by bacterial contact with the host cell. Although the general structure of the T3SS apparatus is similar among the gramnegative pathogens that harbor these systems, the biological function of the secreted proteins (effector molecules) varies. Requirements of T3SS protein substrates include (i) the absence of a Sec-dependent signal sequence, (ii) an amino terminal signal that facilitates secretion, and (iii) export through a specialized conduit spanning the inner and outer membranes of gram-negative bacteria (Cornelis, 2006). Several animal and plant pathogens have evolved unique strategies that alter recipient host cell signaling pathways with effector proteins. Ultimately, the effector proteins modulate the host to facilitate intracellular
324
LARSON ET AL..
FliD
Pseudaminic Acid
- FIaNFlaB (Filament)
OM 1
Peptidog lycan
FlgB,C,F,G
CM
FliF
1
FliM I
FIIH, I, 0, P, Q, R Secretion Apparatus (T3SS) Figure 4. C. jejuni type I11 secretion system (T3SS) is the flagellum that secretes the Cumpylobucter secreted proteins (Csp). A subset of the Csps, termed the Curnpylobucter invasion antigens (Cia), are required for maximal invasion. Both the Csp and Cia proteins harbor nonconsensus secretion signals, which are required for export.
survival, bacterial multiplication, and/ or immune evasion (Journet et al., 2005). Throughout the remainder of the text, when necessary for clarity, we refer to the flagellar T3SS as “flagellar T3SS” and the classical T3SS that is dedicated to the secretion of proteins as “T3SS.”
Yersinia enterocolitica is an intensely studied gastrointestinal pathogen that harbors three T3SS. The three Y. enterocolitica protein export pathways are termed the flagellar, Ysa, and Ysc T3SS, which respond to known environmental stimuli of temperature and salt concentration (Young and Young,
CHAPTER 18 ~~
C. TETUNI PROTEIN SECRETION
325
~~
2002). The flagellar outer proteins (or Fops) are secreted by the flagellar T3SS; the Yersinia secreted proteins (or Ysps) are secreted by the Ysa T3SS, and the Yersinia outer proteins (or Yops) are secreted by the Ysc T3SS. The Ysc system is encoded on the p W plasmid, whereas the former two systems are chromosomally encoded. The Ysc system is induced at 37°C in LB base medium supplemented with 0.2 M sodium oxalate (high salt, high temperature); the Ysa system is induced at 26°C in LB base medium supplemented with 0.29 M NaCl (high salt, low temperature), and the flagellar system is induced at 26°C in base medium without supplemented NaCl (low salt, low temperature). Both the Ysa and Ysc T3SS are capable of secreting the YopE, YopN, and YopP effector proteins, indicating the promiscuous nature of the export pathways (Lee and Galan, 2004; Matsumoto and Young, 2006; Young and Young, 2002). In C. jejuni, the flagellum is required for motility and secretion of the Csp. On the basis of the hypothesis that CiaB is a T3SS protein, we reasoned that it would be recognized in a heterologous system and secreted in a T3SS-dependent manner. To test this hypothesis, the full-length ciaB gene was cloned into the inducible expression vector pMMB207 and transformed into Y. enterocolitica wild-type and flagellar mutant isolates. After induction of the Y. enterocolitica flagellar T3SS, supernatant proteins were harvested and probed with polyclonal rabbit anti-CiaB serum. The 73-kDa CiaB protein was detected in the supernatant fluids of the Y. enterocolitica wild-type isolate, whereas the supernatant fluids of the Y. enterocolitica flhDC flagellar mutant lacked a band of equivalent mass (S. A. Pacheco and M. E. Konkel, unpublished data). When the whole-cell lysates of both the wild-type and flagellar mutant harboring a copy of the ciaB gene were probed with the polyclonal rabbit anti-CiaB serum, a band at 73 kDa was observed in both isolates. This demonstrated that the CiaB protein was synthesized in the Y. enterocolitica flagellar mutant cytosol but was not secreted. Silverstained Fop profiles revealed that only the Y. enterocolitica wild-type isolate secretes the Fop proteins. The Y. enterocolitica flagellar mutant cannot secrete the Fop proteins. This result indicated that the CiaB protein was secreted along with the Fops in a T3SSdependent manner. It also suggested that the CiaB protein harbors a T3SS signal that is recognized for export via the flagellar pathway. This latter point was verified by a new assay (described below), whereby the amino terminus of ciaB was found to harbor a T3SS signal. In summary, two lines of evidence support the hypothesis that the C. jejuni flagellar T3SS is used to secrete virulence proteins in the host environment.
First, a functional flagellum is a prerequisite for Cia protein export. This assertion is based on extensive genetic studies showing that flagellar regulatory and structural gene mutations abolish both motility and Cia export but not cia expression. Second, the Nterminal coding sequence of CiaB contains a T3SS export signal that is recognized by the Y. enterocolitica flagellar T3SS. Identification of C. jejuni-Secreted Proteins To facilitate the identification of C. jejuni genes that harbor a T3SS signal, the pCSP50 shuttle vector was generated (Fig. 5). The pCSP50 shuttle vector incorporates a constitutive promoter (cat) upstream of cloning sites for the encoded signal sequences, the 150-nt amino terminus truncated yplA gene (eliminating the first 50 amino acids including the T3SS signal), and the yplB chaperone gene. An aminoterminal deletion of the Y. enterocolitica YplA enzyme abolishes its secretion. However, if a T3SS signal is fused to the truncated yplA gene, YplA is secreted and detected on phospholipase indicator plates. More specifically, a fatty acid precipitate, which results from cleavage of Tween 80, can be visualized as a halo surrounding the YplA secretion competent colonies. The pCSP50 vector with C. jejuni gene fragments were used to transform Y. enterocolitica yplAB (phospholipase A and the YplB chaperone) mutants also lacking the p W plasmid. By means of this approach, we detected secretion and extracellular phospholipase activity by the ciaB-yplA fusion construct expressed in a yplA chromosomal deletion background. This finding further indicated that the ciaB 5’ end (N terminus) encodes a T3SS signal. In contrast, fusing the N-terminal coding region of C. jejuni genes encoding cytosolic proteins with yplA did not promote YpU export. We then selected genes to screen for T3SS signal sequences from the C. jejuni NCTC 11168 sequence database (Parkhill et al., 2000). A total of 359 of 1,625 ORFs was chosen for analysis after the elimination of genes encoding proteins with known functions or containing membrane-spanning domains, periplasmic domains, Sec-dependent signal sequences, or Tat-dependent signal sequences. No genes were identified with known type I Secindependent motifs. Not surprisingly, most of the 359 ORFs encoded proteins of unknown function and conserved hypothetical proteins. Primers were designed to amplify the first 108 encoding bases of all the ORFs and to facilitate directional cloning into the shuttle vector pCSP50 to generate translational fusions with the truncated YplA. Thus far, 329 of the 359 ORFs have been analyzed. Each vector was se-
326
LARSON ET AL.
m
Ndel-Bglll site for cloning of fragments for fusion with yplA
A
B
C
D
E
Figure 5. Depiction of the pCSP50 shuttle vector and phospholipase indicator plate results for controls. The pCSP50 vector includes a tet cassette, a constitutive promoter (cat), a 5' truncated yplA (lacking 150 nt encoding the native T3SS signal), and the yplB gene (cognate chaperone). The NdeI and BglII sites facilitate directional cloning of C. jejuni sequences as fusions with the truncated yplA. (A to E) Scans of phospholipase indicator plates under flagellar T3SS induction. (A) Y. enterocolitica JB580v (wild type); (B) Y. enterocolitica yplAB mutant; (C) Y. enterocolitica yplAB mutant with pCSP50; (D) Y. enterocolitica yplAB mutant with pCSP5O:ciaB1_,,,; (E) Y. enterocolitica yplAB mutant with pCSP5O:cysM1_,,,. All strains show strong growth; (A) and (D) show strong secretion of YplA and CiaB,-,,:YplA fusion protein, respectively, resulting in a zone of precipitate due to hydrolysis of fatty acids by the phospholipase.
quence confirmed and used to transform the Y. enterocolitica yplAB (phospholipase A and the YplB chaperone) mutant also lacking the p W plasmid. The transformants were then screened for both flagellar and Ysa T3SS secretion by spot inoculation onto phospholipase indicator plates (low and high salt, respectively). A total of 42 of 329 C. jejuni ORFs were identified that strongly drive the export of the C. jejuniYplA fusion proteins via the Y. enterocolitica flagellar and Ysa T3SS (J. E. Christensen and M. E. Konkel, unpublished data). Included within this list are signal sequences from two known C. jejuni-secreted virulence factors, CjO914c (CiaB) and CjO859c (FspA). In addition, the signal sequences from two putative flagellar related proteins, Cj1463 (FlgJ) and Cj1464 (FlgM), resulted in strong secretion from both T3SS. The functional categories (Gundogdu et al., 2007) of 26 of the strongly secreted proteins are either conserved hypothetical proteins or proteins of unknown
function (Fig. 6). From the remaining list of 16 proteins, all but 4 are annotated as putative functions.
C. jejuni-Secreted Proteins (CiaB, FlaA, FlaB, FlaC, and FspA) We have found that CiaB is a T3SS protein and hypothesize that the remaining Cia proteins are also T3SS substrates. Proteins secreted via this pathway will fall into two general categories: proteins that comprise the flagellum, and proteins that are not part of the flagellar structural apparatus, some of which serve as effector proteins. Evidence indicates that the Cia proteins enhance C. jejuni invasion of host cells, and we speculate that this occurs through their modification of host cell behavior. FlaA, FlaB, FlaC, and FspA were also found to harbor a T3SS signal, as evidenced by our Y. enterocolitica phospholipase screen (not shown). Song et
CHAPTER 18
2
C. lElUNI PROTEIN SECRETION
327
I Conserved hypothetical proteins Unknown function Miscellaneous Broad regulatory functions Small molecule metabolism
8
Cell processes
0Macromolecule metabolism 12 Figure 6 . Functional categories of C. jejuni secreted proteins. Shown are the numbers of C. jejuni proteins, categorized according to function, encoding signal sequences that elicited strong secretion from both the Y. enterocolitica flagellar and Ysa T3SS.
al. (2004) reported that FlaC (CjO72Oc), a 26-kDa protein, shares N and C terminus homology to the flagellar FlaA and FlaB filament proteins in the C. jejuni TGH9011 strain. This group also detected the FlaC protein in the supernatant milieu of isolates harboring an intact flagellum. Studies revealed that the flaC mutant was less invasive for INT 407 cells compared with the wild-type isolate, suggesting that the FlaC protein may play a role in C. jejuni pathogenesis. Poly et al. (2007) determined that the C. jejuni FspA gene encoded a 15.5-kDa protein designated FspA, for Flagellar secreted protein A. Similar to FlaC, FspAwas onl; detected-in supernatant fluids harvested from isolates harboring a functional flagellum. Additionally, Poly et al. (2007) found two variant forms of the FspA protein (A1 and A2) within a variety of C. jejuni strains, including C. jejuni 8486 and C. jejuni 81-176. Only FspA2 was associated with the host cell monolayer and induced apoptosis. The FspA protein does not appear to play a role in cell adherence and invasion, as differences were not observed between the wild-type strain and a fspAl mutant. Interestingly, the FlaC protein and two FspA protein variants did not require an external stimulus for protein secretion (i.e., host contact, or coculture with conditioned medium) and were not required for
motility (Poly et al., 2007). Collectively, the data further support our hypothesis that the C. jejuni flagellum is essential for motility and also facilitates the secretion of nonflagellar proteins. Evolutionary Relatedness of the Flagellum and Virulence T3SS Pathways The bacterial flagellum and T3SS injectisome share an evolutionary and functional relationship (Gophna et al., 2003). The first indication of this relationship was the predicted protein similarities between the Caulobacter crescentus flagellar gene flbB (now flhA) and the Y. enterocolitica virulence plasmid-encoded lcrD (Ramakrishnan et al., 199 1; Sanders et al., 1992). After these two reports, additional flagellar protein and virulence protein similarities were found. Because Y. enterocolitica phase varied flagellin and virulence protein secretion, and because both systems were Sec-independent processes, Minnich and Rohde (2007) postulated that the simplest explanation was that the Yersinia basal body might be reorganized for virulence protein secretion in the host environment (Harshey and Toguchi, 1996). As such, this broadened the traditional view of the flagellum as a motility organelle to one that
328
LARSON ET AL.
could also be viewed as a stand-alone, highly efficient protein secretory device (i.e., depending on the circumstances, the basal body could serve a dual purpose). Studies performed in the late 1990s demonstrated for Yersinia and other enteric pathogens that the flagellar T3SS and the classical T3SS, which is dedicated to the secretion of virulence proteins, were separate but parallel systems. As described in the preceding section, C. jejuni is a unique example in which the flagellum is the sole channel to export flagellar components and virulence proteins. Despite the varying roles T3SS serve for pathogens, the mechanism of assembly and protein secretion and the apparatus itself are remarkably similar to the flagellar basal body. About 25 proteins are required to build the nonflagellar T3SS. Eight of these proteins are conserved with proteins that comprise the flagellum (Cornelis, 2006; Galan and Wolf-Watz, 2006). Nonflagellar T3SS harbor basal body structures that serve to anchor the apparatus into the bacterial inner and outer membranes and peptidoglycan layer. Basal body assembly must occur before the integration of the inner rod, needle, and regulatory proteins and involves the incorporation of an ATPase that is involved in the recognition and unfolding of T3SS apparatus substrates. Once these structures have been inserted, conformational changes on the cytoplasmic side of the bacterial membrane allow for protein selection and secretion. Flagellar T3SS are also composed of a similar basal body structure that serves to anchor the flagellar engine to the bacterial membrane (Macnab, 2004). There has been speculation about the evolutionary phylogeny of the flagellar and T3SS systems. Saier (2004) proposed three possibilities, which include the following: (i) both the T3SS and flagellum are derived from a common ancestral protein secretory system; (ii) the T3SS is the precursor for the more complex flagellum, or (iii) the flagellum served as the precursor for the T3SS. Pallen and Matzke (2006) have argued for the first possibility. Gophna et al. (2003) argued for the second possibility, in part because Chlamydia, which is deeply rooted in the phylogenetic tree, contains only a virulence T3SS. Saier favors the third possibility because all groups of eubacteria contain flagella, but the T3SS are thus far limited to a small subset of gram-negative bacteria. Further, Saier (2004) points out that bacterial motility more than likely predated the appearance of eukaryotes and the resulting opportunity for symbiosis or parasitism. Finally, mutation densities in flagellar proteins are much greater than those associated with T3SS, suggesting that the latter are evolutionarily more recent. Our findings demonstrate that the C. jejuni flagellum can provide motility and secretes virulence
proteins. Because of this observation, postulating a common ancestral system giving rise to both flagellar T3SS and classical T3SS may not be necessary. In fact, the evidence from C. jejuni argumentatively favors the flagellum being the ancestral organelle from which the classical T3SS developed. The fact that the CiaB flagellar-dependent T3SS signal of C. jejuni (Epsilonproteobacteria) is recognized by the Yersinia (Gammaproteobacteria) flagellar T3SS and classical T3SS highlights a remarkable conservation. The evidence that the T3SS is the ancestral apparatus inferred from observations with Chlamydia is not disputed. However, given that Chlamydia spp. are obligate intracellular pathogens and that the transition to such a specialized lifestyle often involves a significant loss of genetic information, one could speculate that these organisms may have lost the locomotion functionality of a flagellum while retaining the ability to secrete proteins to facilitate their intracellular survival. Such examples of system losses have been observed for nonmotile Yersinia pestis, Shigella spp., and Bordetella pertussis and recent isolates of nonmotile Escherichia coli 0157, all of which require a T3SS injectisome for virulence and demonstrate clear evidence of mutationally inactivated flagellar systems. In fact, expressing flagella in the host may be a liability for pathogens because flagellin is a potent inducer of innate immunity via Toll-like receptors specifically positioned to recognize these specific pathogen-associated molecular patterns. In summary, our findings on the dual nature of the C. jejuni flagellar T3SS add an interesting twist to the possible origin of these two systems. Motility, Cia Protein Secretion, Host Cell Invasion, and C. jejuni-Mediated Enteritis The role of motility in C. jejuni colonization and subsequent disease production has been intensely studied. Predictably, motility was found to be important in promoting the colonization of animals by C. jejuni (Nachamkin et al., 1993; Pavlovskis et al., 1991). Also expected was the finding that motility, and the expression of the fZaA gene, is necessary for the maximal invasion of C. jejuni into host mammalian cells and for the translocation of polarized cell monolayers (Grant et al., 1993; Wassenaar et al., 1991). However, when these early studies were performed, it was observed that the C. jejuni fZaA @&+) strain is more invasive than a C. jejuni fZaA flaB strain (Konkel et al., 2004). Also interesting was that the invasiveness of a C. jejuni fZaA NUB+)strain was enhanced 10-fold by promoting bacteria-host cell contact via centrifugation. In contrast, the centrifugation step did not change the invasive potential of the C.
CHAPTER 18
jejuni wild-type strain (Wassenaar et al., 1994). On the basis of these findings, Grant et al. (1993) concluded that the flagellar structure played a role in internalization that was independent of motility. Subsequently, we found that the Cia proteins were secreted from a C. jejuni 81116 flaA @aB+) mutant but not from a C. jejuni 81116 flaA fzdB mutant (Grant et al., 1993). Thus, it was clear that a C. jejuni nonmotile strain can be either secretion positive (i.e., the fluA fluB+ mutant) or secretion negative (i.e., the fluA flaB mutant), and the ability of the bacterium to secrete proteins resulted in an increase in its invasive potential. Although these data helped clarify the relationship between C. jejuni motility, secretion, and host cell invasion, the significance of protein secretion and host cell invasion in C. jejuni-mediated gastroenteritis was not known. This question was addressed directly with a C. jejuni ciaB mutant, which is motile but secretion negative. Importantly, in vitro assays revealed that the C. jejuni F38011 wild-type strain (motile and secretion positive) was SO-fold more invasive than the C. jejuni ciaB mutant (motile and secretion negative). Moreover, inoculation of piglets with the C. jejuni wild-type and complemented ciaB strains resulted in diarrhea 24 h after infection, whereas diarrhea was not observed in piglets infected with the C. jejuni ciaB mutant until 3 days after infection (Konkel et al., 2001). More severe histological lesions (i.e., shortening of the villi and the production of an exudate in the lumen of the intestine) were also observed in piglets inoculated with the C. jejuni-complemented ciuB isolate when compared with the C. jejuni ciuB mutant. Although the undifferentiated cells in the crypt remained normal in appearance, C. jejuni destroyed the fully differentiated epithelial cells located at the tips of the villi. Collectively, the infection studies performed with the newborn piglets and the C. jejuni Cia secretioncompetent and secretion-deficient isolates revealed that host cell invasion and the secreted Cia proteins are major contributors to the pathology of C. jejunimediated enteritis.
SUMMARY C. jejuni is an interesting pathogen that has evolved a unique set of virulence mechanisms to cause disease. On the basis of the work described herein, we submit that C. jejuni utilizes its flagellar T3SS for the secretion of virulence proteins that contribute to C. jejuni-mediated enteritis. To our knowledge, this is first example where the flagellum functions in both motility and as the primary conduit for
C. TEIUNI PROTEIN SECRETION
329
virulence protein secretion. Our current focus is to identify the secreted proteins and to gain an understanding of how these proteins modulate host cell functions. Further, C. jejuni strains show significant genetic and phenotypic diversity. How this diversity contributes to the spectrum of disease profiles in susceptible hosts is of keen interest. The question of why C. jejuni elicits disease in humans but establishes a commensal relationship with other animals is critical in understanding C. jejuni pathogenesis. The dynamic interplay of C. jejuni virulence determinants and host responses is necessary to understand why disease develops in susceptible hosts. A more accurate and comprehensive understanding of C. jejuni-mediated enteritis will emerge as researchers unravel the virulence attributes unique to particular C. jejuni strains. Acknowledgments. We thank Phil Mixter (School of Molecular Biosciences, Washington State University) for critical review. Work in M.E.K.’s laboratory is funded with federal funds from NIAID, NIH, and DHHS, under contract N01-AI-30055, by the USDA National Research Initiative’s Food Safety 32.0 program (200635201-17305), and USDA NRI through the Food Safety Research Response Network (2005-35212-15287).
REFERENCES Andersen, M. T., L. Brondsted, B. M. Pearson, F. Mulholland, M. Parker, C. Pin, J. M. Wells, and H. Ingmer. 2005. Diverse roles for HspR in Campylobacter jejuni revealed by the proteome, transcriptome and phenotypic characterization of an hspR mutant. Microbiology 15 1:905-9 15. Ashgar, S. S., N. J. Oldfield, K. G . Wooldridge, M. A. Jones, G . J. Irving, D. P. Turner, and D. A. Ala’Aldeen. 2007. CapA, an autotransporter protein of Campylobacter jejuni, mediates association with human epithelial cells and colonization of the chicken gut. J. Bacteriol. 189:1856-1865. Biswas, D., K. Itoh, and C. Sasakawa. 2003. Role of microfilaments and microtubules in the invasion of INT-407 cells by Campylobacter jejuni. Microbiol. Immunol. 47:469-473, Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Blaser, M. J., P. F. Smith, and P. F. Kohler. 1985. Susceptibility of Carnpylobacter isolates to the bactericidal activity of human serum. J. Infect. Dis. 151:227-235. Blaser, M. J., J. G. Wells, R. A. Feldman, R. A. Pollard, and J. R. Allen. 1983. Campylobacter enteritis in the United States. A multicenter study. Ann. Intern. Med. 98:360-365. Blikslager, A. T., A. J. Moeser, J. L. Gookin, S. L. Jones, and J. Odle. 2007. Restoration of barrier function in injured intestinal mucosa. Physiol. Rev. 87545-564. Bolla, J. M., E. Loret, M. Zalewski, and J. M. Pages. 1995. Conformational analysis of the Campylobacter jejuni porin. J. Bacteriol. 177:4266-4271. Bruce-Staskal, P. J., C. L. Weidow, J. J. Gibson, and A. H. Bouton. 2002. Cas, Fak and Pyk2 function in diverse signaling cascades to promote Yersinia uptake. J. Cell Sci. 115:2689-2700. Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R. Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K. Amoako, S. Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Genome-wide ex-
330
LARSON ET AL.
pression analyses of Campylobacter jejuni NC TC lll68 reveals coordinate regulation of motility and virulence by flhA. J. Biol. Chem. 279:20327-20338. Chen, M. L., Z. Ge, J. G. Fox, and D. B. Schauer. 2006. Disruption of tight junctions and induction of proinflammatory cytokine responses in colonic epithelial cells by Campylobacter jejuni. Infect. Immun. 74:6581-6589. Cornelis, G. R. 2006. The type I11 secretion injectisome. Nut. Rev. Microbiol. 4:8 11-825. De, E., M. Jullien, G. Labesse, J. M. Pages, G. Molle, and J. M. Bolla. 2000. MOMP (major outer membrane protein) of Campylobacter jejuni; a versatile pore-forming protein. FEBS Lett. 469~93-97. De Melo, M. A., G. Gabbiani, and J. C. Pechere. 1989. Cellular events and intracellular survival of Campylobacter jejuni during infection of HEp-2 cells. Infect. Immun. 57:2214-2222. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. Al-Ghusein, B. G . Barrell, J. Parkhill, N. G. Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Campylobacter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11:1706-1715. Eto, D. S., T. A. Jones, J. L. Sundsbak, and M. A. Mulvey. 2007. Integrin-mediated host cell invasion by type 1-piliated uropathogenic Escherichia coli. PLoS Pathog. 3:elOO. Everest, P. H., A. T. Cole, C. J. Hawkey, S. Knutton, H. Goossens, J. P. Butzler, J. M. Ketley, and P. H. Williams. 1993a. Roles of leukotriene B4, prostaglandin E2, and cyclic AMP in Campylobacter jejuni-induced intestinal fluid secretion. Infect. Immun. 61:4885-4887. Everest, P. H., H. Goossens, P. Sibbons, D. R. Lloyd, S. Knutton, R. Leece, J. M. Ketley, and P. H. Williams. 1993b. Pathological changes in the rabbit ileal loop model caused by Campylobacter jejuni from human colitis. J. Med. Microbiol. 38:316-321. Finlay, B. B., F. Heffron, and S. Falkow. 1989. Epithelial cell surfaces induce Salmonella proteins required for bacterial adherence and invasion. Science 243:940-943. Friedman, C. R., R. M. Hoekstra, M. Samuel, R. Marcus, J. Bender, B. Shiferaw, S. Reddy, S. D. Ahuja, D. L. Helfrick, F. Hardnett, M. Carter, B. Anderson, and R. V. Tauxe. 2004. Risk factors for sporadic Campylobacter infection in the United States: a case-control study in FoodNet sites. Clin. Infect. Dis. 38(S~ppl.3):S285-S296. Galan, J. E., and H. Wolf-Watz. 2006. Protein delivery into eukaryotic cells by type I11 secretion machines. Nature 444567573. Gilcrease, M. Z. 2007. Integrin signaling in epithelial cells. Cancer Lett. 247:l-25. Gillespie, I. A., J. O'Brien S, J. A. Frost, C. Tam, D. Tompkins, K. R. Neal, Q. Syed, and M. J. Farthing. 2006. Investigating vomiting and/or bloody diarrhoea in Campylobucter jejuni infection. J. Med. Microbiol. 55:741-746. Gophna, U., E. Z. Ron, and D. Graur. 2003. Bacterial type I11 secretion systems are ancient and evolved by multiple horizontaltransfer events. Gene 312:151-163. Grant, C. C., M. E. Konkel, W. Cieplak, Jr., and L. S. Tompkins. 1993. Role of flagella in adherence, internalization, and translocation of Campylobacter jejuni in nonpolarized and polarized epithelial cell cultures. Infect. Immun. 61:1764-1771. Guerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Pattarini, G. Majam, P. Thibault, and S. Logan. 2006. Changes in flagellin glycosylation affect Campylobacter autoagglutination and virulence. Mol. Microbiol. 60:299-3 11. Gundogdu, O., S. D. Bentley, M. T. Holden, J. Parkhill, N. Dorrell, and B. W. Wren. 2007. Re-annotation and re-analysis of
the Campylobacter jejuni N C T C l l l 6 8 genome sequence. BMC Genomics 8:162. Hale, T. L., and P. F. Bonventre. 1979. Shigella infection of Henle intestinal epithelial cells: role of the bacterium. Infect. Immun. 24~879-886. Harshey, R. M., and A. Toguchi. 1996. Spinning tails: homologies among bacterial flagellar systems. Trends Microbiol. 4:226-23 1. Headley, V. L., and S. M. Payne. 1990. Differential protein expression by Shigellu flexneri in intracellular and extracellular environments. Proc. Natl. Acad. Sci. USA 87:4179-4183. Hendrixson, D. R. 2006. A phase-variable mechanism controlling the Campylobacter jejuni FlgR response regulator influences commensalism. Mol. Microbiol. 61:1646-1659. Hepworth, P. J., H. Leatherbarrow, C. A. Hart, and C. Winstanley. 2007. Use of suppression subtractive hybridisation to extend our knowledge of genome diversity in Campylobacter jejuni. BMC Genomics 8:llO. Hu, L., J. P. McDaniel, and D. J. Kopecko. 2006. Signal transduction events involved in human epithelial cell invasion by Campylobacter jejuni 81-176. Microb. Pathog. 40:91-100. Jeon, B., K. Itoh, N. Misawa, and S. Ryu. 2003. Effects of quorum sensing on flaA transcription and autoagglutination in Campylobacter jejuni. Microbiol. Immunol. 473333439. Jin, S., A. Joe, J. Lynett, E. K. Hani, P. Sherman, and V. L. Chan. 2001. JlpA, a novel surface-exposed lipoprotein specific to Campylobacter jejuni, mediates adherence to host epithelial cells. Mol. Microbiol. 39: 1225-1236. Journet, L., K. T. Hughes, and G. R. Cornelis. 2005. Type 111 secretion: a secretory pathway serving both motility and virulence. Mol. Membr. Biol. 22:41-50. (Review.) Kiehlbauch, J. A., R. A. Albach, L. L. Baum, and K. P. Chang. 1985. Phagocytosis of Campylobacter jejuni and its intracellular survival in mononuclear phagocytes. Infect. Immun. 48:446451. Kierbel, A., A. Gassama-Diagne, C. Rocha, L. Radoshevich, J. 01son, K. Mostov, and J. Engel. 2007. Pseudomonas aeruginosa exploits a PIP3-dependent pathway to transform apical into basolateral membrane. J. Cell Biol. 177:21-27. Konkel, M. E., J. E. Christensen, A. M. Keech, M. R. Monteville, J. D. Klena, and S. G. Garvis. 2005. Identification of a fibronectin-binding domain within the Campylobacter jejuni CadF protein. Mol. Microbiol. 57:1022-1035. Konkel, M. E., and W. Cieplak, Jr. 1992. Altered synthetic response of Campylobacter jejuni to cocultivation with human epithelial cells is associated with enhanced internalization. Infect. Immun. 60:4945-4949. Konkel, M. E., S. G. Garvis, S. L. Tipton, D. E. Anderson, Jr., and W. Cieplak, Jr. 1997. Identification and molecular cloning of a gene encoding a fibronectin-binding protein (CadF) from Campylobacter jejuni. Mol. Microbiol. 24:953-963. Konkel, M. E., S. F. Hayes, L. A. Joens, and W. Cieplak, Jr. 1992. Characteristics of the internalization and intracellular survival of Campylobacter jejuni in human epithelial cell cultures. Microb. Pathog. 13:357-370. Konkel, M. E., and L. A. Joens. 1989. Adhesion to and invasion of HEp-2 cells by Campylobacter spp. Infect. Immun. 57:29842990. Konkel, M. E., B. J. Kim, J. D. Klena, C. R. Young, and R. Ziprin. 1998. Characterization of the thermal stress response of Campylobacter jejuni. Infect. Immun. 66:3666-3672. Konkel, M. E., B. J. Kim, V. Rivera-Amill, and S. G. Garvis. 1999. Bacterial secreted proteins are required for the internaliztion of Campylobacter jejuni into cultured mammalian cells. Mol. MiCrObiOl. 32:691-701. Konkel, M. E., J. D. Klena, V. Rivera-Amill, M. R. Monteville, D. Biswas, B. Raphael; and J. Mickelson. 2004. Secretion of viru-
CHAPTER 18
lence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. J. Bacteriol. 186:32963303. Konkel, M. E., D. J. Mead, and W. Cieplak, Jr. 1993. Kinetic and antigenic characterization of altered protein synthesis by Campylobacter jejuni during cultivation with human epithelial cells. J. Infect. Dis. 168:948-954. Konkel, M. E., M. R. Monteville, V. Rivera-Amill, and L. A. Joens. 2001. The pathogenesis of Campylobacter jejuni-mediated enteritis. Curr. Issues Intest. Microbiol. 255-71. Kopecko, D. J., L. Hu, and K. J. Zaal. 2001. Campylobacter jejuni-microtubule-dependent invasion. Trends Microbiol. 9:3 89396. Kostakioti, M., C. L. Newman, D. G. Thanassi, and C. Stathopoulos. 2005. Mechanisms of protein export across the bacterial outer membrane. J. Bacteriol. 187:4306-4314. Krause-Gruszczynska, M., M. Rohde, R. Hartig, H. Genth, G. Schmidt, T. Keo, W. Konig, W. G. Miller, M. E. Konkel, and S. Backert. 2007. Role of the small Rho GTPases Racl and Cdc42 in host cell invasion of Campylobacter jejuni. Cell. Microbiol. 9:243 1-2444. Lee, C. A., and S. Falkow. 1990. The ability of Salmonella to enter mammalian cells is affected by bacterial growth state. Proc. Natl. Acad. Sci. USA 87:4304-4308. Lee, S. H., and J. E. Galan. 2004. Salmonella type I11 secretionassociated chaperones confer secretion-pathway specificity. Mol. Microbiol. 5 1:483-495. Lin, J., C. Cagliero, B. Guo, Y. W. Barton, M. C. Maurel, S. Payot, and Q. Zhang. 2005. Bile salts modulate expression of the CmeABC multidrug efflux pump in Campylobacter jejuni. J. Bacteriol. 187:74 17-7424. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Chemother. 46:2124-2131. Lloyd, S. A., M. Sjosaom, S. Anderson, and H. Wolf-Watz. 2002. Molecular characterization of type 111 secretion signals via analysis of synthetic N-terminal amino acid sequences. Mol. Microbiol. 4351-59. MacCallum, A., S. P. Hardy, and P. H. Everest. 2005. Campylobacter jejuni inhibits the absorptive transport functions of Caco2 cells and disrupts cellular tight junctions. Microbiology 151: 245 1-2458. Macnab, R. M. 2004. Type I11 flagellar protein export and flagellar assembly. Biochim. Biophys. Acta 1694:207-217. Magalhaes, J. G., I. Tattoli, and S. E. Girardin. 2007. The intestinal epithelial barrier: how to distinguish between the microbial flora and pathogens. Semin. Immunol. 19:106-115. Malik-Kale, P., B. H. Raphael, C. T. Parker, L. A. Joens, J. D. Klena, B. Quinones, A. M. Keech, and M. E. Konkel. 2007. Characterization of genetically matched isolates of Campylobacter jejuni reveals that mutations in genes involved in flagellar biosynthesis alter the organism’s virulence potential. Appl. Environ. Microbiol. 73:3123-3136. Manning, G., C. G. Dowson, M. C. Bagnall, I. H. Ahmed, M. West, and D. G. Newell. 2003. Multilocus sequence typing for comparison of veterinary and human isolates of Campylobacter jejuni. Appl. Environ. Microbiol. 69:6370-6379. Matsumoto, H., and G. M. Young. 2006. Proteomic and functional analysis of the suite of Ysp proteins exported by the Ysa type 111 secretion system of Yersinia enterocolitica biovar IB. Mol. Microbiol. 59:689-706. McCarthy, N. D., F. M. Colles, K. E. Dingle, M. C. Bagnall, G. Manning, M. C. Maiden, and D. Falush. 2007. Host-associated genetic import in Campylobacter jejuni. Emerg. Infect. Dis. 13: 267-272.
C. 7E7uNI PROTEIN SECRETION
331
McCormick, B. A., A. Nusrat, C. A. Parkos, L. D’Andrea, P. M. Hofman, D. Carnes, T. W. Liang, and J. L. Madara. 1997. Unmasking of intestinal epithelial lateral membrane beta 1 integrin consequent to transepithelial neutrophil migration in vitro facilitates inu-mediated invasion by Yersinia pseudotuberculosis. Infect. Immun. 65:1414-1421. Minnich, S. A., and H. N. Rohde. 2007. A rationale for repression and/or loss of motility by pathogenic Yersinia in the mammalian host. Adu. Exp. Med. Biol. 603:298-310. Monteville, M. R., and M. E. Konkel. 2002. Fibronectin-facilitated invasion of T84 eukaryotic cells by Campylobacter jejuni occurs preferentially at the basolateral cell surface. Infect. Immun. 70: 6665-6671. Monteville, M. R., J. E. Yoon, and M. E. Konkel. 2003. Maximal adherence and invasion of INT 407 cells by Campylobacter jejuni requires the CadF outer-membrane protein and microfilament reorganization. Microbiology 149: 153-165. Moser, I., W. Schroeder, and J. Salnikow. 1997. Campylobacter jejuni major outer membrane protein and a 59-kDa protein are involved in binding to fibronectin and INT 407 cell membranes. FEMS Microbiol. Lett. 157:233-238. Myszewski, M. A., and N. J. Stern. 1991. Phagocytosis and intracellular killing of Campylobacter jejuni by elicited chicken peritoneal macrophages. Avian Dis. 35:750-755. Nachamkin, I., X. H. Yang, and N. J. Stern. 1993. Role of Campylobacter jejuni flagella as colonization factors for three-dayold chicks: analysis with flagellar mutants. Appl. Environ. Microbiol. 59:1269-1273. Naikare, H., K. Palyada, R. Panciera, D. Marlow, and A. Stintzi. 2006. Major role for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infect. Immun. 745433-5444. Newell, D. G., H. McBride, and J. M. Dolby. 1985a. Investigations on the role of flagella in the colonization of infant mice with Campylobacter jejuni and attachment of Campylobacter jejuni to human epithelial cell lines. J. Hyg. (Lond.) 95:217-227. Newell, D. G., H. McBride, F. Saunders, Y. Dehele, and A. D. Pearson. 198513. The virulence of clinical and environmental isolates of Campylobacter jejuni. J. Hyg. (Lond.) 94:45-54. Oelschlaeger, T. A., P. Guerry, and D. J. Kopecko. 1993. Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proc. Natl. Acad. Sci. USA 90:6884-6888. O’Hara, A. M., and F. Shanahan. 2006. The gut flora as a forgotten organ. EMBO Rep. 7:688-693. Pallen, M. J., and N. J. Matzke. 2006. From The Origin of Species to the origin of bacterial flagella. Nut. Rev. Microbiol. 4:784790. Panigrahi, P., G. Losonsky, L. J. DeTolla, and J. G. Morris, Jr. 1992. Human immune response to Campylobacter jejuni proteins expressed in vivo. Infect. Immun. 60:4938-4944. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pavlovskis, 0. R., D. M. Rollins, R. L. Haberberger, Jr., A. E. Green, L. Habash, S. Strocko, and R. I. Walker. 1991. Significance of flagella in colonization resistance of rabbits immunized with Campylobacter spp. Infect. Immun. 59:2259-2264. Pei, Z., and M. J. Blaser. 1993. PEB1, the major cell-binding factor of Campylobacter jejuni, is a homolog of the binding component in gram-negative nutrient transport systems. J. Biol. Chem. 268: 18717-1 8725.
332
LARSON ET AL.
Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. Z. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. lmmun. 66:938-943. Perdomo, J. J., P. Gounon, and P. J. Sansonetti. 1994a. Polymorphonuclear leukocyte transmigration promotes invasion of colonic epithelial monolayer by Shigella flexneri. J. Clin. Invest. 93: 633-643. Perdomo, 0.J., J. M. Cavaillon, M. Huerre, H. Ohayon, P. Gounon, and P. J. Sansonetti. 1994b. Acute inflammation causes epithelial invasion and mucosal destruction in experimental shigellosis. J. Exp. Med. 180:1307-1319. Perez-Perez, G. I., D. L. Cohn, R. L. Guerrant, C. M. Patton, L. B. Reller, and M. J. Blaser. 1989. Clinical and immunologic significance of cholera-like toxin and cytotoxin production by Campylobacter species in patients with acute inflammatory diarrhea in the USA. J. Infect. Dis. 160:460-468. Poly, F., C. Ewing, S. Goon, T. E. Hickey, D. Rockabrand, G. Majam, L. Lee, J. Phan, N. J. Savarino, and P. Guerry. 2007. Heterogeneity of a Carnpylobacter jejuni protein that is secreted through the flagellar filament. Infect. lmmun. 75:3859-3867. Ramakrishnan, G., J. L. Zhao, and A. Newton. 1991. The cell cycle-regulated flagellar gene flbF of Caulobacter crescentus is homologous to a virulence locus (IcrD) of Yersinia pestis. J. Bacteriol. 173:7283-7292. Ramos, H. C., M. Rumbo, and J. C. Sirard. 2004. Bacterial flagellins: mediators of pathogenicity and host immune responses in mucosa. Trends Microbiol. 12509-517. Raphael, B. H., S. Pereira, G. A. Flom, Q. Zhang, J. M. Ketley, and M. E. Konkel. 2005. The Campylobacter jejuni response regulator, CbrR, modulates sodium deoxycholate resistance and chicken colonization. J. Bacteriol. 187:3662-3670. Richardson, W. P., and J. C. Sadoff. 1988. Induced engulfment of Neisseria gonowhoeae by tissue culture cells. Infect. lmmun. 56: 2512-2514. Rivera-Amill, V., and M. E. Konkel. 1999. Secretion of Campylobacter jejuni Cia proteins is contact dependent. Adv. Exp. Med. Biol. 473:225-229. Saier, M. H., Jr. 2004. Evolution of bacterial type I11 protein secretion systems. Trends Microbiol. 12:113-115. Sanders, L. A., S. Van Way, and D. A. Mullin. 1992. Characterization of the Caulobacter crescentus flbF promoter and identification of the inferred FlbF product as a homolog of the LcrD protein from a Yersinia enterocolitica virulence plasmid. J. Bacteriol. 174: 857-8 66. Sansonetti, P. J., J. Arondel, M. Huerre, A. Harada, and K. Matsushima. 1999. Interleukin-8 controls bacterial transepithelial translocation at the cost of epithelial destruction in experimental shigellosis. Infect. Immun. 67:1471-1480. Schlumberger, M. C., and W. D. Hardt. 2006. Salmonella type I11 secretion effectors: pulling the host cell's strings. Cur. Opin. Microbiol. 9:46-54. Schroder, W., and I. Moser. 1997. Primary structure analysis and adhesion studies on the major outer membrane protein of Campylobacter jejuni. FEMS Microbiol. Lett. 150:141-147. Schwerer, B. 2002. Antibodies against gangliosides: a link between preceding infection and immunopathogenesis of Guillain-Barre syndrome. Microbes Infect. 4:373-384.
Shi, J., and J. E. Casanova. 2006. Invasion of host cells by Salmonella typhimurium requires focal adhesion kinase and pl30Cas. Mol. Biol. Cell 17:4698-4708. Siegesmund, A. M., M. E. Konkel, J. D. Klena, and P. F. Mixter. 2004. Campylobacter jejuni infection of differentiated THP-1 macrophages results in interleukin 1 beta release and caspase-lindependent apoptosis. Microbiology 150:561-569. Song, Y. C., S. Jin, H. Louie, D. Ng, R Lau, Y. Zhang, R. Weerasekera, S. A1 Rashid, L. A. Ward, S. D. Der, and V. L. Chan. 2004. FlaC, a protein of Campylobacter jejuni TGH9011 (ATCC43431) secreted through the flagellar apparatus, binds epithelial cells and influences cell invasion. Mol. Microbiol. 53: 541-553. Sorg, J. A., N. C. Miller, and 0. Schneewind. 2005. Substrate recognition of type I11 secretion machines-testing the RNA signal hypothesis. Cell. Microbiol. 7: 1217-1225. St Geme, J. W., 111, and S. Falkow. 1990. Haemophilus influenzae adheres to and enters cultured human epithelial cells. Infect. lmmun. 5 8 :403 6-4 044. Szymanski, C. M., S. M. Logan, D. Linton, and B. W. Wren. 2003. Campylobacter-a tale of two protein glycosylation systems. Trends Microbiol. 11:233-238. Szymanski, C. M., and B. W. Wren. 2005. Protein glycosylation in bacterial mucosal pathogens. Nut. Rev. Microbiol. 3:225-237. Thornley, J. P., D. Jenkins, K. Neal, T. Wright, J. Brough, and R. C. Spiller. 2001. Relationship of Campylobacter toxigenicity in vitro to the development of postinfectious irritable bowel syndrome. J. Infect. Dis. 184:606-609. Walker, R. I., E. A. Schmauder-Chock, J. L. Parker, and D. Burr. 1988. Selective association and transport of Campylobacter jejuni through M cells of rabbit Peyer's patches. Can. J. Microbiol. 34:1142-1147. Walker, T. S., and H. H. Winkler. 1978. Penetration of cultured mouse fibroblasts (L cells) by Rickettsia prowazeki. Infect. Immun. 22:200-208. Wassenaar, T. M., N. M. Bleumink-Pluym, D. G. Newell, P. J. Nuijten, and B. A. van der Zeijst. 1994. Differential flagellin expression in a flaA P U B f mutant of Campylobacter jejuni. lnfect. lmmun. 62:3901-3906. Wassenaar, T. M., N. M. Bleumink-Pluym, and B. A. van der Zeijst. 1991. Inactivation of Campylobacter jejuni flagellin genes by homologous recombination demonstrates that flaA but not fluB is required for invasion. EMBO J. 10:2055-2061. Wassenaar, T. M., M. Engelskirchen, S. Park, and A. Lastovica. 1997. Differential uptake and killing potential of Campylobacter jejuni by human peripheral monocytes/macrophages. Med. Microbiol. lmmunol. 186:139-144. Yao, R., D. H. Burr, and P. Guerry. 1997. CheY-mediated modulation of Campylobacter jejuni virulence. Mol. Microbiol. 23: 1021-1031. Young, B. M., and G. M. Young. 2002. YpL4 is exported by the Ysc, Ysa, and flagellar type I11 secretion systems of Yersinia enterocolitica. J. Bacteriol. 184: 1324-1334. Ziprin, R. L., C. R. Young, L. H. Stanker, M. E. Hume, and M. E. Konkel. 1999. The absence of cecal colonization of chicks by a mutant of Campylobucter jejuni not expressing bacterial fibronectin-binding protein. Avian Dis. 43586-589.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 19
Innate Immunity in Campylobacter Infections NICOLEM. IOVINE
Immunocompetent adults who are naike to Campylobacter jejuni usually resolve the infection without systemic spread and before the adaptive immune response is mounted (Blaser and Allos, ZOOS), indicating a key role of the innate immune system in clearing the infection. Even in the unusual case of bacteremic C. jejuni infection, patients can recover completely in about a week without antimicrobial therapy (Shandera et al., 1992). Despite the central role of innate immunity in clearance of Cumpylobacter, less is known about the interaction of many innate immune components with Cumpylobacter compared with other enteric organisms. In this chapter, the current knowledge pertaining to innate immunity and Cumpylobacter will be addressed, and important areas of future research will be highlighted. Reflecting the epidemiology of C. jejuni as the most common Cumpylobacter species implicated in human disease (Blaser and Allos, ZOOS), the preponderance of the literature reviewed herein describes host-pathogen interactions with this species. However, published reports describing innate immune interactions with other Campylobacter species will be included when possible.
ponents of this system before the mobilization of specific immune mechanisms, innate immunity constitutes a crucial first line of defense against potential pathogens. The rapid mobilization of innate immune defenses is of particular importance in the nayve host in whom specific immunity has not yet developed. The innate immune system alone may not allow a microorganism to take hold and cause an infection or may limit its growth and spread sufficiently long enough to allow the development of specific immunity. An important strategy of innate immune defense involves recognition of conserved components of microorganisms, called pathogen-associated molecular patterns, by host pattern recognition molecules (PFWs) (Albiger et al., 2007). By this means, a limited number of PRMs can rapidly engage a broad array of potential pathogens. For example, recognition of the pathogen-associated molecular pattern lipopolysaccharide (LPS) by the PRM Toll-like receptor 4 enables the early detection of essentially all gramnegative bacteria (Fitzgerald et al., 2004; Poltorak et al., 1998). With a better understanding of the intricacies of immunity comes the realization that our host defenses cannot be strictly divided into innate and adaptive systems. However, in this chapter, we will adhere to the definition that innate immunity is defined as those host defense mechanisms that do not require previous exposure to a microorganism. Such mechanisms relevant to protection against a gastrointestinal infection such as Campylobacter therefore include the multiple cellular and soluble factors listed in Table 1. Although the interplay between many of these components and Cumpylobacter are not fully understood, this chapter will review the current knowledge
OVERVIEW OF IMMUNE SYSTEM COMPONENTS RELEVANT T O INNATE DEFENSE AGAINST GASTROINTESTINAL INFECTION The innate immune system can be broadly defined as those responses that require neither the specific prior induction characteristic of acquired immunity nor the lag time associated with its induction. Because microorganisms are likely to encounter com-
-
~~
Nicole M. Iovine Department of Medicine, New York University School of Medicine, Veteran’s Administration Medical Center, 423 East 23rd St., Room 6005 West, New York, NY 10010.
333
IOVINE
334
Table 1. Immune components relevant to innate defense against gastrointestinal infections ~~~
~~~
~
Anatomic site
Innate immune component
Oral cavity
Salivary nitrate
Breast
Fucosylated sugars in milk
Stomach Gallbladder
Low pH and nitric oxide Bile acids
Small intestine
Intact epithelium Defensins and cathelicidins Toll-like receptors
NOD proteins
Large intestine
Blood/Tissue: Soluble components
Cellular and cell-derived components
Normal biota
Reference(s)
Property Antimicrobial; substrate for production of gastric nitric oxide Chemical barrier interfering with Campylobacter binding to intestinal cells Antimicrobial Antimicrobial via a detergent-like effect of amphipathic molecules Physical barrier that restricts microbes to outside the host Antimicrobial, cationic peptides Intra- and extracellular recognition of pathogen-associated molecular patterns Intracellular recognition of pathogenassociated molecular patterns (peptidoglycan) Occupies a niche otherwise available to pathogens; source of natural antimicrobials
Acute-phase proteins and comp1ement
Antimicrobial activity and opsonization
Collectins and mannosebinding lectin Iron sequestration
Recognition of pathogen-associated molecular patterns Chemical barrier that decreases availability of key microbial nutrient and signaling molecule Autocrine/paracrine proinflammatory signals Antigen processing Phagocytic; production of molecular defenses (ROS/RNS) and cytokines Phagocytic; production of ROS and antimicrobial peptides/proteins Antimicrobial
Cytokines and chemokines Dendritic cells Macrophages Neutrophils ROS / RNS
and outline areas in need of further study. Taking an anatomic approach, we will first discuss the innate immune components present in the oral cavity and upper gastrointestinal tract (including breast milk) that are relevant to defense against Campylobacter, followed by the contribution of bile and the lower gastrointestinal tract, and finally the systemic defenses that come into play when Campylobacter organisms break through these defenses. In many cases, the individual components of innate immunity are present in multiple sites within the body; for example, cationic antimicrobial peptides are produced by many cells throughout the alimentary tract as well as by professional phagocytes in the tissues and circulation (Hancock and Diamond, 2000). To avoid redundancy, each component will be discussed once in the location where it is likely to have the greatest role in defense against Campylobacter.
Duncan et al. (1995) Newburg (2005)
Muller et al. (2005) Reynoso-Paz et al. (1999) Muller et al. (2005) Muller et al. (2005) Akira et al. (2006), Leaver et al. (2007) Bourhis and Werts (2007), Fritz et al. (2006), Murray (2005) Abrams and Bishop (1966)
Diffenbach and Tramont (2005), Walport (2001a, 2001b) Takahashi and Ezekowitz (2005) Barclay (1985), Pickett (1992) Iwasaki (2007), Johanesen and Dwinell (2006), Muller et al. (2005) Taylor et al. (2005) Haslett et al. (1989) Nathan and Shiloh (2000)
INNATE IMMUNE DEFENSES PRESENT IN THE GASTROINTESTINAL TRACT Gastric Acid Gastric acidity has long been known to provide defense against potential bacterial enteric pathogens such as nontyphoidal salmonellae and Vibrio cholerae (Giannella et al., 1972; Martinsen et al., 2005). Several lines of evidence suggest that the low-pH environment of the stomach similarly provides a means of host defense against Campylobacter infection. For example, in vitro studies have documented the acid sensitivity of C. jejuni (Rotimi et al., 1990; Waterman and Small, 1998). Also, a case-control study of 21 1 patients currently receiving a potent medication to inhibit gastric acid secretion (omeprazole) documented a nearly 12-fold increased risk of developing Campylobacter infection compared with patients no
CHAPTER 19
longer receiving this drug (Neal et al., 1996). A lesser but similar effect was noted in a smaller study as well (Garcia Rodriguez and Ruigomez, 1997). Finally, patients with an impaired ability to secrete gastric acid were found to be at increased risk for Campylobacter infection (Black et al., 1988). All these studies suggest that intact production of gastric acid is an innate immune defense mechanism against Campylobacter infection. Gastric Reactive Nitrogen Species Initially recognized as a vasodilator, the antimicrobial property of nitric oxide ( N O ) is well established (Fang, 1997; Nathan, 1997). The interaction of NO' with other molecules such as reactive oxygen species (ROS) yields a family of antimicrobial molecules collectively referred to as reactive nitrogen species (RNS) (Fang, 2004; Nathan and Shiloh, 2000). N O can be produced enzymatically by the activity of three differentially expressed nitric oxide synthases (Knowles and Moncada, 1994), or nonenzymatically from nitrite under acidic conditions (Zweier et al.,
Nitrate present in foods...
INNATE IMMUNITY
1999). In the stomach, N O is produced from dietary nitrate via enterosalivary circulation (Duncan et al., 1995) (Fig. 1). As the amount of dietary nitrate increases, serum levels also increase, as do levels of salivary nitrite (Spiegelhalder et al., 1976). Therefore, it has also been suggested that the enterosalivary production of N O might constitute an element of innate host defense against ingested bacteria (Benjamin et al., 1994; Dykhuizen et al., 1996, 1998; McKnight et al., 1999; Xu et al., 2001). That C. jejuni is susceptible to NO'/RNS generated enzymatically by macrophages (Iovine et al., 2008) supports the notion that nonenzymatic production of gastric NO'/RNS might also contribute to host defense against Campylo bacter. Breast Milk Oligosaccharides Human milk contains a complex array of compounds that assist in protecting the infant from enteric diseases (Newburg, 2005; Newburg et al., 2004). Its first opportunity to interact with Campylobacter is in the stomach of the nursing infant. Fu-
and circulated
Salivary glands
absorbed,
concentrate nitrate,
and bacteria on
low nH m-----paqtrir . .. I--eInvironment.
335
NO,
+ -?.--L-
iuiraie
NO,
reductase Figure 1. Production of nitric oxide from dietary nitrate. Nitrate in foods such as meat is ingested, is absorbed in the intestine, and enters the systemic circulation. The salivary glands concentrate the nitrate from blood and secrete it into the saliva. On the tongue, particularly in the area of the posterior papillae, facultative microbes that express nitrate reductase convert nitrate to nitrite. This nitrite is swallowed and interacts with hydrogen ions in the stomach to produce nitric oxide. Although dietary nitrate may be directly reduced on the tongue to form nitrite, the high nitrate concentration in saliva is the major source of the nitrite that enters the stomach.
336
IOVINE
cosylated oligosaccharides in breast milk might bind to Campylobacter in the infant stomach, thereby preventing bacterial binding to intestinal H ( 0 ) antigen, also a fucosylated oligosaccharide (Bereswill and Kist, 2003). This hypothetical mechanism is supported by a study that used Chinese hamster ovary (CHO) cells, which do not bind Campylobacter but did so avidly after transfection with the human a1,2-fucosyltransferase gene. Similarly, preincubation of either the transfected CHO cells or C. jejuni itself with an a1,2linked sugar abrogated bacterial binding to CHO cells. Suckling mice of dams transgenic for human al,2-fucosyltransferase completely cleared an orally delivered C. jejuni clinical isolate under conditions that permitted colonization of 70% of control mice (Ruiz-Palacios et al., 2003). Surveillance data from the Foodborne Diseases Active Surveillance Network (FoodNet) of the CDC showed that Campylobacter infection was less likely to occur in breast-fed infants up to 6 months of age (Fullerton et al., 2007). Furthermore, in a human case-control study, breastfeeding appeared to have a protective effect against Campylobacter infection (Megraud et al., 1990). In summary, breast-fed infants appear to derive protection against Campylobacter infection, possibly via maternal production of alY2-linkedsugars. Intestinal Defenses The diverse defense mechanisms of intestinal epithelium as well as other innate components not produced by intestinal cells but that are active in the intestine (such as bile) comprise the major barrier against enteric infections and are shown schematically in Fig. 2. The cells of the intestinal epithelium, which are the first to have prolonged contact with Campylobacter, are well positioned to detect and respond to Campylobacter- and pathogen-associated molecular patterns. These cells express a variety of PRMs, including Toll-like receptors and NOD (nucleotide-binding oligomerization domain) proteins involved in cellular responses culminating in the production of proinflammatory cytokines, chemokines, and antimicrobial peptides. Bile Acids Bile acids are amphipathic molecules that possess the antimicrobial properties of detergents; therefore, tolerance to bile is required for bacteria to survive transit through the gastrointestinal tract (ReynosoPaz et al., 1999). In general, Campylobacter is considered bile resistant; C. jejuni NCTC 11168 has been shown to survive in 5% ox bile (Fox et al., 2007). A bioinformatic study of the effects of bile on
Campylobacter revealed a complicated response comprising proteins homologous to those found in other bile-tolerant microorganisms, as well as several unique to C. jejuni (Okoli et al., 2007). Specifically, the multidrug efflux pump CmeABC plays a key role in mediating bile tolerance (Lin et al., 2002, 2003) and is induced on exposure to bile salts (Lin et al., 2005). The CbrR gene product also confers resistance to the bile salt deoxycholate and permits colonization of strain F38011 in chickens (Raphael et al., 2005). In C. coli, ox bile stimulated the expression of the Campylobacter invasion antigen (CiaB) (Kovach et al., 2007), which has been linked to colonization and virulence (Konkel et al., 1999; Ziprin et al., 2001). Bile salts also triggered CiaB expression in C. jejuni (Rivera-Amill et al., 2001). This complex response suggests that bile may protect the host from Campylobacter species that do not express these genes. Conversely, one might predict that bile-resistant Campylobacter would be a cause of cholecystitis (gallbladder infection). However, there is no strong evidence that this is the case, probably because antibiotics are usually provided to such patients before surgery, and the special conditions required to support Campylobacter growth are not routinely used with such specimens (Dakdouki et al., 2003). Mucins Intestinal mucus, long recognized as a chemoattractant to C. jejuni (Hugdahl et al., 1988), also promotes the adhesion and internalization of C. jejuni (de Melo and Pechere, 1988) and C. upsaliensis (Sylvester et al., 1996) in tissue culture cells. Two wellcharacterized constituents of intestinal mucus are the transmembrane mucin-1 (MUC1) and secretory mucin-2 (MUC2) (Dekker et al., 2002). MUC2 is the major secretory mucin in the intestine and may account for the chemoattractant property of intestinal mucus (Mendz et al., 2007). MUC2 also may trigger the upregulation of genes association with virulence and invasion, including the cytolethal distending toxin and CiaB (Mendz et al., 2007). In contrast, the transmembrane mucin MUC 1 may oppose infection with C. jejuni; gastrointestinal expression of MUC 1 increased in mice after oral challenge with C. jejuni (McAuley et al., 2007). Bacteria were found in the spleen, liver, and lungs of most Mucl-’- mice (strain 129/SvJ) but never in the wild-type mice, indicating that MUCl contributes to innate defense against C. jejuni infection in mice (McAuley et al., 2007). Other studies found that culture medium containing mucus produced by HT-29-MTX intestinal cells expressing high MUC1, -3, and -5 levels but low levels of MUC2
CHAPTER19
Systemic Circulation
INNATEIMMUNITY
337
-
Complement and other Acute Phase Proteins
Figure 2. Innate defenses active in the intestinal tract. Defenses present in the small and large intestine are depicted schematically. Particularly in the lumen of the large intestine, the normal biota occupies a niche that otherwise might be available to Cumpylobucter. Mucous layer mucins may exert an antimicrobial effect. Coating of Curnpylobucter with breast milk fucosylated sugars may impede interaction of Cumpylobucter with H(0) antigens on intestinal epithelium. Bile acids exert an antimicrobial, detergent-like effect. Dendritic cells (DC) send cellular extensions into the intestinal lumen to sample Cumpylobucter antigens, leading to DC activation. Cumpylobucter organisms that evade these mechanisms are challenged by defenses present in the epithelium itself. Toll-like receptor 4 (TLR4) senses Campylobucter LPS and triggers an NF-KB-dependent cascade, resulting in production of proinflammatory molecules such as IL-8. Upon epithelial cell invasion, interaction of Cumpylobucter peptidoglycan with NOD 1 augments induction of P-defensins with known antimicrobial activity against Cumpylobucter. Cumpylobacter organisms that survive these defenses may enter the submucosa, where phagocytes recruited and activated by IL-8 produce potent molecular defenses: NOS2-derived nitric oxide (NO) and other reactive nitrogen species principally from macrophages (Me) and NADPH oxidase (NADPH ox)-derived superoxide (02-) and other ROS principally from polymorphonuclear neutrophils (PMN). The latter also produce cationic antimicrobial peptides and proteins (CAPPs) including defensins, and may exert Curnpylobucter killing after phagocytosis or in the extracellular space after degranulation. Similarly, highly diffusible N O may effect killing outside of the Me. Finally, Cumpylobucter organisms that enter the systemic circulation are faced with the potent antimicrobial activity of acute-phase proteins, as well as complement, in addition to circulating PMNs.
(Lesuffleur et al., 1993) supported growth of C. jejuni better than culture medium alone (Stanley et al., 2007). Therefore, the ultimate effect of human mucin on Campylobacter infection remains unclear. Interleukin-8 (IL-8) IL-8 is a prototypic chemokine produced by myeloid and epithelial cells (Baggiolini et al., 1999). The importance of IL-8 in host defense derives from its potent chemoattractant properties for lymphocytes
and neutrophils and its ability to activate neutrophils. IL-8 production triggered by enteric pathogens including Salmonella and Shigella requires signaling through Nf-KB (Hobbie et al., 1997; Philpott et al., 2000), as does Campylobacter (Mellits et al., 2002). Evidence for the role of Nf-KB in innate defense against Campylobacter is suggested by a study of Nf&-deficient mice (strain 3X) that were orally infected with C. jejuni 81-176 and developed gastritis and duodenitis, and that were persistently colonized, in contrast to wild-type mice (Fox et al., 2004).
338
IOVINE
Campylobacter elicits IL-8 production from intestinal epithelial cells as well as myeloid and other cell types (Abram et al., 2000; Al-Salloom et al., 2003; Bakhiet et al., 2004; Chen et al., 2006; Hickey et al., 1999; Hu and Hickey, 2005; Jones et al., 2003; MacCallum et al., 2006; Mellits et al., 2002; Watson and Galan, 2005). Although there is no animal model for IL-8 deficiency, the outcome of oral Campylobacter inoculation in mice unable to respond to IL-8 due to an IL-8 receptor deficiency (Broxmeyer et al., 1996) should be studied. Defensins Defensins, a major family of antimicrobial peptides produced by neutrophils and mammalian epithelia, are defined as small, cysteine-containing cationic peptides with three disulfide bonds (Lehrer et al., 1993). The three types of defensins--a (aka murine cryptdin), p, and 0-all exhibit antimicrobial activity against an array of bacteria, protozoa, viruses, and fungi, probably by disturbing membrane integrity (Lehrer et al., 1993). Human a defensin is constitutively expressed and stored in secretory granules of Paneth cells of the small intestine, which also secrete p defensins, lysozyme, and type I1 phospholipase A (Lehrer et al., 1993). Human p defensin 1 (HBD1) is constitutively expressed in cells throughout the gastrointestinal tract (Lehrer et al., 1993), whereas expression of the inducible p defensins 2, 3, and 4 is triggered by proinflammatory cytokines acting through Nf-KB (O’Neil et al., 1999). Theta defensins, which are primarily involved in defense against viral pathogens (Nguyen et al., 2003), will not be discussed further. Defensins are believed to reach levels up to millimolar concentrations under inflammatory conditions (Ganz, 2003). Enteric gram-negative organisms including Salmonella and Escherichia coli are susceptible to the antimicrobial effects of defensins and also induce their expression (Ogushi et al., 2001; Schlee et al., 2007; Takahashi et al., 2001; Wehkamp et al., 2004). Similarly, C. jejuni 81-176 and 11168 induced expression of HBD2 and HBD3 in the human intestinal epithelial cell lines CaCo2 and HT-29 but did not affect expression of the constitutively expressed HBDl (Zilbauer et al., 2005). Strain 81-176 was a more potent inducer of HBD2 and HBD3, but recombinant HBD2 and HBD3 were bactericidal against both C. jejuni strains (Zilbauer et al., 2005). Defensins contribute to innate defense against Campylobacter infection in poultry. Chickens do not show signs of illness despite high levels of intestinal colonization (up to 108 organisms/g) (Wempe et al., 1983), suggesting important differences between in-
nate defenses of poultry and humans, possibly involving the potency and/or level of expression of avian defensins. For example, the avian defensins, called gallinacins, demonstrate bactericidal activity approximately 10-fold more potent compared with human defensin against E. coli (Campylobacter was not tested) (Harwig et al., 1994). In another study, three gallinacins from chicken and turkey heterophils (neutrophil-like avian phagocytes) effected 2-log killing of C. jejuni 81-176 in vitro (Evans et al., 1995). Levels of gallinacin expression also contribute to the resistance of chickens to Campylobacter infection. In the chicken proximal digestive tract, the fi defensin gallinacin-6 is highly expressed constitutively, and analysis of the region upstream of the Gal6 gene revealed potential transcription factor binding sites associated with inflammation, including Nf-KB and AP-1 (van Dijk et al., 2007). In this study, recombinant gallinacin-6 also effected a 3-log kill of C. jejuni (van Dijk et al., 2007). The role of other intestinal epithelia-derived cationic antimicrobial peptides such as cathelicidins (Wehkamp et al., 2007) and proteins such as the bactericidal/permeability-increasing protein (Levy et al., 2003) in defense against Campylobacter have not been studied.
NOD proteins NOD proteins, cytoplasmic PRMs for peptidoglycan, are expressed primarily in antigen-presenting cells and epithelia, including intestinal epithelium (Strober et al., 2006; Uehara et al., 2007). Both N O D l and NOD2 bind peptidoglycan from gramnegative bacteria (Strober et al., 2006). That NODl serves as an intracellular PRM for C. jejuni in intestinal epithelial cells was shown by siRNA knockdown of NOD1, which led to a 6- to 12-fold increased survival of C. jejuni compared with untreated or control siRNA-treated Caco-2 cells (Zilbauer et al., 2007). Cells treated with NOD 1-specific siRNA showed decreased expression of HBD2, known to be bactericidal against C. jejuni (Zilbauer et al., 2005). Therefore, knockdown of N O D l was linked to a biologically relevant outcome via decreased HBD2 expression. Also, NODl expression remained unchanged after exposure to C. jejuni, indicating its inability to modulate NODl expression to its advantage (Zilbauer et al., 2007). Finally, these studies did not support a role of NOD2 as a PRM for C. jejuni in Caco-2 cells (Zilbauer et al., 2007), although a role for NOD2 in other cell types cannot be excluded. The existence of NOD1- and NOD2-deficient mice unresponsive to their respective peptidoglycan ligands provides excellent tools for the study of cel-
CHAPTER 19
INNATE IMMUNITY
339
lular and in vivo responses to Campylobacter (Chamaillard et al., 2003; Pauleau and Murray, 2003).
in C. jejuni flagellin may allow it to circumvent the role of TLR5 in host defense.
Toll-like receptors
RNS
Toll receptors were first described in Drosophila from a genetic screen to identify developmental mutants, and serendipitously they were found to play a role in resistance to fungi (Lemaitre et al., 1996). Since then, 10 human Toll-like receptors (TLRs) have been described, most of which signal through the adaptor protein MyD88 (Rock et al., 1998). Although the cognate ligands for all TLRs have not been identified (Medzhitov and Janeway, 2000), LPS serves as the ligand for TLR4 (Hoshino et al., 1999; Poltorak et al., 1998). The delivery of LPS to TLR4 triggers a signaling cascade involving Nf-KB, ultimately leading to the production of proinflammatory cytokines including tumor necrosis factor (TNF)-a, IL-6, and IL-1 (Kopp and Medzhitov, 2003). In humans, two polymorphisms have been identified in the extracellular domain of TLR4 (Asp299Gly and Thr399Ile) that lead to an attenuated response to LPS, decreased cytokine production, and an increased risk of gram-negative infection (Arbour et al., 2000; Kuhns et al., 1997). Although no increase in C. jejuni infections was detected in a cohort of patients expressing these polymorphisms, the high prevalence of Campylobacter seropositivity in normal hosts may have precluded finding such a difference (Geleijns et al., 2004). TLR4 is expressed on myeloid cells and a variety of human intestinal epithelial cell lines (Fitzgerald et al., 2004; Uehara et al., 2007). Animal studies suggest a role for TLR pathways in innate defense against Campylobacter infection. Whereas mice are usually refractory to Campylobacter infection (Newell, 2001), 80% of MyD88-deficient mice, unable to elicit TLR-dependent proinflammatory signaling, continued to shed C. jejuni 81-176 one week after both oral and intraperitoneal injection compared with 30% of wild-type mice (Watson et al., 2007). Bacterial flagellin has been identified as the ligand which triggers signaling via TLR5 (Hayashi et al., 2001). Specific, conserved residues as positions 89 through 96 in bacterial flagellin are required for TLR5-depending signaling (Andersen-Nissen et al., 2005). TLRS is expressed in intestinal epithelium but only on the basolateral side, providing a mechanism by which invasive but not commensal bacteria can be recognized by the host (Gewirtz et al., 2001). However, C. jejuni and Helicobacter use alternative amino acids at these positions, rendering Campylobacter flagellin unrecognizable by TLR5 (Andersen-Nissen et al., 2005; Watson and Galan, 2005). This alteration
Inducible nitric oxide synthase, also known as NOS2, catalyzes the production of NO/RNS and is expressed in macrophages (Fang, 1997; Nathan, 1997) as well as a variety of intestinal cell lines and in primary human intestinal tissue (Inaba et al., 1999; Krieglstein et al., 2001; Resta-Lenert and Barrett, 2002; Salzman et al., 1996). NOS2 expression can be induced by various gram-negative organisms, including Helicobacter pylori, E. coli, and Salmonella (Kim et al., 2002; Roberts et al., 2001; Salzman et al., 1998; Witthoft et al., 1998). Evidence for an antiCampylobacter role of macrophage-derived N O / RNS will be discussed later in this chapter. Normal Biota of Large Intestine The importance of the normal intestinal biota in innate defense against potential enteric pathogens has long been recognized. It presumably occurs via the occupation of a niche that would otherwise be available to a potential pathogen (Abrams and Bishop, 1966). Prior antibiotic use, which perturbs the normal intestinal biota, has not been identified as a risk factor for Campylobacter infection, but when C. jejuni infection does occur after a course of antibiotics, it is more likely to be drug resistant (Gupta et al., 2004; Smith et al., 1999). It remains plausible, however, that the presence of the normal biota provides some protection against Carnpylobacter infection. At least in mice that are relatively resistant to Campylobacter colonization, those with a limited gut flora could be efficiently colonized by C. jejuni 81-176 compared with mice with normal flora (Chang and Miller, 2006), indicating that the normal biota occupies a niche that might otherwise be available to C. jejuni. Summary of Gastrointestinal Defenses All the above innate immune defenses found in the gastrointestinal tract appear extremely effective in limiting C. jejuni to the gut. Campylobacter organisms that evade these multiple host defense mechanisms by means of enhanced bacterial virulence, depressed host defenses, or a combination of both, may enter the deeper tissues of the intestine and gain systemic access.
340
IOVINE
INNATE IMMUNE DEFENSES PRESENT IN THE INTESTINAL SUBMUCOSA AND SYSTEMIC CIRCULATION Phagocytes
A robust inflammatory response characterized by the influx of neutrophils and mononuclear cells/macrophages accompanies human Campylobacter infection (Lambert et al., 1979; van Spreeuwel et al., 1985). The direct antimicrobial activities of these phagocytes is attributable to production of antimicrobial peptides/proteins, ROSY and RNS (the latter mainly by mononuclear phagocytes and/or macrophages) (Fang, 1997; Tosi, 2005). The ability of phagocytes to ingest and kill Campylobacter varies widely, likely due to Campylobacter strain differences, the host source of the phagocyte, and its activation state (Banfi et al., 1986; Bar et al., 1991; Wassenaar et al., 1997; Wooldridge and Ketley, 1997). In general, ingestion of Campylobacter by phagocytes is enhanced by opsonins such as complement, leading to efficient bacterial killing (Blaser et al., 1985). Neutrophil Production of ROS The antimicrobial arsenal of neutrophils can be divided into oxygen-dependent and oxygen-independent systems. The oxygen-dependent system relies on the production of ROS via the activity of the enzyme NADPH oxidase (“respiratory burst”) (Babior et al., 2002). The oxygen-independent system is composed of cationic antimicrobial peptides and proteins such as defensins and the bactericidal permeability-increasing protein (Ganz and Weiss, 1997). Neutrophil ROS production depends on the Campylobacter species and strain (Bar et al., 1991; Bernatowska et al., 1989; Ullmann and Krausse, 1987; Walan et al., 1992). The importance of neutrophil production of ROS to effect control of Campylobacter infection can be inferred from studies describing the increased susceptibility of various Campylobacter strains in which genes required for resistance to oxidative stress (i.e., sod& ahpC, katA, and dps) were mutated (Baillon, 1999; Day et al., 2000; Grant and Park, 1995; Ishikawa et al., 2003; Pesci et al., 1994; Purdy et al., 1999; Purdy and Park, 1994; van Vliet et al., 2001; Wai et al., 1996; Yamasaki et al., 2004). For example, a C. coli sodB mutant could not grow in the presence of methyl viologen (a superoxide-generating agent), whereas the parent strain (UA585) grew well (Purdy et al., 1999). The sodB mutant strain also showed a 3-log reduction in colonization of 1-day-old chicks compared
with the wild-type strain (Purdy et al., 1999). Interestingly, a katA (catalase-deficient) mutant of C. coli UA585 colonized as well as wild type, suggesting that hydrogen peroxide is a less potent ROS compared with superoxide, at least in this model (Purdy et al., 1999). In contrast, a katA C. jejuni mutant showed 6-log lower survival compared with wild-type 24 h after coculture with the murine macrophage cell line 5774, an effect reversed by the addition of the respiratory burst inhibitor apocynin (Day et al., 2000). Given the differences between the Campylobacter species and strains tested against various ROSgenerating cell types, the relative importance of individual oxidative species varies, but as a whole, ROS likely contribute to innate defense against Campylobacter. Neutrophil Production of Antimicrobial Peptides and Proteins There is no clear relationship between the level of ROS production detected as chemiluminescence and Campylobacter killing. For example, a clinical C. jejuni strain that was most resistant to neutrophilmediated killing nonetheless elicited the largest chemiluminescence response; the remaining three isolates triggered chemiluminescence responses of varying degrees but were efficiently killed (Walan et al., 1992). These findings imply an important role of the oxygen-independent antimicrobial system composed of cationic peptides and proteins. In human neutrophils, these include the bactericidal permeability-increasing protein (Weiss et al., 1978), cathelicidins and defensins (Wehkamp et al., 2007). Macrophages Blood monocytes differentiate into macrophages after migration into host tissues and concomitantly acquire enhanced antimicrobial abilities, a phenomenon further boosted by cytokines to yield activated macrophages (Hume et al., 2002). Like neutrophils, monocytes and macrophages express important PRMs, including LPS-binding protein, CD 14, collectin receptors and Toll-like receptors, but in addition, they are important sources of proinflammatory cytokines that link the activation of the innate and adaptive immune systems (Taylor et al., 2005). The importance of macrophages in innate defense against Campylobacter infection was initially suggested by the >50% mortality in mice depleted of macrophages (via injection of silica dust) compared with mock-treated animals that remained healthy (Bar, 1988). In another early study, ingestion of clinical isolates by 5774 murine macrophages and elicited
CHAPTER 19
murine peritoneal macrophages from BALB/c mice was reported to be -10% of the added inoculum, whereas human peripheral blood monocytes ingested 48% after 30 min (although their methodology did not allow discrimination between adherent and intracellular bacteria) (Kiehlbauch et al., 1985). Despite this apparently efficient uptake, Campylobacter survived intracellularly for up to 6 days (Kiehlbauch et al., 1985). However, these experiments were performed in the absence of gentamicin; therefore, substantial numbers of C. jejuni may have remained and proliferated after washing the cells with media. Also, these monocytes and/or macrophages were not activated. In a later study examining several clinical isolates of C. jejuni and C. coli UA585 in the absence of opsonins, no colony-forming units were recovered after 48 h (Wassenaar et al., 1997). This study differed from the 1985 study in that a gentamicin wash was included, and steps were taken to distinguish adherent from intracellular bacteria (Wassenaar et al., 1997). Efficient macrophage killing of Campylobacter without the assistance of opsonins also has been reported by others, indicating the importance of activated macrophages in defense against Campylobacter infection (Field et al., 1991; Myszewski and Stern, 1991; Pancorbo et al., 1994). The resistance of avian species to illness caused by Campylobacter disease, in contrast to human susceptibility to Campylobacter, suggests that comparative studies of avian and human macrophages could be instructive. In this regard, the avian cell line H D l l is a useful tool to explore macrophage responses to Campylobacter. Perhaps surprisingly, one study reported a robust and complex proinflammatory response of H D l l cells exposed to C. jejuni 11168, including production of IL-lp, IL-6, and K60, a homolog of IL-8 that is chemotactic for heterophils and mononuclear cells (Smith et al., 2005). The significance of K60 production is puzzling because cellular inflammatory response is not characteristic of avian Campylobacter colonization, indicating perhaps another role for K60 in avian immune defense. After 24 h, no viable Campylobacter could be recovered from the macrophages (Smith et al., 2005). Macrophage Production of Nitric Oxide and RNS Macrophages derived from many species including humans, mice, and fowl produce nitric oxide (NO) in response to inflammatory stimuli such as LPS and interferon (1FN)-y via induction of nitric oxide synthase (NOS2) (Ding et al., 1988; Drapier et al., 1988; Lyons et al., 1992; MacMicking et al., 1997). Evidence exists that Campylobacter also may induce NOS2 expression in H D l l cells, an avian
-
INNATE IMMUNITY
341
macrophage-like cell line, and murine macrophages (Iovine et al., 2008; Smith et al., 2005). Studies of NOS2-deficient mice (MacMicking et al., 1995) reveal an important role of the nitrosative stress imposed by NO/RNS in innate defense against many pathogens, including Mycobacterium tuberculosis and Salmonella enterica (Shiloh and Nathan, 2000). The relevance of nitrosative stress in innate defense against bacterial infection is implied by the multiple gene products expressed in a variety of bacteria aimed at the detoxification of N O . These include the globins, cytochrome reductases, and peroxiredoxins in Salmonella, E. coli, Erwinia, and Mycobacterium species, among others (Poole, 2005). The best-studied molecule involved in resistance to nitrosative (and oxidative stress) is the flavohemoglobin Hmp from E. cob. Expression of this protein is induced by nitric oxide (Poole et al., 1996), upregulates the soxS and so& promoters in response to oxidative stress (Membrillo-Hernandez et al., 1999), and confers resistance to nitrosative stress (Gardner et al., 1998; Hausladen et al., 1998). A related microbial globin, Vgb from Vitreoscilla (Wakabayashi et al., 1986), participates in oxygen utilization, but studies of this gene when transferred to E. coli suggest that Vgb also may provide resistance to nitrosative stress (Frey et al., 2002). In C. jejuni and C. coli, a related globin, Cgb, is induced on exposure to the nitric oxide donor S-nitrosoglutathione and confers resistance to nitrosative stress (Elvers et al., 2004). Unlike other microbial globins, which also can be induced by oxidative stress, Cgb is specifically induced by nitrosative stress (Elvers et al., 2004; Poole, 2005). A second, constitutively expressed gene product, NrfA, also confers resistance to exogenously added nitric oxide donors (Pittman et al., 2007). A link between the antimicrobial activities of NO1RNS against C. jejuni was shown in studies that used bone marrow-derived murine macrophages stimulated with IFN--y/LPS to produce NO’/RNS detected as nitrite, a stable RNS. A 2-log decrease in viability of C. jejuni 81-176 after coculture with IFNy/LPS-stimulated wild-type macrophages compared with similarly treated NOS2-’- macrophages provides evidence that NO/RNS participate in innate defense against Campylobacter infection (Iovine et al., 2008). An important question to be addressed in future studies is to determine how Campylobacter genes involved in resistance to nitrosative stress affect natural infection in which production of NO/RNS is expected. Siglecs on Phagocytes and Natural Killer Cells Sialic acid-binding immunoglobulin-like lectins (Siglecs) are surface transmembrane proteins ex-
342
IOVINE
pressed on immune cells, including phagocytes, that bind sialylated glycans on host cells and bacteria (Crocker et al., 1998). Tyrosine-based inhibition motifs located in the cytoplasmic tail of some siglecs relay inhibitory signals to downstream molecules, ultimately diminishing or delaying immune responses (Ravetch and Lanier, 2000). The clinical significance of infection with Campylobacter-expressing sialylated lipooligosaccharide (LOS) (such as HS: 19) is usually discussed in terms of its propensity to trigger the Guillain-Barrt syndrome via molecular mimicry with host gangliosides (Aspinall et al., 1994). Therefore, it is interesting to note that these HS-19 strains, which express a2, 8-linked sialylated LOS, were found to interact specifically with Siglec-7 on monocytes and natural killer cells (Avril et al., 2006). Although host responses were not assessed, it is intriguing to speculate that perhaps Campylobacter benefits from expression of sialylated LOS by siglec-dependent modulation of the host innate immune response. In further support of the possible biological significance of siglec-Campylobacter interactions, sialylated C. jejuni strains bind to siglec-1 when expressed on CHO cells (Heikema et al., 2007). In that study, monocyte phagocytosis of sialylated LOS-expressing C. jejuni was twofold higher than that of a mutant strain lacking expression of the sialyltransferase gene (cstll)necessary for production of sialylated LOS (Heikema et al., 2007). This suggests that not all siglec-mediated cellular events are inhibitory and that the host may respond to the presence of sialyl residues on Campylobacter by initiating phagocytosis. Dendritic Cells Dendritic cells are at the crossroads between the innate and adaptive immune systems: they are a source of proinflammatory cytokines that assist in the activation of macrophages and T cells, and they mature into professional antigen-presenting cells upon interaction with bacteria (Iwasaki, 2007) and LPS (Franchi et al., 2003). The presence of dendritic cells in the intestinal lamina propria places them in close proximity to gut microorganisms. The dendrites of dendritic cells penetrate the epithelial layer to directly sample the contents of the intestinal lumen (Rescigno et al., 2001). Therefore, intestinal dendritic cells are well positioned to interact with both noninvasive and invasive potential pathogens, including Yersinia, Shigella, and Salmonella (Niess and Reinecker, 2006), and may similarly engage Campylobacter. Although the study of interactions between dendritic cells and Campylobacter is in its infancy, Campylobacter also triggers dendritic cell maturation and increases their production of IL-lp, IL-8, IL-6,
and TNF-a (Hu et al., 2006). That this effect is broadly applicable to Campylobacter species is supported by a study showing dendritic cell maturation triggered by C. jejuni strains associated with the development of Guillain-Barrt syndrome and mutants lacking a functional sialyltransferase (cstll) gene (Kuijf et al., 2007). Furthermore, dendritic cell recruitment by several clinical isolates of C. jejuni as well as 81-176 and 11168 was triggered by secretion of the dendritic cells chemokine CCL20 by polarized intestinal epithelia cells (T84 cells) in a manner that directly correlated with the invasive capability of each isolate (Johanesen and Dwinell, 2006). However, the ultimate role played by dendritic cells in innate (and adaptive) immune defense against Campylobacter infection requires further study. Acute-Phase Response The acute-phase response can be defined as the physiologic processes that occur as a result of the onset of infection and includes nonspecific antimicrobial factors produced mainly by the liver (Table 2). Transcription of acute-phase genes is induced primarily by TNF-a, IL-1, and IL-6 secreted by resident macrophages (Kupffer cells) in the liver (Baumann et al., 1987; Dinarello, 1984) in response to Toll-like receptor activation (Diffenbach and Tramont, 2005). For example, C-reactive protein level increases in response to infection, whereas serum albumin decreases. Acute-phase proteins may inhibit or promote inflammation, but the majority promote phagocytosis (Diffenbach and Tramont, 2005). With the exception of the complement system, the effect of most acutephase components on Campylobacter has not been well studied. Complement The importance of complement activation in host defense against bacterial infection is well established: a 1960 study of the relationship between the bactericidal effect of serum on clinical isolates of E. coli and Klebsiella and a laboratory strain of Shigella dysenteriae revealed that whereas only 2 of 14 blood isolates were serum sensitive, 27 of 35 strains isolated from stool or urine were similarly sensitive to serum (Roantree and Rantz, 1960). These authors hypothesized that “resistance to serum is of importance to a strain’s ability to cause bacteremia. . .this distribution may be related to the frequency of cases of bacteremia caused by strains of Salmonella and the rarity of such cases caused by Shigella.” Twenty-five years later, a similar analogy was made between C. fetus, which is responsible for the majority of bacteremic
CHAPTER 19
-
INNATE IMMUNITY
343
Table 2. Acute-phase proteins relevant to gastrointestinal infections" Acute-phase protein al-Antitrypsin C-reactive protein Collectins and mannose-binding lectin Complement and complement inhibitors Iron- and heme-binding proteins
LPS-binding protein and CD14 Serum amyloid A Serum amyloid P
Proposed role in GI innate immunity
Reference(s)
Binds to virulence-associated proteins of Yersinia and E. coli Pattern recognition molecule that protects transgenic mice from fatal Salmonella infection Pattern recognition molecule that opsonizes many bacteria, including gram-negative enteric bacteria, and also activates complement Opsonizes and is directly bactericidal to many bacteria, including enteric gram-negative bacteria and Campylobacter Restricts availability of essential nutrient to many bacteria, including enteric gram-negative bacteria and Campylobacter Pattern recognition molecules that effect the transfer of LPS to Toll-like receptor 4 Opsonizes enteric gram-negative bacteria via binding to OmpA Opsonizes enteric gram-negative bacteria via binding to LPS
Heusipp et al. (2006), Knappstein et al. (2004) Szalai et al. (2000) Takahashi and Ezokowitz (2005), Takahashi et al. (2006) Blaser et al. (1985), Roantree and Rantz (1960) Barclay (1985), Pickett et al. (1992) Kitchens and Thompson (2005) Hari-Dass et al. (2005) de Haas et al. (1998)
"Items in boldface are discussed further in the text.
Campylobacter infections, and other, primarily diarrheal isolates of Campylobacter (Blaser et al., 1985). In that study, clinical isolates of C. jejuni and C. coli showed 90% killing by fresh human serum, an effect mitigated by heating serum to 56"C, which destroys complement activity (Blaser et al., 1985). In a subsequent clinical study, it was noted that isolation of C. jejuni and C. coli from blood was more commonly associated with defects in innate (and adaptive) immunity, whereas blood isolates from immunocompetent persons were likely to be serum resistant (Blaser et al., 1986). C. upsaliensis, a weakly catalase-positive species often isolated from the feces of dogs and cats, is an unusual cause of disease in humans (Bourke et al., 1998). Case reports describe a wide spectrum of disease in humans, from gastroenteritis to bacteremia in both previously healthy and immunocompromised persons and infants (Lastovica et al., 1989; Patton et al., 1989). Interestingly, the majority of the blood isolates showed resistance to complement-dependent killing, although the virulence properties of the organism that permitted survival are not known (Patton et al., 1989).
ceruloplasmin, prealbumin, and procalcitonin) have not been studied in relation to gastrointestinal infection. However, for several gram-negative organisms, including Salmonella, Shigella, E. coli, Yersinia, and Vibrio, evidence exists for a role in host defense of al-antitrypsin, collectins including mannose-binding lectin, LPS-binding protein, CD14, serum amyloid A, and serum amyloid P. None has been examined in connection with Campylobacter.
ACUTE-PHASE PROTEINS RELEVANT T O CAMPYLOBACTER INFECTION IN NEED OF FURTHER STUDY
Mannose-Binding Lectin
Most acute-phase proteins (including al-acid glycoprotein (orosomucoid), a1-antichymotrypsin (SERPINA3), a2-macroglobulin, angiotensinogen,
Alpha 1-Antitrypsin Alpha l-antitrypsin (AAT) is a serine protease inhibitor that protects the lower respiratory epithelium from innocent bystander proteolytic damage caused by release of elastase from degranulating neutrophils (Crystal, 1990). AAT is produced mainly by the liver, but also by Paneth cells of the small intestine (Molmenti et al., 1993). AAT may play a role in mitigating enteric disease caused by Yersinia and E. coli by binding to the virulence-associated proteins YopM and EspB, respectively (Heusipp et al., 2006; Knappstein et al., 2004). A better understanding of the Campylobacter virulence factors may enable studies of the possible role of AAT in host defense.
Mannose-binding lectin (MBL) is a PRM that distinguishes differences in the spatial geometry of surface-exposed sugars typical of microorganisms versus host cells (Takahashi and Ezekowitz, 2005; Takahashi et al., 2006). MBL bound to a target mi-
344
IOVINE
crobe can activate the complement cascade (Matsushita and Fujita, 1992; Selander et al., 2006) as well as act as an opsonin in its own right via recognition by cellular receptors known as collectin receptors (Kuhlman et al., 1989; Tenner et al., 1995). Binding of MBL to sugars present on LPS decreases with sialylation (Devyatyarova-Johnson et al., 2000), raising the question of whether MBL can bind to Campylobacter strains expressing sialylated LOS. Clarification of the interaction between Campylobacter and MBL is potentially significant because MBL deficiency and possibly gene polymorphisms are associated with increased risk of a variety of infections, including those caused by gram-negative organisms such as Klebsiella pneumoniae and Neisseria meningitidis (Summerfield et al., 1995; Takahashi and Ezekowitz, 2005).
Typhimurium, Klebsiella pneumoniae, Shigella flexneri, Pseudomonas aeruginosa, and Vibrio cholerae via binding to bacterial outer membrane protein A (OmpA) (Hari-Dass et al., 2005). The increased serum amyloid A levels present during the acute-phase response promote ingestion of E. coli by human neutrophils and monocyte-derived macrophages as efficiently as when 20% human serum was used as the opsonin (Shah et al., 2006). The Campylobacter virulence-associated protein CadF, which mediates binding to host cell fibronectin, is homologous to Enterobacteriaceae OmpA (Konkel et al., 1997; Mamelli et al., 2006) and therefore also might be a target of serum amyloid A.
CONCLUDING REMARKS Serum Amyloid A Serum amyloid A opsonizes many gram-negative bacteria, including Salmonella enterica serovar
Complement, CRP, Fe limitation, phagocytes and their products limit systemic spread Fucosylated sugars present in breast mil
Our understanding of the interplay between innate immunity and enteric pathogens such as Salmonella, Shigella, Yersinia, and E. coli far outstrips
salivary nitrite derived from dietary nitrate?
Low stomach pH Role of NO*/RNS?
Acute phase proteins produced in the liver
Bile acids stored in the gallbladder
Defensins produced in stomach, small and large
Role of cathelicidins, NO*/RNS,NOD proteins? Dendritic cells?
Figure 3. Key and proposed elements of innate immune defense against Cumpylobucter infection. Innate immune components for which good evidence exists of their contribution to defense against Cumpylobucter infection in humans are shown in bold. Those components that are proposed to play a role in defense against Cumpylobucter and that warrant further study are shown in italics. CRP, C-reactive protein; Fe, iron; NO/RNS, nitric oxide/reactive nitrogen species.
CHAPTER 19
our knowledge of the relationships between the innate immune system and Campylobacter. However, the lessons learned from these other enteric organisms as well as Helicobacter provide the framework for studies with Campylobacter. The sequencing of several Campylobacter genomes and improved mutagenesis techniques will facilitate molecular studies of particular Campylobacter genes with specific innate immune components. From the point of view of the host, genomewide association studies of genetic polymorphisms associated with alterations in host defense functions may provide insight into those genes that are important for defense against Campylobacter infection. In particular, the role of cellular and soluble PRMs in defense against Campylobacter infection should take high priority. The key and proposed players in innate defense against Campylobacter infection are shown in Fig. 3 .
REFERENCES Abram, M., D. Vukovic, B. Wraber, and M. Doric. 2000. Plasma cytokine response in mice with bacterial infection. Mediators Inflamm. 9:229-234. Abrams, G. D., and J. E. Bishop. 1966. Effect of the normal microbial flora on the resistance of the small intestine to infection. J. Bacteriol. 92:1604-1608. Akira, S., S. Uematsu, and 0. Takeuchi. 2006. Pathogen recognition and innate immunity. Cell 124:783-801. Al-Salloom, F. S., A. A1 Mahmeed, A. Ismaeel, G. A. Botta, and M. Bakhiet. 2003. Campylobacter-stimulated INT407 cells produce dissociated cytokine profiles. J. Infect. 47:217-224. Albiger, B., S. Dahlberg, B. Henriques-Normark, and S . Normark. 2007. Role of the innate immune system in host defence against bacterial infections: focus on the Toll-like receptors. J. Intern. Med. 261511-528. Andersen-Nissen, E., K. D. Smith, K. L. Strobe, S. L. Barrett, B. T. Cookson, S . M. Logan, and A. Aderem. 2005. Evasion of Tolllike receptor 5 by flagellated bacteria. Proc. Natl. Acad. Sci. USA 102~9247-9252. Arbour, N. C., E. Lorenz, B. C. Schutte, J. Zabner, J. N. Kline, M. Jones, K. Frees, J. L. Watt, and D. A. Schwartz. 2000. TLR4 mutations are associated with endotoxin hyporesponsiveness in humans. Nut. Genet. 25:187-191. Aspinall, G. O., A. G. McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1994. Lipopolysaccharides of Campylobacter jejuni serotype 0:19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-BarrC syndrome. Biochemistry 33:241-249. Avril, T., E. R. Wagner, H. J. Willison, and P. R. Crocker. 2006. Sialic acid-binding immunoglobulin-like lectin 7 mediates selective recognition of sialylated glycans expressed on Campylobacter jejuni lipooligosaccharides. Infect. Immun. 74:41334141. Babior, B. M., J. D. Lambeth, and W. Nauseef. 2002. The neutrophil NADPH oxidase. Arch. Biochem. Biophys. 397:342-344. Baggiolini, M., B. Dewald, and B. Moser. 1999. Chemokines, p. 419-431. In J. Gallin and R. Snyderman (ed.), Inflammation: Basic Principles and Clinical Correlates, 3rd ed. Lippincott Williams & Wilkins, Philadelphia. Bakhiet, M., F. S. Al-Salloom, A. Qareiballa, K. Bindayna, I. Farid, and G. A. Botta. 2004. Induction of alpha and beta che-
INNATE IMMUNITY
345
mokines by intestinal epithelial cells stimulated with Campylobacter jejuni. J. Infect. 48:236-244. Banfi, E., M. Cinco, and G. Zabucchi. 1986. Phagocytosis of Campylobacter jejuni and C. coli by peritoneal macrophages. J. Gen. Microbiol. 132:24 09-24 12. Bar, W. 1988. Role of murine macrophages and complement in experimental Campylobacter infection. 1. Med. Microbiol. 26: 55-59. Bar, W., E. Glenn-Calvo, and R. Krausse. 1991. Phagocytosis of enteric Campylobacter by human and murine granulocytes. FEMS Microbiol. Immunol. 3:143-149. Barclay, R. 1985. The role of iron in infection. Med. Lab. Sci. 42: 166-177. Baumann, H., V. Onorato, J. Gauldie, and G. P. Jahreis. 1987. Distinct sets of acute phase plasma proteins are stimulated by separate human hepatocyte-stimulating factors and monokines in rat hepatoma cells. J. Biol. Chem. 262:9756-9768. Benjamin, N., F. O’Driscoll, H. Dougall, C. Duncan, L. Smith, M. Golden, and H. McKenzie. 1994. Stomach NO synthesis. Nature 368502. Bereswill, S., and M. Kist. 2003. Recent developments in Campylobacter pathogenesis. Curr. Opin. Infect. Dis. 16:487-491. Bernatowska, E., P. Jose, H. Davies, M. Stephenson, and D. Webster. 1989. Interaction of Campylobacter species with antibody, complement and phagocytes. Gut 30:906-911. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans.]. Infect. Dis. 157:472-479. Blaser, M., and B. M. Allos. 2005. Campylobacter and related species, p. 2548-2556. In G. Mandell, J. E. Bennett, and R. D o h (ed.), Mandell, Douglas and Bennett’s Principles and Practice of Infectious Diseases, 6th ed., vol. 2. Elsevier Churchill Livingstone, Philadelphia. Blaser, M. J., G. P. Perez, P. F. Smith, C. Patton, F. C. Tenover, A. J. Lastovica, and W. I. Wang. 1986. Extraintestinal Campylobacter jejuni and Campylobacter coli infections: host factors and strain characteristics. J. Infect. Dis. 153552-559. Blaser, M. J., P. F. Smith, and P. F. Kohler. 1985. Susceptibility of Campylobacter isolates to the bactericidal activity of human serum. J. Infect. Dis. 151:227-235. Bourhis, L. L., and C. Werts. 2007. Role of Nods in bacterial infection. Microbes Infect. 9:629-63 6. Bourke, B., V. L. Chan, and P. Sherman. 1998. Campylobacter upsaliensis: waiting in the wings. Clin. Microbiol. Rev. 11:440449. Broxmeyer, H. E., S. Cooper, G. Cacalano, N. L. Hague, E. Bailish, and M. W. Moore. 1996. Involvement of Interleukin (IL) 8 receptor in negative regulation of myeloid progenitor cells in vivo: evidence from mice lacking the murine IL-8 receptor homologue. J. Exp. Med. 184:1825-1832. Chamaillard, M., M. Hashimoto, Y. Horie, J. Masumoto, S. Qiu, L. Saab, Y. Ogura, A. Kawasaki, K. Fukase, S. Kusumoto, M. A. Valvano, S. J. Foster, T. W. Mak, G. Nunez, and N. Inohara. 2003. An essential role for NOD1 in host recognition of bacterial peptidoglycan containing diaminopimelic acid. Nut. Immunol. 4:702-707. Chang, C., and J. F. Miller. 2006. Cumpylobacter jejuni colonization of mice with limited enteric flora. Infect. Immun. 74: 5261-5271. Chen, M. L., 2. Ge, J. G. Fox, and D. B. Schauer. 2006. Disruption of tight junctions and induction of proinflammatory cytokine responses in colonic epithelial cells by Campylobacter jejuni. Infect. Immun. 74:658 1-6589. Crocker, P. R., E. A. Clark, M. Filbin, S. Gordon, Y. Jones, J. H. Kehrl, S. Kelm, N. Le Douarin, L. Powell, J. Roder, R. L. Schnaar, D. C. Sgroi, K. Stamenkovic, R. Schauer, M. Schach-
346
IOVINE
ner, T. K. van den Berg, P. A. van der Merwe, S. M. Watt, and A. Varki. 1998. Siglecs: a family of sialic-acid binding lectins. Glycobiology 8:v. Crystal, R. G. 1990. Alpha 1-antitrypsin deficiency, emphysema, and liver disease. Genetic basis and strategies for therapy. J. Clin. Invest. 85 :1343-1352. Dakdouki, G. K., G. F. Araj, and M. Hussein. 2003. Campylobacter jejuni: unusual cause of cholecystitis with lithiasis. Case report and literature review. Clin. Microbiol. Infect. 9:970-972. Day, W. A., Jr., J. L. Sajecki, T. M. Pitts, and L. A. Joens. 2000. Role of catalase in Carnpylobacter jejuni intracellular survival. Infect. Immun. 68:6337-6345. de Haas, C. J., M. E. van der Tol, K. P. Van Kessel, J. Verhoef, and J. A. Van Strijp. 1998. A synthetic lipopolysaccharidebinding peptide based on amino acids 27-39 of serum amyloid P component inhibits lipopolysaccharide-induced responses in human blood. J. lmmunol. 161:3607-3615. Dekker, J., J. W. Rossen, H. A. Buller, and A. W. Einerhand. 2002. The MUC family: an obituary. Trends Biochem. Sci. 27: 126-131. de Melo, M. A., and J. C. Pechere. 1988. Effect of mucin on Campylobacter jejuni association and invasion on HEp-2 cells. Microb. Pathog. 5:71-76. Devyatyarova-Johnson, M., I. H. Rees, B. D. Robertson, M. W. Turner, N. J. Klein, and D. L. Jack. 2000. The lipopolysaccharide structures of Salmonella enterica serovar Typhimurium and Neisseria gonorrhoeae determine the attachment of human mannose-binding lectin to intact organisms. Infect. Immun. 68: 3 894-3 899. Diffenbach, C., and E. C. Tramont. 2005. Innate (general or nonspecific) host defense mechanisms, p. 34-42. In G. Mandell, J. E. Bennett, and R. D o h (ed.), Mandell, Douglas and Bennett’s Principles and Practice of Infectious Diseases, 6th ed., vol. 1. Elsevier Churchill Livingstone, Philadelphia. Dinarello, C. A. 1984. Interleukin-1 and the pathogenesis of the acute-phase response. N. Engl. J. Med. 311:1413-1418. Ding, A. H., C. F. Nathan, and D. J. Stuehr. 1988. Release of reactive nitrogen intermediates and reactive oxygen intermediates from mouse peritoneal macrophages. Comparison of activating cytokines and evidence for independent production. J. Immunol. 141:2407-2412. Drapier, J. C., J. Wietzerbin, and J. B. Hibbs, Jr. 1988. Interferongamma and tumor necrosis factor induce the L-argininedependent cytotoxic effector mechanism in murine macrophages. Eur. J. Immunol. 18:1587-1592. Duncan, C., H. Dougall, P. Johnston, S. Green, R. Brogan, C. Leifert, L. Smith, M. Golden, and N. Benjamin. 1995. Chemical generation of nitric oxide in the mouth from the enterosalivary circulation of dietary nitrate. Nut. Med. 1546-551. Dykhuizen, R. S., R. Frazer, C. Duncan, C. C. Smith, M. Golden, N. Benjamin, and C. Leifert. 1996. Antimicrobial effect of acidified nitrite on gut pathogens: importance of dietary nitrate in host defense. Antimicrob. Agents Chernother. 40: 1422-1425. Dykhuizen, R S., A. Fraser, H. McKenzie, M. Golden, C. Leifert, and N. Benjamin. 1998. Helicobacter pylori is killed by nitrite under acidic conditions. Gut 42:334-337. Elvers, K. T., G. Wu, N. J. Gilberthorpe, R. K. Poole, and S. F. Park. 2004. Role of an inducible single-domain hemoglobin in mediating resistance to nitric oxide and nitrosative stress in Campylobacter jejuni and Campylobacter coli. J. Bacteriol. 186: 5332-5341. Evans, E. W., F. G. Beach, K. M. Moore, M. W. Jackwood, J. R Glisson, and B. G. Harmon. 1995. Antimicrobial activity of chicken and turkey heterophil peptides CHP1, CHP2, THP1, and THP3. Vet. Microbiol. 47:295-303.
Fang, F. C. 2004. Antimicrobial reactive oxygen and nitrogen species: concepts and controversies. Nut. Rev. Microbiol. 2:820832. Fang, F. C. 1997. Perspectives series: host/pathogen interactions. Mechanisms of nitric oxide-related antimicrobial activity. J. Clin. Invest. 99:2818-2825. Field, L. H., J. L. Underwood, S. M. Payne, and L. J. Berry. 1991. Virulence of Campylobacter jejuni for chicken embryos is associated with decreased bloodstream clearance and resistance to phagocytosis. Infect. Immun. 59: 1448-1456. Fitzgerald, K. A., D. C. Rowe, and D. T. Golenbock. 2004. Endotoxin recognition and signal transduction by the TLR4/MD2complex. Microbes Infect. 6:1361-1367. Fox, E. M., M. Raftery, A. Goodchild, and G. L. Mendz. 2007. Campylobacter jejuni response to ox-bile stress. FEMS Immunol. Med. Microbiol. 49:165-172. Fox, J. G., A. B. Rogers, M. T. Whary, Z. Ge, N. S. Taylor, S. Xu, B. H. Horwitz, and S. E. Erdman. 2004. Gastroenteritis in NF-kappaB-deficient mice is produced with wild-type Camplyobacter jejuni but not with C. jejuni lacking cytolethal distending toxin despite persistent colonization with both strains. Infect. Immun. 72: 1116-1 125. Franchi, L., I. Condo, B. Tomassini, C. Nicolo, and R. Testi. 2003. A caspaselike activity is triggered by LPS and is required for survival of human dendritic cells. Blood 102:2910-2915. Frey, A. D., J. Farces, C. J. Bollinger, and P. T. Kallio. 2002. Bacterial hemoglobins and flavohemoglobins for alleviation of nitrosative stress in Escherichia coli. Appl. Environ. Microbiol. 68:4835-4840. Fritz, J. H., R. L. Ferrero, D. J. Philpott, and S. E. Girardin. 2006. Nod-like proteins in immunity, inflammation and disease. Nut. Immunol. 7:1250-1257. Fullerton, K. E., L. A. Ingram, T. F. Jones, B. J. Anderson, P. V. McCarthy, S. Hurd, B. Shiferaw, D. Vugia, N. Haubert, T. Hayes, S. Wedel, E. Scallan, 0. Henao, and F. J. Angulo. 2007. Sporadic Campylobacter infection in infants: a population-based surveillance case-control study. Pediatr. Infect. Dis. J. 26: 19-24. Ganz, T. 2003. Defensins: antimicrobial peptides of innate immunity. Nut. Rev. Immunol. 3:710-720. Ganz, T., and J. Weiss. 1997. Antimicrobial peptides of phagocytes and epithelia. Semin. Hematol. 34:343-354. Garcia Rodriguez, L. A., and A. Ruigomez. 1997. Gastric acid, acid-suppressing drugs, and bacterial gastroenteritis: how much of a risk? Epidemiology 8571-574. Gardner, P. R., A. M. Gardner, L. A. Martin, and A. L. Salzman. 1998. Nitric oxide dioxygenase: an enzymic function for flavohemoglobin. Proc. Natl. Acad. Sci. USA 95:10378-10383. Geleijns, K., B. C. Jacobs, W. Van Rijs, A. P. Tio-Gillen, J. D. Laman, and P. A. van Doorn. 2004. Functional polymorphisms in LPS receptors CD14 and TLR4 are not associated with disease susceptibility or Campylobacter jejuni infection in Guillain-BarrC patients. J. Neuroimmunol. 150:132-138. Gewirtz, A. T., T. A. Navas, S. Lyons, P. J. Godowski, and J. L. Madara. 2001. Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J. Immunol. 167:1882-1885. Giannella, R A., S. A. Broitman, and N. Zamcheck. 1972. Gastric acid barrier to ingested microorganisms in man: studies in vivo and in vitro. Gut 13:251-256. Gupta, A., J. M. Nelson, T. J. Barrett, R. V. Tauxe, S. P. Rossiter, C. R. Friedman, K. W. Joyce, K. E. Smith, T. F. Jones, M. A. Hawkins, B. Shiferaw, J. L. Beebe, D. J. Vugia, T. RabatskyEhr, J. A. Benson, T. P. Root, and F. J. Angulo. 2004. Antimicrobial resistance among Campylobacter strains, United States, 1997-2001. Emerg. Infect. Dis. 10:1102-1109.
CHAPTER 19
INNATE IMMUNITY
347
~
Hancock, R. E., and G. Diamond. 2000. The role of cationic antimicrobial peptides in innate host defences. Trends Microbiol. 8~402-410. Hari-Dass, R., C. Shah, D. J. Meyer, and J. G. Raynes. 2005. Serum amyloid A protein hinds to outer membrane protein A of gram-negative bacteria. J. Biol. Chem. 280:18562-18567. Harwig, S. S., K. M. Swiderek, V. N. Kokryakov, L. Tan, T. D. Lee, E. A. Panyutich, G. M. Aleshina, 0. V. Shamova, and R. I. Lehrer. 1994. Gallinacins: cysteine-rich antimicrobial peptides of chicken leukocytes. FEBS Lett. 342:281-285. Haslett, C., J. S. Savill, and L. Meagher. 1989. The neutrophil. Curr. Opin. Immunol. 2:lO-18. Hausladen, A., A. J. Gow, and J. S. Stamler. 1998. Nitrosative stress: metabolic pathway involving the flavohemoglobin. Proc. Natl. Acad. Sci. USA 95:14100-14105. Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R. Goodlett, J. K. Eng, S. Akira, D. M. Underhill, and A. Aderem. 2001. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5 . Nature 410:1099-1103. Heikema, A. P., M. Bergman, W. Van Rijs, M. Kuijf, H. Endtz, and A. Van Belkum. 2007. Role of siglecisialic acid interaction in phagocytosis of Campylobacter jejuni in human monocytes, abstr. 0.23, p. 8. 14th Int. Workshop CampylobacterHelicobacter Relat. Organisms. Blackwell Publishing, Rotterdam, The Netherlands. Heusipp, G., K. Spekker, S. Brast, S. Falker, and M. A. Schmidt. 2006. YopM of Yersinia enterocolitica specifically interacts with alphal-antitrypsin without affecting the anti-protease activity. Microbiology 152:1327-1335. Hickey, T. E., S. Baqar, A. L. Bourgeois, C. P. Ewing, and P. Guerry. 1999. Campylobacter jejuni-stimulated secretion of interleukin-8 by INT407 cells. Infect. lmmun. 67:88-93. Hobbie, S., L. M. Chen, R J. Davis, and J. E. Galan. 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. lmmunol. 159:5550-5559. Hoshino, K., 0. Takeuchi, T. Kawai, H. Sanjo, T. Ogawa, Y. Takeda, K. Takeda, and S. Akira. 1999. Cutting edge: Toll-like receptor 4 (TLR4)-deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product. J. Immunol. 162:3749-3752. Hu, L., M. D. Bray, M. Osorio, and D. J. Kopecko. 2006. Campylobacter jejuni induces maturation and cytokine production in human dendritic cells. Infect. lmmun. 74:2697-2705. Hu, L., and T. E. Hickey. 2005. Campylobacter jejuni induces secretion of proinflammatory chemokines from human intestinal epithelial cells. Infect. Immun. 73:4437-4440. Hugdahl, M. B., J. T. Beery, and M. P. Doyle. 1988. Chemotactic behavior of Campylobacter jejuni. Infect. Immun. 56:15601566. Hume, D. A., I. L. Ross, S. R. Himes, R. T. Sasmono, C. A. Wells, and T. Ravasi. 2002. The mononuclear phagocyte system revisited. J. Leukoc. Biol. 72:621-627. Inaba, T., J. W. Alexander, J. D. Ogle, and C. K. Ogle. 1999. Nitric oxide promotes the internalization and passage of viable bacteria through cultured Caco-2 intestinal epithelial cells. Shock 11:276-282. Iovine, N., S. Pursnani, A. Voldman, A. Wasserman, M. J. Blaser, and Y. Weinrauch. Reactive nitrogen species contribute to innate host defense against Campylobacter jejuni. Infect. Immun., in press. Iwasaki, A. 2007. Mucosal dendritic cells. Annu. Rev. Immunol. 25~381-418. Johanesen, P. A., and M. B. Dwinell. 2006. Flagellin-independent regulation of chemokine host defense in Campylobacter jejuniinfected intestinal epithelium. Infect. lmmun. 743437-3447.
Jones, M. A., S. Totemeyer, D. J. Maskell, C. E. Bryant, and P. A. Barrow. 2003. Induction of proinflammatory responses in the human monocytic cell line THP-1 by Campylobacter jejuni. Infect. lmmun. 71:2626-2633. Kiehlbauch, J. A., R. A. Albach, L. L. Baum, and K. P. Chang. 1985. Phagocytosis of Campylobacter jejuni and its intracellular survival in mononuclear phagocytes. Infect. Immun. 48:446451. Kim, J. M., J. S. Kim, H. C. Jung, I. S. Song, and C. Y. Kim. 2002. Up-regulation of inducible nitric oxide synthase and nitric oxide in Helicobacter pylori-infected human gastric epithelial cells: possible role of interferon-gamma in polarized nitric oxide secretion. Helicobacter 7:116-128. Kitchens, R. L., and P. A. Thompson. 2005. Modulatory effects of sCD14 and LBP on LPS-host cell interactions. J. Endotoxin. Res. 11:225-229. Knappstein, S., T. Ide, M. A. Schmidt, and G. Heusipp. 2004. Alpha 1-antitrypsin binds to and interferes with functionality of EspB from atypical and typical enteropathogenic Escherichia coli strains. Infect. Immun. 72:4344-4350. Knowles, R. G., and S. Moncada. 1994. Nitric oxide synthases in mammals. Biochem. J. 298(Pt. 2):249-258. Konkel, M. E., S. G. Garvis, S. L. Tipton, D. E. Anderson, Jr., and W. Cieplak, Jr. 1997. Identification and molecular cloning of a gene encoding a fibronectin-binding protein (CadF) from Campylobacter jejuni. Mol. Microbiol. 24:953-963. Konkel, M. E., B. J. Kim, V. Rivera-Amill, and S. G. Garvis. 1999. Bacterial secreted proteins are required for the internaliztion of Campylobacter jejuni into cultured mammalian cells. Mol. Microbiol. 32:691-701. Kopp, E., and R Medzhitov. 2003. Recognition of microbial infection by Toll-like receptors. Curr. Opin. Immunol. 15:396401. Kovach, A., Q. V. Tu, G. H. K. Chang, A. S. Okoli, W. G. Miller, and G. L. Mendz. 2007. Changes in the expression of virulencerelated genes in response to ox-bile stress of six species of Campylobacterales. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms. Blackwell Publishing, Rotterdam, The Netherlands. Krieglstein, C. F., W. H. Cerwinka, F. S. Laroux, J. W. Salter, J. M. Russell, G. Schuermann, M. B. Grisham, C. R. Ross, and D. N. Granger. 2001. Regulation of murine intestinal inflammation by reactive metabolites of oxygen and nitrogen: divergent roles of superoxide and nitric oxide. J. Exp. Med. 194: 1207-1218. Kuhlman, M., K. Joiner, and R. A. Ezekowitz. 1989. The human mannose-binding protein functions as an opsonin. J. Exp. Med. 169:1733-1745. Kuhns, D. B., D. A. Long Priel, and J. I. Gallin. 1997. Endotoxin and IL-1 hyporesponsiveness in a patient with recurrent bacterial infections. J. Immunol. 158:3959-3964. Kuijf, M. L., J. N. Samsom, W. Van Rijs, A. P. Heikema, P. A. Van Doorn, H. P. Endtz, E. E. S. Nieuwenhuis, and B. C. Jacobs. 2007. Maturation of dendritic cells by Guillain-Barrt syndrome-associated Campylobacter jejuni is related to expression of sialic acid. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms. Blackwell Publishing, Rotterdam, The Nether 1and s. Lambert, M. E., P. F. Schofield, A. G. Ironside, and B. K. Mandal. 1979. Carnpylobacter colitis. Br. Med. 1. 153574359. Lastovica, A. J., E. Le Roux, and J. L. Penner. 1989. “Campylobacter upsaliensis” isolated from blood cultures of pediatric patients. J. Clin. Microbiol. 27:657-659. Leaver, S. K., S. J. Finney, A. Burke-Gaffney, and T. W. Evans. 2007. Sepsis since the discovery of Toll-like receptors: disease
348
IOVINE
concepts and therapeutic opportunities. Crit. Care Med. 35: 1404-1410. Lehrer, R. I., A. K. Lichtenstein, and T. Ganz. 1993. Defensins: antimicrobial and cytotoxic peptides of mammalian cells. Annu. Rev. Immunol. 11:105-128. Lemaitre, B., E. Nicolas, L. Michaut, J. M. Reichhart, and J. A. Hoffmann. 1996. The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86:973-983. Lesuffleur, T., N. Porchet, J. P. Aubert, D. Swallow, J. R. Gum, Y. S . Kim, F. X. Real, and A. Zweibaum. 1993. Differential expression of the human mucin genes MUCl to MUCS in relation to growth and differentiation of different mucus-secreting HT-29 cell subpopulations. J. Cell. Sci. 106(Pt. 3):771-783. Levy, O., G. Canny, C. N. Serhan, and S . P. Colgan. 2003. Expression of BPI (bactericidal/ permeability-increasing protein) in human mucosal epithelia. Biochem. SOC. Trans. 31:795-800. Lin, J., C. Cagliero, B. Guo, Y. W. Barton, M. C. Maurel, S . Payot, and Q. Zhang. 2005. Bile salts modulate expression of the CmeABC multidrug efflux pump in Campylobacter jejuni. J. Bacteriol. 187:7417-7424. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Cbemother. 46:2124-213 1. Lin, J., 0. Sahin, L. 0. Michel, and Q. Zhang. 2003. Critical role of multidrug efflux pump CmeABC in bile resistance and in vivo colonization of Campylobucter jejuni. Infect. lmmun. 71:42504259. Lyons, C. R, G. J. Orloff, and J. M. Cunningham. 1992. Molecular cloning and functional expression of an inducibie nitric oxide synthase from a murine macrophage cell line. J. Biol. Chem. 267~6370-6374. MacCallum, A. J., D. Harris, G. Haddock, and P. H. Everest. 2006. Campylobacter jejuni-infected human epithelial cell lines vary in their ability to secrete interleukin-8 compared to in vitro-infected primacy human intestinal tissue. Microbiology 152: 3661-3665. MacMicking, J., Q. W. Xie, and C. Nathan. 1997. Nitric oxide and macrophage function. Annu. Rev. Immunol. 15:323-350. MacMicking, J. D., C. Nathan, G. Horn, N. Chartrain, D. S. Fletcher, M. Trumbauer, K. Stevens, Q. W. Xie, K. Sokol, N. Hutchinson, H. Chen, and J. S . Mudget. 1995. Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase. Cell 81:641-650. Mamelli, L., J. M. Pages, M. E. Konkel, and J. M. Bolla. 2006. Expression and purification of native and truncated forms of CadF, an outer membrane protein of Campylobacter. Int. J. Biol. Macromol. 39:135-140. Martinsen, T. C., K. Bergh, and H. L. Waldum. 2005. Gastric juice: a barrier against infectious diseases. Basic Clin. Pbamacol. Toxicol. 96:94-102. Matsushita, M., and T. Fujita. 1992. Activation of the classical complement pathway by mannose-binding protein in association with a novel Cls-like serine protease. J. Exp. Med. 176:14971502. McAuley, J. L., S. K. Linden, C. W. Png, R. M. King, H. L. Pennington, s. J. Gendler, T. H. Florin, G. R. Hill, V. Korolik, and M. A. McGuckin. 2007. MUCl cell surface mucin is a critical element of the mucosal barrier to infection. J. Clin. Invest. 117: 2313-2324. McKnight, G. M., C. W. Duncan, C. Leifert, and M. H. Golden. 1999. Dietary nitrate in man: friend or foe? Br. J. Nutr. 81:349358. Medzhitov, R., and C. Janeway, Jr. 2000. The Toll receptor family and microbial recognition. Trends Microbiol. 8:452-456.
Megraud, F., G. Boudraa, K. Bessaoud, S . Bensid, F. Dabis, R. Soltana, and M. Touhami. 1990. Incidence of Campylobacter infection in infants in western Algeria and the possible protective role of breast feeding. Epidemiol. Infect. 105:73-78. Mellits, K. H., J. Mullen, M. Wand, G. Armbruster, A. Patel, P. L. Connerton, M. Skelly, and I. F. Connerton. 2002. Activation of the transcription factor NF-kappaB by Campylobacterjejuni. Microbiology 148:2753-2763. Membrillo-Hernandez, J., M. D. Coopamah, M. F. Anjum, T. M. Stevanin, A. Kelly, M. N. Hughes, and R. K. Poole. 1999. The flavohemoglobin of Escherichiu coli confers resistance to a nitrosating agent, a “nitric oxide releaser,” and paraquat and is essential for transcriptional responses to oxidative stress. J. Biol. Chem. 274:748-754. Mendz, G. L., Q. V. Tu, and M. A. McGuckin. 2007. Effects of human mucin MUC2 on the expression of Campylobacter jejuni virulence factors. 14th Int. Workshop Campylobacter Helicobucter Relat. Organisms. Blackwell Publishing, Rotterdam, The Netherlands. Molmenti, E. P., D. H. Perlmutter, and D. C. Rubin. 1993. Cellspecific expression of alpha 1-antitrypsin in human intestinal epithelium. J. Clin. Invest. 92:2022-2034. Muller, C. A., I. B. Autenrieth, and A. Peschel. 2005. Innate defenses of the intestinal epithelial barrier. Cell. Mol. Life Sci. 62: 1297-1307. Murray, P. J. 2005. NOD proteins: an intracellular pathogenrecognition system or signal transduction modifiers? Cum Opin. Immunol. 17:352-358. Myszewski, M. A., and N. J. Stern. 1991. Phagocytosis and intracellular killing of Campylobacter jejuni by elicited chicken peritoneal macrophages. Avian Dis. 35:750-755. Nathan, C. 1997. Inducible nitric oxide synthase: what difference does it make? J. Clin. Invest. 100:2417-2423. Nathan, C., and M. U. Shiloh. 2000. Reactive oxygen and nitrogen intermediates in the relationship between mammalian hosts and microbial pathogens. PYOC.Nutl. Acad. Sci. USA 97:8841-8848. Neal, K. R, H. M. Scott, R. C. Slack, and R. F. Logan. 1996. Omeprazole as a risk factor for Cumpylobucter gastroenteritis: case-control study. BMJ 312:414-415. Newburg, D. S. ZOOS. Innate immunity and human milk. J. Nutr. 135:1308-1312. Newburg, D. S., G . M. Ruiz-Palacios, M. Altaye, P. Chaturvedi, J. Meinzen-Derr, L. Guerrero Mde, and A. L. Morrow. 2004. Innate protection conferred by fucosylated oligosaccharides of human milk against diarrhea in breastfed infants. Glycobiology 14:2S3-263. Newell, D. G. 2001. Animal models of Campylobacter jejuni colonization and disease and the lessons to be learned from similar Helicobacter pylori models. Symp. Ser. SOC.Appl. Microbiol. 2001 57S-67S. Nguyen, T. X., A. M. Cole, and R. I. Lehrer. 2003. Evolution of primate theta-defensins: a serpentine path to a sweet tooth. Peptides 24:1647-1654. Niess, J. H., and H. C. Reinecker. 2006. Dendritic cells: the commanders-in-chief of mucosal immune defenses. Curr. Opin. Gustroenterol. 22:354-360. Ogushi, K., A. Wada, T. Niidome, N. Mori, K. Oishi, T. Nagatake, A. Takahashi, H. Asakura, S . Makino, H. Hojo, Y. Nakahara, M. Ohsaki, T. Hatakeyama, H. Aoyagi, H. Kurazono, J. Moss, and T. Hirayama. 2001. Salmonella enteritidis FliC (flagella filament protein) induces human beta-defensin-2 mRNA production by Caco-2 cells. J. Biol. Chem. 276:30521-30526. Okoli, A. S., T. Wadstrom, and G. L. Mendz. 2007. MiniReview: bioinformatic study of bile responses in Campylobacterales. FEMS Immunol. Med. Microbiol. 49:lOl-123.
CHAPTER19
O’Neil, D. A., E. M. Porter, D. Elewaut, G. M. Anderson, L. Eckmann, T. Ganz, and M. F. Kagnoff. 1999. Expression and regulation of the human beta-defensins hBD-1 and hBD-2 in intestinal epithelium. J. lmmunol. 163:6718-6724. Pancorbo, P. L., A. M. Gallego, M. de Pablo, C. Alvarez, E. Ortega, and G. Alvarez de Cienfuegos. 1994. Inflammatory and phagocytic response to experimental Campylobacter jejuni infection in mice. Microbiol. Immunol. 38:89-95. Patton, C. M., N. Shaffer, P. Edmonds, T. J. Barrett, M. A. Lambert, C. Baker, D. M. Perlman, and D. J. Brenner. 1989. Human disease associated with “Campylobacter upsaliensis” (catalasenegative or weakly positive Campylobacter species) in the United States. J. Clin. Microbiol. 27:66-73. Pauleau, A. L., and P. J. Murray. 2003. Role of Nod2 in the response of macrophages to toll-like receptor agonists. Mol. Cell. Biol. 23:753 1-753 9. Philpott, D. J., S. Yamaoka, A. Israel, and P. J. Sansonetti. 2000. Invasive Shigella flexneri activates NF-kappa B through a lipopolysaccharide-dependent innate intracellular response and leads to IL-8 expression in epithelial cells.J. Immunol. 165:903914. Pickett, C. L., T. Auffenberg, E. C. Pesci, V. L. Sheen, and S. S. Jusuf. 1992. Iron acquisition and hemolysin production by Campylobacter jejuni. Infect. Immun. 60:3872-3877. Pittman, M. S., K. T. Elvers, L. Lee, M. A. Jones, R. K. Poole, S . F. Park, and D. J. Kelly. 2007. Growth of Campylobacter jejuni on nitrate and nitrite: electron transport to NapA and NrfA via NrfH and distinct roles for NrfA and the globin Cgb in protection against nitrosative stress. Mol. Microbiol. 6357.5590. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, E. Alejos, M. Silva, C. Galanos, M. Freudenberg, P. Ricciardi-Castagnoli, B. Layton, and B. Beutler. 1998. Defective LPS signaling in C3H/HeJ and C57BL/lOScCr mice: mutations in Tlr4 gene. Science 282:2085-2088. Poole, R. K. 2005. Nitric oxide and nitrosative stress tolerance in bacteria. Biochem. SOL. Trans. 33:176-180. Poole, R. K., M. F. Anjum, J. Membrillo-Hernandez, S . 0. Kim, M. N. Hughes, and V. Stewart. 1996. Nitric oxide, nitrite, and Fnr regulation of hmp (flavohemoglobin) gene expression in Escherichia coli K-12. J. Bacteriol. 1785487-5492. Purdy, D., S. Cawthraw, J. H. Dickinson, D. G. Newell, and S . F. Park. 1999. Generation of a superoxide dismutase (SOD)deficient mutant of Campylobacter coli: evidence for the significance of SOD in Campylobacter survival and colonization. Appl. Environ. Microbiol. 65:2540-2546. Purdy, D., and S. F. Park. 1994. Cloning, nucleotide sequence and characterization of a gene encoding superoxide dismutase from Campylobacter jejuni and Campylobacter coli. Microbiology 140~1203-1208. Raphael, B. H., S . Pereira, G. A. Flom, Q. Zhang, J. M. Ketley, and M. E. Konkel. 2005. The Campylobacter jejuni response regulator, CbrR, modulates sodium deoxycholate resistance and chicken colonization. J. Bacteriol. 187:3662-3670. Ravetch, J. V., and L. L. Lanier. 2000. Immune inhibitory receptors. Science 290534-89. Rescigno, M., M. Urbano, B. Valzasina, M. Francolini, G. Rotta, R. Bonasio, F. Granucci, J. P. Kraehenbuhl, and P. RicciardiCastagnoli. 2001. Dendritic cells express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria. Nat. lmmunol. 2:361-367. Resta-Lenert, S., and K. E. Barrett. 2002. Enteroinvasive bacteria alter barrier and transport properties of human intestinal epithelium: role of iNOS and COX-2. Gastroenterology 122: 10701087.
INNATEIMMUNITY
349
Reynoso-Paz, S., R. L. Coppel, I. R. Mackay, N. M. Bass, A. A. Ansari, and M. E. Gershwin. 1999. The immunobiology of bile and biliary epithelium. Hepatology 30:351-357. Rivera-Amill, V., B. J. Kim, J. Seshu, and M. E. Konkel. 2001. Secretion of the virulence-associated Campylobacter invasion antigens from Campylobacter jejuni requires a stimulatory signal. 1.Infect. Dis. 183:1607-1616. Roantree, R. J., and L. A. Rantz. 1960. A study of the relationship of the normal bactericidal activity of human serum to bacterial infection. J. Clin. Invest. 39:72-8 1. Roberts, P. J., G. P. Riley, K. Morgan, R. Miller, J. 0. Hunter, and S. J. Middleton. 2001. The physiological expression of inducible nitric oxide synthase (iNOS) in the human colon./. Clin. Pathol. 54:293-297. Rock, F. L., G. Hardiman, J. C. Timans, R. A. Kastelein, and J. F. Bazan. 1998. A family of human receptors structurally related to Drosophila Toll. Proc. Natl. Acad. Sci. USA 95:588-593. Rotimi, V. O., L. Egwari, and B. Akande. 1990. Acidity and intestinal bacteria: an in-vitro assessment of the bactericidal activity of hydrochloric acid on intestinal pathogens. Afr. J. Med. Med. Sci. 19:275-280. Ruiz-Palacios, G. M., L. E. Cervantes, P. Ramos, B. ChavezMunguia, and D. S. Newburg. 2003. Campylobacter jejuni binds intestinal H ( 0 ) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J. Biol. Chem. 278:14112-14120. Salzman, A., A. G. Denenberg, I. Ueta, M. O’Connor, S. C. Linn, and C. Szabo. 1996. Induction and activity of nitric oxide synthase in cultured human intestinal epithelial monolayers. Am. J. Physiol. 270:G565-573. Salzman, A. L., T. Eaves-Pyles, S. C. Linn, A. G. Denenberg, and C. Szabo. 1998. Bacterial induction of inducible nitric oxide synthase in cultured human intestinal epithelial cells. Gastroenterology 114~93-102. Schlee, M., J. Wehkamp, A. Altenhoefer, T. A. Oelschlaeger, E. F. Stange, and K. Fellermann. 2007. Induction of human betadefensin 2 by the probiotic Escherichia coli Nissle 1917 is mediated through flagellin. Infect. Immun. 75:2399-2407. Selander, B., U. Martensson, A. Weintraub, E. Holmstrom, M. Matsushita, S. Thiel, J. C. Jensenius, L. Truedsson, and A. G. Sjoholm. 2006. Mannan-binding lectin activates C3 and the alternative complement pathway without involvement of C2. J. Clin. Invest. 116:1425-1434. Shah, C., R Hari-Dass, and J. G. Raynes. 2006. Serum amyloid A is an innate immune opsonin for gram-negative bacteria. Blood 108:1751-1757. Shandera, W. X., M. P. Tormey, and M. J. Blaser. 1992. An outbreak of bacteremic Campylobacter jejuni infection. Mt. Sinai /. Med. 5953-56. Shiloh, M. U., and C. F. Nathan. 2000. Reactive nitrogen intermediates and the pathogenesis of Salmonella and Mycobacteria. Curr. Opin. Microbiol. 3:35-42. Smith, C. K., P. Kaiser, L. Rothwell, T. Humphrey, P. A. Barrow, and M. A. Jones. 2005. Campylobacter jejuni-induced cytokine responses in avian cells. Infect. Immun. 73:2094-2100. Smith, K. E., J. M. Besser, C. W. Hedberg, F. T. Leano, J. B. Bender, J. H. Wicklund, B. P. Johnson, K. A. Moore, and M. T. Osterholm. 1999. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992-1998 Investigation Team. N.Engl. 1. Med. 340:1525-1532. Spiegelhalder, B., G. Eisenbrand, and R Preussmann. 1976. Influence of dietary nitrate on nitrite content of human saliva: possible relevance to in vivo formation of N-nitroso compounds. Food Cosmet. Toxicol. 14545-548, Stanley, K. N., S. H. Cabry, C. F. Dalton, and N. Jordan-Mahy. 2007. Use of mucin-overproducing cell line HT29-MTX to
350
IOVINE
study the interaction between Campylobacter jejuni and commensal micro-organisms at the mucin interface of the gastrointestinal tract. 14th Int. Workshop Campylobacter Helicobacter Relat. Organisms. Blackwell Publishing, Rotterdam, The Netherlands. Strober, W., P. J. Murray, A. Kitani, and T. Watanabe. 2006. Signalling pathways and molecular interactions of NODl and NOD2. Nut. Rev. Immunol. 6:9-20. Summerfield, J. A., S. Ryder, M. Sumiya, M. Thursz, A. Gorchein, M. A. Monteil, and M. W. Turner. 1995. Mannose binding protein gene mutations associated with unusual and severe infections in adults. Lancet 345:886-889. Sylvester, F. A., D. Philpott, B. Gold, A. Lastovica, and J. F. Forstner. 1996. Adherence to lipids and intestinal mucin by a recently recognized human pathogen, Campylobacter upsaliensis. Infect. Immun. 64:4060-4066. Szalai, A. J., J. L. VanCott, J. R. McGhee, J. E. Volanakis, and W. H. Benjamin, Jr. 2000. Human C-reactive protein is protective against fatal Salmonella enterica serovar typhimurium infection in transgenic mice. Infect. Immun. 685652-5656. Takahashi, A., A. Wada, K. Ogushi, K. Maeda, T. Kawahara, K. Mawatari, H. Kurazono, J. Moss, T. Hirayama, and Y. Nakaya. 2001. Production of beta-defensin-2 by human colonic epithelial cells induced by Salmonella enteritidis flagella filament structural protein. FEBS Lett. 508:484-488. Takahashi, K., and R A. Ezekowitz. 2005. The role of the mannose-binding lectin in innate immunity. Clin. Infect. Dis. 41(S~ppl.7):S440-S444. Takahashi, K., W. E. Ip, I. C. Michelow, and R. A. Ezekowitz. 2006. The mannose-binding lectin: a prototypic pattern recognition molecule. Cum. Opin. Immunol. 18:16-23. Taylor, P. R., L. Martinez-Pomares, M. Stacey, H. H. Lin, G. D. Brown, and S. Gordon. 2005. Macrophage receptors and immune recognition. Annu. Rev. Immunol. 23:901-944. Tenner, A. J., S. L. Robinson, and R. A. Ezekowitz. 1995. Mannose binding protein (MBP) enhances mononuclear phagocyte function via a receptor that contains the 126,000 M(r) component of the C l q receptor. Immunity 3:485-493. Tosi, M. F. 2005. Innate immune responses to infection. J. Allergy Clin. lmmunol. 116:24 1-249. Uehara, A., Y. Fujimoto, K. Fukase, and H. Takada. 2007. Various human epithelial cells express functional Toll-like receptors, NODl and NOD2 to produce anti-microbial peptides, but not proinflammatory cytokines. Mol. Immunol. 44:3 100-3 111. Ullmann, U., and R. Krausse. 1987. The reaction of Campylobacter species on the chemiluminescence, chemotaxis and hemagglutination. Zentralbl. Bakteriol. Mikrobiol. Hyg. A 266:178190. van Dijk, A., E. J. Veldhuizen, S. I. Kalkhove, J. L. Tjeerdsma-van Bokhoven, R. A. Romijn, and H. I?. Haagsman. 2007. The betadefensin gallinacin-6 is expressed in the chicken digestive tract and has antimicrobial activity against food-borne pathogens. Antimicrob. Agents Chemother. 51:912-922. van Spreeuwel, J. P., G. C. Duursma, C. J. Meijer, R. Bax, P. C. Rosekrans, and J. Lindeman. 1985. Campylobacter colitis: histological immunohistochemical and ultrastructural findings. Gut 26~945-951. Wakabayashi, S., H. Matsubara, and D. A. Webster. 1986. Primary sequence of a dimeric bacterial haemoglobin from Vitreoscilla. Nature 322:481-483. Walan, A., C. Dahlgren, E. Kihlstrom, 0. Stendahl, and R Lock. 1992. Phagocyte killing of Campylobacter jejuni in relation to oxidative activation. APMlS 100:424-430.
Walport, M. J. 2001a. Complement. First of two parts. N. Engl. J. Med. 344:1058-1066. Walport, M. J. 2001b. Complement. Second of two parts. N. Engl. J. Med. 344:1140-1144. Wassenaar, T. M., M. Engelskirchen, S. Park, and A. Lastovica. 1997. Differential uptake and killing potential of Campylobacter jejuni by human peripheral monocytes/macrophages. Med. Microbiol. Immunol. 186:139-144. Waterman, S. R, and P. L. Small. 1998. Acid-sensitive enteric pathogens are protected from killing under extremely acidic conditions of pH 2.5 when they are inoculated onto certain solid food sources. Appl. Environ. Microbiol. 64:3882-3886. Watson, R O., and J. E. Galan. 2005. Signal transduction in Campylobacter jejuni-induced cytokine production. Cell. Microbiol. 7~655-665. Watson, R. O., V. Novik, D. Hofreuter, M. Lara-Tejero, and J. E. Galan. 2007. A MyD88-deficient mouse model reveals a role for Nrampl in Campylobacter jejuni infection. Infect. Immun. 75: 1994-2003. Wehkamp, J., J. Harder, K. Wehkamp, B. Wehkamp-von Meissner, M. Schlee, C. Enders, U. Sonnenborn, S. Nuding, S. Bengmark, K. Fellermann, J. M. Schroder, and E. F. Stange. 2004. NF-kappaB- and AP-1-mediated induction of human beta defensin-2 in intestinal epithelial cells by Escherichia coli Nissle 1917: a novel effect of a probiotic bacterium. Infect. Immun. 725750-5758. Wehkamp, J., J. Schauber, and E. F. Stange. 2007. Defensins and cathelicidins in gastrointestinal infections. Cum. Opin. Gastroenterol. 23:32-38. Weiss, J., P. Elsbach, I. Olsson, and H. Odeberg. 1978. Purification and characterization of a potent bactericidal and membrane active protein from the granules of human polymorphonuclear leukocytes. J. Biol. Chem. 253:2664-2672. Wempe, J. M., C. A. Genigeorgis, T. B. Farver, and H. I. Yusufu. 1983. Prevalence of Campylobacter jejuni in two California chicken processing plants. Appl. Environ. Microbiol. 45:355359. Witthoft, T., L. Eckmann, J. M. Kim, and M. F. Kagnoff. 1998. Enteroinvasive bacteria directly activate expression of iNOS and N O production in human colon epithelial cells. Am. J. Physiol. 275 :G564-571. Wooldridge, K. G., and J. M. Ketley. 1997. Campylobacter-host cell interactions. Trends Microbiol. 5:96-102. Xu, J., X. Xu, and W. Verstraete. 2001. The bactericidal effect and chemical reactions of acidified nitrite under conditions simulating the stomach. J. Appl. Microbiol. 90523-529. Zilbauer, M., N. Dorrell, P. K. Boughan, A. Harris, B. W. Wren, N. J. Klein, and M. Bajaj-Elliott. 2005. Intestinal innate immunity to Campylobacter jejuni results in induction of bactericidal human beta-defensins 2 and 3. Infect. Immun. 73:72817289. Zilbauer, M., N. Dorrell, A. Elmi, K. J. Lindley, S. Schuller, H. E. Jones, N. J. Klein, G. Nunez, B. W. Wren, and M. Bajaj-Elliott. 2007. A major role for intestinal epithelial nucleotide oligomerization domain 1 (NOD1) in eliciting host bactericidal immune responses to Campylobacter jejuni. Cell. Microbiol. 9:24042416. Ziprin, R. L., C. R. Young, J. A. Byrd, L. H. Stanker, M. E. Hume, S. A. Gray, B. J. Kim, and M. E. Konkel. 2001. Role of Campylobucter jejuni potential virulence genes in cecal colonization. Avian Dis. 45549-557. Zweier, J. L., A. Samouilov, and P. Kuppusamy. 1999. Nonenzymatic nitric oxide synthesis in biological systems. Biochim. Biophys. Acta 1411:250-262.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 20
Chemosensory Signal Transduction Pathway of Campylobacter jejuni VICTORIA KOROLIKAND JULIAN KETLEY
Bacteria have evolved a range of mechanisms to enable them to move within their environment. They sense changes in environmental conditions and change their temporal position in order to avoid unfavorable conditions and maneuver toward nutrient sources and new niches. Chemotaxis, where the bacterial cell moves along a chemical gradient, involves the coordination of motility by a signal transduction pathway. In this chapter, we will describe the composition of the Campylobacter jejuni chemotaxis system in the context of pathways found in other bacterial species and derive models of the mechanism of signal transduction in the C. jejuni chemotaxis pathway. In recent years, our knowledge of chemotaxis pathways has progressed from what we now know to be a simple, host-adapted Escherichia coli paradigm to a much more complex scenario in other bacteria where similar pathways may lead to twitching or darting motility and may be intricately interwoven with quorum sensing and cytoplasmic communication (Mattick, 2002; Morgan et al., 2006; O’Toole and Kolter, 1998). The ability of bacterial cells to move toward their target environment or host cells via chemotaxis has long been implicated in virulence of pathogenic bacteria and is particularly thought to play an important role in invasion of host tissues (Josenhans and Suerbaum, 2002; Ottemann and Miller, 1997). The most prevalent mechanism involves chemotactic movement along a chemical gradient via the flagellum, a thin, whiplike structure that extends from the cell (reviewed in Eisenbach, 1996; Falke et al., 1997). The chemotaxis pathway, governed by a two-component regulatory system involving a phosphorelay from a histidine kinase protein (HK) to a response regulatory protein (RR), transduces a signal detected by receptors to direct changes in flagellar
motor behavior to enable movement toward an attractant and away from a repellent. Chemotactic control of motility also involves a process of adaptation, a rudimentary chemical memory, whereby the bacterial cell responds in a matter of milliseconds to temporal changes in ligand levels and not the absolute concentration of ligand (Adler, 1965). The bacterial cell responds to a chemical ligand by alternating a swimming or tumbling motion. The cell therefore combines rapid reoriention by tumbling with prolonged swimming periods to move toward or away from particular ligands (Eisenbach, 1996). The relative frequency of tumbling versus straight swimming is governed by the direction of flagellar rotation. The fundamental components of the chemotaxis signal transduction pathway are the sensory receptor, the HK CheA, a scaffold protein Chew, and the RR CheY that ultimately acts on the flagellar motor to switch rotation either clockwise or counterclockwise. Chemotaxis signal transduction pathways of prokaryotes are now known to be as varied as the environments they can transverse, hosts they can occupy, and diseases they can cause (Szurmant and Ordal, 2004). The natural habitat of campylobacters is the intestine of warm-blooded animals, and therefore chemotactic motility is likely to be an important mechanism involved in colonization and pathogenicity; this has been shown experimentally for colonization of mice (Takata et al., 1992), ferrets (Yao et al., 1997) and chickens (Hendrixson and DiRita, 2004). Although chemotaxis has been demonstrated for Campylobacter (Hugdahl et al., 1988; Takata et al., 1992), the mechanisms involved in the sensory control of motility are yet to be fully elucidated. Chemoattractants of C. jejuni include components of mucin, such as L-fucose and L-serine, as well as certain organic acids, such as pyruvate and succinate (Hug-
Victoria Korolik * Institute for Glycomics, Griffith University, Gold Coast, QLD 4222, Australia. Genetics, University of Leicester, Leicester, United Kingdom.
35 1
Julian Ketley * Department of
352
KOROLIK AND KETLEY
dahl et al., 1988). C. jejuni was reported to colonize the intestinal mucus itself rather than exclusively adhering to epithelial cells with campylobacters observed to actively swim up and down the cecal crypts (Lee et al., 1986). Thus, chemotaxis toward mucin combined with efficient motility in a viscous milieu (Ferrero and Lee, 1988; Szymanski et al., 1995) would enable the bacteria to remain localized within the mucus. In Helicobacter pylori, which is phenotypically and genetically similar to campylobacters, the chemotactic response was demonstrated to be involved in establishing and maintaining gastric infection, as well as reaching high levels of bacterial load in the infected animals (Terry et al., 2005). Other evidence demonstrates the importance of chemosensory receptors in colonization and disease in H. pylori, where the sensory receptors TlpA and TlpC were shown to be required for full infection potential in the stomach of laboratory mice and Tlp B appears to be essential for pH taxis and colonization of gastric mucosa (Andermann et al., 2002; Croxen et al., 2006). Similarly, in C. jejuni, mutants of chemoreceptor Tlp9 (CjllS9 or CetB) were deficient in their ability to invade human tissue culture cells (Golden and Acheson, 2002), and TlplO (CjOOl9) was found to be a determinant of chicken cecal colonization (Hendrixson and DiRita, 2004). With respect to the signal transduction pathway, CheA mutants were unable to establish colonization in mice (Chang and Miller, 2006) and invade tissue culture cells (Golden and Acheson, 2002). CheY mutants are attenuated for chicken colonization (Hendrixson and DiRita, 2004), but appear to be hyperinvasive (Golden and Acheson, 2002; Yao et al., 1997). In a recent report of basolateral invasion of cells via a subcellular route (subvasion), the phenotype was promoted in the absence of Chew and therefore a functional chemotaxis pathway (van Alphen et al., 2008). Clearly, there are some interesting connections between the C. jejuni chemotactic response and host cell interactions that require further investigation. A detailed understanding of the mechanisms and determinants involved in intestinal colonization and disease is essential to the development of intervention strategies that target the presence of campylobacters in food animals and disease in humans. Chemotaxis is a fundamental aspect of intestinal colonization by campylobacters and therefore merits further investigation.
THE CHEMOSENSORY PATHWAY: THE E. COLI PARADIGM The bacterial chemotaxis signal transduction pathway has been extensively studied in the enteric
bacteria E. coli and Salmonella spp. and can be considered a paradigm of bacterial two-component regulatory systems in general and chemotaxis signal transduction in particular (Eisenbach, 1996; Falke et al., 1997) (Fig. 1).The peritrichously flagellated E. coli moves toward attractants or away from repellants by way of a biased random walk. This walk pattern consists of two forms of movement: a tumbling mode induced by clockwise-rotating dissociated flagella, and a smooth, straight swimming mode induced by counterclockwise flagellar rotation. The former tumbling mode enables the cell to reorient itself, and the counterclockwise rotation of the latter smooth swimming mode results in the bundling of flagella at one cell pole. As the concentration of chemoattractant increases, the frequency of tumbling subsequently decreases, leading to a greater proportion of smooth swimming and therefore movement up the concentration gradient. The signal transduction pathway is initiated by sensory receptor proteins that belong to the family of transmembrane methyl-accepting chemotaxis proteins (MCPs). The MCPs are a part of a unit complex of proteins that consists of a homodimeric MCP receptor, two molecules of a linker protein Chew, and two molecules of HK CheA. Binding of CheA into the MCP complex is dependent on Chew, which acts as a scaffolding protein with no catalytic activity. These protein complex units form mixed-receptor higher-order aggregates (Sourjik, 2004; Zhang et al., 2007) of several thousand MCP complexes, becoming large 100-nm rafts of mixed receptors at the cell poles. Receptor clustering into signaling organellelike structures appears to produce signal amplification and greater sensitivity as a result of cooperative interactions between the mixed receptors (Ames et al., 2002; Grebe and Stock, 1998; Lybarger and Maddock, 2000). Five sensory receptors have been characterized in E. coli: aspartate receptor Tar, serine receptor Tsr, ribose/galactose receptor Trg, dipeptide receptor Tap, and redox potential sensor Aer receptor. Additional ligand specificity is provided by periplasmic binding proteins that interact with MCP receptors (Grebe and Stock, 1998). Tar and Tsr form the major MCPs in the polar receptor rafts (Grebe and Stock, 1998). All MCPs are large transmembrane proteins resembling bunches of hairlike a helices with spatially and functionally defined periplasmic (sensory), transmembrane, and cytoplasmic domains (Stock and Levit, 2000). In the absence of chemoattractants, the interaction of the HK CheA with the cytoplasmic domain results in the autophosphorylation of a histidine residue (His-48) (Fig. 1). The phosphoryl group is subsequently transferred to the second cytoplasmic
CHAPTER 20
Any RR domain
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
353
0 ( . ,
CheY CheYP CheA CheB CheR
0
CheZ
(I
Chew Figure 1. Diagrammatic overview of the E. coli chemotaxis signal transduction pathway. In E. coli, the receptor complex consists of MCP, Chew, and CheA. In the absence of chemoattractant, the CheA kinase domain is active, and after autophosphorylation of CheA, the phosphate is transferred to CheY. Phospho-CheY then binds to FliM on the flagellum motor. Sufficient binding of phospho-CheY to the flagellum motor leads to reversal from counterclockwise to clockwise rotation (not shown). Signal termination occurs by the action of the phosphatase CheZ on phospo-CheY. System adaptation, which resets the signaling properties of the receptor, occurs by reversible methylation by CheB and CheR; the level of methylation is controlled by phosphorylation of the response regulator domain on CheB. For further details, see text.
component, the RR CheY. Phosphorylation of Asp 57 of CheY induces a conformational change that decreases binding affinity for CheA, but that increases the binding affinity of CheY to its distant target: the flagellar motor protein, FliM. In the presence of an attractant, autophosphorylation of CheA is inhibited, and hence CheY is not phosphorylated. Interactions between CheY and FliM, and FliM induction of rotational direction are complex and not well understood. In E. coli, the response regulator, CheY, binds to FliM, and the proportion of phosphorylated molecules will determine whether the tumbling motion will be induced (Bren and Eisenbach, 2001). It appears that phosphorylated CheY
molecules need to occupy at least 70% of the available FliM molecules in the basal body for change of rotational direction to occur. Therefore, the multiple phosphorylated CheY binding events induce a tumbling motion by switching the counterclockwise motion of the flagella (swimming in one direction) to clockwise. The bound-phosphorylated CheY protein then autodephosphorylates (a slow reaction of 10 to 20 s) or is dephosphorylated by the phosphatase CheZ (TIi2,approximately 200 ms) (Fig. l),and the flagella rapidly returns to swimming, counterclockwise motion. By alternating the swimming and tumbling motion, the bacterial cells can rapidly reorient themselves in the environment and induce prolonged
354
KOROLIK AND KETLEY
swimming periods toward the chemoattractants (Falke et al., 1997). Methylation and demethylation of chemosensory receptor molecules control the level of response to the stimuli and the rate of adaptation (Szurmant and Ordal, 2004). The methytransferase CheR constitutively adds methyl groups from S-adenosyl methionine to MCP glutamate residues, and the methylesterase CheB removes methyl groups (Fig. 1). In E. coli, the activity of CheB is upregulated by CheA phosphorylation of a RR domain. When ligand is bound to the MCP, CheA activity decreases, reducing the level of CheY and CheB phosphorylation. The reduction in the level of phospho-CheY results in counterclockwise rotation and smooth swimming, and the decrease in phospho-CheB leads to increasing levels of MCP methylation until a threshold is reached whereby CheA is reactivated despite the presence of bound ligand. This resetting to the prestimulus state is changed if more ligand is encountered and binds to additional MCPs in the complex. Aggregation of the MCPs into mixed clusters enables cooperability in adaptation by methylation as minor MCPs do not have CheR docking motifs, and therefore methylation in trans occurs in the tightly packed aggregates. It has been suggested (Stock and Levit, 2000) that clustering of the MCP receptors at the cellular poles and the methylation level of MCPs produced by competitive influence of CheR and CheB on chemoreceptors serves bacterial cells as rudimentary chemical memory that records chemical gradients over time, so that the clustered chemosensory organelle is a prokaryotic brain located at a polar end of the cell-the head! The balance between the response to stimulus and adaptation to existing conditions in bacteria is remarkable: the tiny single cell can sense and respond to changes in environmental levels of chemicals, whether they are chemoattractants or repellants, in a matter of milliseconds. It only takes further milliseconds for it to adapt to the local levels and bring the sensory apparatus back to prestimulus state and ready to respond to any further changes. VARIATION IN CHEMOTAXIS SYSTEMS The well-researched E. coli chemotaxis system serves as an excellent model to understand chemotaxis and reference for the characterization of chemotaxis in other bacteria (Fig. 1).However, comparative genomics and experimental analysis of other systems highlight its relative simplicity and the degree of variation in chemotaxis systems. In comparison, Rhodobacter sphaeroides possesses multiple chemosensory pathways, each separately localized in the
cell and responsible for different functions (Armitage and Schmitt, 1997; Szurmant and Ordal, 2004). A similar multiple pathway scenario exists in Vibrio cholerae where there are three chemotaxis pathways, and it has been suggested that at least one of these is involved in virulence with one of the sensory receptors implicated in regulation of the cholera toxin (CT) and toxin-coregulated pilus (TCP) cascade (Butler and Camilli, 2005; Hyakutake et al., 2005; Lee et al., 2001). On another level, a major difference in chemotaxis mechanism is illustrated by Bacillus subtilis, where, in contrast to E. coli, attractant binding to a receptor results in an increase and not a decrease in CheA kinase activity; therefore, the decrease, rather than increase, of CheY-P binding to the flagellar motor induces tumbling motion (Garrity and Ordal, 1995; Kirby et al., 2001; Kristich and Ordal, 2002, 2004). In E. coli, the complement of chemosensory receptors is relatively small, with five MCPs, and H. pylori appears to encode only four receptors (Andermann et al., 2002; Terry et al., 2006; Williams et al., 2007). Contrast this with bacteria such as Pseudomonas aeruginosa, which has 26 MCPs (Guvener et al., 2006; Shitashiro et al., 2005), subsets of which are differentially expressed. Receptor adaptation systems also show variation; for example, in B. subtilis, other components contribute, such as CheC and CheD (Kirby et al., 2001; Kristich and Ordal, 2002, 2004). Other additional components are present in B. subtilis, with CheV, a protein consisting of both a Chew domain and RR domain being a pertinent example (see below); here, CheV, along with CheC, has been shown to be involved in a methylationindependent adaptation of receptors (Rosario et al., 1994; Saulmon et al., 2004). There are examples of RR domains attached to CheA proteins, for example, H. pylori (Foynes et al., 2000), and redundancy in component content in comparison to E. coli is well illustrated in R sphaeroides, which has multiple copies of chew, cheA, and cheY that encode multiple chemotaxis systems adapted to different niches (Martin et al., 2001; Porter et al., 2006; Shah et al., 2000; Szurmant and Ordal, 2004). Here different CheA proteins appear to possess variant subsets of domains to cluster with particular Chew proteins and sensory receptors to phosphorylate specific CheYs. Both R sphaeroides and Sinorhizobium meliloti lack a CheZ phosphatase, and it is thought that the apparently redundant CheYs act as “phosphate sinks” to reduce the level of phosphorylation of the CheY that interacts with the flagellar motor (Szurmant and Ordal, 2004). CheZ is also absent in B. subtilis, where an alternative mechanism involving CheC, CheD, CheX, and FliY (see below) is thought to operate (Szurmant et al., 2004, Szurmant and Ordal, 2004).
CHAPTER 20
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
CHEMOTAXIS PATHWAY IN CAMPYLOBACTER Genome Content and Organization of C. jejuni Chemotaxis Pathway Components The unique complement of chemotaxis pathway components found in C. jejuni highlights the important differences between the C. jejuni chemotaxis system and those found in other bacteria (Marchant et al., 2002). The specific characteristics of the chemotaxis mechanism and the exact role of the various sensory and signal transduction components need to be elucidated if we are to fully appreciate intestinal colonization by C. jejuni. Study of the genome of C. jejuni NCTC 11168 (Marchant et al., 2002; Parkhill et al., 2000) revealed that directed motility in Campylobacterspp., like that of other bacteria, appears to be governed by the universal two-component, regulator-based backbone. This signal transduction pathway backbone involves a number of MCP-like genes, termed Tlps (transducer-like -proteins) that encode the receptors that form a complex with the products of single cheA and chew genes. The complex controls the phosphorylation status of the single CheY encoded by the genome that interacts with the flagellar motor. Diversity in the ligand binding domains of the Tlps and notable differences in the complement of RR domains indicate that the overall mechanism of chemotaxis signal transduction in campylobacters is likely to be unique. The complexity of the signal transduction pathway in campylobacters appears to lie somewhere between the simpler E. coli pathway and the more complex pathways such as that of R. sphaeroides and V. cholerae (Szurmant and Ordal, 2004), each genus displaying some minor variations. The genome encodes 10 possible Tlp receptor proteins and two aerotaxis (Aer) orthologues that potentially feed signals into the single CheA-CheWCheY signal transduction pathway backbone. The presence of orthologues of cheB and cheR genes indicates that receptor adaptation is likely to occur in some form in campylobacters, although with some differences because the methylesterase CheB lacks the regulatory RR domain. The C. jejuni CheA, however, does contain a RR domain with similarity to CheY, and another such RR domain is encoded by cheV, the B. subtilis orthologue, present in C. jejuni. Like other nonenterobacterial chemotaxis systems, no CheZ appears to be produced by C. jejuni, and therefore the mechanism by which CheY is dephosphorylated must be addressed. No other genes known to be directly involved in chemotaxis were found in the C. jejuni genome (Marchant et al., 2002; Parkhill et al., 2000).
355
The six C. jejuni chemotaxis signal transduction pathway genes are located in three separate regions of the genome. In two regions, the che genes are located with genes of apparently unrelated function, and the distances between open reading frames and the strand-specific grouping of open reading frames suggests an operon arrangement. Thus, as noted with other apparent operons (Parkhill et al., 2000), some chemotaxis genes may be cotranscribed and coregulated with genes of unrelated function. The cheY gene is located adjacent to the Pgl protein glycosylation gene cluster (Parkhill et al., 2000) and possibly at the start of an operon, where no other gene is likely to be involved in chemotaxis. The second region includes the genes cheV, cheA, and chew, which are located next to one another, and the genes flanking the three che genes in the operon do not appear to be associated with chemotaxis. The third region of the C. jejuni chromosome in which che genes are located contains the genes cheR and cheB, which are likely to be cistronic and not cotranscribed with the flanking genes, rpiB and pebC. The overall organization of the che genes in C. jejuni is similar to that seen in H. pylori, except that in H. pylori, the presence of any cheR and cheB homologues is yet to be identified (Tomb et al., 1997). The chemosensory receptor homologues revealed by the genome sequence of C. jejuni 11168 (Parkhill et al., 2000) are likely to be important in how C. jejuni monitors the environment. It is not possible from sequence data alone to determine which receptors are directly involved in chemotaxis. Nevertheless, the Tlps in strain 11168 can be classified into several groups (Marchant et al., 2002) according to a putative function assigned by their homology to known proteins in other organisms, particularly the high sequence similarity encoding the transmembrane and cytoplasmic coiled-coil signaling domains of predicted Tlps. Group A receptors Group A consists of Tlpl, Tlp2, Tlp3, Tlp4, Tlp7, and TlplO. These receptors contain three distinct domains, periplasmic, cytoplasmic, and transmembrane, consistent with membrane-spanning chemoreceptor proteins (Falke et al., 1997). The Cterminal cytoplasmic regions are highly homologous to the signaling domains of other chemoreceptors, and the C-terminal signaling domains of Tlp2, Tlp3, and Tlp4 form the three identical genome copies of repeat 1, which begin at the second transmembrane domain. The less conserved periplasmic N-terminal regions are likely to contain the ligand-binding domains. Group A Tlps also show similarity between
356
KOROLIK AND KETLEY
some of the ligand-binding domains with H. pylori receptor proteins TlpA and TlpC, as well as putative TlpA proteins in Helicobacter and Wolinella spp. (Baar et al., 2003; Tomb et al., 1997; Suerbaum et al., 2003). Group A chemoreceptors are most probably responsible for sensing ligands external to the cell. Recent availability of full and partial sequences of nine C. jejuni strains further enables comparison of the receptor sets (Table 1).The group A Tlps that are represented in these strains appear to have high sequence identity at both the DNA and the amino acid levels. Tlpl and TlplO appear to be the most conserved at 98 to 100% identity at the amino acid level, and Tlp2, Tlp3, and Tlp4 slightly less conserved, with the amino acid identity dropping to 92% in some strains. This is an indication that similar to other organisms, group A receptors are highly conserved among C. jejuni strains. There is some diversity in sensor content and therefore in how campylobacters monitor their environment because different subsets of group A receptors are present in each of the sequenced strains (Table l), with Tlpl being the only receptor universally represented. Strain 81116 appears to have an identical group A Tlp receptor set to 11168, and 81-176 shows a similar receptor set to 11168 and 8 1116, with only Tlp3 not represented. This type of receptor-set variation has been recently reported for some uropathogenic strains of E. coli that appear to be missing functional receptors Trg (ribose and galactose) and Tap (dipepides) usually present in fecal isolates (Lane et al., 2006). The genome sequences of strains C. jejuni 8425 and C. jejuni subspecies doylei 269.97 show a limited set of receptors; however, this may be because the sequences of these strains were not complete at the time of analysis. C. jejuni strain 84-25, which was isolated from a patient with meningitis, encodes an additional Tlp, T l p l l , similar to a hyperinvasive C. jejuni strain 520 (Khieongoen and Korolik, 2003; C. Day, personal
communication). This Tlp shares sequence similarity with TcpI, a chemoreceptor involved in stimulating expression of the CT and TCP pathway of V. cholerae and is considered to be involved in virulence. It is yet to be established which subset of C. jejuni strains has this additional receptor or whether it enhances virulence of these strains. Tlp7 (CjO95lc) contains a typical signaling domain but no transmembrane regions or ligand-binding domain. However, the orf may continue upstream of CjO95lc (Parkhill et al., 2000) into CjO952c, which shows similarity to other MCP proteins and consists of a transmembrane domain only. Although in strain 11168 and a number of other strains (81176,81116, CF93-6, and 260-94) Cj0951/52 appear to represent a pseudogene, in strain 93-13, the sequence is not interrupted by a stop codon and is likely to encode a fully functional protein, indicating that at least in some C. jejuni strains, Tlp7 is a functional part of their chemoreceptor set. Groups ByC, and Aer receptors Group B contains one receptor homologue (Tlp9 or CetA) with a predicted structure of a cytoplasmic protein anchored to the membrane by a single Nterminal transmembrane region. The absence of a specific ligand-binding domain implied that Tlp9 may interact with another receptor protein or proteins (Marchant et al., 2002), and this has been found to be the case. Aerl and Aer2 are homologues of cytoplasmic Aer proteins, which play a role in redox sensing (Bibikov et al., 1997; Taylor et al., 1999). Given that aerl and aer2 are adjacent to tlp9, this might indicate that Tlp9 is the cognate signal transducer. Aer2 and Tlp9 were found to be required for energy taxis and were named CetB and CetA for Campylobacter energy taxis, respectively (Hendrixson et al., 2001);the relationship between Aerl and CetA remains unknown.
Table 1. Content of group A chemoreceptors in C. jejuni strains Presence of chemoreceptor in C. jejuni strain: Chemoreceptor Tlpl Tlp2 Tlp3 Tlp4 Tlp7 Tlp 10 T lpll “p, pseudogene.
11168
81116
RM1221
81-176
CG8486
HB93-13
CF93-6
84-25
260.94
+ + + +
+ + + +
+ + +
+ + +
+ +
+ + + +
+ +
+ + + +
+Pa
+P
-
+P
-
+P
-
-
-
-
+
+
-
+
+
+
+ + + + +
-
-
-
-
-
+
-
-
-
+ +
269.97
+ -
+
+P
-
-
-
-
-
CHAPTER 20
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
The final group (group C, comprising Tlp5, Tlp6, and Tlp8) is likely to contain receptor proteins that detect cytoplasmic signals and are similar to family C transducers in H. salinarum (Zhang et al., 1996). These potential receptors have homology to the C-terminal signaling domains of a variety of chemoreceptors, with one (Tlp8) having additional unique N-terminal sequence. The existence of cytoplasmic chemotaxis receptor-like proteins suggests that sensing the internal physiological state of the cell may play an important role in C. jejuni chemotaxis. Sensory Receptor Complex in C . jejuni The E. coli chemotaxis sensory receptor complex contains clustered receptors in signaling organelle-like structures at the cell poles (see above). The complex also includes the scaffolding protein Chew and HK CheA (Fig. 1).A similar complex, which also probably includes CheV, is likely to occur in campylobacters, although the composition and polar organization of the complexes have yet to be verified experimentally. Receptor specificity On the basis of structure, the group A receptors are likely to be the primary chemotaxis receptors for
external ligands and therefore involved in colonization. TlplO (CjOOl9) is the only receptor so far known to be a determinant of chicken cecal colonization (Hendrixson and DiRita, 2004). It is not possible from sequence data alone to predict which receptors are directly involved in chemotactic responses and determine receptor specificity. Little experimental progress has been made on this front, but some advances have been made with Tlpl and CetAB. A role for Tlpl in chemotaxis is supported by the observation that mutation affects chemotaxis (Sandhu et al., 2007). More significantly, the ligand binding specificity of Tlpl has recently been shown to include L-aspartate (Hartley et al., 2007). To date, this is the only receptor-ligand binding specificity identified for an organism other than E. coli and Salmonella spp. The specific effector molecules for other C. jejuni Tlps are currently unknown. The presence of redox sensing proteins may serve to enable C. jejuni to locate optimal oxygen concentrations, and such a mechanism would be especially important for microaerophilic C. jejuni, where oxygen is essential at low concentrations but toxic at the levels found in air. As in E. coli, Aer, Aerl, and Aer2 contain PAS (an acronym of the genes first found to contain the repeat sequences) domains
357
and probably bind the cofactor flavin adenine dinucleotide (Bibikov et al., 1997; Repik et al., 2000), supporting a role in sensing oxygen-related changes in redox potential. PAS domains are involved in signal transduction by cytoplasmic proteins, possibly involving a second transducer protein, and PAS has also been associated with protein-protein interactions in some systems (Taylor et al., 1999). Characterization of nonmotile transposon mutants experimentally verified some of these predictions (Hendrixson and DiRita, 2001). Aer2 and Tlp9 were found to be required for energy taxis (CetB and CetA). However, the mutation of aerl did not affect motility or energy taxis, and thus the role for this receptor remains to be determined. Thus, campylobacters are able to sense energy levels through Aer2/ CetB and transduce the signal into the chemotaxis pathway via Tlp9/ CetA. CheA histidine kinase The basic role of the complexed CheA in the chemotaxis signal transduction backbone is that of a HK capable of phosphorylating CheY. In silico analysis of CheA homologues from C. jejuni and a range of other bacterial species indicates that CheA contains the known HK functional domains (Dutta et al., 1999; Szurmant and Ordal, 2004). The HK region from C. jejuni CheA is 26% identical to E. coli CheA. The P1 phosphorylation domain, the site of histidine phosphorylation, the catalytic kinase domain, including the residues that form the N, G1, F, and G2 boxes and the receptor-docking region, which is probably integral to its function of receptor-Chew interaction (see below), are all conserved in C. jejuni CheA. The P2 response regulator docking domain and the flexible linkers flanking it do not show any conserved regions in C. jejuni, but this region does not appear to be generally conserved between different CheA proteins; this is perhaps to be expected because the domain has no catalytic function, and hence no active site to be conserved. There is a conserved methionine residue at the alternative start site of transcription (Met98), suggesting that the short form of CheA, CheA, (Djordjevic and Stock, 1998), similar to E. coli, could also feature in C. jejuni. This short form, if produced at all in C. jejuni, would lack the P1 phosphorylation domain and could act to modulate phosphorylation of full-size CheA in the complex. Alternatively, CheA, may recruit any CheZ orthologue (see below) to the cluster (Armitage, 1999). The CheA RR domain is found at the C terminus of the protein and appears to be fully functional; the possible role of this domain is discussed below.
358
KOROLIK AND KETLEY
Analysis of the C. jejuni CheA sequence therefore indicates that it is similar to other CheA proteins, including E. coli CheA, along the whole of the length of the HK domain. It seems probable that C. jejuni CheA has a similar function to other CheA in chemotaxis pathways already studied, i.e., it is a histidine kinase, located as part of the ternary complex, that autophosphorylates (with kinase activity modulated by the receptor) and passes the phosphate group onto CheY. As part of the signal complex, CheA will interact with Chew, and evidence of such an interaction was seen in a high-throughput yeast two-hybrid analysis of C. jejuni protein interactions (Parrish et al., 2007). CheA mutants have been shown to have a reduced chemotaxis phenotype on semisolid agar (Colegio et al., 2001; Hendrixson and DiRita, 2001) while retaining motility (Bridle et al., 2007). Moreover, the consequences of the disabled chemotaxis signal transduction system of CheA mutants are shown by an impairment of intestinal colonization and cell invasion (Chang and Miller, 2006; Golden and Acheson, 2002). It is likely that the less conserved P2 domain does interact with CheY because CheA is the only protein to contain a HK domain. However, given the lack of a RR domain in CheB and the presence of RR domains on CheA itself and CheV, the P2 domain in CheA may reflect the unique aspects of the C. jejuni chemotaxis system. Whether CheA is capable of phosphorylating all of the CheY-like RR domains in C. jejuni is currently under investigation. Interaction studies in a yeast-based (Parrish et al., 2007) and bacterial systems (Bridle et al., 2007) show that CheA can bind to CheY and the latter system has revealed an interaction with CheV (see below). Chew and CheV In E. coli, the signaling complex depends on the presence of the Chew scaffolding protein (Fig. 1). Chew is a small (18 kDa) cytoplasmic protein with no known regulatory or catalytic function (Armitage, 1999), but despite the lack of active catalytic sites, the amino acid sequences from C. jejuni and other bacteria still show many conserved residues throughout the protein. This suggests that the C. jejuni Chew is likely to have a similar structure and probably also acts as an essential structural component of the ternary complex. Indeed, mutation of chew does affect chemotaxis (Bridle et al., 2007). The C. jejuni Chew does, however, have three insertions when compared with the E. coli sequence, two of 6 amino acids and one of 7 amino acids. Two of these are located in the N-terminal region of the protein and are shared by H. pylori (Tomb et al., 1997). These differences may
possibly reflect the presence of a RR domain fused to CheA and/or the additional incorporation of CheV into the complex. CheV is a composite protein that consists of an N-terminal Chew domain fused to a C-terminal RR domain. The function of CheV is not fully understood, but B. subtilis Chew and CheV are, in part, functionally redundant because the W domain of CheV can partly substitute for the lack of Chew (Rosario et al., 1994). Chew and CheV were thought to function together in the same receptor bound multiprotein complex. A comparison of the CheV amino acid sequences from C. jejuni, H. pylori, and B. subtiEis showed that they are very similar, with several highly conserved clusters, as well as many shared identical or similar residues spaced throughout the whole sequence. The high degree of conservation between the sequences of W domains of CheV and Chew proteins suggests that CheV, in addition to Chew, is likely to be somehow incorporated into the sensory complex. In a yeast two-hybrid (Y2H) analysis of C. jejuni (Parrish et al., 2007), Chew interacted with CheA, but importantly was also found to interact with receptors. Chew interactions were only observed with Tlp4 and Tlp8 (group C Tlp lacks a periplasmic sensory domain), and not with other Tlps. With respect to CheV, receptor interactions were observed with Tlp4, Tlp6 (also a group C Tlp), and Tlp8. In contrast, other Y2H data (Shewell and Korolik, 2007) has shown Tlpl interacts with CheV, but not Chew. Keeping in mind the issues of using a yeast-based system, these observations suggest that only one group A chemoreceptor, Tlp4, may be the receptor that incorporates both CheV and Chew into the sensory receptor complex. All of the C. jejuni chemoreceptor homologues share the highly conserved signaling domain (HCD). Multiple alignment of over 120 chemotaxis transducer sequences from different evolutionary branches of Bacteria as well as Archaea generated a 90% consensus for a signature motif in the core of this signaling domain or HCD (Zhulin, 2001). Because the HCD is believed to interact physically with Chew, the presence of this sequence suggests that all of the receptors are involved in forming a signaling complex (Le Moual and Koshland, 1996). Clearly there may be variation in the relative levels of Chew and CheV in sensory complexes with some possibly specific to either Chew or CheV. Different incorporation into the complex may result from Chew and CheV competition for position, which may further depend on the phosphorylation state of CheV, introducing a further level of regulation into the system. Given the fact that chew, cheV, and cheA appear under the same transcriptional control, over-
CHAPTER 20
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
all levels of incorporation may not reflect the regulation of expression unless posttranscriptional regulation is a factor. Until further analysis of Tlp-Chew and Tlp-CheV interactions are performed, it is difficult to discern whether the results of the Y2H system can be attributed to unusual composition of C. jejuni ternary signaling complexes or limitations of a highthroughput yeast-based system where some interactions may remain undetected because of the lack of phosphorylation affecting the protein-protein interactions. Nevertheless, it is likely that both types of complexes may be present in C. jejuni where some receptors will bind only Chew or CheV and others may bind both. Therefore, a model proposing large arrays of mixed receptors would be compatible with the involvement of both Chew and CheV in a cluster, interacting with specific receptors and CheA. If CheV is present in sensory complexes, then the role of the protein in C. jejuni, and specifically this function of the RR domain, must be considered. Adaptation: CheR and CheB The relative levels of methylation by CheR and demethylation by CheB of MCP receptors over time forms the basis for a rudimentary chemical or molecular memory that enables a chemotactic cell to detect changing levels of ligand. C. jejuni contains both CheR and CheB (see below), and the presence of possible near-consensus methylation sites (Le Moual and Koshland, 1996) in Tlpl, Tlp2, Tlp3, Tlp4, and Tlp9 leads one to expect that chemoreceptors in C. jejuni can be methylated. Absence of sufficiently close consensus sequences in Tlp5, Tlp6, Tlp7, Tlp8, and TlplO indicates that not all chemoreceptors are methylated, although other currently unrecognized modification mechanisms may exist or perhaps no adaptation occurs, reflecting a role in nonchemotaxis signaling. The lack of CheR and CheB in H. gylori (Tomb et al., 1997) may provide evidence for adaptation by mechanisms not involving methylation, and these may be available in C. jejuni. C. jejuni CheR shows highest identity to CheR from B. subtilis (18% identity) than to CheRs from Salmonella enterica serovar Typhimurium (15% identity) and E. coli (14% identity). This is a lower level of similarity than for some of the other chemotaxis components, but there are shared residues throughout the CheR sequences, indicating similarity of both of the domains. Residues associated with the positioning and hydrogen bonding of the Sadenosylmethionine (AdoMet) substrate (Djordjevic and Stock, 1997) are identical or similar in C. jejuni, and therefore C. jejuni CheR is likely to have a sim-
359
ilar methyltransferase activity to its S. enterica serovar Typhimurium equivalent. CheB consists of a C-terminal methylesterase domain that is only active when the N-terminal RR domain is phosphorylated by phospho-CheA. Comparison of the C. jejuni CheB sequence to the amino acid sequences of E. coli, S. enterica serovar Typhimurium, and B. subtilis CheBs makes it immediately clear that the C. jejuni CheB has no RR domain and consists of the methylesterase domain only; this CheB structure was unique to C. jejuni. All of the residues known to be necessary for methylesterase function are present in C. jejuni CheB. The sequence of the C. jejuni methylesterase domain itself is similar to other CheB sequences (e.g., 25% identity to E. coli CheB), with the catalytic triad and residues implicated in active site positioning being conserved. Both of the nucleotide binding fold consensus sequences from the other CheBs (GXSGG and GXGXXG) are present and are in the appropriate positions (West et al., 1995). In other CheBs, the unphosphorylated receiver domain inhibits the methylesterase activity by packing against the C-terminal domain, directly inhibiting access to the active site (Djordjevic and Stock, 1998). The isolated S. enterica serovar Typhimurium methylesterase domain retains catalytic activity and specificity for the transmembrane chemoreceptors (West et al., 1995). The absence of the receiver domain in C. jejuni will relieve the inhibition of methylesterase activity, and a likely prediction for C. jejuni would be Tlp modification involving a constitutively active methylesterase and methyltransferase. The consequence for C. jejuni would suggest a functionally redundant constant rate of CheR and CheB methylation or demethylation and therefore an absence of adaptation. Given the apparent absence of CheB control, there may be some other mechanism for regulating the activity of CheR and CheB, thus providing the necessary link with the chemotaxis signal in C. jejuni. One possibility is that regulation is built into the chemoreceptor itself. It has been proposed that in response to ligand binding to the periplasmic domain of the receptor, conformational flexibility (Djordjevic and Stock, 1998) (Mowbray and Sandgren, 1998) in the cytoplasmic domain could either preferentially sequester or expose methylation sites (Stock and Koshland, 1981; Terwilliger et al., 1986). In C. jejuni Tlp, conformational change after ligand binding might limit esterase access to certain residues, and there might also be asymmetry in the rates of the methylating and demethylating enzymes. Experimental evidence from E. coli indicates that the C-terminal domain of CheB alone is not enough to allow wild-type chemotaxis (Djordjevic and Stock, 1998). It is pos-
360
KOROLIK AND KETLEY
sible that the RR domain in CheB has other functions, for example, to modulate the level of phosphoCheA, which in C. jejuni is mediated by other RR domains in CheA and CheV (see below). One can speculate, therefore, that the methylesterase domain alone, mediated by conformational changes in the chemoreceptor, could be sufficient for efficient chemotaxis in organisms with the extra RR domains, such as C. jejuni. However, this does not explain why the absence of the RR domain in C. jejuni CheB is so unusual in chemotaxis systems. An alternative possibility is that one of the extra RR domains in either CheA or CheV could serve to inhibit methylesterase activity when phosphorylated. The structure of the catalytic domain of CheB does present many potential targets for inhibitory interactions (Lupas and Stock, 1989). In B. subtilis, the CheV-RR domain is thought to have a role in methylation (Rosario et al., 1994). CheV could either inhibit CheB directly, as part of the sensory complex, or it could simply be involved in causing
Any RR
0
CheY
@
phosphorylation-dependent changes in the conformation of the complex that regulate CheB access to the methylation sites. The former scenario would be supported by evidence of an interaction between CheV and CheB; however, this was not forthcoming in the recent high-throughput Y2H analysis of C. jejuni protein interactions (Parrish et al., 2007). Keeping in mind the caveats of the use of a yeast-based system, both CheA-CheV and CheB-CheV interactions have been observed in an alternative system (Bridle et al., 2007), which would support a role for CheV in the control of CheB activity in C. jejuni. A model proposing the control of CheB by CheV is presented in Fig. 2. CheY, Dephosphorylation, and RR Domains The CheY amino acid sequence shows all the characteristics of a RR as well as regions of homology generally conserved among CheYs. Most importantly, Asp57 (all numbers as in E . coli), the site of phos-
Figure 2. Model for the control of CheB activity in campylobacters. (1) CheA-mediated kinase activity is inhibited when chemoreceptors bind chemoattractants. In consequence, phosphorylation of the response regulator domain in CheV is reduced (probably involving the phosphatase CheZ; Fig. 3). Unphosphorylated CheV inhibits the methylesterase activity of CheB, resulting in increased methylation of receptors due to CheR. (2) In the absence of chemoattractant CheA phosphorylates the response regulator domain present in CheV. Phospho-CheV ceases to inhibit CheB methylesterase activity, leading to increased demethylation of chemoreceptors.
CHAPTER 20
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
phorylation, is present, as well as those residues associated with catalyzing CheY auto-dephosphorylation (Tomb et al., 1997). Such conservation of residues indicates that CheY, when phosphorylated by CheA, is capable of interacting with the flagellar motor to reverse the direction of rotation. This hypothesis was confirmed because a CheY null mutant is unable to reverse the direction of flagellar rotation or to form a chemotactic response (Marchant et al., 1998; Yao et al., 1997). This phenotype suggests that CheY is the only component of the chemotaxis pathway to interact with the motor, and therefore (at least under the conditions tested), all tactic responses must be integrated at a point higher in the pathway (possibly CheA, as in E. coli) and pass through CheY. Given that CheY may be the only RR domain that interacts with the flagellar motor, this raises the question as to the function or functions of the other chemotaxis-associated RR domains found in C. jejuni. The RR domains in CheA and CheV must have different roles in C. jejuni chemotaxis, and this is supported by the fact that CheY is less similar to CheARR and CheV-RR than to comparable RRs from other bacteria. Without prompt removal of the phosphate from CheY, there will be a persistent rotational bias by the flagellar motor and a slow response to changing chemotactic signals. In E. coli, the phosphatase CheZ removes the phosphate from CheY and therefore terminates the signal transmitted to the motor (Huang and Stewart, 1993; Szurmant and Ordal, 2004) (Fig. 1). In addition, CheY is capable of autodephosphorylation, albeit at a rate much slower (and thus functionally incompatible with a rapid response) than that mediated by CheZ. However, in the majority of other chemotaxis systems, there is no close homologue to CheZ, which is limited to the Betaproteobacteria and Gammaproteobacteria, despite the need to dephosphorylate CheY. It would seem that different bacteria may have evolved distinct mechanisms to remove excess phosphorylated CheY. Multiple RR domains in chemotaxis signal transduction pathways are often found in species that lack CheZ or another phosphatase. Sinorhizobium meliloti, for example, also lacks CheZ and has two chemotaxis proteins containing RR domains (Armitage and Schmitt, 1997). CheY2 interacts with the motor to influence flagellar rotation. In contrast, CheYl is thought to act as a phosphate sink (Armitage, 1999; Szurmant and Ordal, 2004), whereby phosphate groups are passed from CheY2 back to CheA, and then to CheYl (Armitage and Schmitt, 1997; Sourjik and Schmitt, 1998). In acting as a sink, or phosphate store, CheYl removes phosphate from the system, and therefore the pool of phosphorylated CheY2 is
361
limited. A similar situation has been proposed for Rhodobacter sphaeroides (Shah et al., 2000), which has four cheY genes. In S. meliloti and R. sphaeroides, the CheYs are separate proteins, but it is possible that in C. jejuni the RR domains fused to CheA and CheV could also fulfill such a function; the fact that the CheV-RR and CheA-RR domains are more similar to S. meliloti CheYl than CheY2 might support such speculation. As far as CheV is concerned, a role in the control of CheB activity has been proposed here (see above). Returning to the theme of dephosphorylation, there is another group of CheY-P-hydrolyzing proteins, which are distinct as a result of a different structure, that are found in a wide range of eubacteria and archaea (Szurmant and Ordal, 2004). One example of this second group of CheY dephosphorylating proteins is the flagellar motor assembly protein FliY in B. subtilis (Muff et al., 2007; Park et al., 2004). The protein contains a FliN domain and Nterminal region with a domain that binds CheY and motifs, assumed to have a role in CheY-P hydrolysis, that are conserved among this group of proteins. Two other B. subtilis proteins, CheC and CheX, contain a FliY-like region, although neither contains the CheY binding domain and CheX appears to be a truncated version of CheC (Park et al., 2004). In B. subtilis, CheC alone only poorly dephosphorylates CheY, but this activity is enhanced by CheD (Kirby et al., 2001; Saulmon et al., 2004). In Borrelia burgdorferi, CheX was found to be required for chemotaxis and able to dephosphorylate CheY in vitro (Motaleb et al., 2005). FLY homologues are found in gram-positive bacteria, some spirochetes, and Thermatoga, while the CheC/X type is widespread among bacterial phyla (Szurmant and Ordal, 2004). C. jejuni does not contain a CheC or CheX homologue; however, the annotated fliY (11168:CjOO59c) is a poor match to B. subtilis FLY, and the protein product is not able to interact with CheY (Bridle et al., 2007). Until recently, a phosphate sink mechanism for the control of CheY phosphorylation levels was the most likely scenario in C. jejuni; however, recent observations on chemotaxis in H. pylori may suggest a wider distribution of the CheZ group of CheY-P hydrolyzing proteins (Terry et al., 2006). Further analysis of revertants of chew mutants led to the identification of HPO170. Sequence analysis revealed that HP0170 appears to be a distant homologue of CheZ, and the expressed protein was found to be able to dephosphorylate CheY-P in vitro. C. jejuni was one of a range of other bacteria containing this gene, and our preliminary studies have shown Cj0700 to have a role in chemotaxis and capable of interacting with chemotaxis-related RR domains (0. Bridle and J.
362
KOROLIK AND KETLEY
Ketley, unpublished data). Therefore, in C. jejuni, signal termination by CheY dephosphorylation is likely to involve the CheZ orthologue, Cj0700 (Fig. 3).
CONCLUSIONS The genome sequences of C. jejuni strains reveal a range of putative chemotaxis protein orthologues scattered throughout the genome. C. jejuni can be predicted to encode between three and six membrane-spanning, methyl-accepting chemotaxis receptors in a strain-specific manner (Table 1).These receptors consist of two membrane-spanning homologues, which may or may not be involved in chemotaxis, one membrane-bound MCP predicted to
Any RR CheY
mediate aerotactic and energy-related responses, and three cytoplasmic receptors, which may sense the internal physiological and metabolic state of the cell. All of these proteins (except Tlp7 in some strains) appear to be capable of forming a ternary (quaternary?) complex and therefore of relaying signals into the chemotaxis signal transduction pathway to affect flagellar rotation and cause a behavioral response to the signal. Tlp7 is present in all but one sequenced strain as a pseudogene and has a slightly altered HCD, suggesting that it may not form a ternary complex with Chew in these strains. The gene content and the nature of the domains encoded by the C. jejuni chemotaxis gene complement, combined with recent mutational and twohybrid protein interaction analyses, can be used to
U 0
CheB CheR CheZ
Figure 3. Model for the dephosphorylation of CheY and other chemotaxis-related response regulator domains in campylobacters. Cj0700, the proposed CheZ, dephosphorylates phospho-CheY, terminating the signal transmitted to the motor by CheY. In the absence of attractant, CheA is proposed to phosphorylate CheV, and CheZ will dephosphorylate CheV to terminate continued CheB methylesterase activity. Phosphorylation of the CheA response regulator domain CheA-RR is also proposed to be reversed by CheZ, but the regulatory role of the CheA-RR domain remains to be determined.
Any RR domain
CheY
0 0
0
CheR
0
CheZ Ligand
0
Tlp group C
Figure 4. Model of chemotaxis signal transduction in campylobacters. Three pathways are proposed to be associated with different chemosensory complexes focused on the structure-based grouping of predicted chemoreceptors. Group A receptors, possibly arranged in large array clusters of mixed receptor specificity and positioned at the cell poles, are predicted to sense periplasmic signals. Binding of ligands to periplasmic domains of group A receptors might involve accessory periplasmic ligand-binding proteins (not shown). The specificity of Tlpl has been shown to be aspartate. For the group B receptor CetA, a chemotactic response to redox potential changes is mediated by interaction with CetB. The group C receptors might play a role in chemotactic responses to intracellular signals, possibly by other cytoplasmic ligand-binding proteins (not shown). Some receptors, for example Aerl, might transduce signals into other nonchemotaxis systems. Signal loss, for example decreasing binding of attractant to the receptor, is transduced to CheA in the complex activating the CheA kinase domain. Complexes may contain Chew alone or both Chew and CheV (shown left and right of group A receptors, respectively). After autophosphorylation of CheA, the phosphate is transferred to CheY, and phospho-CheY then binds to FliM on the flagellum motor. Sufficient binding of phosphoCheY to the flagellum motor leads to reversal from counterclockwise to clockwise rotation (not shown). System adaptation, which resets the signaling properties of the receptor, is presumed to occur via reversible methylation by CheB and CheR. The CheA kinase domain might also pass phosphate to CheV that in turn controls CheB activity. Signal termination by dephosphorylation of CheY is mediated by CheZ, which also may dephosphorylate CheV and the response regulator domain on CheA.
364
KOROLIK AND KETLEY
predict a role for the chemotaxis proteins expressed in C. jejuni. Some of the elements of the C. jejuni signaling pathway such as CheY, CheR, and Chew are similar to those found in E. coli, whereas others are similar to components of chemotaxis pathways of other bacteria, including, for example, CheV in B. subtilis, and at least one protein, CheB, lacking a RR domain, is unique to C. jejuni. CheA and Chew appear to form a ternary complex with the chemoreceptors, functioning to phosphorylate CheY, which then interacts with the flagellar motor to cause clockwise flagellar rotation. But there are some important departures from the E. coli paradigm that affect the adaptation and termination of the chemotactic response in C. jejuni. There is evidence that methylation- and demethylation-based adaptation of the chemoreceptors also occurs in C. jejuni. CheR and CheB probably act to methylate and demethylate at least some of the receptors, but in the absence of a CheB-RR domain, the nature of the control mechanism regulating CheB activity is not clear in C. jejuni. CheV is probably involved in at least some of the ternary complexes, and CheV-W is probably at least partly functionally redundant with Chew. There is preliminary evidence of a CheV-RR domain role in methylation-dependent adaptation by phosphorylation-dependent interaction with CheB (Fig. 2). The first appraisals of the likely chemotaxis signal transduction pathway in C. jejuni recognized the apparent lack of CheZ in C. jejuni (Parkhill et al., 2000) and highlighted the need for CheY-associated signal termination (Marchant et al., 2002). An initial hypothesis of a phosphate sink mechanism was proposed on the basis of the presence of additional RR domains and lack of alternative CheY-P hydrolases, such as CheC. However, the recent identification in H. pylori of a distant homologue of CheZ (Terry et al., 2006) has led to the new model (Fig. 3) for CheY dephosphorylation and decoupling of CheY from the flagellum motor that is presented here, which may involve the distant homologue of CheZ in C. jejuni. Chemotactic motility is central to the intestinal lifestyle of C. jejuni and an essential prerequisite to pathogenesis in human disease. The chemoreceptor signal transduction pathway may even have input into the expression of other virulence determinants in campylobacters. An overall model illustrating the possible chemoreceptor signal transduction pathways discussed in this review that are present in C. jejuni is summarized diagrammatically in Fig. 4. Clearly, this model for the most part is speculative, being largely based on the exploitation of genomic sequence data, but ongoing investigation of chemotaxis in C. jejuni is providing experimental support. A
thorough understanding of the chemotaxis system in campylobacters will be important in addressing the problem of the intestinal colonization of poultry and livestock animals and the initiation of disease in humans. REFERENCES Adler, J. 1965. Chemotaxis in Escherichia coli. Cold Spring Harb. Symp. Quant. Biol. 30:289-292. Ames, P., C. A. Studdert, R. H. Reiser, and J. S. Parkinson. 2002. Collaborative signaling by mixed chemoreceptor teams in Escherichia coli. PNAS 99:7060-7065. Andermann, T. M., Y. T. Chen, and K. M. Ottemann. 2002. Two predicted chemoreceptors of Helicobacter pylori promote stomach infection. Infect. Immun. 705877-5881. Armitage, J. P. 1999. Bacterial tactic responses. Adu. Microb. Physiol. 4 1:229-28 9. Armitage, J. P., and R. Schmitt. 1997. Bacterial chemotaxis: Rhodobacter sphaeroides and Sinorhizobium meliloti-variations on a theme? Microbiology 143(Pt. 12):3671-3682. Baar, C., M. Eppinger, G. Raddatz, J. Simon, C. Lanz, 0. Klimmek, R Nandakumar, R. Gross, A. Rosinus, H. Keller, P. Jagtap, B. Linke, F. Meyer, H. Lederer, and S . C. Schuster. 2003. Complete genome sequence and analysis of Wolinella succinogenes. PNAS 100:11690-11695. Bibikov, S. I., R. Biran, K. E. Rudd, and J. S. Parkinson. 1997. A signal transducer for aerotaxis in Escherichia coli. J. Bacteriol. 179~4075-4079. Bren, A., and M. Eisenbach. 2001. Changing the direction of flagellar rotation in bacteria by modulating the ratio between the rotational states of the switch protein F1iM. J. Mol. Biol. 312: 699-709. Bridle, O., R. Sandhu, and J. Ketley. 2007. Identifiing Proteinprotein interactions in the chemotaxis system of Campylobacter jejuni. Zoonoses Public Health 5454. Butler, S. M., and A. Camilli. 2005. Going against the grain: chemotaxis and infection in Vibrio cholerae. Nut. Rev. Microbiol. 3: 611-620. Chang, C., and J. F. Miller. 2006. Campylobacter jejuni colonization of mice with limited enteric flora. Infect. Immun. 74: 5261-5271. Colegio, 0. R., T. J. T. Griffin, N. D. Grindley, and J. E. Galan. 2001. In vitro transposition system for efficient generation of random mutants of Campylobacter jejuni. J. Bacteriol. 183: 2384-23 88. Croxen, M. A., G. Sisson, R. Melano, and P. S. Hoffman. 2006. The Helicobacter pylori chemotaxis receptor TlpB (HP0103) is required for pH taxis and for colonization of the gastric mucosa. J. Bacteriol. 188:2656-2665. Djordjevic, S., and A. M. Stock. 1997. Crystal structure of the chemotaxis receptor methyltransferase CheR suggests a conserved structural motif for binding S-adenosylmethionine. Structure 5545-558. Djordjevic, S., and A. M. Stock. 1998. Structural analysis of bacterial chemotaxis proteins: components of a dynamic signaling system. J. Struct. Biol. 124:189-200. Dutta, R., L. Qin, and M. Inouye. 1999. Histidine kinases: diversity of domain organization. Mol. Microbiol. 34:633-640. Eisenbach, M. 1996. Control of bacterial chemotaxis. Mol. Microbiol. 20:903-910. Falke, J. J., R. B. Bass, S . L. Butler, S. A. Chervitz, and M. A. Danielson. 1997. The two-component signaling pathway of bacterial chemotaxis: a molecular view of signal transduction by
CHAPTER 20
CHEMOSENSORY SIGNAL TRANSDUCTION PATHWAY
receptors, kinases, and adaptation enzymes. Annu. Rev. Cell. Dev. Biol. 13:457-512. Ferrero, R. L., and A. Lee. 1988. Motility of Campylobacterjejuni in a viscous environment: comparison with conventional rod shaped bacteria.]. Gen. Microbiol. 13453-59. Foynes, S., N. Dorrell, S. J. Ward, R. A. Stabler, A. A. McColm, A. N. Rycroft, and B. W. Wren. 2000. Helicobacter pylori possesses two CheY response regulators and a histidine kinase sensor, CheA, which are essential for chemotaxis and colonization of the gastric mucosa. Infect. Immun. 68:2016-2023. Garrity, L. F., and G. W. Ordal. 1995. Chemotaxis in Bacillus subtilis: how bacteria monitor environmental signals. Pharmacol. Ther. 68 :8 7- 104. Golden, N. J., and D. W. K. Acheson. 2002. Identification of motility and autoagglutination Carnpylobacter jejuni mutants by random transposon mutagenesis. Infect. Immun. 70:1761-1771. Grebe, T. W., and J. Stock. 1998. Bacterial chemotaxis: the five sensors of a bacterium. Curr. Biol. 8:R154-157. Guvener, Z. T., D. F. Tifrea, and C. S. Harwood. 2006. Two different Pseudomonas aeruginosa chemosensory signal transduction complexes localize to cell poles and form and remould in stationary phase. Mol. Microbiol. 61:106-118. Hartley, L. E., J. C. Wilson, and V. Korolik. 2007. Expression, purification and characterisation of the sensory domain of Tlpl of C. jejuni. Zoonoses Publ. Health 54:64. Hendrixson, D. R., B. J. Akerley, and V. J. DiRita. 2001. Transposon mutagenesis of Campylobacter jejuni identifies a bipartite energy taxis system required for motility. Mol. Microbiol. 40: 214-224. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Huang, C., and R. C. Stewart. 1993. CheZ mutants with enhanced ability to dephosphorylate CheY, the response regulator in bacterial chemotaxis. Biochim. Biophys. Acta. 1202:297-304. Hugdahl, M. B., J. T. Beery, and M. P. Doyle. 1988. Chemotactic behavior of Campylobacter jejuni. Infect. Immun. 56:15601566. Hyakutake, A., M. Homma, M. J. Austin, M. A. Boin, C. C. Hase, and I. Kawagishi. 2005. Only one of the five CheY homologs in Vibrio cholerae directly switches flagellar rotation. J. Bacteriol. 187~8403-8410. Josenhans, C., and S . Suerbaum. 2002. The role of motility as a virulence factor in bacteria. Int. J. Med. Microbiol. 291:605614. Khieongoen, A., and V. Korolik. 2003. Characterisation of Campylobacter jejuni chemoreceptor protein, MCP. Int. J. Med. Microbiol. 293:69. Kirby, J. R., C. J. Kristich, M. M. Saulmon, M. A. Zimmer, L. F. Garrity, I. B. Zhulin, and G. W. Ordal. 2001. CheC is related to the family of flagellar switch proteins and acts independently from CheD to control chemotaxis in Bacillus subtilis. Mol. Microbiol. 42573-5 85. Kristich, C. J., and G. W. Ordal. 2002. Bacillus subtilis CheD is a chemoreceptor modification enzyme required for chemotaxis. J. Biol. Chem. 277:25356-25362. Kristich, C. J., and G. W. Ordal. 2004. Analysis of chimeric chemoreceptors in Bacillus subtilis reveals a role for CheD in the function of the McpC HAMP domain. J. Bacteriol. 18659505955. Lane, M. C., A. L. Lloyd, T. A. Markyvech, E. C. Hagan, and H. L. Mobley. 2006. Uropathogenic Escherichiu coli strains generally lack functional Trg and Tap chemoreceptors found in the majority of E. coli strains strictly residing in the gut. J. Bacteriol. 188:5618-5625.
365
Lee, A., J. L. O'Rourke, P. J. Barrington, and T. J. Trust. 1986. Mucus colonization as a determinant of pathogenicity in intestinal infection by Campylobacter jejuni: a mouse cecal model. Infect. Immun. 51536-546. Lee, S. H., S. M. Butler, and A. Camilli. 2001. Selection for in vivo regulators of bacterial virulence. PNAS 98:6889-6894. Le Moual, H., and D. E. Koshland, Jr. 1996. Molecular evolution of the C-terminal cytoplasmic domain of a superfamily of bacterial receptors involved in taxis. J. Mol. Biol. 261568-585. Lupas, A., and J. Stock. 1989. Phosphorylation of an N-terminal regulatory domain activates the CheB methylesterase in bacterial chemotaxis. J. Biol. Chem. 264:17337-17342. Lybarger, S. R., and J. R. Maddock. 2000. Differences in the polar clustering of the high- and low-abundance chemoreceptors of Escherichia coli. PNAS 97:8057-8062. Marchant, J., J. Henderson, B. Wren, and J. Ketley. 1998. Role of the cheY gene in the chemotaxis of Campylobacter jejuni, p. 306-311. In A. Lastovica, D. Newel1 and E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. University of Cape Town, Cape Town. Marchant, J., B. Wren, and J. Ketley. 2002. Exploiting genome sequence: predictions for mechanisms of Campylobacter chemotaxis. Trends Microbiol. 10:155-159. Martin, A. C., G. H. Wadhams, and J. P. Armitage. 2001. The roles of the multiple Chew and CheA homologues in chemotaxis and in chemoreceptor localization in Rhodobacter sphaeroides. Mol. Microbiol. 40:1261-1272. Mattick, J. S. 2002. Type IV pili and twitching motility. Annu. Rev. Microbiol. 56:289-3 14. Morgan, R., S . Kohn, S. H. Hwang, D. J. Hassett, and K. Sauer. 2006. BdlA, a chemotaxis regulator essential for biofilm dispersion in Pseudomonas aeruginosa. J. Bacteriol. 188:7335-7343. Motaleb, M. A., M. R. Miller, C. Li, R. G. Bakker, S . F. Goldstein, R E. Silversmith, R. B. Bourret, and N. W. Charon. 2005. CheX is a phosphorylated CheY phosphatase essential for Borrelia burgdorferi chemotaxis. J. Bacteriol. 187:7963-7969. Mowbray, S . L., and M. 0. Sandgren. 1998. Chemotaxis receptors: a progress report on structure and function. J. Struct. Biol. I24 :25 7-275. Muff, T. J., R. M. Foster, P. J. Liu, and G. W. Ordal. 2007. CheX in the three-phosphatase system of bacterial chemotaxis. J. Bacteriol. 189:7007-7013. O'Toole, G. A., and R. Kolter. 1998. Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol. Microbiol. 30:295-304. Ottemann, K. M., and J. F. Miller. 1997. Roles for motility in bacterial-host interactions. Mol. Microbiol. 24:1627-163 1. Park, S. Y., X. Chao, G. Gonzalez-Bonet, B. D. Beel, A. M. Bilwes, and B. R. Crane. 2004. Structure and function of an unusual family of protein phosphatases: the bacterial chemotaxis proteins CheC and CheX. Mol. Cell 16563-574. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S . Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Parrish, J. R., J. Yu, G. Liu, J. A. Hines, J. E. Chan, B. A. Mangiola, H. Zhang, s. Pacifico, F. Fotouhi, V. J. Dirita, T. Ideker, P., Andrews, and R. L. Finley, Jr. 2007. A proteome-wide protein interaction map for Campylobacter jejuni. Genome Biol. 8: R130. Porter, S. L., G. H. Wadhams, A. C. Martin, E. D. Byles, D. E. Lancaster, and J. P. Armitage. 2006. The CheYs of Rhodobacter sphaeroides. J. Biol. Chem. 281:32694-32704.
366
KOROLIK AND KETLEY
Repik, A., A. Rebbapragada, M. S. Johnson, J. 0. Haznedar, I. B. Zhulin, and B. L. Taylor. 2000. PAS domain residues involved in signal transduction by the Aer redox sensor of Escherichia coli. Mol. Microbiol. 36:806-816. Rosario, M. M., K. L. Fredrick, G. W. Ordal, and J. D. Helmann. 1994. Chemotaxis in Bacillus subtilis requires either of two functionally redundant Chew homo1ogs.J. Bacteriol. 176:27362739. Sandhu, R., 0. Bridle, and J. Ketley. 2007. Characterisation of the Methyl-Accepting Chemotaxis Protein (MCP) Receptors in Campylobacter jejuni. Zoonoses Public Health 54:92. Saulmon, M. M., E. Karatan, and G. W. Ordal. 2004. Effect of loss of CheC and other adaptational proteins on chemotactic behaviour in Bacillus subtilis. Microbiology 150581-589. Shah, D. S., S. L. Porter, D. C. Harris, G. H. Wadhams, P. A. Hamblin, and J. P. Armitage. 2000. Identification of a fourth cheY gene in Rhodobacter sphaeroides and interspecies interaction within the bacterial chemotaxis signal transduction pathway. Mol. Microbiol. 35:lOl-112. Shelwell, L. K., and V. Korolik. 2007. Investigation of proteinprotein interactions involved in the chemotaxis pathway of Campylobacter jejuni. Zoonoses Publ. Health 54:94. Shitashiro, M., H. Tanaka, C. S. Hong, A. Kuroda, N. Takiguchi, H. Ohtake, and J. Kato. 2005. Identification of chemosensory proteins for trichloroethylene in Pseudomonas aeruginosa. J. Biosci. Bioeng. 99:396-402. Sourjik, V. 2004. Receptor clustering and signal processing in E. coli chemotaxis. Trends Microbiol. 12569-576. Sourjik, V., and R. Schmitt. 1998. Phosphotransfer between CheA, CheY1, and CheY2 in the chemotaxis signal transduction chain of Rhizobium meliloti. Biochemistry 37:2327-2335. Stock, J., and M. Levit. 2000. Signal transduction: hair brains in bacterial chemotaxis. Curr. Biol. 10:Rll-R14. Stock, J. B., and D. E. Koshland, Jr. 1981. Changing reactivity of receptor carboxyl groups during bacterial sensing.J. Biol. Chem. 256:10826-10833. Suerbaum, S., C. Josenhans, T. Sterzenbach, B. Drescher, P. Brandt, M. Bell, M. Droge, B. Fartmann, H.-P. Fischer, 2. Ge, A. Horster, R. Holland, K. Klein, J. Konig, L. Macko, G. L. Mendz, G. Nyakatura, D. B. Schauer, Z. Shen, J. Weber, M. Frosch, and J. G. Fox. 2003. The complete genome sequence of the carcinogenic bacterium Helicobacter hepaticus. PNAS 100: 7901-7906. Szurmant, H., T. J. Muff, and G. W. Ordal. 2004. Bacillus subtilis CheC and FliY are members of a novel class of CheY-Phydrolyzing proteins in the chemotactic signal transduction cascade. J. Biol. Chem. 279:21787-21792. Szurmant, H., and G . W. Ordal. 2004. Diversity in chemotaxis mechanisms among the Bacteria and Archaea. Microbiol. Mol. Biol. Rev. 68:301-319. Szymanski, C. M., M. King, M. Haardt, and G. D. Armstrong. 1995. Campylobacter jejuni motility and invasion of Caco-2 cells. Infect. Immun. 63:4295-4300.
Takata, T., S. Fujimoto, and K. Amako. 1992. Isolation of nonchemotactic mutants of Campylobacter jejuni and their colonization of the mouse intestinal tract. Infect. Immun. 60:35963600. Taylor, B. L., I. B. Zhulin, and M. S. Johnson. 1999. Aerotaxis and other energy-sensing behavior in bacteria. Annu. Rev. Microbiol. 53 :103-1 28. Terry, K., A. C. Go, and K. M. Ottemann. 2006. Proteomic mapping of a suppressor of non-chemotactic chew mutants reveals that Helicobacter pylori contains a new chemotaxis protein. Mol. Microbiol. 61:871-882. Terry, K., S. M. Williams, L. Connolly, and K. M. Ottemann. 2005. Chemotaxis plays multiple roles during Helicobacter p y lori animal infection. Infect. Immun. 73:803-8 11. Tenvilliger, T. C., J. Y. Wang, and D. E. Koshland, Jr. 1986. Surface structure recognized for covalent modification of the aspartate receptor in chemotaxis. €"AS 83:6707-6710. Tomb, J. F., 0. White, A. R. Kerlavage, R. A. Clayton, G. G. Sutton, R D. Fleischmann, K. A. Ketchum, H. P. Klenk, S. Gill, B. A. Dougherty, K. Nelson, J. Quackenbush, L. Zhou, E. F. Kirkness, S. Peterson, B. Loftus, D. Richardson, R. Dodson, H. G. Khalak, A. Glodek, K. McKenney, L. M. Fitzegerald, N. Lee, M. D. Adams, E. K. Hickey, D. E. Berg, J. D. Gocayne, T. R. Utterback, J. D. Peterson, J. M. Kelley, M. D. Cotton, J. M. Weidman, C. Fujii, C. Bowman, L. Watthey, E. Wallin, W. S. Hayes, M. Borodovsky, P. D. Karp, H. 0. Smith, C. M. Fraser, and J. C. Venter. 1997. The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388539547. van Alphen, L. B., N. M. C. Bleumink-Pluym, K. D. Rochat, B. W. M. van Balkom, M. M. S. M. Wosten, and J. P. M. van Putten. 2008. Active migration into the subcellular space precedes Campylobacter jejuni invasion of epithelial cells. Cell. Microbiol. 1053-66. West, A. H., E. Martinez-Hackert, and A. M. Stock. 1995. Crystal structure of the catalytic domain of the chemotaxis receptor methylesterase, CheB. J. Mol. Biol. 250:276-290. Williams, S. M., Y. T. Chen, T. M. Andermann, J. E. Carter, D. J. McGee, and K. M. Ottemann. 2007. Helicobacter pylori chemotaxis modulates inflammation and bacterium-gastric epithelium interactions in infected mice. Infect. Immun. 75:37473757. Yao, R., D. H. Burr, and P. Guerry. 1997. CheY-mediated modulation of Campylobacter jejuni virulence. Mol. Microbiol. 23: 1021-1031. Zhang, P., C . M. Khursigara, L. M. Hartnell, and S. Subramaniam. 2007. Direct visualization of Escherichia coli chemotaxis receptor arrays using cryo-electron microscopy. PNAS 104: 3777-3781. Zhang, W., A. Brooun, J. McCandless, P. Banda, and M. Alam. 1996. Signal transduction in the archaeon Halobacterium salinarium is processed through three subfamilies of 13 soluble and membrane-bound transducer proteins. PNAS 93:4649-4654. Zhulin, I. B. 2001. The superfamily of chemotaxis transducers: from physiology to genomics and back. Adv. Microb. Physiol. 45:157-198.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 23
Animal Models of Campylobacter jejuni Infections LINDAS. MANSFIELD, DAVIDB. SCHAUER,AND
JAMES
G. Fox
1980; Fox et al., 1987), and swine (Mansfield and Urban, 1996), but limitations exist with these models. Monkeys may provide the best naturally occurring disease models, but monkey experimentation is prohibitively expensive and not available to most institutions. Neonatal swine-either conventionally reared, gnotobiotic, or colostrum deprived (Babakhani et al., 1993; Eaton et al., 1989; Mansfield et al., 2003)-and weaning aged ferrets (Bell and Manning, 1990; Poly et al., 2007), all declared to be specific pathogen free before inoculation, have been most widely used as disease models. These models have allowed many advances but have limitations for delineation of host factors because all of these animals are genetically outbred. The wide variety of factors that likely interact to produce disease after infection with C. jejuni makes development of any animal model difficult. In humans, disease symptoms and the course of infection as a result of C. jejuni infection can vary dramatically (Pazzaglia et al., 1991; Skirrow and Blaser, 2000). Epidemiological and other data suggest that the disease outcome depends on a variety of factors, such as the dose and genetics of the C. jejuni strain, innate and adaptive immune status of the patient, the composition of the patient’s enteric microflora, and the number of repeat challenges (Blaser, 1997; Oberhelman et al., 2003; Pazzaglia et al., 1991). Factors such as passive immunity from the dam, age at challenge, and exposure to endemic strains from animals in the environment (e.g., chickens) can be especially important in determining the outcome of exposure in children in developing countries (Blaser et al., 1980; Oberhelman et al., 2003; Pazzaglia et al., 1991). With such a complex interplay of factors, reliable murine infection models that use the widely available array of inbred and outbred strains and genetic knockouts
Many studies have sought to establish tractable animal models for study of C. jejuni colonization and enteritis. Large animal models and ferrets enabled early progress in the study of C. jejuni pathogenesis, transmission, and vaccines (Babakhani et al., 1993; Bell and Manning, 1990; Boosinger and Powe, 1988; Caldwell et al., 1983; Fox, 1992; Russell et al., 1992; Walker et al., 1988). Chickens (Biswas et al., 2006; Hendrixson and DiRita, 2004; Jones et al., 2004; Knudsen et al., 2006; Ringoir et al., 2007), hamsters (Humphrey et al., 1985), ferrets (Bell and Manning, 1990; Fox, 1992), dogs (Olson and Sandstedt, 1987), primates (now including New World monkeys Aotus nancyrnae Uones et al., 20061) (Fernando et al., 2007; Islam et al., 2006; Russell et al., 1989, 1993; Sestak et al., 2003), rabbits (Caldwell et al., 1983; Pavlovskis et al., 1991; Walker et al., 1988), mice (Ziprin et al., 2001), and pigs (Baqar et al., 1993; Hu and Kopecko, 2000) have been inoculated experimentally with C. jejuni by various routes to mimic the course of infection in humans. Significant progress has been made with these models to understand pathogenic mechanisms of C. jejuni strains having targeted or spontaneous mutations in genes encoding such functions as chemotaxis, motility, adherence, invasion of eukaryotic cells, membrane transport, heat-shock response, cytolethal distending toxin (CDT) production, phospholipase activity, lipopolysaccharide synthesis, and two-component signal transduction (AbuOun et al., 2005; Fox et al., 2004; Fry et al., 2000; Hendrixson and DiRita, 2004; Konkel et al., 1998; MacKichan et al., 2004; Purdy et al., 2000; Wu et al., 1994; Ziprin et al., 2001). Naturally occurring infections with campylobacters, including C. jejuni, have been reported in juvenile rhesus monkeys (Fernando et al., 2007; Sestak et al., 2003), ferrets (Forbes and Gros, 2001), dogs (Bruce et al.,
Linda S. Mansfield Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI 48824. David B. Schauer and James G . Fox Division of Comparative Medicine and Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139.
367
368
MANSFIELD ET AL.
to reproduce human disease outcomes are essential
to facilitate research into the mechanisms of pathogenesis and into the development of new therapies and vaccines. Other advantages are small size, comparatively inexpensive housing costs, and amenability to manipulation of murine models to diet and microbiota. Important early studies have been conducted to test various inbred and outbred strains of mice as colonization and disease models. In many experimental trials, outbred and inbred mice have been investigated as C. jejuni disease models, including BALB/c mice, C57BL/6 mice, CBA mice, DBA/2 mice, ddY mice, HA-ICR mice, Hsd:ICR mice, and Swiss or Swiss Webster mice (Baqar et al., 1993; Berndtson et al., 1994; Hu and Kopecko, 2000; Jesudason et al., 1989; Lee et al., 1999; Manninen et al., 1982; RuizPalacios et al., 2003; Szymanski et al., 2002; Vuckovic et al., 1998). Mice with limited enteric flora and/ or with spontaneous or targeted alterations in immune function have also been explored as possible models for C. jejuni infection, including nude BALB/c mice, C3H and SCID-C3H limited flora mice, RAG-2 mice, SCID-Beige, and C.B-17-SCIDBeige mice (Hodgson et al., 1998; MacKichan et al., 2004; Purdy et al., 2000). Infant and adult mice have been inoculated intragastrically, intranasally, and intraperitoneally with C. jejuni in efforts to produce colonization and disease models (Abimiku and Dolby, 1988; Abimiku et al., 1989; Hodgson et al., 1998; Newel1 and Pearson, 1984). It should be noted that the infection conditions and C. jejuni strains used varied among these studies, so most are not directly comparable. Mouse challenge with C. jejuni has been conducted for a variety of purposes, including elucidation of colonization and/or virulence mechanisms and host responses (Hu and Kopecko, 2000; Jesudason et al., 1989; Kita et al., 1990; Ruiz-Palacios et al., 2003; Ueki et al., 1987; Vuckovic et al., 1998), screening of natural isolates or laboratory strains carrying spontaneous or targeted mutations thought to affect colonization and/or virulence (Berndtson et al., 1994; Hu and Kopecko, 2000; Konkel et al., 2000; MacKichan et al., 2004; Manninen et al., 1982; Purdy et al., 2000; Vuckovic et al., 2006), and evaluation of the efficacy of vaccines or therapeutic agents (Baqar et al., 1993; Lee et al., 1999; Rollwagen et al., 1993). What can be concluded from these studies is that until recently, the majority of mouse models of Carnpylobacter infection are colonization models; if disease develops, it is inconsistent or atypical. So although these experiments represent a start to understanding what is needed for development of a useable mouse model for C. jejuni pathogenesis studies, more experiments are needed to implement tractable models.
CRITERIA FOR SELECTING A MODEL Major Uses and Desirable Characteristics Animal models for study of campylobacters are essential for the following reasons: (i) to act as surrogates to study the genetic basis of virulence by screening putative virulence gene knockouts, (ii) to validate in vitro cell challenge model studies testing virulence factors, (iii) to act as surrogates for determining which bacterial genes are expressed only in the host (in vivo expression technology [IVET] etc.), (iv) to determine the basis of host immunity and immunopathology, (v) to act as surrogate hosts for initiating protective immunity through vaccines, and (vi) to act as surrogates for biological and pharmaceutical interventions for prevention and treatment. It has been well established that desirable characteristics of a model include a course of infection that accurately mimics that seen in humans for some aspect of infection (e.g., colonization or disease with susceptibility through the same route of infection as humans), repeatability, reliability, availability, low cost, ease of maintenance, ease of manipulation, specific-pathogen-free status, and known immune status. Also, depending on the chosen function for the model, genetic homogeneity and a specific aspect of pathophysiology similar to humans can be essential features. Specific-Pathogen-Free Status Recent work in murine models has demonstrated the need for strict adherence to the specific-pathogenfree status of animals used for study of campylobacters. IL-10-’- mice kept in housing without barriers were shown to spontaneously develop a form of chronic inflammatory bowel disease, whereas ILlo-’- mice kept in barrier facilities with animals specific pathogen free for known pathogens did not (Berg et al., 1996; Kiihn et al., 1993; Kullberg et al., 1998, 2001). Studies on these mouse models of inflammatory bowel disease have shown that helicobacters, campylobacters, Citrobacter rodentiurn, and possibly some species of Enterococcus are associated with inflammatory changes in the GI tract (Higgins et al., 1999; Taylor et al., 2007). Similar carriage of these organisms by ferrets and pigs has necessitated the use of cesarean rederivation to assure specificpathogen-free status. The recent report that enterohepatic Helicobacter species are prevalent in mice from commercial and academic institutions in Asia, Europe, and North America stresses the importance of ensuring that experimental animal are free of colitogenic bacteria before experimental infections (Taylor et al., 2007). Many of these organisms appear to
CHAPTER 21
occupy the same ecological niche as Campylobacter spp. and share phenotypic traits with campylobacters (Newell, 2001; Taylor et al., 2007). In addition, the presence of Helicobacter spp. in intestines of animals may influence the immunological profiles being described in the species in question (Taylor et al., 2007). When screening animals before experimental infection, one should also consider newly recognized rare Helicobacter species, some of which do not give positive results with commonly used all-Helicobacter PCR primers (Erdman et al., 2003; Johansson et al., 2006; Whary et al., 2006; Zhang et al., 2005). For mouse models, screening for murine viruses, endoand ectoparasites, and potentially pathogenic epsilon proteobacteria is recommended. For most of these agents, rapid tests are available, including genus- or species level PCR, serological screening for antibodies, or fecal diagnostics (Erdman et al., 2003; Taylor et al., 2007). Housing of mice in closed colonies with strict infection control protocols is recommended. Reagents and Resources Availability of the mouse genome sequence, various inbred mouse strains with different genetic backgrounds, numerous gene knockouts and knockins, and availability of immunological reagents have all facilitated the use of mice as animal models for bacterial pathogenesis studies. Ease of model development is further enhanced by readily available immune reagents such as monoclonal antibodies to cell surface markers, conjugated immunoglobulins, and assays for immune response elements (e.g., enzyme-linked immunosorbent assay, immunohistochemistry, and real-time PCR analysis of gene expression related to immune function). Although these reagents are routinely available for mice, many swine reagents are now also commercially available (Dawson et al., 2005; Kringel et al., 2006). Development of advanced immunological reagents for ferrets would greatly contribute to the use of this model for vaccine work. C. jejuni Strain Variation
The infecting strain of C. jejuni will influence the outcome of infection even in inbred mice (Bell and Manning, 1990; Mansfield et al., 2003). Therefore, C. jejuni strain and history in the laboratory should be taken into consideration. It has been recognized that fresh clinical isolates that have undergone minimal passage in the laboratory have a greater likelihood of producing pathology that resembles infection in humans (Hodgson et al., 1998). Conversely, strains with a long history of passage on in vitro me-
ANIMAL MODELS OF c. r E r m r INFECTIONS
369
dia show less ability to colonize and cause disease (Gaynor et al., 2004). However, repeated passage in the host is reported to enhance the ability to cause disease although the underlying basis for this has not been critically evaluated (Hodgson et al., 1998). The nine C. jejuni strains with completed genome sequences are a resource for animal model work. However, with regard to screening of various C. jejuni strains in mice, there is a compelling need for standardization of procedures and mouse strains used so that the results can be compared between laboratories.
WHY ARE IMMUNOCOMPETENT MICE REFRACTORY TO C . JEJUNIINDUCED DISEASE? Murine models of colonization and enteritis induced by primary oral C. jejuni challenge have long been needed to explore the genetics of resistance to this pathogen and the genetics of virulence of the pathogen; however, the basis for why most immunocompetent mice are refractory to C. jejuni induced disease is still unknown. Although wild rodents have been reported to harbor C. jejuni (Pacha et al., 1987), until recently, the majority of inbred mouse models of C. jejuni infection described are colonization models. Disease has been only rarely observed in mice after C. jejuni challenge. When observed, it was inconsistent or unlike that seen in humans and other animals. It is possible that a necessary pathogenic host cellular receptor is missing in laboratory mice. For example, Listeria rnonocytogenes invaded enterocytes and crossed the intestinal barrier in transgenic mice expressing the human E-cadherin receptor (Lecuit et al., 2001). Lecuit and colleagues found that murine E-cadherin did not interact with the L. monocytogenes surface protein internalin and could not mediate enterocyte invasion in vivo. It is also possible that a number of technical issues could have obscured susceptibility of inbred mice to C. jejuni in early studies. A detailed discussion of these appears below. The presence of cross-reactive immunity generated against colonization with other Epsilonproteobacteria (e.g., Helicobacter spp.) could possibly prevent colonization with C. jejuni and subsequent disease (Russell et al., 1989, 1992). The strain of C. jejuni used for murine challenge and the number of in vitro passages could also have affected early challenge studies (Gaynor et al., 2004; Hodgson et al., 1998). Studies using mice with limited GI flora (Chang and Miller, 2006) or using iron supplementation along with the C. jejuni dose (Coker and Obi, 1991; Stanfield et al., 1987) did improve results of primary challenge stud-
370
MANSFIELD ET AL.
ies with C. jejuni somewhat. However, it seems clear from current data that some degree of immune compromise is needed to cause disease in mice due to C. jejuni and the basis for this is not completely known. Additional studies are needed to explore the interaction of C. jejuni with the intestinal epithelium of mice. Exploration of polymorphisms in immune response elements in patients with severe campylobacteriosis would also be interesting.
COLONIZATION MODELS The use of colonization models, particularly the chick, has demonstrated the role of colonization factors such as flagella (Diker et al., 1992), the cellbinding factor, peblA (Pei et al., 1998), a methyl-accepting chemotaxis protein (Hendrixson and DiRita, 2004), a putative cytochrome c peroxidase that may function to reduce periplasmic hydrogen peroxide stress during in vivo growth (Hendrixson and DiRita, 2004), polyphosphate metabolism linked with stringent response mechanisms (Candon et al., 2007), the response regulator, CbrR, thought to modulate resistance to bile salts (Raphael et al., 2005), and elements of a type I11 secretion system (Feng et al., 2005). A number of investigators have shown that oral dosing of C. jejuni in mice of different strains (both inbred and outbred) results in intestinal colonization and in some cases bacteremia, but the organism does not usually cause clinical diarrhea (Ziprin et al., 2001). It was originally thought that mice are refractory to C. jejuni infection and show only transient colonization. However, several inbred strains of mice have now been reported with robust colonization with a variety of c. jejuni strains. C57BL/6 myd88-’- (Watson et al., 2007), C57BL/6 (Mansfield et al., 2007), and C3H-Limited Flora (C3H-LF) (Chang and Miller, 2006) inbred mice serve as efficient colonization models, with uniform colonization seen in the majority of mice. In these models, colonization is observed in the absence of clinical signs and lesions. C57BL/6 (Mansfield et al., 2007) and C3H-Limited Flora (C3H-LF) (Chang and Miller, 2006) will be covered in other sections paired with their respective disease models. It has been shown that C57BL/6 myeloid differentiation factor 88 (MyD88)-deficient mice (C57BLl 6 myd88-’-) serve as a useful model to study C. jejuni colonization (Watson et al., 2007). C57BL/6 myd88-I- mice, deficient for TLR signaling, were efficiently and persistently colonized by C. jejuni 81176 when inoculated orally and intraperitoneally with lo9 and lo6 CFU, respectively. Although few
C57BL/6 myd88+lf mice were colonized by either route between 1 and 3 weeks after infection, C57BL/ 6 myd88-I- mice were robustly colonized by both routes between 1 and 3 weeks after infection. This laboratory also used double knockout mice with an additional loss-of-function mutant in nrampl , a gene that encodes a divalent cation transporter essential for the control of infection by intracellular pathogens such as Salmonella spp., Mycobacterium spp., and Leishmania spp. (Fitzgeorge et al., 1981). It is expressed exclusively in cells of the reticuloendothelial system. These C57BL/6 rnyd88-/- nrampl-/- mice had increased susceptibility to C. jejuni 81-176 compared with those with the MyD88 deficiency alone when dosed intraperitoneally, but not when dosed orally, suggesting that Nrampl plays a role in controlling systemic infection. C. jejuni survived longer in nrampl -deficient macrophages, although killing was nearly complete by 8 h. Finally, the model was used to screen putative C. jejuni virulence gene mutants and both pglF (encoding a general glycosylation system) and Cj1418 (a member of a gene cluster involved in synthesis of polysaccharide capsule) were defective for colonization in the model. These results show that MyD88-deficient mice are useful as a Campylobacter spp. colonization model.
DISEASE MODELS Murine C. jejuni disease models have now been developed that will allow for further dissection of genetic elements encoding the ability of the bacterium to cause enteritis with pathological lesions (Fox et al., 1983; Mansfield et al., 2007). In these models, clinical signs, including ruffled hair coat, hunched posture, decreased appetite, diarrhea, and depression, have been observed that can be used to track the course of infection (Mansfield et al., 2007). Typhlocolitis is the main pathological lesion that can be observed by either gross or histopathological examination, although inflammation is observed in other GI tissues, including the stomach and liver. A commonality of these models is that host defects in genetic elements that regulate inflammation after challenge (e.g., absence of downmodulatory IL-10, absence of functional NF-KB needed for clearance) lead to disease and enteritis (Fox et al., 1983; Mansfield et al., 2007).
C57BLI6 IL-lO+/+and Congenic IL-lO-’- Mice Campylobacter jejuni NCTC 11168 colonized the GI tract of both C57BL/6 mice and congenic C57BL/6 IL-lO-/- mice. The IL-lO-/- mice devel-
CHAFTER 21
oped C. jejuni NCTC 11168-mediated clinical signs and lesions, thus providing linked colonization and disease models in mice of the same genetic background (Mansfield et al., 2007). In a series of timecourse and dose-response experiments, where individually housed, specific-pathogen-free mice of these genotypes were challenged with C. jejuni NCTC 11168, all mice were colonized, most at high rates. The ileocecocolic junction was the GI site with the highest rates of colonization, although the bacterium was isolated from most other GI compartments as well. Colonization of the mouse GI tract by C. jejuni NCTC 11168 was necessary but not sufficient for development of enteritis. Only C57BL/6 IL-lO-/- mice developed typhlocolitis 2 to 35 days after infection with C. jejuni NCTC 11168. The expression of disease in these mice was biphasic; some mice experienced acute infection severe enough to require euthanasia, whereas most went on to develop chronic typhlocolitis. Infected C57BL/6 IL-lo-’- mice developed clinical signs of disease, with diarrhea, depression, anorexia, ruffled hair coat, and hunched posture characteristic of gastroenteritis. Lesions were most severe at the ileocecocolic junction, and immunohistochemical staining demonstrated C. jejuni antigens within GI tissues of these infected mice. In the time-course experiment, the percentage of infected mice with histopathology scores of 2 1 0 increased as a function of time, with 12% of mice having high scores at 7 days, 38% at 14 days, 47% at 21 days, 50% at 28 days, and 53% at 35 days after infection. In the main dose-response experiment, the percentage of infected mice with histopathology scores of >10 ranged from 67 to 90%. The higher rates of lesions in this experiment probably resulted because mice were killed at 35 to 39 days after infection. No differences in the severity of histopathology were observed between male and female mice in any experiment. In comparison, histopathology scores of 97% of the uninfected control mice fell into grade 0; in four experiments combined, only 2 (3%) of 62 exhibited signs of spontaneous colitis. Therefore, disease and pathology detected in these experiments were associated with the presence of C. jejuni and were not attributable to the development of spontaneous colitis. Infected C57BL/6 IL-lO-’- mice had lesions at the ileocecocolic junction with marked crypt hyperplasia, extensive mononuclear cell infiltrates, crypt abscesses, and neutrophilic exudates (Mansfield et al., 2007). In immunohistochemically stained sections, Campylobacter jejuni was seen in crypts and crypt abscesses, was associated with epithelial cells, and was among and apparently inside of inflammatory cells underlying the sloughing epithelium. In the sub-
ANIMAL MODELS OF C. JEJUNI INFECTIONS
371
mucosal layer underlying the ileocecocolic junction, there were extensive perivascular mononuclear cell infiltrates, edema, and fibrosis. IL-lO-’- mice with the highest histopathology scores had breakdown of gut architecture. In comparison, infected C57BL/6 IL-lO+/+mice exhibited cellularity in the lamina propria, but they did not have either mononuclear cells or neutrophils in the lumen, crypts, or submucosa. However, both wild-type and IL-lO-’- mice developed significant anti-C. jejuni plasma immunoglobulin levels (predominantly T-helper cell 1-associated subtypes) by day 28 after infection. These C. jejunispecific antibodies neither cleared the bacterium from the GI tract of C57BL/6 mice or C57BL/6 IL-lo-’mice nor protected against pathological lesions in the IL-10-/- animals. In another experiment, C57BL/6 IL-lo-’- mice provided lo8 CFU C. jejuni NCTC 11168 remained colonized for 104 days, but rates of enteritis remained consistent with that seen at 28 to 45 days after infection in earlier experiments. The infected mice still exhibited increased histopathological scores 104 days after infection, indicating that the immune dysregulation resulting from IL-10 deficiency and C. jejuni infection together led to persistent chronic inflammation. Finally, because human disease has been associated with doses as low as 500 to 1,000 CFU of C. jejuni, doses from 1 X l o 2 to 1 X l o 6 CFU C. jejuni NCTC 11168 were tested to determine their ability to initiate disease and pathology in C57BL/6 IL-lO-/- mice. This study showed that in IL-lo-/mice, colonization and enteritis occurred at high rates with doses from l o 2 to l o 6 CFU, suggesting that C57BL/6 IL-lO-’- mice acquire the bacterium and express disease in a manner similar to humans. The pathogenesis of this disease in IL-lO-’- mice likely represents an enhanced T h l response to C. jejuni antigens as a result of lack of IL-10 that normally downregulates such T-cell reactivity. The C57BL/6 IL-lo+/+ and congenic IL-lo-/- mouse models should prove useful to meet some of the major needs for studies of C. jejuni infectivity and pathogenesis (Mansfield et al., 2007). NF-KB-Deficient Mice Fox and colleagues based their C. jejuni disease model on the observation that mice homozygously deficient in NF-KB subunits (p50-/-) and heterozygous for p65 (+’-), referred to as 3X mice, on a C5 7BL/ 129 background were particularly susceptible to colitis induced by another closely related enterohepatic microaerobe, Helicobacter hepaticus (Erdman et al., 2001; Fox et al., 2004). This model was used to explore the role of CDT because C. jejuni pro-
372
MANSFIELD ET AL.
duces a toxin that is functionally similar to that of H. hepaticus (Chien et al., 2000). Two experimental designs were used. In the first experiment, NF-KBdeficient mice were challenged orally with wild-type C. jejuni 81-176 and then killed and necropsied at 10 and 15 weeks after infection to determine susceptibility of 3X mice to colonization and enteritis. At 10 weeks after infection, 3X mice were colonized with C. jejuni in both stomach and colon; no C. jejuni could be isolated from mesenteric lymph nodes and liver. By 15 weeks after infection, a single rectal prolapse and stomach inflammation with associated Warthin-Starry stained C. jejuni were seen. In experiment #2, six groups of 16 of each mice were used, as follows: group 1, 3X mice challenged with C. jejuni 81-176 wild type; group 2, 3X mice challenged with C. jejuni CDT mutant; group 3, 3X mice provided brucella broth alone; group 4, C57BL/129 mice challenged with C. jejuni 81-176 wild type; group 5 , C57BL/129 mice challenged with C. jejuni CDT mutant; and group 6, C57BL/129 mice provided brucella broth alone. Half of the mice were examined at 2 months after infection and the remainder at 4 months after infection. Wild-type C. jejuni colonized 29 and 50% of the C57BL/ 129 mice at 2 and 4 months after infection, respectively, whereas the C. jejuni cdtB mutant colonized 50% of the C57BL/129 mice at 2 months after infection but none of the mice at 4 months after infection. Seventyfive to 100% of the 3X mice were colonized with both wild-type C. jejuni and the cdtB mutant at similar levels at both times examined (Fox et al., 1983). Pathological lesions were seen in mice of both genotypes, but only the C. jejuni wild-type strain with functional CDT produced typhlocolitis. C. jejuni was associated with histopathological lesions in the stomach, small intestines, and colon, along with periportal and nodular liver lesions. Analysis of the ileocecocolic junction revealed typhlocolitis with small to moderate numbers of lymphocytes and polymorphonuclear cells in the lamina propria and submucosal cellular infiltrates. Stomach and duodena of 3X mice had inflammatory cells diffusely infiltrating the mucosa and submucosa, with the highest concentrations near the muscularis mucosa. In this region, leukocytic infiltrates were composed of lymphocytes, polymorphonuclear cells, and eosinophils. There was also an expansion of mucous neck-type cells in the midglandular region of the stomach with a concomitant loss of parietal and chief cells. C57BL/129 mice developed mild gastritis and typhlocolitis, and produced significant immunoglobulin G (IgG) and Thl-promoted IgG2a humoral responses to both the wildtype strain and the C. jejuni cdtB mutant. Much more severe disease was observed in the 3X mice. Lesions
associated with C. jejuni were seen throughout the GI tract. Wild-type C. jejuni caused moderately severe gastritis and proximal duodenitis in 3X mice that were both more severe than the corresponding lesions caused by the C. jejuni cdtB mutant. Persistent colonization of NF-KB-deficient mice with both wildtype C. jejuni and the C. jejuni cdtB mutant was associated with significantly impaired IgG and IgG2a humoral responses, consistent with an innate or adaptive immune system defect or defects. These experiments suggest that the mechanism of clearance of C. jejuni is NF-KB dependent and that CDT can play a role in eliciting inflammation and lesions in the host and may aid in evading immune clearance. Additionally, these experiments show that murine disease models are feasible for screening C. jejuni mutants in putative virulence genes. Taken together, the observations of Mansfield et al. (2007) that C57BL/6 IL-lO-’- but not C57BL/6 IL-lO+’+ mice exhibited histopathological changes when infected with C. jejuni NCTC 11168 and the observations of Fox et al. (2004) about C. jejuni 81176-infected NF-KB-deficient mice suggest that multiple defects in regulation of host inflammatory processes can lead to disease and significant pathological lesions due to enteric infection with C. jejuni, and that the primary mediator or mediators determining colonization and disease outcomes in these mouse models are likely to be anti-inflammatory regulators such as IL-10, not circulating antibodies.
MODELS DEFECTIVE FOR ADAPTIVE IMMUNITY OR WITH ALTERED FLORA Screening Immunodeficient Mice for Susceptibility to C. jejuni When SCID-Beige, C.B-17-SCID-Beige, and RAG-2 mice were inoculated intragastrically with lo9 CFU C. jejuni NCTC 11168, all of these immunodeficient mice became heavily colonized for up to 20 weeks compared with immunocompetent BALB/c mice, which rapidly cleared the infection (Hodgson et al., 1998). Furthermore, C.B-17-SCID-Beige mice were more heavily colonized for up to 5 months when gavaged with a fresh clinical isolate of C. jejuni, and 10 to 20% of these mice developed clinical signs, mainly diarrhea. Lesions were restricted to the large intestines with mucosal hyperplasia, elongated crypts, and infiltration of polymorphonuclear and mononuclear cells. The authors concluded that these mice would be useful for pathogenesis studies, particularly for various cell reconstitution experiments to determine the basis of adaptive immunity to C. jejuni.
CHAPTER 21
C3H SCID Limited Flora Mouse Model Severe combined immunodeficient mice have been tested for colonization and enteritis induced by C. jejuni (Chang and Miller, 2006; MacKichan et al., 2004; Purdy et al., 2000). Chang and Miller (2006) showed that wild-type C3H mice with normal enteric flora were colonized inconsistently and inefficiently by C. jejuni 81-176. However, a 10’ CFU oral challenge of the same mice with limited gut flora resulted in efficient, high-level colonization that persisted for several weeks with no observable clinical signs or lesions. Dose-response experiments showed that 2 X lo2 CFU was the 50% infectious dose for this strain, approximately that seen for the C57BL/6 and C57BL/6 IL-lO-/- models. The C3H mice were obtained from a mouse colony where germ-free founders were inoculated with several nonpathogenic Clostridium spp. and thereafter were maintained in a barrier facility. It is now recognized that these mice harbor mainly members of the Firmicutes, but also have Lactobacillus and Acinetobacter taxa (Chang and Miller, 2006). Further work showed that C3H SCID mice with the same limited flora remained persistently colonized at a consistently high level up to 8 months after inoculation, when they were humanely killed (Chang and Miller, 2006). These mice had moderate to severe inflammation of the cecum and colon at 7 and 28 days after infection and inflammatory changes of the stomach at 28 days after infection. In the lower bowel, lesions were found in the mucosa, submucosa, and adjacent to the muscularis. Inflammatory cells, cryptitis, epithelial cell hyperplasia, and mucosal edema were present; inflammatory cells were primarily granulocytes with some mononuclear cells and leaky oligoclonal lymphocytes. These findings showed that although the innate response alone cannot block long-term colonization, it was sufficient to induce marked gut inflammation. It is likely that the absence of adaptive immune response is necessary for this host response. This model was also validated for screening C. jejuni insertion-deletion mutants, including mutations in motB and fliI, which encode products essential for motility and flagellar assembly, and the presumptive chemotaxis gene cheA (histidine kinase). None of the mutants colonized the C3H SCID-LF mice, showing the model’s utility for identifying and characterizing virulence determinants required for colonization. MacKichan and colleagues (2004) also used this SCID-LF model to show that a C. jejuni 8 1-176 two-component signal transduction system (dccR-dccS) (diminished capacity to colonize) was important for in vivo colonization but not for in vitro growth of the bacterium. In these experiments, the
-
ANIMAL MODELS OF c. T E r m r INFECTIONS
373
dccR mutant failed to produce lesions, including edema, effacement of the epithelial layer, and the influx of inflammatory cells seen in the lower bowel of mice receiving wild-type C. jejuni 8 1-176. Thus, SCID-LF mice are validated as both C. jejuni colonization and enteritis models. Widespread use will await further dissemination of this colony of mice and better characterization of the exact bacteria comprising the limited flora, but use of these mice will always be hampered by the extensive barrier housing required to keep these mice otherwise germ free and the flora defined and truly limited.
VACCINE MODELS Models for C. jejuni vaccine development are dependent on animals capable of expressing disease after challenge and mounting an adaptive immune response after vaccination. Ferrets have been used extensively for this purpose because they are completely refractory to diarrhea after a secondary challenge infection even though they show no resistance to colonization with the homologous strain of C. jejuni (Bell and Manning, 1990). Further work with this model has shown that a killed whole-cell vaccine prepared from C. jejuni 81-176 is capable of inducing protection against disease and that cross protection against some of the major serotypes of the organism is possible (Burr et al., 2005). Murine colonization models have also been used in a number of studies of natural- and vaccineacquired immunity developed to C. jejuni. Investigators have demonstrated protection against colonization in infant offspring of female mice immunized with the homologous strain of C. jejuni via immunity conferred through IgG antibodies in the milk of the dam (Abimiku and Dolby, 1988). Similarly, 39 of 68 offspring of female mice vaccinated with a particular Lior serotype were protected from intestinal colonization when challenged orally with a heterologous strain (Abimiku et al., 1989). Mice have been used to test the immunogenicity and protective efficacy of a killed whole-cell Campylobacter vaccine (Lee et al., 1999). Oral administration of a 1:l mixture of heatand formalin-inactivated C. jejuni 81-176 with an adjuvant (Escherichiu coli heat-labile enterotoxin) resulted in the generation of significant levels of secretory IgA. Moreover, vaccinated animals were protected from colonization and systemic spread when orally challenged with C. jejuni 81-176. No clinical disease was observed during the course of this experiment.
374
MANSFIELD ET AL.
Murine Intranasal Challenge Model Some researchers have used Helicobacter-free BALB/c mice vaccinated by intranasal inoculation to assess immunogenicity and protective efficacy of a C. coli-based flaA flagellin subunit vaccine (maltosebinding protein-FlaA [MBP-FlaA]) (Lee et al., 1999). The vaccine was administered to mice under anesthesia by intranasal inoculation of 30 to 50 p g of MBPFlaA vaccine mixed with or without E. coli heat-labile enterotoxin (LT) in 30- to 35-pl volumes of phosphate-buffered saline. Challenge was performed by intranasal instillation of 2 X lo9 C. jejuni 81-176 per mouse 26 days after the second vaccination or by oral challenge (Baqar et al., 1995), and the mice were monitored for clinical signs for 5 days (Baqar et al., 1996). Clinical signs were monitored daily and an illness index established to quantify these observations (Baqar et al., 1996). After intranasal challenge 26 days after the last vaccination, the best protection was seen in animals provided 50 p g of MBP-FlaA plus LT. The protective efficacies of this dose against clinical signs and intestinal colonization by C. jejuni 81-176 were 81.1 and 84%, respectively. All doses of MBP-FlaA were effective in eliciting antigenspecific serum IgG responses that were enhanced by LT adjuvant use, except in the highest dosing group. FlaA-specific intestinal IgA was only elicited in the high-dose vaccinates provided with 2 2 5 pg MBPFlaA. The effect of dose of MBP-FlaA on protective efficacy after oral challenge was also established. When challenged orally with 8 X 1010, 8 X lo9, or 8 X l o 8 CFU C. jejuni 81-176, mice immunized with 50 p g of MBP-FlaA plus LT intranasally had protective efficacies against intestinal colonization at 7 days after infection of 71.4, 71.4, and loo%, respectively. Therefore, these experiments show that this vaccine based on a FlaA subunit provided with LT adjuvant is capable of eliciting protective immunity against a heterologous strain of C. jejuni in mouse models by both oral and intranasal challenge. Future work is needed to produce murine disease models capable of providing an understanding of the ability of candidate C. jejuni vaccines to elicit innate and adaptive immunity without excessive immunopathology or autoimmunity.
TECHNICAL ASPECTS OF SUCCESSFUL MUFUNE MODELS OF C . JEJUnrI
of very small size. Many studies did not establish dose-response parameters, which masked the fact that mice are susceptible to low numbers of C. jejuni of particular strains. It is also important to evaluate the time course of disease. Both acute and chronic manifestations of disease can be missed if only a single point for evaluation is used. In mice, determining colonization by isolation of the organism from 1 or 2 fresh fecal pellets is not the most sensitive measure of colonization. The greater the volume of feces that is cultured, the higher the probability will be that an accurate identification of colonization status is made. This fact has been demonstrated repeatedly with large animal models like swine, where culture of a minimum of 10 g of feces is required for reliable isolation when animals are colonized (Nielsen et al., 1997). In most cases, the ileocecocolic junction provides the best place for isolation of C. jejuni from mice; however, this site selection should be determined experimentally. Basing colonization status on culture alone will always introduce some bias into the results.
Individual Housing During challenge experiments, mice are often housed together according to sex to prevent fighting and breeding, and to reduce cost and space. In early studies of C57BL/6 IL-lO-’- mice challenged orally with C. jejuni NCTC 11168, five infected female mice designated for euthanasia and necropsy on day 40 and housed together in a single cage all exhibited severe illness and pathology at the end of the experiment, unlike any other mice (Mansfield et al., 2007). Thus, one could not distinguish among the following possibilities: that female mice are more susceptible than male mice to C. jejuni NCTC 11168, that evolution toward increased virulence occurred in the C. jejuni NCTC 11168 populations in one mouse in that cage, and that the hypervirulent strain was then transmitted to the other mice in the cage. Thereafter, individual housing was adopted for these studies, after which the rates of disease observed in the mice after C. jejuni challenge were more uniform across experiments and no sex-associated differences were observed (Mansfield et al., 2007). Stress associated with single versus group housing may also contribute to susceptibility, however. Therefore, individual housing of mice for Campylobacter studies may be recommended in some studies.
Experimental Design for Mouse Studies It is likely that some early studies with C. jejuni murine colonization and disease models failed as a result of technical problems of working with animals
Diet Effects
It is known that immunodeficient mice such as IL-lO-’- mice develop spontaneous colitis associated
CHAPTER 21
with enzootic infection (Bristol et al., 2000; Kiihn et al., 1993) as an unregulated response to stimulation of the immune system by resident intestinal microbiota (Sellon et al., 1997); this response can also be influenced by environmental factors such as diet and water (Mahler and Leiter, 2002) and manipulation of the enteric flora (O’Mahoney et al., 2001). Therefore, environmental factors such as handling, diet and feeding, cage changing, and screening for enteric pathogens should be standardized to limit the exposure of mice to extraneous microorganisms and to increase the chance that the gut flora of the mice will be stable throughout the course of the study. Quantifying Clinical Signs of Disease and Pathology Quantitative or semiquantitative measures of C. jejuni colonization or disease are useful for comparing results to other animal models and ultimately to human patients. Scoring methods are available for assessing clinical signs in mice so that a quantitative measure can be achieved (Baqar et al., 1996; Mansfield et al., 2007). Most were developed for the purpose of establishing humane end points for mouse studies. A weighted score is given for each clinical sign, with a particular threshold triggering automatic euthanasia. Trained personnel are needed, and mice must be observed at least twice daily to monitor clinical signs. It is likely that in the past, disease manifested in mice due to C. jejuni challenge was overlooked as a result of the need for frequent observation and training in interpreting clinical signs in mice. It is now clear that even with low doses of C. jejuni, significant negative changes in mouse health take place in as little as 12 h after infection. Similarly, pathological lesions can be readily quantified. Morphometric analysis and statistical analyses to compare values between experimental groups can be conducted to provide estimates of GI crypt height and/or hyperplasia, mucus cell hyperplasia, specific cell numbers in stained sections, or other relevant values (Fox et al., 1983; Mansfield et al., 2007). Of course, care must be taken to use a fixation method that preserves antigenic sites for immunohistochemical studies and standardizes the anatomical location evaluated, because C. jejuni is often found in low numbers throughout tissues, making it difficult to accurately assess invasion. Another key issue is using a large enough experimental group size because variation in response does occur even in susceptible inbred mouse strains.
ANIMAL MODELS OF C. 7E7UNI INFECTIONS
375
enteritis, gaps in model development still exist. First, serious disease sequelae can follow GI infections with Campylobacter, and few models for these syndromes exist. The acute neuropathy Guillain-BarrC syndrome (GBS) and reactive arthritis can both be associated with recent infection with Campylobacter (Nachamkin et al., 1998). In the United States, the mean annual incidence rate of GBS is 1.3 cases per 100,000 population (Takahashi et al., 2005). Likewise, Miller Fisher syndrome (MFS) is an anatomically localized variant of GBS that is characterized by acute onset of ophthalmoplegia (paralysis of the eye muscles), ataxia, and areflexia (lack of normal reflexes). It is considered a variant of GBS because some patients who manifest MFS progress to GBS (Takahashi et al., 2005). The estimated annual incidence rate of MFS is 0.09 per 100,000 population. C. jejuni is the most frequently identified antecedent pathogen to MFS or GBS. It is believed that molecular mimicry is also a possible cause of MFS. Second, improved murine vaccine models are needed. Here, immunosufficient mouse strains are needed in which disease occurs naturally after inoculation by the oral route and enteritis occurs, followed by resolution and healing. CONCLUSIONS Taken together with previous studies on Helicobacter-associated colitis, these data on new murine models of C. jejuni colonization and enteritis suggest that multiple defects in regulation of host inflammatory processes can lead to disease and significant pathological lesions due to enteric Epsilonproteobacteria (Chang and Miller, 2006; Erdman et al., 2001; Fox et al., 2004; Mansfield et al., 2007; Young et al., 2004). These new disease models should prove to be useful as surrogates for study of the genetic basis of virulence, for screening putative virulence gene knockouts, for validation of in vitro cell challenge model studies testing virulence factors, for determining bacterial genes expressed only in the host (in vivo expression technology, IVET, etc.), and for determining the basis of host immunity and immunopathology (Fox et al., 1983; Mansfield et al., 2007). Further studies are needed to develop better animal models to act as surrogate hosts for initiating protective immunity through vaccines and to act as surrogates for biologic and pharmaceutical interventions for prevention and treatment.
FUTURE GOALS
MURINE ENTEFUC DISEASES PHENOME DATABASE (MEDPeD)
Despite progress in development of tractable murine models for study of C. jejuni colonization and
As part of the National Institutes of Health Food and Waterborne Diseases Integrated Research Net-
376
MANSFIELD ET AL.
work (FWD IRN), a first version of the MEDPeD has been posted at the Michigan State University Microbiology Research Unit website for access by scientists studying enteric diseases (Mansfield et al., 2007). Standard operating procedures and links to published articles about mouse infection are available at the database. Microbiology Research Unit members encourage scientists working with murine models of enteric infections to contact our research unit to create links from their websites to the MEDPeD to build the value of this resource. Acknowledgments. We thank Julia Bell for critical evaluation of the chapter. This work was supported by NIH grants R01 0K052413, PO1 CA026731, and P30 EJ002190. Mansfield et al. studies were funded in whole with federal funds from NIAID, NIH, Department of Health and Human Services, under contract no. NO 1-AI-30058.
REFERENCES Abimiku, A. G., and J. M. Dolby. 1988. Cross-protection of infant mice against intestinal colonisation by Campylobacter jejuni: importance of heat-labile serotyping (Lior) antigens.]. Med. Microbiol. 26:265-268. Abimiku, A. G., J. M. Dolby, and S . P. Borriello. 1989. Comparison of different vaccines and induced immune response against Campylobacter jejuni colonization in the infant mouse. Epidemiol. Infect. 102:27 1-28 0. AbuOun, M., G. Manning, S. A. Cawthraw, A. Ridley, I. H. Ahmed, T. M. Wassenaar, and D. G. Newell. 2005. Cytolethal distending toxin (CDT)-negative Campylobacter jejuni strains and anti-CDT neutralizing antibodies are induced during human infection but not during colonization in chickens. Infect. Immun. 73 ~3053-3062. Babakhani, F. K., G. A. Bradley, and L. A. Joens. 1993. Newborn piglet model for campylobacteriosis. Infect. lmmun. 61:34663475. Baqar, S., L. A. Applebee, and A. L. Bourgeois. 1995. Immunogenicity and protective efficacy of a prototype Campylobacter killed whole-cell vaccine in mice. Infect. Immun. 63:3731-3735. Baqar, S., A. L. Bourgeois, L. A. Applebee, A. S. Mourad, M. T. Kleinosky, Z. Mohran, and J. R. Murphy. 1996. Murine intranasal challenge model for the study of Campylobacter pathogenesis and immunity. Infect. Immun. 64:4933-4939. Baqar, S., N. D. Pacheco, and F. M. Rollwagen. 1993. Modulation of mucosal immunity against Campylobacter jejuni by orallyadministered cytokines. Antimicrob. Agents Chemother. 37: 2688-2692. Bell, J. A., and D. D. Manning. 1990. A domestic ferret model of immunity to Campylobacter jejuni-induced enteric disease. lnfect. lmmun. 58:1848-1852. Berg, D. J., N. Davidson, R. Kiihn, W. Miiller, S. Menon, G. Holland, L. Thompson-Snipes, M. W. Leach, and D. Rennick. 1996. Enterocolitis and colon cancer in interleukin-10-deficient mice are associated with aberrant cytokine production and CD4( +) TH1-like responses. 1.Clin. Invest. 98:lOlO-1020. Berndtson, E., M. L. Danielsson-Tham, and A. Engvall. 1994. Experimental colonization of mice with Campylobacter jejuni. Vet. Microbiol. 4 1:183-1 88. Biswas, D., U. Fernando, C. Reiman, P. Willson, A. Potter, and B. Allan. 2006. Effect of cytolethal distending toxin of Campylobacter jejuni on adhesion and internalization in cultured
cells and in colonization of the chicken gut. Avian Dis. 50586593. Blaser, M. J. 1997. Epidemiologic and clinical features of Campylobacter jejuni infections. ]. Infect. Dis. 176:S103-S105. Blaser, M. J., R. I. Glass, M. I. Hug, B. Stoll, G. M. Kibrya, and A. R. M. A. Alim. 1980. Isolation of Campylobacter fetus subsp. jejuni from Bangladeshi children. 1. Clin. Microbiol. 12:744747. Boosinger, T. R., and T. A. Powe. 1988. Campylobacter jejuni infections in gnotobiotic pigs. Am. 1.Vet. Res. 49:456-458. Bristol, I. J., M. A. Farmer, Y. 2. Cong, X. X. Zheng, T. B. Srom, C. 0. Elson, J. P. Sundberg, and E. H. Leiter. 2000. Heritable susceptibility for colitis in mice induced by IL-10 deficiency. Inflamm. Bowel Dis. 6:290-302. Bruce, D., W. Zochowski, and G. A. Fleming. 1980. Campylobacter infections in cats and dogs. Vet. Rec. 107:200-201. Burr, D. H., D. Rollins, L. H. Lee, D. L. Pattarini, S. S. Walz, J. H. Tian, J. L. Pace, A. L. Bourgeois, and R. I. Walker. 2005. Prevention of disease in ferrets fed an inactivated whole cell Campylobacter jejuni vaccine. Vaccine 23 :43 15-4321. Caldwell, M. B., R. I. Walker, S. D. Stewart, and J. E. Rogers. 1983. Simple adult rabbit model for Campylobacter jejuni enteritis. Infect. Immun. 42:1176-1182. Candon, H. L., B. J. Allan, C. D. Fraley, and E. C. Gaynor. 2007. Polyphosphate kinase 1 (PPK1) is a pathogenesis determinant in Campylobacter jejuni. 1.Bacteriol. 189:8099-8108. Chang, C., and J. F. Miller. 2006. Campylobacter jejuni colonization of mice with limited enteric flora. Infect. lmmun. 74: 5261-5271. Chien, C. C., N. S. Taylor, 2. Ge, D. B. Schauer, V. B. Young, and J. G. Fox. 2000. Identification of cdtB homologues and cytolethal distending toxin activity in enterohepatic Helicobacter spp. 1.Med. Microbiol. 49525-534. Coker, A. O., and C. L. Obi. 1991. Effect of iron-dextran on lethality of Nigerian isolates of Campylobacter jejuni in mice. Cent. Afr. 1.Med. 37:20-23. Dawson, H. D., E. Beshah, S . Nishi, G. Solano-Aguilar, M. Morimoto, A. Zhao, K. B. Madden, T. K. Ledbetter, J. P. Dubey, T. Shea-Donohue, J. K. Lunney, and J. F. Urban, Jr. 2005. Localized multigene expression patterns support an evolving T h l / Th2-like paradigm in response to infections with Toxoplasma gondii and Ascaris suum. Infect. Immun. 73:1116-1128. Diker, K. S., G. Hascelik, and S. Diker. 1992. Colonization of infant mice with flagellar variants of Campylobacter jejuni. Acta Microbiol. Hung. 39: 133-1 36. Eaton, K. A., D. R Morgan, and S. Krakowka. 1989. Campylobacter pylori virulence factors in gnotobiotic piglets. Infect. Immun. 57:1119-1125. Erdman, S. E., J. G. Fox, C. A. Dangler, D. Feldman, and B. Horwitz. 2001. Cutting edge: typhlocolitis in NF-KB deficient mice. 1.lmmunol. 66:1443-1447. Erdman, S. E., V. P. Rao, T. Poutahidis, M. M. Ihrig, Z. Ge, Y. Feng, M. Tomczak, A. B. Rogers, B. H. Horwitz, and J. G. Fox. 2003. CD4(+)CD25(+) regulatory lymphocytes require interleukin 10 to interrupt colon carcinogenesis in mice. Cancer Res. 63~6042-6050. Feng, S., K. Ku, E. Hodzic, E. Lorenzana, K. Freet, and S. W. Barthold. 2005. Differential detection of five mouse-infecting Helicobacter species by multiplex PCR. Clin. Diagn. Lab. Immun. 12531-536. Fernando, U., D. Biswas, B. Allan, P. Willson, and A. A. Potter. 2007. Influence of Campylobacter jejuni fliA, rpoN and flgK genes on colonization of the chicken gut. Int. J. Food Microbiol. 118:194-200. Fitzgeorge, R. B., A. Baskerville, and K. P. Lander. 1981. Experimental infection of Rbesus monkeys with a human strain of Campylobacter jejuni. 1.Hyg. (Lond.) 86:343-351.
CHAPTER 2 1
Forbes, J. R., and P. Gros. 2001. Divalent-metal transport by NRAMP proteins at the interface of host-pathogen interactions. Trends Microbiol. 9:397-403. Fox, J. G. 1992. In vivo models of enteric campylobacteriosis: natural and experimental infections, p. 131-138. In I. Nachamkin, M. J. Blaser, and L. S . Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. ASM Press, Washington, DC. Fox, J. G., J. I. Ackerman, N. Taylor, M. Claps, and J. C. Murphy. 1987. Campylobacter jejuni infection in the ferret: an animal model of human campylobacteriosis. Am. J. Vet. Res. 48:85-90. Fox, J. G., R. Moore, and J. I. Ackerman. 1983. Campylobacter jejuni-associated diarrhea in dogs. J. Am. Vet. Med. Assoc. 183: 1430-1433. Fox, J. G., A. B. Rogers, M. T. Whary, Z. Ge, N. S. Taylor, S. Xu, B. H. Horwitz, and S. E. Erdman. .2004. Gastroenteritis in NF-KB-deficient mice is produced with wild-type Camplyobacter jejuni but not with C. jejuni lacking cytolethal distending toxin despite persistent colonization with both strains. Infect. Immun. 72: 1116-1 125. Fry, B. N., S. Feng, Y. Y. Chen, D. G. Newell, P. J. Coloe, and V. Korolik. 2000. The galE gene of Campylobacter jejuni is involved in lipopolysaccharide synthesis and virulence. Infect. Immun. 68:2594-2601. Gaynor, E. C., S. Cawthraw, G. Manning, J. K. MacKichan, S. Falkow, and D. G . Newell. 2004. The genome-sequenced variant of Campylobacter jejuni NCTC 11168 and the original clonal clinical isolate differ markedly in colonization, gene expression, and virulence-associated phenotypes. J. Bacteriol. 186: 503-517. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Higgins, L. M., G. Frankel, G. Douce, G . Dougan, and T. T. MacDonald. 1999. Citrobacter rodentium infection in mice elicits a mucosal T h l cytokine response and lesions similar to those in murine inflammatory bowel disease. Infect. Immun. 67:303 13039. Hodgson, A. E., B. W. McBride, M. J. Hudson, G. Hall, and S. A. Leach. 1998. Experimental Campylobacter infection and diarrhoea in immunodeficient mice. J. Med. Microbiol. 47:799-809. Hu, L., and D. J. Kopecko. 2000. Interactions of Campylobacter with eukaryotic cells: gut luminal colonization and mucosal invasion mechanisms, p. 191-216. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Humphrey, C. D., D. M. Montag, and F. E. Pittman. 1985. Experimental infection of hamsters with Campylobacter jejuni. J. Infect. Dis. 151:485-493. Islam, D., M. D. Lewis, A. Srijan, L. Bodhidatta, A. Aksomboon, M. Gettayacamin, S. Baqar, D. Scott, and C. J. Mason. 2006. Establishment of a non-human primate Campylobacter disease model for the pre-clinical evaluation of Campylobacter vaccine formulations. Vaccine 24:3762-3771. Jesudason, M. V., D. J. Hentges, and P. Pongpech. 1989. Colonization of mice by Campylobacter jejuni. Infect. Immun. 57: 2279-2282. Johansson, S. K., R. E. Feinstein, K. E. Johansson, and A. V. Lindberg. 2006. Occurrence of Helicobacter species other than H. hepaticus in laboratory mice and rats in Sweden. Comp. Med. 56~110-113. Jones, F. R., S. Baqar, A. Gozalo, G. Nunez, N. Espinoza, S. M. Reyes, M. Salazar, R. Meza, C. K. Porter, and S. E. Walz. 2006. New World monkey Aotus nancymae as a model for Campylobacter jejuni infection and immunity. Infect. Immun. 74:790793.
-
ANIMAL MODELS OF C. TETUNI INFECTIONS
377
Jones, M. A., K. L. Marston, C. A. Woodall, D. J. Maskell, D. Linton, A. V. Karlyshev, N. Dorrell, B. W. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni NCTClll68 to high-level colonization of the avian gastrointestinal tract. Infect. Immun. 72:3769-3776. Kita, E., D. Oku, A. Hamuro, F. Nishikawa, M. Emoto, Y. Yagyu, N. Katsui, and S. Kashiba. 1990. Hepatotoxic activity of Campylobacter jejuni. ]. Med. Microbiol. 33:171-182. Knudsen, K. N., D. D. Bang, L. 0. Andresen, and M. Madsen. 2006. Campylobacter jejuni strains of human and chicken origin are invasive in chickens after oral challenge. Avian Dis. 5O:lO14. Konkel, M. E., L. A. Joens, and P. F. Mixter. 2000. Molecular characterization of Campylobacter jejuni virulence determinants, p. 217-240. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Konkel, M. E., B. J, Kim, J. D. Klena, C. R. Young, and R. Ziprin. 1998. Characterization of the thermal stress response of Campylobacter jejuni. Infect. Immun. 66:3666-3672. Kringel, H., T. Iburg, H. Dawson, B. Aasted, and A. Roepstorff. 2006. A time course study of immunological responses in Trichuris suis infected pigs demonstrates induction of a local type 2 response associated with worm burden. Int. J. Parasitol. 36: 9 15-924. Kiihn, R., J. Lohler, D. Rennick, K. Rajewsky, and W. Miiller. 1993. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75:263-274. Kullberg, M. C., A. G. Rothfuchs, D. Jankovic, P. Caspar, T. A. Wynn, P. L. Gorelick, A. W. Cheever, and A. Sher. 2001. Helicobacter bepaticus-induced colitis in interleukin-10-deficient mice: cytokine requirements for the induction and maintenance of intestinal inflammation. Infect. Immun. 69:4232-4241. Kullberg, M. C., J. M. Ward, P. L. Gorelick, P. Caspar, S. Hieny, A. Cheever, D. Jankovic, and A. Sher. 1998. Helicobacter hepaticus triggers colitis in specific-pathogen-free interleukin-10 (IL-10)-deficient mice through an IL-12- and gamma interferondependent mechanism. Infect. Immun. 66:5 157-5 166. Lecuit, M., S. Vandormael-Pournin, J. Lefort, M. Huerre, P. Gounon, C. Dupuy, C. Babinet, and P. Cossart. 2001. A transgenic model for listeriosis: role of internalin in crossing the intestinal barrier. Science 292:1722-1725. Lee, L. H., E. Burg, 3rd, S. Baqar, A. L. Bourgeois, D. H. Burr, C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evaluation of a truncated recombinant flagellin subunit vaccine against Campylobacter jejuni. Infect. Immun. 675799-5805. MacKichan, J. K., E. C. Gaynor, C. Chang, S. Cawthraw, D. G. Newell, J. F. Miller, and S. Falkow. 2004. The Campylobacter jejuni dccRS two-component system is required for optimal in vivo colonization but is dispensable for in vitro growth. Mol. Microbiol. 54: 1269-1 28 6. Mahler, M., and E. H. Leiter. 2002. Genetic and environmental context determines the course of colitis developing in IL-10deficient mice. Inflamm. Bowel Dis. 8:347-355. Manninen, K. I., J. F. Prescott, and I. R. Dohoo. 1982. Pathogenicity of Carnpylobacter jejuni isolates from animals and humans. Infect. Immun. 38:46-52. Mansfield, L. S., J. A. Bell, D. L. Wilson, A. J. Murphy, H. M. Elsheikha, V. A. Rathinam, B. R. Fierro, J. E. Linz, and V. B. Young. 2007. C57BL/6 and congenic interleukin-10-deficient mice can serve as models of Campylobacter jejuni colonization and enteritis. Infect. Immun. 75:1099-1115. Mansfield, L. S., D. T. Gauthier, S. R. Abner, K. M. Jones, S. R. Wilder, and J. F. Urban. 2003. Enhancement of disease and pathology by synergy of Trichuris suis and Campylobacter jejuni in the colon of immunologically naive swine. Am. J. Trop. Med. Hyg. 68:70-80.
378
MANSFIELD ET AL.
Mansfield, L. S., and J. F. Urban. 1996. The pathogenesis of necrotic proliferative colitis in swine is linked to whipworm induced suppression of mucosal immunity to resident bacteria. Vet. Immunol. Immunop. 5O:l-17. Nachamkin, I., B. M. Allos, and T. Ho. 1998. Campylobacter species and Guillain-Barrt syndrome. Clin. Microbiol. Rev. 11: 555-567. Newell, D. G. 2001. Animal models of Campylobacter jejuni colonization and disease and the lessons to be learned from similar Helicobacter pylori models. 1.Appl. Microbiol. 90:57S-67S. Newell, D. G., and A. Pearson. 1984. The invasion of epithelial cell lines and the intestinal epithelium of infant mice by Campylobacter jejunilcoli. ]. Diarrhoeal. Dis. Res. 2: 19-26. Nielsen, E. M., J. Engberg, and M. Madsen. 1997. Distribution of serotypes of Campylobacter jejuni and C. coli from Danish patients, poultry, cattle and swine. FEMS Immunol. Med. Mic. 19: 47-56. Oberhelman, R. A., R. H. Gilman, P. Sheen, J. Cordova, D. N. Taylor, M. Zimic, R. Meza, J. Perez, C. LeBron, L. Cabrera, F. G. Rodgers, D. L. Woodward, and L. J. Price. 2003. Campylobacter transmission in a Peruvian shantytown: a longitudinal study using strain typing of Campylobacter isolates from chickens and humans in household clusters. ]. Infect. Dis. 187:260269. Olson, P., and K. Sandstedt. 1987. Campylobacter in the dog: a clinical and experimental study. Vet. Rec. 121:99-101. O’Mahoney, L., M. Feeney, S. O’Halloran, L. Murphy, B. Kiely, J. Fitzgibbon, G. Lee, G. O’Sullivan, F. Shanahan, and J. K. Collins. 2001. Probiotic impact on microbial flora, inflammation, and tumour development in IL-10 knockout mice. Aliment. Pharmacol. Ther. 15: 1219-1225. Pacha, R. E., G. W. Clark, E. A. Williams, A. M. Carter, J. J. Scheffelmaier, and P. Debusschere. 1987. Small rodents and other mammals associated with mountain meadows as reservoirs of Giardia spp. and Campylobacter spp. Appl. Environ. Microb. 53:1574-1579. Pavlovskis, 0. R., D. M. Rollins, R. L. Haberberger, Jr., A. E. Green, L. Habash, S. Strocko, and R. I. Walker. 1991. Significance of flagella in colonization resistance of rabbits immunized with Campylobacter spp. Infect. Immun. 59:2259-2264. Pazzaglia, G., A. L. Bourgeois, K. el Diwany, N. Nonr, N. Badran, and R. Hablas. 1991. Campylobacter diarrhoea and an association of recent disease with asymptomatic shedding in Egyptian children. Epidemiol. Infect. 106:77-82. Pei, Z. H., C. Burucoa, B. Grignon, S. Baqar, X. 2. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943. Poly, F., T. Read, D. R. Tribble, S. Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. Immun. 75:3425-3433. Purdy, D., C. M. Buswell, A. E. Hodgson, K. McAlpine, I. Henderson, and s. A. Leach. 2000. Characterisation of cytolethal distending toxin (CDT) mutants of Campylobacter jejuni. J. Med. Microbiol. 49:473-479. Raphael, B. H., S. Pereira, G. A. Flom, Q. Zhang, J. M. Ketley, and M. E. Konkel. 2005. The Campylobacter jejuni response regulator, CbrR, modulates sodium deoxycholate resistance and chicken colonization. J. Bacteriol. 187:3662-3670. Ringoir, D. D., D. Szylo, and V. Korolik. 2007. Comparison of 2day-old and 14-day-old chicken colonization models for Campylobacter jejuni. FEMS Immunol. Med. Microbiol. 49:155-158. Rollwagen, F. M., N. D. Pacheco, J. D. Clements, 0. Pavlovskis, D. M. Rollins, and R I. Walker. 1993. Killed Campylobacter
elicits immune response and protection when administered with an oral adjuvant. Vaccine 11:1316-1320. Ruiz-Palacios, G. M., L. E. Cervantes, P. Ramos, B. ChavezMunguia, and D. S. Newburg. 2003. Campylobacter jejuni binds intestinal H(0) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J. Biol. Chem. 278:14112-14120. Russell, R. G., M. J. Blaser, J. I. Sarmiento, and J. Fox. 1989. Experimental Campylobacter jejuni infection in Macaca nemestrina. Infect. Immun. 57:1438-1444. Russell, R. G., J. A. Kiehlbauch, C. J. Gebhart, and L. J. DeTolla. 1992. Uncommon Campylobacter species in infant Macaca nemestrina monkeys housed in a nursery. 1. Clin. Microbiol. 30: 3 024-3 027. Russell, R. G., M. Odonnoghue, D. C. Blake, J. Zulty, and L. J. Detolla. 1993. Early colonic damage and invasion of Campylobacter jejuni in experimentally challenged infant Macaca mulatta.]. Infect. Dis. 168:210-215. Sellon, R. K., S. L. Tonkonogy, H. Grable, J. Kwon, M. Schultz, E. Balish, D. Rennick, and R. B. Sartor. 1997. Development of spontaneous colitis in IL-10 knockout mice requires normal enteric bacterial flora. Gastroenterology 112:A1088-A1088. Sestak, K., C. K. Merritt, J. Borda, E. Saylor, S. R. Schwamberger, F. Cogswell, E. S. Didier, P. J. Didier, G. Plauche, R. P. Bohm, P. P. Aye, P. Alexa, R. L. Ward, and A. A. Lackner. 2003. Infectious agent and immune response characteristics of chronic enterocolitis in captive rhesus macaques. Infect. lmmun. 71: 4079-4086. Skirrow, M. B., and M. J. Blaser. 2000. Clinical aspects of Campyiobacter infection, p. 69-88. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Stanfield, J. T., B. A. McCardell, and J. M. Madden. 1987. Campylobacter diarrhea in an adult mouse model. Microb. Pathog. 3:155-165. Stern, N. J. 1992. Reservoirs for Campylobacter jejuni and approaches for intervention in poultry, p. 49-60. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. ASM Press, Washington, DC. Szymanski, C. M., D. H. Burr, and P. Guerry. 2002. Campylobacter protein glycosylation affects host cell interactions. Infect. Immun. 70:2242-2244. Takahashi, M., M. Koga, K. Yokoyama, and N. Yuki. 2005. Epidemiology of Campylobacter jejuni isolated from patients with Guillain-Barrt and Fisher syndromes in Japan.]. Clin. Microbiol. 43:335-339. Taylor, N. S., S. Xu, P. Nambiar, F. E. Dewhirst, and J. G. Fox. 2007. Enterohepatic Helicobacter species are prevalent in mice from commercial and academic institutions in Asia, Europe, and North America. 1. Clin. Microbiol. 45:2166-2172. Ueki, Y., A. Umeda, S. Fujimoto, M. Mitsuyama, and K. Amako. 1987. Protection against Campylobacter jejuni infection in suckling mice by anti-flagellar antibody. Microbiol. Immunol. 3 1: 1161-1171. Vuckovic, D., M. Abram, M. Bubonja, B. Wraber, and M. Doric. 2006. Host resistance to primary and secondary Campylobacter jejuni infections in C57BLI 6 mice. Microb. Pathogenesis 40:3539. Vuckovic, D., M. Abram, and M. Doric. 1998. Primary Campylobacter jejuni infection in different mice strains. Microb. Pathogenesis 24:263-268. Walker, R. I., E. A. Schmauder-Chock, J. L. Parker, and D. Burr. 1988. Selective association and transport of Campylobacter jejuni through M cells of rabbit Peyer’s patches. Can. J. Microbiol. 34~1142-1147. Watson, R. O., V. Novik, D. Hofreuter, M. Lara-Tejero, and J. E. Galan. 2007. A MyD88-deficient mouse model reveals a
CHAPTER 21
role for Nrampl in Campylobacter jejuni infection. Infect. Immun. 75:1994-2003. Whary, M. T., S. J. Danon, Y. Feng, Z. Ge, N. Sundina, V. Ng, N. S. Taylor, A. B. Rogers, and J. G . Fox. 2006. Rapid onset of ulcerative typhlocolitis in B6.12YP2-ILlOtmlCgn (IL-lO-’-) mice infected with Helicobacter trogontum is associated with decreased colonization by altered Schaedler’s flora. Infect. Immun. 74:6615-6623. Wu, Y. L., L. H. Lee, D. M. Rollins, and W. M. Ching. 1994. Heat shock- and alkaline pH-induced proteins of Campylobacter jejuni: characterization and immunological properties. Infect. Immun. 62:4256-4260. Young, V. B., K. A. Knox, J. S. Pratt, J. S. Cortez, L. S. Mansfield, A. B. Rogers, J. G . Fox, and D. B. Schauer. 2004. In vitro and in vivo characterization of Helicobacter hepaticus cytolethal distending toxin mutants. Infect. Immun. 72:2521-2527.
ANIMAL MODELS OF C. TETUNI INFECTIONS
379
Young, V. B., D. B. Schauer, and J. G. Fox. 2000. Animal models of Campylobacter infection, p. 287-302. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Young, C. R, R. L. Ziprin, M. E. Hume, and L. H. Stanker. 1999. Dose response and organ invasion of day-of-hatch Leghorn chicks by different isolates of Campylobacter jejuni. Avian Dis. 43:763-767. Zhang, L., S. J. Danon, M. Grehan, V. Chan, A. Lee, and H. Mitchell. 2005. Natural colonization with Helicobacter species and the development of inflammatory bowel disease in interleukin-10-deficient mice. Helicobacter 10:223-230. Ziprin, R L., C. R. Young, J. A. Byrd, L. H. Stanker, M. E. Hume, S. A. Gray, B. J. Kim, and M. E. Konkel. 2001. Role of Campylobacter jejuni potential virulence genes in cecal colonization. Avian Dis. 45549-557.
Campylobacfer, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chatxer 22
Rabbit Model of Guillain-Barre Syndrome NOBUHIRO YUKI
1995). That study established an epidemiological association between C. jejuni infection and GBS. C. jejuni-associated GBS peaked in 10- to 30-year-old individuals, and the male-to-female ratio was 1.7 to 1 (Takahashi et al., 2005). The median latent period between antecedent symptoms and the onset of neuropathy was 10 days. Autoantibodies against the surface components of peripheral nerves were considered to be pathogenic substances that induce GBS because plasma exchange facilitates the rate of recovery (Hughes et al., 2007). Gangliosides constitute a large family predominantly made up of cell-surface glycosphingolipids bearing a ceramide moiety anchored in the external leaflet of the lipid bilayer and a sialylated oligosaccharide core exposed extracellularly. In AMAN, immunoglobulin (Ig) G is deposited on the axolemma of the spinal anterior roots (Hafer-Macko et al., 1996). This indicates that IgG, which binds effectively with complement, is an important factor in the development of AMAN. Patients who developed AMAN subsequent to C . jejuni enteritis had IgG antibodies against GM1, and their autoantibody titers decreased with the clinical course (Yuki et al., 1990). In contrast, patients who had had C. jejuni enteritis but no neurological disorder did not have the autoantibody. GDla as well as GM1 is an autoantigen for IgG antibodies in patients with AMAN subsequent to C. jejuni enteritis (Ho et al., 1999). Lipo-oligosaccharide (LOS) is a major cellsurface structure expressed by C. jejuni. A C. jejuni strain (CF90-26) isolated from an AMAN patient carrying anti-GM1 IgG antibodies expressed an oligosaccharide structure [Gal pl-3 GalNAc pl-4 (NeuAc a2-3) Galp], which protruded from the LOS core (Fig. 1) (Yuki et al., 1993). This terminal structure was identical to that of the terminal tetrasaccharide of the GM1 ganglioside, indicating definitive
Guillain-BarrC syndrome (GBS), characterized by limb weakness and areflexia, is a typical postinfectious autoimmune disease (Yuki, 2001). Most GBS patients have had gastrointestinal or upper respiratory symptoms 1 to 2 weeks before the onset of neurological symptoms. The gram-negative bacterium Campylobacter jejuni, a leading cause of acute gastroenteritis in humans, is the most frequent antecedent pathogen. GBS was considered to be a demyelinating disease of peripheral nerves. This view of GBS was shaped largely by what is known of its laboratory analogue, an experimental autoimmune neuritis induced by immunization with P2 protein (a component of peripheral nerve myelin). Now the presence of a primary axonal GBS acute motor axonal neuropathy (AMAN) is widely recognized through findings of autopsy studies (McKhann et al., 1993). At least in Japan, C. jejuni infection is associated with AMAN, but not with demyelinating GBS (Kuwabara et al., 2004). Four criteria must be satisfied to conclude that a disease is triggered by molecular mimicry (Ang et al., 2004): (i) the establishment of an epidemiological association between the infectious agent and the immune-mediated disease, (ii) the identification of T cells or antibodies directed against the patient’s target antigens, (iii) the identification of microbial mimics of the target antigen, and (iv) reproduction of the disease in an animal model. As reviewed here, GBS subsequent to C. jejuni enteritis fulfills all the criteria and provides the first verification that molecular mimicry contributes to the development of human autoimmune diseases. A prospective case-control study detected evidence of recent C. jejuni infection in 26% of patients with GBS as compared with 2% of the household controls (a member of the patient’s household) and 1% of the age-matched hospital controls (Rees et al.,
Nobuhiro Yuki
GBS Laboratory, 3-9-5 Kami-ikedai, Ota, Tokyo 145-0064, Japan.
381
YUKI
382
GMI
GDla
Cer
m
c
e
r
Figure 1. Molecular mimicry of gangliosides and Carnpylobacter jejuni LOS. C. jejuni isolate (CF90-26) from a patient with acute motor axonal neuropathy carries GM1-like and GDla-like LOSS.
evidence of molecular mimicry between human nerve tissue and C. jejuni. This strain also carried a GDlalike LOS (Fig. 1) (Koga et al., 2006).
GANGLIOSIDE IMMUNIZATION MODEL There have been only two reports describing neurological dysfunction in GM1-immunized animals. In the first study, GM1-immunized rabbits developed a spastic paralysis, whereas GDla-immunized rabbits had a flaccid paralysis (Nagai et al., 1976). Histologically, phagocytic cells containing myelin debris could be observed. In the other study, rabbits developed a subclinical neuropathy (Thomas et al., 1991). There was mild axonal degeneration in the sciatic nerve and IgM deposits at the nodes of Ranvier. These suggested that failure to induce neuropathy by sensitization with gangliosides might depend on species susceptibility and on the immunization procedure used. On the basis of a procedure that successfully induces experimental sensory ataxic neuropathy by sensitization with GDlb (Kusunoki et al., 1996), an AMAN model was established by sensitization of Japanese white rabbits with a bovine brain ganglioside (BBG) mixture that included GM1 as wcll as with purified GM1 (Yuki, 2001). This model is helpful to clarify the molecular pathogenesis of AMAN and to develop new treatments for it. Male Japanese white rabbits weighing 2.0 to 2.5 kg were obtained. A 2.5-mg portion of a BBG mixture (GM1 21%, GDla 40%, GDlb 16%, GTlb 19%; Cronassial, Fidia, Italy) or 0.5 mg of GM1 isolated from bovine brain (Sygen; Fidia, Padova, Italy) was dissolved in 0.5 ml of keyhole lympet hemocyanin (KLH; 2 mg/ml) in phosphate-buffered saline. A 0.5-ml portion of complete Freund’s adjuvant (CFA) was added, and the mixture was emulsified. A
1-ml sample of the emulsion of the BBG mixture or GM1 was injected subcutaneously to the back and intraperitoneally at 3-week intervals until limb weakness developed or 6 months had passed after the first inoculation. Control rabbits were injected under the same protocol with the same inocula but without gangliosides. The rabbits were checked daily for clinical signs and weighed twice a week. Weekly plasma samples were taken by ear vein puncture. None of the 10 control rabbits inoculated with KLH and CFA showed limb weakness within 6 months after the first inoculation. In contrast, all 13 rabbits immunized with BBG, KLH, and CFA developed flaccid paresis of the hind limbs, the onset ranging from 35 to 57 days (median, 43 days) after the initial inoculation. Eleven of the rabbits began to lose weight 3 to 20 days (median, 6 days) before the onset of limb weakness. Three rabbits had body and limb tremors several days before the onset of limb weakness. Tetraparesis developed rapidly in two rabbits, with respiration becoming labored and gasping (Fig, 2B). Limb weakness progressed for 4 to 13 days (median, 5 days) after onset in nine rabbits, then reached a plateau. One rabbit could lift neither its head nor body (Fig. 2A) but was able to walk 2 weeks after the onset of limb weakness. No significant changes were found in the brains or spinal cords of any of the rabbits immunized with BBG. In contrast, the sciatic nerves showed severe to mild Wallerian-like degeneration with rare demyelination or remyelination (Fig. 2C). Macrophage invasion was prominent in endoneurial-perivascular areas, but no lymphocytic infiltration was found in any of the sciatic nerve regions. Quantitative comparisons showed that the mean degenerative axon density was significantly increased in 11 BBG-immunized rabbits (1289 k 476/mm2 [mean 2 standard deviation]) 152/mm2). than in four adjuvant controls (498
CHAPTER 22
RABBIT MODEL OF GBS
Figure 2 . (A) Rabbit with limb weakness induced by sensitization with BBG mixture 14 days after onset. It could not maintain a normal standing position or lift its head or body. The muscles of the extremities and trunk were weak and slack, offering less than usual resistance to passive movement. (B) Rabbit with limb weakness induced by sensitization with the BBG mixture 4 days after onset. It lay splayed out, all extremities extended, and its head on the floor instead of sitting in the usual compact, hunched posture. This rabbit attempted to stand but could not. (C) Transverse section of the sciatic nerve from a BBGimmunized rabbit. Many myelin ovoids produced by Wallerian-like degeneration of the myelinated fibers are present. Toluidine blue-safranine stain. (D) Transverse section of the anterior root from a BBG-immunized rabbit. Clusters of small myeh a t e d fibers indicative of regenerating sprouted fibers are present (arrowhead). Toluidine blue-safranine stain. (E) Myelin ovoids indicative of Wallerian-like degeneration of myelinated fibers from a BBG-immunized rabbit are present in three sciatic nerve teased fiber preparations. (F) Transverse section of the spinal anterior nerve root from a BBG-immunized rabbit with high anti-GM1 IgG antibody titer. Some axons are deeply stained by peroxidase-conjugated Protein G (arrowhead). Scale bars = 10 pm. From Yuki et al. (2001).
383
384
WKI
The surviving axon density was significantly decreased in the BBG-immunized rabbits (9,597 k 2,419/mm2)than in the adjuvant controls (14,971 +. 3,131/mm2). The anterior spinal nerve roots from a rabbit in the recovery phase of weakness had occasional clusters of small fibers showing various degrees of myelination, indicative of axonal sprout regeneration (Fig. 2D). No pathological changes were seen in the posterior roots. Teased fiber studies showed a significantly higher percentage of fibers that had undergone Wallerian-like degeneration in the BBGimmunized rabbits (12, 21, and 24%; Fig. 2E) than in three adjuvant controls (0, 0, and 3%). In contrast, the paralyzed rabbits had low percentages of fibers that had undergone paranodal demyelination (1, 0, and O%), as did the controls (0, 1, and 2%). Protein G positively stained some axons in spinal anterior roots obtained from BBG-immunized rabbits (Fig. 2F), evidence of the presence of IgG in or around the axons. Similar findings were obtained when anterior roots were immunostained with anti-rabbit IgG antibodies. No pathological changes or immunoglobulin deposits were detected in the sciatic nerves or anterior roots of the normal rabbits and adjuvant controls. Thin-layer chromatogram (TLC) with immunostaining showed that the rabbit plasma IgG strongly bound to GM1 (Fig. 3A). An enzyme-linked immunosorbent assay confirmed that the immune response predominantly was directed against GM1. Neither the IgMs nor IgGs obtained from these rabbits before inoculation nor those from the control rabbits bound
to GM1. Anti-GM1 IgM antibodies were detected in the 13 paralyzed rabbits 2 to 3 weeks after their first sensitization with BBG. Titers increased for 1 to 5 weeks (median, 2 weeks) after detection (range, 2,000 to 64,000; median, 32,000). Anti-GM1 IgG antibodies were detected in these rabbits 3 to 4 weeks after the first inoculation. The titers increased for 1 to 6 weeks (median, 3 weeks) after detection (range, 4,000 to 512,000; median, 32,000). Ten of the 13 developed flaccid paresis of the hind limbs within 1 to 3 weeks (median, 1week) after the peak anti-GM1 IgG titer was reached. Immunoblotting showed that none of the IgGs from the paralyzed rabbits bound to proteins from rabbit peripheral nerves. In contrast, TLC-immunostaining showed that the IgG strongly reacted with a band in the monosialosylganglioside fraction from rabbit peripheral nerve (Fig. 3B). The mobility of the band was similar to that of authentic GM1. Figure 3C shows the TLC blotting and negative secondary ion mass spectrum of the IgG-reactive band. On the TLC plates, cholera toxin B subunit, which specifically recognizes GM1, reacted with a band in the monosialosylganglioside fraction (Fig. 3B), which had mobility similar to that of the IgG-reactive band. These findings suggest that GM1 is present in rabbit peripheral nerves. The cholera toxin B-subunit bound to the axons of normal rabbit sciatic nerves (Fig. 3D) and anterior and posterior roots, evidence that GM1 is expressed on those axons. Immunohistochemical experiments with plasma from a BBG-immunized rabbit who had high anti-GM1 activity clearly
Figure 3. (A) Anti-ganglioside antibody from rabbits who developed limb weakness after sensitization with a BBG mixture. (a) TLC stained with the orcinol reagent for hexose. (b) Immunostained chromatogram that had been overlaid first with plasma from the rabbits then with peroxidase-conjugated anti-rabbit y-chain-specific antibodies. Lanes 1 to 10, plasma from 10 rabbits showing limb weakness. Orcinol reagent stains GM1, GDla, GDlb, and GTlb. Plasma IgGs from the rabbits strongly bind to GM1, and some weakly react with GDlb. (B) Target molecule for IgG autoantibodies among peripheral nerve gangliosides. (a) TLC stained with the orcinol reagent. (b) Immunostained chromatogram incubated first with plasma from a rabbit inoculated with BBG and then with peroxidase-conjugated anti-rabbit ychain-specific antibodies. (c) Binding of the peroxidase-conjugated cholera toxin B subunit. (d) Immunostained chromatogram incubated first with plasma from a rabbit inoculated with GM1 and then with peroxidase-conjugated anti-rabbit y-chain-specific antibodies. Lane 1, BBG mixture (Cronassial). Lane 2, authentic GM1 (Sygen) from bovine brain. Lanes 3 to 5, monosialosyl-, disialosyl-, and polysialosylganglioside fractions from rabbit peripheral nerves. The mobility of the monosialosylganglioside, with which the cholera toxin B subunit and IgGs from the rabbits react, is similar to that of authentic GM1. (C) Negative secondary ion mass spectrum of the monosialosylganglioside that reacted with IgG from a paralyzed rabbit. Cer, ceramide; Hex, hexose; HexNAc, N-acetylhexosamine; NeuAc, Nacetylneuraminic acid. The ion at m / z 1,544 consisted of stearic acid (C18:O) and sphingenine (d18:1), and that at m / z 1,572 of C18:O and icosasphingenine (d20:l). The seven major fragment ions of the sugar sequence ions in the spectrum were representative ceramides (miz 564, 592; out of spectrum), glucosylceramides (a; m / z 726, 754), lactosylceramides (b; m / z 888, 916), gangliotriaosylceramides (c; m / z 1,091, 1,119), gangliotetraosylceramides (d; m / z 1,253, 1,281), I13-N-acetylneuraminosyllactosylceramides(e; m / z 1,179, 1,207), and 113-N-acetylneuraminosylgangliotriaosylceramides(f; m / z 1,382, 1,410). The two fragment ions that corresponded to the nonreducing terminal side of the carbohydrate chain were m / z 833 (g; [(Hex-HexNAc-(NeuAc)Hex-OH)-H,-HI-) and m / z 995 (h; [(Hex-HexHexNAc-(NeuAc)Hex-OH)-H,-HI-). The m / z 290 and 308 ions that corresponded to N-acetylneuraminic acid also were present. Its spectrum, together with its mobility on the TLC plate similar to that of authentic GM1 and its binding with the cholera toxin B subunit, suggest that the structure of this ganglioside is Gal pl-3 GalNAc pl-4 (NeuAc a2-3) Gal pl-4 Glc pl-1’ Cer. (D) Cross section of normal rabbit sciatic nerve stained with the peroxidase-conjugated cholera toxin B subunit. Axons are stained. (E, F) Cross sections (E) and longitudinal (F) of normal rabbit sciatic nerve immunostained with plasma IgG that has anti-GM1 activity from a BBG-immunized rabbit. Axon surfaces are positively stained. Scale bars = 10 pm. From Yuki et al. (2001).
RABBIT MODEL OF GBS
CHAPTER 22
A
a
b
GD1b GTI b 1
6
2
3
4
5
6
a
b
1 2 3 4 5
1 2 3 4 5
7
c
8
9
10
d
GMI
3 1 2 3 4 5
385
386
YUKI
showed that its plasma IgG reacted positively with the axon surfaces of sciatic nerves from a normal rabbit (Fig. 3E, F). Nine of 11 rabbits immunized with GM1, KLH, and CFA developed flaccid paresis of the hind limbs. Quadriparesis and respiratory paresis rapidly progressed in two rabbits, who died, respectively, 3 and 11 days after the onset of limb weakness. Limb weakness progressed for 2, 6, 7, and 27 days, respectively, after onset in four rabbits, then reached a plateau. Weakness lessened in two rabbits 11 and 20 days after onset and disappeared in two rabbits 5 and 29 days after onset. TLC-immunostaining and the enzyme-linked immunosorbent assay showed that anti-GM1 IgM antibodies were induced and that a switch from IgM to IgG occurred in all 11 rabbits with or without development of paralysis. IgG from a paralyzed rabbit reacted with the GM1 of rabbit peripheral nerve (Fig. 3B). Mild to moderate Wallerian-like degeneration was present in Eponembedded sciatic nerve sections obtained from rabbits with limb weakness; the degenerative axon density was increased (929 f 310/mm2). There was mild to extensive fiber loss; the surviving axon density was decreased (12,044 1,680/mm2). Neither lymphocytic infiltration nor demyelination was found. Teased fiber preparations showed Wallerianlike degeneration in 21, 23, and 26% of the large myelinated fibers in three rabbits, and paranodal demyelination was seen in 0, 0, and 2% of these fibers. The clinical features of paresis with an acute, monophasic course in BBG-immunized rabbits were similar to those of patients with AMAN. AMAN has predominant axonal involvement, characterized by Wallerian-like degeneration of nerve fibers, with only minimal demyelination and minimal lymphocytic infiltration (McKhann et al., 1993). The paralyzed rabbits showed similar pathological changes with predominant axonal degeneration of the peripheral nerves. Results of teased fiber studies confirmed a higher percentage of Wallerian-like degeneration of the large myelinated fibers of the sciatic nerves. In addition, there was regenerative sprouting of anterior root fibers taken from a rabbit during the recovery phase. The paralyzed rabbits also had the anti-GM1 IgG antibodies. As in rodents and humans (Sheikh et al., 1999), GM1 is expressed on the axons of rabbit peripheral nerves. The IgG with anti-GM1 activity from a paralyzed rabbit strongly bound to the axon. As in AMAN patients (Hafer-Macko et al., 1996), IgG was deposited on the motor nerve axons in the paralyzed rabbits. The cholera toxin bound with both anterior and posterior roots, indicative of the presence of GM1 on sensory and motor fibers. Why neither the IgG deposition nor the Wallerian-like degeneration was seen in the posterior roots is unknown.
*
Sensitization with GM1 highly induced AMAN similar to that induced by the BBG mixture injection, providing strong support for the notion that GM1 is the immunogen in the BBG mixture and the target molecule for the autoantibodies. GM1 and galactocerebroside (GalC) are expressed in both the peripheral and central nervous system, but sensitization with these molecules produces only peripheral neuropathy (Saida et al., 1979a). The blood-nerve barrier that guards the peripheral nervous system is not as tight as the blood-brain barrier; therefore, small amounts of circulating IgG, which cannot enter the central nervous system, can penetrate the endoneurial space. This relative leakiness may make the peripheral nervous system, especially the roots and nerve terminals, more vulnerable than constituents of the central nervous system to IgG antibody-mediated disorders. Studies on BBG or purified GM1 as the agent for treating a variety of neurological disorders were initially reported with enthusiasm to be successful, but later double-blind controlled studies failed to confirm these findings (Bradley, 1990). Cronassial (BBG mixture) and Sygen (isolated GMl), which had been available on the Italian market, were used. Although it is still unclear whether there is an epidemiological relationship between exogenous gangliosides and GBS (Emilia-Romagna Study Group, 1999; Govoni et al., 1997), the results indicate that in some cases a BBG or GM1 injection may trigger AMAN. As in the patients with Ah4AN after BBG therapy (Illa et al., 1995), the immune response in the BBGinoculated rabbits predominantly directed against GM1. The BBG mixture contains more GDla than GM1, but it is unknown why the immune response against GDla was poor in the rabbits. All the antiGDlb antibody titers were lower than the anti-GM1 antibody titers. High antibody affinity has been proposed as a disease-determining factor in neuropathies associated with anti-GM1 antibodies (Mitzutamari et al., 1998). Plasma from rabbits taken at disease onset and 1 or 2 weeks before onset were tested for the presence of high-affinity anti-GM1 IgG antibodies (Comin et al., 2006). Affinity was estimated from soluble antigenbinding inhibition. High-affinity antibodies (binding inhibition by M GM1) were detected at disease onset but not before. No such difference was found for the other antibody parameters (titer, fine specificity, and population distribution). These findings support high affinity being an important factor in disease induction by anti-GM1 IgG antibodies. Various ganglioside immunization protocols were examined to refine the procedure for establishing an animal model of AMAN (Susuki et al., 2004).
CHAPTER 22
The most effective was subcutaneous injection of an emulsion of 2.5 mg of BBG mixtures, KLH and CFA, to Japanese white rabbits, with repeated injections at 3-week intervals. Under that protocol, all the rabbits developed marked flaccid paralysis associated with increased plasma anti-GM1 IgG antibody levels. This AMAN rabbit model could be reproduced by injection of incomplete Freund’s adjuvant instead of CFA and methylated bovine serum albumin instead of KLH, and by using New Zealand White rabbits instead of Japanese white rabbits.
IMMUNE ATTACK ON NERVE ROOT AXONS Pathological findings in AMAN autopsy cases show noninflammatory axonal degeneration of the motor axons with little demyelination (McKhann et al., 1993), but whether axonal damage is primary or secondary to demyelination was not clear. Secondary breakdown of axons under severe demyelination in GBS patients has been reported (Berciano et al., 1993; Fuller et al., 1992).When high doses of myelin antigens extracted from bovine nerve roots are used to induce experimental autoimmune neuritis, a model of demyelinating GBS, the disease is severe and marked by extensive axonal degeneration (Hahn et al., 1988). A prominent pathological feature of AMAN is the presence of macrophages within the periaxonal space, surrounding or displacing the axon and surrounded by an intact myelin sheath (McKhann et al., 1993). This characteristic finding is supportive evidence of primary immune-mediated attack on the axons in AMAN. The presence of periaxonal macrophages-evidence of primary axonal damagewas not shown in our first study (Yuki, 2001). A pathological study of human AMAN showed the initial lesions to be mainly in the spinal nerve roots (Griffin et al., 1995). In GBS, electrophysiological evidence of nerve dysfunction is most prominent in the distal nerve terminals, nerve roots, or common entrapment sites (Brown and Snow, 1991). In the second study (Susuki et al., 2003), pathological, immunohistochemical, and electrophysiological evaluations were performed to determine the distribution of initial lesions in the rabbit AMAN model. Five of seven rabbits sensitized with GM1 subcutaneously developed flaccid paresis of their limbs, weakness being severe in three and mild in two. All six rabbits immunized with BBG subcutaneously developed flaccid paresis of the limbs; five had severe weakness, and the other mild weakness. All the paralyzed rabbits sensitized with BBG or GM1 had high titers of anti-GM1 IgG antibodies within a week after the onset of limb weakness. Seven of eight rabbits
-
RABBIT MODEL OF GBS
387
immunized with GalC subcutaneously also developed flaccid limb paresis, and all had anti-GalC IgG antibodies. None of the seven control rabbits showed limb weakness until 16 weeks after the first inoculation. Mild or moderate Wallerian-like degeneration was seen in toluidine blue-safranine-stained cross sections of the sciatic nerves from the GM1- or BBGimmunized rabbits. Demyelination and remyelination were rare, and there was no lymphocytic infiltration. There were no significant findings for the rabbits killed 3 or 4 days after onset or in the rabbit with mild weakness. Unlike the sciatic nerves, cauda equina or spinal nerve root specimens from the paralyzed rabbits frequently showed macrophage infiltration in the periaxonal space of the nerve fibers (Fig. 4A to E). These macrophages, present within the nerve fibers, occupied most of the periaxonal space (Fig. 4D) or colocalized with atrophic axons (Fig. 4E). The surrounding myelin sheaths appeared almost normal. This characteristic finding was never present in nerve specimens from GalC-immunized rabbits with limb weakness. Six BBG-immunized rabbits with limb weakness were prepared for immunohistochemical study. Protein G positively stained some axons, evidence of IgG deposits in or around the axons (Fig. 4F to J). In some fibers, the axolemma appeared to be specifically stained (Fig. 41). In adjacent cross and longitudinal sections, lengths of the stained portions of the axons ranged from 40 to 60 pm. This phenomenon was found mainly in the spinal nerve roots or cauda equina and in a few sciatic nerve axons, but never in the tibia1 nerves. No selectively stained axons were seen in cauda equina specimens from a normal, an adjuvant control, and a GalC-sensitized rabbit. Ranvier nodes were specifically stained by protein G in longitudinal sections of the cauda equina from a BBG-immunized rabbit (Fig. 4K). Electrophysiological evaluations were made for three rabbits inoculated with BBG, seven with GM1, eight with GalC, and seven control animals. A nerve conduction study in the GalC-immunized rabbits showed typical demyelination: prolonged distal latency, delayed motor conduction velocity, and severe temporal dispersion. Representative results are shown in Fig. 5B. Gradual improvement paralleled spontaneous recovery of muscle power in a GalCsensitized rabbit (Fig. 5C). Changes in nerve conduction study data for the paralyzed GM1- or BBGimmunized rabbits were minimal during the acute phase of illness. Distal latency was slightly prolonged in four of six rabbits, but the changes were not remarkable as compared with those in GalC-sensitized rabbits. Compound muscle action potential ampli-
388
WKI
CHAPTER 22
tude was slightly decreased (6.4 mV) in a GM1immunized rabbit but unremarkable because the amplitude recorded 14 days before onset was 6.8 mV (Fig. 5D, E). No obvious temporal dispersion of compound muscle action potential was found. Motor conduction velocity was within the normal range in all the rabbits. Detailed evaluations showed abnormal change of late potentials. Late components of F waves were absent in three rabbits, whereas minimal F latencies were preserved. Nerve conduction study results for the control rabbits did not show any remarkable abnormalities. Four of the six paralyzed rabbits showed denervation potentials in the needle electromyography. Hypo- or areflexia was detected in five of the six paralyzed animals. One rabbit who had mild weakness showed neither denervation potentials nor hyporeflexia. The clinical course of a GM1-immunized rabbit was followed; it had severe tetraparesis but subsequently recovered. In the acute phase, there were no obvious changes in distal latency, motor conduction velocity, and compound muscle action potential amplitudes (Fig. 5E). Late F-wave components were absent (Fig. 5H). Amelioration began 8 days after symptom onset. Late F-wave components appeared with slightly prolonged latencies again 14 days after onset (Fig. 51). In the needle electromyography, denervation potentials were detectable 18 days after onset. Muscle power of the limbs gradually improved, but loss of body weight and muscle atrophy continued. Compound muscle action potential amplitudes decreased 6 weeks after onset (Fig. 5F). In some pathological studies of human AMAN, the most conclusive finding has been the presence of the periaxonal macrophages (McKhann et al., 1993). This characteristic was found in the GM1- or BBGimmunized rabbits. Primary axonal damage, not axonal breakdown resulting from severe demyelination, therefore occurred in the AMAN model. Sciatic nerve specimens, which did not show it, were mainly used in our previous study (Yuki et al., 2001). Because this characteristic image was found in the spinal roots of
RABBIT MODEL OF GBS
389
AMAN patients (McKhann et al., 1993), it was investigated in the cauda equina and nerve roots. Retrospectively, this characteristic also was found in the cauda equina of a BBG-immunized rabbit in our previous study (Yuki et al., 2001). Patients who have died in the early stage of AMAN show IgG and the complement activation product C3d frequently bound to the nodal axolemma of motor nerve fibers (Hafer-Macko et al., 1996). In the more severe cases, IgG and C3d have been detected within the periaxonal space of myelinated internodes. Similarly, IgG deposits on axons were present in the AMAN model. The internodal axolemma appeared to be selectively stained in some nerve fibers, and Ranvier nodes were selectively stained in a BBG-immunized rabbit. Like the periaxonal macrophages, IgG deposits on axons primarily were located in the spinal nerve roots or cauda equina, less frequently in the sciatic nerve, and never in the tibia1 nerve. This distribution corresponds well to that of the demyelinative lesions found in GalC-sensitized rabbits (Saida et al., 1979b) and the initial lesions seen in AMAN patients (Griffin et al., 1995). Nerve roots are thought to be vulnerable because of a relative blood-nerve barrier deficiency (Olsson, 1968), and initial lesions in rabbits must mainly involve nerve root axons. Consequently, Wallerian-like degeneration would occur on the peripheral nerves in the AMAN model. Only mild or no Wallerian-like degeneration was found on sciatic nerve specimens in rabbits killed 3 or 4 days after the onset of limb weakness, but it was frequent in AMAN rabbits killed in the advanced or recovery phase. Nerve conduction study results supported this distribution of initial lesions. In the AMAN rabbit model, despite marked limb weakness in the acute phase of the illness, neither compound muscle action potential amplitudes nor motor conduction velocities showed obvious changes, indicating that distal motor nerve conduction was preserved. These nerve conduction study findings are not similar to those for AMAN patients (Tamura et al., 2007) and differ markedly from those for GalC-sensitized rabbits,
Figure 4. (A to E) Macrophages in nerve fibers. Cross sections of the cauda equina from a GM1-immunized rabbit. (A to C) Toluidine blue-safranine stain. Macrophages are present in the nerve fibers. Demyelination and remyelination are rare, and there are no inflammatory cells in the endoneurium. (D) Electron micrograph of the nerve fiber with macrophage infiltration shown in (A). A macrophage (m) occupies the periaxonal space, and the axon has disappeared. (E) Another example of a nerve fiber with macrophage infiltration. Macrophage (m) processes surround the atrophic axon (a). The surrounding myelin sheath appears almost normal. Scale bars = 10 pm. (F to K) IgG deposits on axons. Specimens from a BBG-immunized rabbit stained with serially diluted peroxidase-conjugated protein G. (F, G) Adjacent cross sections (20 p m thick) of ventral roots. (F) Some axons are strongly stained (arrow, arrowhead). (G) Staining intensity gradually decreases (arrow). (H) High-power magnification of the nerve fiber in (B) (arrowhead). The axon is stained diffusely. (I) Cross section of the cauda equina. Selective staining is seen along the axonal membrane. (J, K) Longitudinal sections of the cauda equina. 0) The axon is selectively stained for about 50 pm, the intensity gradually decreasing on both sides. (K) Ranvier nodes are stained selectively. Scale bars = 10 pm. From Susuki et al. (2003).
390
WKI
which indicated typical demyelination. Part of the late potentials was absent in three paralyzed rabbits. The features of these potentials are compatible with the F wave; they are present on supramaximal stimulation of the nerve, their latencies are shorter on stimulation at proximal sites other than the ankle, and the waveform varies. Similarly, the absence of F waves is an isolated conduction abnormality during the acute phase of human AMAN (Kuwabara et al., 2000). These findings point to the presence of nerve conduction failure at proximal sites. Compound muscle action potential amplitude reduction, later de-
E I
H
I
II
tected in a GM1-immunized rabbit, may indicate subsequent Wallerian-like degeneration. Several points, however, differ from F-wave findings in human AMAN. Latencies of the late F wave components reelicited in the recovery phase were slightly prolonged in the GM1-immunized rabbit, whereas F waves that appeared during follow-up studies of AMAN patients showed no prolongation of latency (Kuwabara et al., 2000). F waves with minimal latencies were preserved in the paralyzed rabbits, whereas all F waves were absent or markedly decreased in AMAN patients. Some fast-conducting motor fibers may be
CHAPTER 22
spared throughout their entire lengths. As in those findings, rabbits inoculated with a mixture of canine sciatic nerve and CFA had normal F-wave latencies, but demyelination was detected pathologically in the ventral roots, sciatic nerve, or both (Tuck et al., 1982). These phenomena may be due to differences between rabbits and human beings. In AMAN patients, axonal degeneration appears to develop predominantly in the motor nerve terminal, and only occasionally in the spinal anterior roots (Tamura et al., 2007). The blood-nerve barrier at nerve terminals is deficient because there is no perineurium (Burkel, 1967). Markedly reduced compound muscle action potential amplitudes or inexcitable nerves, often present in AMAN patients (Hiraga et al., 2005), were not found in the six paralyzed AMAN rabbits, including those with severe limb weakness. Why conduction abnormality at the nerve terminals was not detectable in the AMAN model is not clear, but the blood-nerve barrier of rabbit nerve terminals is also deficient. The GalC-sensitized rabbits had remarkable demyelinative nerve conduction study findings, especially at distal sites in the peripheral nervous system, indicative that demyelination occurs on nerve terminals. A previous immunohistochemical study suggested that GM1 is expressed on the nerve terminals of rabbits as well as humans, but the rabbit strain used was not reported (Thomas et al., 1989). Certain technical problems of the nerve conduction study should be considered, and mild conduction failure on nerve terminals may not have been detected. Features of the lesions on the nerve terminals of AMAN rabbits, such as severity or time of onset, may differ from those of AMAN patients.
RABBIT MODEL OF GBS
391
Detailed electrophysiological and pathological examinations are required to evaluate whether immunemediated lesions are present on the nerve terminals of the AMAN rabbits.
LOS IMMUNIZATION MODEL Chickens fed C. jejuni isolated from a GBS patient developed a paralytic neuropathy (Li et al., 1996). Their nerves showed Wallerian-like degeneration similar to that in human AMAN contracted after C. jejuni enteritis. Neither whether the C. jejuni strain carried a GMl-like LOS nor whether the paralyzed chickens had anti-GM1 antibodies was investigated. The most straightforward way to verify whether molecular mimicry between microbes and autoantigens causes GBS would be to establish a GBS model by immunization of animals with components of an antecedent infectious agent. Several research groups have failed to induce neuropathy by sensitization with the GM1-like LOS of C. jejuni. Rats immunized with C. jejuni LOS had only an IgM response to GM1 (Wirguin et al., 1997). Anti-GM1 IgG antibodies were induced in rabbits by sensitization with the LOS from the C. jejuni reference strain obtained from an enteritis patient or LOSSfrom GBSassociated strains, but there was no muscle weakness (Ang et al., 2000). This might have been because the rabbits were immunized with a smaller amount of LOS and no KLH; moreover, the observation period might have been too short. C. jejuni LOS bearing GM1- and GDla-like structures (Fig. 1) (Koga et al., 2006; Yuki et al.,
Figure 5 . (A to F) Distal nerve conduction study results. Stainless steel needles were inserted close to the left sciatic nerve and its tibial branch at the ankle, knee, and sciatic notch to stimulate the nerves. At each level, a needle placed close to the nerve served as the cathode, and a remote subcutaneous one as the anode. Recording from plantar muscles was through subcutaneous needles, one placed transversely over the muscle bellies in the sole of the foot, the other at a distance. Compound muscle action potentials (CMAPs)were recorded after supramaximal stimulation. The upper, middle, and lower traces are for stimulus to the left sciatic nerve and its tibial branch at the three levels ankle, knee, and sciatic notch. (A to C ) Serial nerve conduction studies of a galactocerebroside-sensitizedrabbit. (A) CMAPs recorded before sensitization. CMAP amplitude is 13.0 mV, distal latency (DL) 2.10 ms, and motor conduction velocity (MCV) 48.0 m/s. (B) CMAPs recorded 2 weeks after the onset of limb weakness (nadir of the symptoms). Note that the amplitude scale is 1 mV per division. CMAPs show remarkable temporal dispersion, and their amplitudes are markedly decreased. CMAP amplitude is 1.5 mV, DL is prolonged to 3.40 ms, and MCV is decreased to 37.5 m/s. (C) CMAPs recorded 14 weeks after the onset of limb weakness. Muscle power has recovered almost to normal. CMAP amplitude has increased with the resolution of temporal dispersion (7.0 mV). DL is still prolonged (3.50 ms). MCV has recovered to 51.0 m/s. (D to F) Serial nerve conduction studies of a GM1-immunized rabbit. (D) CMAPs recorded 14 days before the onset of limb weakness. CMAP amplitude is 6.8 mV, DL 2.10 ms, and MCV 51.1 m/s. (E) CMAPs recorded 4 days after the onset of limb weakness. CMAP amplitude (6.4 mV) and DL (2.10 ms) show no marked changes. MCV is 62.5 m/s. (F) CMAPs recorded 6 weeks after the onset of limb weakness. Muscle power has gradually returned, but limb weakness and muscle atrophy are still present. CMAP amplitude has decreased to 2.5 mV, DL 2.30 ms, and MCV 52.3 m/s. (G to I) Serial F-wave recordings from a GMlimmunized rabbit. F waves were recorded after ankle stimulation. Ten consecutive recordings (minimum) were obtained after supramaximal stimulation delivered at the frequency of 1 Hz. (G) F waves before sensitization. (H) F waves recorded 4 days after onset of limb weakness. No late F-wave components have been elicited, whereas minimal F-wave latencies are preserved. (I) F waves recorded 2 weeks after onset of limb weakness. Clinical symptoms have begun to be ameliorated. Late F wave components again are recorded, but their latencies are slightly delayed. From Susuki et al. (2003).
392
WKI
1993) was prepared from the strain (CF90-26) isolated from an AMAN patient. Male Japanese white rabbits weighing 2.0 to 2.5 kg were immunized (Yuki et al., 2004). A 2.5- or 10-mg portion of the C. jejuni LOS was dissolved in 0.5 ml of KLH (2 mg/ml) in phosphate-buffered saline, after which 0.5 ml of CFA was added, and the mixture emulsified. A 1-ml sample of the C. jejuni LOS emulsion was injected subcutaneously to the back at 3-week intervals until limb weakness developed or 12 months had passed since the first inoculation. Control rabbits were injected under the same protocol with 10 mg of Escherichiu coli K-12, D31m4 (Re) LOS, Salmonella entericu serovar Minnesota FU95 (Re) LOS, or without the LOS. The cholera toxin B subunit (a specific ligand for GM1-oligosaccharide) bound strongly to the C. jejuni LOS but did not react with E . coli K-12 LOS or S. enterica serovar Minnesota R595 LOS. None of 10 controls inoculated only with KLH and CFA every 3 weeks showed limb weakness at 12 months after the first inoculation, whereas 4 of 10 rabbits immunized with 2.5 mg of C. jejuni LOS, KLH, and CFA developed limb weakness sooner. Two rabbits developed flaccid paresis of the hind limbs 215 and 216 days after the initial inoculation and tetraparesis the next day. They were unable to lift their heads and bodies, respectively, 5 and 4 days after onset. Two other rabbits developed tetraparesis 133 and 329 days after the initial inoculation, but weakness was mild during the course of the illness. None of control rabbits inoculated with 10 mg of E. coli LOS (n = 5 ) or Salmonella serovar Minnesota LOS (n = 5) showed limb weakness at 12 months after the first inoculation. In contrast, all eight rabbits immunized with 10 mg of the C. jejuni LOS, KLH, and CFA developed limb weakness earlier. One rabbit developed flaccid paresis of the hind limbs 64 days after the initial inoculation and tetraparesis 7 days after its onset (Fig. 6A). The others respectively developed tetraparesis 40 to 227 days (median, 128 days) after the initial inoculation. Quadriparesis and respiratory paresis developed and
progressed rapidly in three rabbits, causing death 1, 2, and 1.5 days after onset of limb weakness. One rabbit, however, had a monophasic course like that of patients with AMAN. After onset, limb weakness worsened for 8 days, reached a plateau, then lessened from day 16. TLC with immunostaining showed that as to rabbit peripheral nerve gangliosides, plasma IgG antibodies from the paralyzed rabbits reacted strongly with the GM1 (Fig. 6B) and with the C. jejuni LOS, whereas neither E. coli LOS nor Salmonella serovar Minnesota LOS induced anti-GM1 antibodies in the rabbits. Six of the eight rabbits developed flaccid paresis within 3 weeks (median, l week) after the peak anti-GM1 IgG titer was reached (range, 2,000 to 32,000; median, 8,000), but the titers did not correlate with the severity. The paralyzed rabbits did not develop anti-GDla antibodies, although GDla-like LOS as well as GM1-like LOS was included in the immunogen. The spinal nerve roots of the paralyzed rabbit killed 11 days after onset had few nerve fibers showing Wallerian-like degeneration, whereas the nerve roots showed occasional macrophages within the periaxonal spaces surrounded by almost intact myelin sheaths (Fig. 6C to F). Axons of these nerve fibers had various degrees of degeneration. Sciatic nerve specimens from the paralyzed rabbits showed more severe Wallerian-like degeneration than did the proximal nerve roots (Fig. 6G). Demyelination and remyelination were rare, and no inflammatory cells were present in the endoneurium. Furthermore, protein G bound selectively to some axons in the cauda equina of one of the paralyzed rabbits, indicative that there were IgG deposits on those axons. In contrast, no significant changes were found in the brains or spinal cords of any of the paralyzed rabbits. These findings, which are compatible with the features of human AMAN (Hafer-Macko et al., 1996; McKhann et al., 1993), provide evidence that the rabbits inoculated with C. jejuni LOS constitute a valid AMAN model.
Figure 6 . (A) Rabbit with flaccid limb weakness induced by sensitization with Cumpylobucter jejuni LOS. It lies splayed out, all extremities extended, head on the floor, instead of sitting upright in the usual compact, hunched posture. (B) Anti-ganglioside antibody from rabbits that developed limb weakness after sensitization with C. jejuni LOS. Of the BBGs, plasma IgG from a paralyzed rabbit binds to GM1 (lane l), isolated GM1 from bovine brain (lane 2), and GMl from rabbit peripheral nerve (lane 3). IgGs from other three paralyzed rabbits and the GM1 from rabbit peripheral nerve (lane 3). (C to F) Macrophages in nerve fibers. Cross sections of the cauda equina from a paralyzed rabbit. (C, D) Toluidine blue stain. Macrophages are present in the nerve fibers (arrowheads). The initial degenerated axon stage also is shown (D, arrow). Demyelination and remyelination are rare, and there are no inflammatory cells in the endoneurium. (E, F) Electron micrographs of nerve fibers with macrophage infiltration. The nerve fiber in (E) is the same as in (C). Macrophages (m) occupy the periaxonal space between the atrophic axons (a) and surrounding myelin sheaths which appear almost normal. (G) Wallerianlike degeneration of nerve fibers. Cross sections of the sciatic nerve from a rabbit killed 39 days after onset. Toluidine blue stain. Myelin ovoids produced by Wallerian-like degeneration of the myelinated fibers are present (arrowheads). Scale bars = 10 pm. From Yuki et al. (2004).
CHAPTER 22
-
RABBIT MODEL OF GBS
1 2 3 1 2 3 3 3 3
C
E
G
F
393
394
WKI
New Zealand White rabbits were immunized with both the C. jejuni LOS isolated from a GBS patient and Freund’s adjuvant (group I) or with Freund’s adjuvant plus KLH (group 11) (Caporale et al., 2006). The LOS expressed the GM1 and GDla epitopes. Both rabbit groups had high anti-LOS and anti-GM1 antibody titers and lower anti-GDlb and anti-GDla antibody titers. Weakness and sciatic nerve axonal degeneration was present in 1 of 11 in group I and 6 of 7 in group 11, evidence that KLH has an additional function in neuropathy induction. LOS is composed of oligosaccharide and lipid A. Several lines of evidence support a GM1 oligosaccharide rather than lipid A structure for the C. jejuni LOS that has importance in the development of AMAN. E . coli K-12 LOS and Salmonella serovar Minnesota R595 LOS do not carry the GM1 epitope. Sensitization of those LOSs that carry lipid A did not induce anti-GM1 antibodies in the rabbits, evidence that anti-GM1 IgG is not a result of polyclonal B-cell stimulation. None of the rabbits developed limb weakness. In contrast, the AMAN model was established that uses inoculation with GM1, which carries a ceramide moiety but not lipid A (Yuki, 2001), indirect evidence that the nature of the GM1oligosaccharide structure is important for the development of AMAN. Direct evidence could be obtained by the use of knockout C. jejuni (CF90-26) lacking the GM1-like LOS. The cst-II gene encodes an enzyme that transfers sialic acid to LOS, and neuAl encodes an enzyme that synthesizes the donor (CMP-sialic acid) used by CstI1 sialyltransferase (Gilbert et al., 2000). Because both genes function in LOS sialylation, they are essential for ganglioside-like LOS synthesis. Mutants of C. jejuni that lack these genes were produced and analyzed (Godschalk et al., 2004). Whereas GM1-like and GDla-like LOSs were identified in wild-type C. jejuni strains isolated from GBS patients, neither was present in the corresponding cst-II and neuAl knockout mutants. cst-II and neuAl knockout mutants, unlike the wild types, had decreased reactivity to GBS patients’ sera. GM2/ GD2 synthase knockout mice, which lack GM1 and GDla, are immune-naYve hosts that can be used to obtain high-titer anti-ganglioside antibody responses. Immunization with the wild-type strain induced an anti-GDla IgG antibody response in the mice, although it did not induce an anti-GM1 IgG antibody response. In contrast, mutant strain immunization did not. These mean that the genes involved in ganglioside-like LOS biosynthesis are essential for the induction of anti-ganglioside antibodies, as confirmed by the following studies. The neuBl gene that encodes sialic acid synthetase also is required for the synthesis of ganglioside-like LOS, and
the gulE gene functions in the biosynthesis of LOS outer-core oligosaccharide structures (Shu et al., 2006; Xiang et al., 2006). The C. jejuni (HB9313) parental strain expresses an LOS with a GM1 epitope, and knockout mutants of neuB1 and gulE express a truncated LOS without that epitope. AntiGM1 IgG antibodies were induced in guinea pigs immunized with the wild-type strain but not in those immunized with mutants. Axonal degeneration of sciatic nerves of guinea pigs sensitized with the wildtype strain was more frequent than of animals sensitized with both mutants.
COMPLEMENT-MEDIATED NERVE INJURY AND ITS TREATMENT Whether anti-GM1 antibodies affect sodium channels at the nodes of Ranvier is controversial (Hirota et al., 1997; Takigawa et al., 1995), but anti-GM1 antibodies have been shown to mediate complement-dependent destruction of sodium channel clusters in peripheral motor nerves (Susuki et al., 2007). Japanese white rabbits sensitized with BBG that included GM1 had anti-GM1 IgG antibodies and acute flaccid paralysis. These AMAN rabbits were killed when in the acute progressive (a few days after onset), early recovery (2 weeks after onset), and late recovery (more than 4 weeks after onset) phases, and immunohistochemical studies were performed on spinal anterior roots obtained from each rabbit. As in human AMAN (Griffin et al., 1996), lengthening of the nodes of Ranvier was present in AMAN rabbits in the acute phase (Susuki et al., 2007). IgG was deposited at some nodes, where GM1 was expressed. The bound autoantibodies activated complement, and C3 components were deposited at the nodes. Thereafter, the final product of complement activation, the membrane attack complex, was formed at the nodal axolemma. Sodium channel clusters disappeared at the lengthened nodes with complement deposition. Sodium channels, PIV spectrin, and neurofascin 186 were localized at the nodes of Ranvier in ventral roots obtained from control rabbits. When complement-deposited nodes were lengthened in the AMAN rabbits, sodium channel clusters and related molecules disappeared. Schwann cell microvilli are important in the stabilization of nodal sodium channels (Yang et al., 2004). In the AMAN rabbits, the moesin at the microvilli disappeared. Paranodal junctions act as a diffusion barrier restricting lateral mobility of nodal sodium channels (Bhat et al., 2001). Autoimmune attack caused the disappearance of contactin-associated protein at the paranodes. In the advanced stage of paranodal dis-
CHAPTER 22
ruption, potassium channels at the juxtaparanodes were disrupted and disappeared. Complement deposition was prominent in the acute progressive phase but decreased with the clinical course. Macrophages were not always recruited at complement deposit nodes, where the nodal structure was disrupted. Macrophage invasion was not prominent in the acute progressive phase but was in the early recovery phase. These findings suggest that complement has a crucial role in nerve injury and that macrophages scavenge injured nerve fibers. In human AMAN, axonal degeneration appears to develop predominantly in the motor nerve terminals, and only occasionally more proximally in the nerve roots (Tamura et al., 2007). The possible pathogenesis of AMAN subsequent to C. jejuni enteritis is as follows. (i) Infection by C. jejuni bearing GM1-like LOS induces production of anti-GM1 IgG antibodies. (ii) The autoantibody binds to GM1 at the nodes of Ranvier in the motor nerve terminals. (iii) Activated complement recruited by the anti-GM1 antibodies then forms membrane attack complex (Fig. 7). (iv) Autoimmune attack disrupts the nodal cytoskeleton, Schwann cell microvilli, and paranodal axo-glial junctions. Sodium channels disappear, producing muscle weakness in the early phase of illness, and in severe cases, Wallerian-like degeneration subsequently occurs.
ACTION MECHANISM OF IMMUNOGLOBULIN TREATMENT Intravenous immunoglobulin (IVIG) is effective for GBS in shortening recovery time (Hughes et al., 2007), but the mechanism of action has yet to be clarified. IVIG treatment of GBS patients with antiGM1 antibodies was reported as superior to plasmapheresis in terms of time to recovery (Jacobs et al., 1996; Kuwabara et al., 2001), although this could be not confirmed in another study (Hadden et al., 1998). Clinical, histological, and immunological effects in a disease model of AMAN treated by IVIG were evaluated (Nishimoto et al., 2004). Rabbits were sensitized with BBG including GM1, and divided randomly into two groups at disease onset. One received homologous immunoglobulin (400 mg/kg per day) intravenously for 5 days (n = 15); the other received saline (n = 15). Disease severity was scored (0 to 13 points) daily. Between both groups at onset, there was no difference in any characteristics, including clinical score. The IVIG group had faster recovery than the saline group, and the percentage of rabbits that improved by a score of 4 or less was higher in the IVIG (53%) than in the saline (13%) group 60
RABBIT MODEL OF GBS
395
days after onset. The therapeutic efficacy of IVIG in the AMAN model was confirmed. In the AMAN rabbits 60 days after onset, there was no significant difference in anti-GM1 IgG titers between the IVIG and saline-treated groups. Whether IVIG can effectively reduce anti-GM1 IgG titer in GBS patients has yet to be investigated, but it did not reduce the anti-GM1 IgM titers of multifocal motor neuropathy patients whose strength improved (Chaudhry et al., 1993). Anti-idiotypic antibodies in IVIG, discussed below, may affect antibody production by sending negative signals to B cells (Dalakas, 2002). The large amount of IgG in an IVIG preparation may saturate the neonatal Fc receptor, accelerate the catabolism of endogenous pathogenic IgG, and reduce autoantibody levels (Yu and Lennon, 1999). The findings, however, suggest that neither suppression of anti-GM1 IgG production nor accelerated catabolism of the autoantibody occurred in AMAN rabbits treated with WIG, and neither IVIG mechanism has importance for the effective treatment of GBS associated with anti-GM1 IgG. The anterior roots of rabbits surviving 60 days after onset showed lower frequency of axonal degeneration in the IVIG-treated rabbits (n = 11, mean 4.5%) than that in the saline-treated rabbits (n = 8, mean 11.1%). Histological findings for the AMAN rabbits treated with IVIg suggest partial prevention of axonal degeneration in the ventral root. These findings are important for understanding the action mechanism of IVIG, which cannot be obtained from clinical trials and in vitro experiments, although there are potential differences between rabbit and human IgG. IVIG produced less degeneration of motor axons, causing the faster clinical improvement in the AMAN rabbits. As mentioned, activated complement recruited by the anti-GM1 IgG antibodies form membrane attack complex, resulting in motor axonal injury (Susuki et al., 2007). IVIG may inhibit complement activation and prevent axonal degeneration. Safer therapy than blood products can be obtained by clarifying the pharmacological mechanisms of IVIG. Intact-type immunoglobulin was superior to F(ab'), of the IgG molecule in decreasing the clinical score of experimental autoimmune neuritis rats treated before onset (Miyagi et al., 1997). This indicates that the Fc portion of the IgG molecule is more important for the therapeutics. Whether the F(ab'), or Fc portion is as effective as intact-type immunoglobulin used in the AMAN model will need to be examined. This will provide more information about the mechanisms of therapeutic action and facilitate development of new drugs. As an example, if the Fc portion of IgG is effective in the AMAN model, its active part can be identified. Moreover, if
396
YUKI
A Normal
Juxtaparanode Paranode Q Kv 0 Caspr
Node (INav
Cytoskeleton
6 AMAN rabbits
Anti-GM1 IgG Ab
oC3
@ MAC
Figure 7. Schematic presentation of the nodal disruption in peripheral motor nerve fibers in AMAN model. (A) Structures and molecular organizations at and near normal nodes of Ranvier. Nodal voltage-gated Na+ (NAV) channels are located at nodes and make multiprotein complexes, including cytoskeletal proteins such as PIV spectrin. Nav channel clusters are further stabilized by interaction between the complexes and Schwann cell microvilli. Contactin-associated protein (Caspr) forms axoglial junction at paranodes, which act as diffusion barrier to restrict lateral mobility of nodal Nav channels. Voltage-gated Kf (Kv) channels are localized to juxtaparanodes. (B) Autoimmune-mediated disruption of Nav channel clusters and nodes of Ranvier. Anti-GM1 IgG antibodies cause complement-mediated attack with membrane attack complex (MAC) formation at the nodal and paranodal axolemma. Nav channel clusters are altered by destruction of structures mediating their stabilization, including axonal cytoskeleton at nodes, Schwann cell microvilli, and paranodal junctions. As the autoimmune-mediated destruction extends, Nav channels and other components at and near nodes disappear. Kv channel clusters are preserved unless the immune attack extends to juxtaparanodes. From Susuki et al. (2007).
treatment with synthetic peptides of the active part in the Fc portion is successful in our AMAN model, then cheaper and safer peptide treatments should prove to be of future clinical use.
PASSIVE TRANSFER MODEL GM1 is the autoantigen for IgG antibodies in some patients with AMAN subsequent to C. jejuni enteritis (Yuki et al., 1990). IgG class of the autoan-
tibody was produced against GM1 and GM1-like LOS in experimental animals (Yuki et al., 2001; Yukie et al., 2004). Pathological changes appeared in the peripheral nerves that are identical to those seen in human AMAN. The data needed to satisfy Witebsky’s postulates (the equivalent of Koch’s postulates for an autoimmune pathogenic sequence) are induction of clinical and pathological disease by passive transfer of anti-GM1 IgG antibodies (Sheikh and Griffin, 2001). According to the procedure described elsewhere (Saida et al., 1978), we performed intraneu-
CHAPTER 22
ral injection of the sciatic nerves in rats (Yuki et al., 2002). Intraneural injection of rabbit antigalactocerebroside plasma produced focal demyelinative lesions, whereas injection of rabbit anti-GM1 plasma induced predominant axonal degeneration. Moreover, purified IgG from a patient with AMAN, which has anti-GM1 activity, produced predominant axonal degeneration when homologous complement was given. Systemic transfer of anti-GM1 IgG antibodies from AMAN patients did not induce paralysis in mice, but the autoantibody did block muscle action potentials in a rat muscle-spinal cord coculture (Yuki et al., 2004). Passive transfer attempted with systemically administered mouse anti-GDla IgG antibody did not cause nerve fiber degeneration despite high circulating autoantibody titers (Sheikh et al., 2004). Half of a population of mice given an intraperitoneal implant of anti-GDla IgG antibody-secreting hybridoma did, however, develop a patchy, predominantly axonal neuropathy that affected a small number of nerve fibers. Mice implanted with the hybridoma had a leaky blood-nerve barrier compared with those that received systemically administered anti-GDla IgG antibody. These findings intimate that in addition to circulating anti-ganglioside antibodies, such factors as antibody accessibility and nerve fiber resistance to antibody-mediated injury are important in the development of a neuropathy. Ex vivo nerve-muscle preparations from GDla-overexpressing, GD3 synthase knockout mice were exposed to mouse anti-GDla IgG antibody in the presence of a complement source (Goodfellow et al., 2005). Dense antibody and complement deposits were present only on presynaptic motor axons and were accompanied by severe ultrastructural damage and electrophysiological blockade of motor nerve terminal functions. Identical paralyzing effects were found in human anti-GDla-positive sera tests. IgG antibodies to GQlb ganglioside are believed to be pathogenic in Fisher syndrome, a variant of GBS, characterized by acute ophthalmoplegia, ataxia, and areflexia (Willison and Yuki, 2002). G Q l b is enriched in presynaptic membranes, and some clinical and experimental evidences suggest that the neuromuscular junction is affected in Fisher syndrome. Mouse models demonstrated that anti-GQlb antibodies activate the classical complement pathway with membrane attack complex formation, resulting in severe neuronal and perisynaptic Schwann cell membrane injury (Halstead et al., 2004). In these Fisher syndrome murine models, an inhibitor of complement activation completely prevents membrane attack complex formation, and thereby has a major neuroprotective effect at the nerve terminal (Halstead
RABBIT MODEL OF GBS
397
et al., 2005). Despite the frequent use of immunomodulating therapies, GBS still carries considerable mortality and residual disability (Hughes et al., 2007). The modulation of complement pathway during the acute phase might be expected to be an effective therapeutic approach to attenuate the disease process caused by anti-ganglioside antibodies.
SUMMARY Establishment of the rabbit AMAN model by sensitization with GM1 or C. jejuni LOS has provided conclusive evidence that AMAN can be caused by molecular mimicry. This model is helpful to clarify the molecular pathogenesis of AMAN and to develop new treatments for it. REFERENCES Ang, C. W., H. P. Endtz, B. C. Jacobs, J. D. Laman, M. A. de Klerk, F. G. A. van der Mecht, and P. A. van Doorn. 2000. Cumpylobacter jejuni lipopolysaccharides from Guillain-Barrt syndrome patients induce IgG anti-GM1 antibodies in rabbits. J. Neuroimmunol. 104: 133-138. Ang, C. W., B. C. Jacobs, and J. D. Laman. 2004. The GuillainBarrt syndrome: a true case of molecular mimicry. Trends Immunol. 25:61-66. Berciano, J., F. Coria, F. Monton, J. Calleja, J. Figols, and M. LaFarga. 1993. Axonal form of Guillain-Barrt syndrome: evidence for macrophage-associated demyelination. Muscle Nerve 16:744-75 1. Bhat, M. A., J. C. Rios, Y. Lu, G. P. Garcia-Fresco, W. Ching, M. St Martin, J. Li, S. Einheber, M. Cheder, J. Rosenbluth, J. L. Salzer, and H. J. Bellen. 2001. Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/ Caspr/ Paranodin. Neuron 30:369-3 83. Bradley, W. G. 1990. Critical review of gangliosides and thyrotropin-releasing hormone in peripheral neuromuscular diseases. Muscle Nerve 13:833-842. Brown, W. F., and R. Snow. 1991. Patterns and severity of conduction abnormalities in Guillain-Barrt syndrome. J. Neurol. Neurosurg. Psychiatry 54:768-774. Burkel, W. E. 1967. The histological fine structure of perineurium. Anut. Rec. 158:177-189. Caporale, C. M., M. Capasso, M. Luciani, V. Prencipe, B. Creati, P. Gandolfi, M. V. de Angelis, A. Di Muzio, V. Caporale, and A. Uncini. 2006. Experimental axonopathy induced by immunization with Cumpylobacter jejuni lipopolysaccharide from a patient with Guillain-Barrt syndrome. J. Neuroimmunol. 174: 12-20. Chaudhry, V., A. M. Corse, D. R. Cornblath, R. W. Kuncl, D. B. Drachman, M. L. Freimer, R. G. Miller, and J. W. Griffin. 1993. Multifocal motor neuropathy: response to human immune globulin. Ann. Neurol. 33:237-242. Comin, R., N. Yuki, P. H. Lopez, and G. A. Nores. 2006. High affinity of anti-GM1 antibodies is associated with disease onset in experimental neuropathy. J. Neurosci. Res. 84: 1085-1090. Dalakas, M. C. 2002. Mechanisms of action of IVIg and therapeutic considerations in the treatment of acute and chronic demyelinating neuropathies. Neurology 59:S13-S21.
398
WKI
Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology. 1998. Guillain-Bard syndrome variants in Emilia-Romagna, Italy, 1992-3: incidence, clinical features, and prognosis. J. Neurol. Neurosurg. Psychihtry 65:218-224. Fuller, G. N., J. M. Jacobs, P. D. Lewis, and R. J. Lane. 1992. Pseudoaxonal Guillain-Bard syndrome: severe demyelination mimicking axonopathy: a case with pupillary involvement. J. Neurol. Neurosurg. Psychiatry 55:1079-1083. Gilbert, M., J. R. Brisson, M. F. Karwaski, J. Michniewicz, A. M. Cunningham, Y. Wu, N. M. Young, and W. W. Wakarchnk. 2000. Biosynthesis of ganglioside mimics in Campylobacter jejuni OH4384: identification of the glycosyltransferasegenes, enzymatic synthesis of model compounds, and characterization of nanomole amounts by 600-MHz ‘H and I3C NMR analysis. J. Biol. Chem. 275:3896-3906. Godschalk, P. C. R., A. P. Heikema, M. Gilbert, T. Komagamine, C. W. Ang, J. Glerum, D. Brochu, J. Li, N. Yuki, B. C. Jacobs, A. van Belkum, and H. P. Endtz. 2004. The crucial role of Campylobacter jejuni genes in anti-ganglioside antibody induction in Guillain-Barrt syndrome. J. Clin. Invest. 114:1659-1665. Goodfellow, J. A., T. Bowes, K. Sheikh, M. Odaka, S . K. Halstead, P. D. Humphreys, E. R. Wagner, N. Yuki, K. Furukawa, K. Furukawa, J. J. Plomp, and H. J. Willison. 2005. Overexpression of GDla ganglioside sensitizes motor nerve terminals to anti-GDla antibody-mediated injury in a model of acute motor axonal neuropathy. J. Neurosci. 25:1620-1628. Govoni, V., E. Granieri, M. R. Tola, E. Paolino, I. Casetta, E. Fainardi, and V. C. Monetti. 1997. Exogenous gangliosides and Guillain-Barrt syndrome: an observational study in the local health district of Ferrara, Italy. Brain 120:1123-1130. Griffin, J. W., C. Y. Li, T. W. Ho, P. Xue, C. Macko, C. Y. Gao, C. Yang, M. Tian, B. Mishu, and D. R. Cornblath. 1995. Guillain-BarrC syndrome in northern China: the spectrum of neuropathological changes in clinically defined cases. Brain 118: 577-5 95. Griffin, J. W., C. Y. Li, C. Macko, T. W. Ho, S . T. Hsieh, P. Xue, F. A. Wang, D. R. Cornblath, G. M. McKhann, and A. K. Asbury. 1996. Early nodal changes in the acute motor axonal neuropathy pattern of the Guillain-BarrC syndrome. J. Neurocytol. 25 :3 3-5 1. Hadden, R. D., D. R. Cornblath, R. A. C. Hughes, J. Zielasek, H. P. Hartung, K. V. Toyka, A. V. Swan, and Plasma Exchange/ Sandoglobulin Guillain-Barri Syndrome Trial Group. 1998. Electrophysiological classification of Guillain-BarrC syndrome: clinical associations and outcome. Ann. Neurol. 44:780-788. Hafer-Macko, C., S.-T. Hsieh, C. Y. Li, T. W. Ho, K. Sheikh, D. R. Cornblath, G. M. McKhann, A. K. Asbury, and J. W. Griffin. 1996. Acute motor axonal neuropathy: an antibodymediated attack on axolemma. Ann. Neurol. 40:635-644. Hahn, A. F., T. E. Feasby, A. Steele, D. S . Lovgren, and J. Berry. 1988. Demyelination and axonal degeneration in Lewis rat experimental allergic neuritis depend on the myelin dosage. Lab. Invest. 59:115-125. Halstead, S . K., P. D. Humphreys, J. A. Goodfellow, E. R. Wagner, R. A. Smith, and H. J. Willison. 2005. Complement inhibition abrogates nerve terminal injury in Miller Fisher syndrome. Ann. Neurol. 58:203-210. Halstead, S . K., G. M. O’Hanlon, P. D. Humphreys, D. B. Morrison, B. P. Morgan, A. J. Todd, J. J. Plomp, and H. J. Willison. 2004. Anti-disialoside antibodies kill perisynaptic Schwann cells and damage motor nerve terminals via membrane attack complex in a murine model of neuropathy. Brain 127:2109-2123. Hiraga, A., S . Knwabara, K. Ogawara, S . Misawa, T. Kanesaka, M. Koga, N. Yuki, T. Hattori, and M. Mori. 2005. Patterns and serial changes in electrodiagnostic abnormalities of axonal Guillain-Barrt syndrome. Neurology 64:856-860.
Hirota, N., R. Kaji, H. Bostock, K. Shindo, T. Kawasaki, K. Mizutani, N. Oka, N. Kohara, T. Saida, and J. Kimura. 1997. The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 120:2159-2169. Ho, T. W., H. J. Willison, I. Nachamkin, C. Y. Li, J. Veitch, H. Ung, G. R. Wang, R. C. Liu, D. R. Cornblath, A. K. Asbury, J. W. Griffin, and G. M. McKhann. 1999. Anti-GDla antibody is associated with axonal but not demyelinating forms of Guillain-Barre syndrome. Ann. Neurol. 45:168-173. Hughes, R. A. C., A. V. Swan, J. C. Raphael, D. Annane, R. van Koningsveld, and P. A. van Doorn. 2007. Immunotherapy for Guillain-Barrt syndrome: a systematic review. Brain 130:22452257. Illa, I., N. Ortiz, E. Gallard, C. Juarez, J. M. Grau, and M. C. Dalakas. 1995. Acute axonal Guillain-Barrt syndrome with IgG antibodies against motor axons following parenteral gangliosides. Ann. Neurol. 38:218-224. Jacobs, B. C., P. A. van Doorn, P. I. M. Schmitz, A. P. Tio-Gillen, P. Herbrink, L. H. Visser, H. Hooijkass, and F. G. A. van der MechC. 1996. Campylobacter jejuni infections and anti-GM1 antibodies in Guillain-BarrC syndrome. Ann. Neurol. 40:181-187. Koga, M., M. Gilbert, M. Takahashi, J. Li, S . Koike, K. Hirata, and N. Yuki. 2006. Comprehensive analysis of bacterial risk factors for the development of Guillain-Barrt syndrome after Campylobacter jejuni enteritis. J. Infect. Dis. 193547-555. Kusunoki, S., J. Shimizu, A. Chiba, Y. Ugawa, S . Hitoshi, and I. Kanazawa. 1996. Experimental sensory neuropathy induced by sensitization with ganglioside GDlb. Ann. Neurol. 39:424-43 1. Kuwabara, S., M. Mori, K. Ogawara, T. Hattori, S . Oda, M. Koga, and N. Yuki. 2001. Intravenous immunoglobulin therapy for Guillain-Barrt syndrome with IgG anti-GM1 antibody. Muscle Nerve 2454-58. Kuwabara, S., K. Ogawara, S . Misawa, M. Koga, M. Mori, A. Hiraga, T. Kanesaka, T. Hattori, and N. Yuki. 2004. Does Cumpylobucter jejuni infection elicit “demyelinating” GuillainBarrC syndrome? Neurology 63529-533. Kuwabara, S., K. Ogawara, K. Mizobuchi, M. Koga, M. Mori, T. Hattori, and N. Yuki. 2000. Isolated absence of F waves and proximal axonal dysfunction in Guillain-Barrt syndrome with antiganglioside antibodies. J. Neurol. Neurosurg. Psychiatry 68: 191-195. Li, C. Y., P. Xue, W. Q. Tian, R. C. Liu, and C. Yang. 1996. Experimental Campylobacter jejuni infection in the chicken: an animal model of axonal Guillain-Barrt syndrome. J. Neurol. Neurosurg. Psychiaty 61:279-284. McKhann, G. M., D. R. Cornblath, J. W. Griffin, T. W. Ho, C. Y. Li, Z. Jiang, H. S. Wu, G. Zhaori, Y. Liu, L. P. Jou, T. C. Lin, C. Y. Gao, J. Y. Mao, M. J. Blaser, B. Mishu, and A. K. Asbury. 1993. Acute motor axonal neuropathy: a frequent cause of acute flaccid paralysis in China. Ann. Neurol. 33:333-342. Mitzutamari, R. K., L. J. Kremer, E. A. Basile, and G. A. Nores. 1998. Anti-GM1 ganglioside IgM-antibodies present in human plasma: affinity and biological activity changes in a patient with neuropathy. J. Neurosci. Res. 5 1:237-242. Miyagi, F., H. Horiuchi, I. Nagata, S . Kitahara, M. Kiyoki, K. Komoriya, and N. Yuki. 1997. Fc portion of intravenous immunoglobulin suppresses the induction of experimental allergic neuritis. J. Neuroimmunol. 78:127-13 1. Nagai, Y., T. Momoi, M. Saito, E. Mitsuzawa, and S . Ohtani. 1976. Ganglioside syndrome, a new autoimmune neurologic disorder, experimentally induced with brain gangliosidesNeurosci. Lett. 2: 107-1 11. Nishimoto, Y., Koga, M., Kamijo, M., Hirata, K., and N. Yuki. 2004. Immunoglobulin improves a model of acute motor axonal
CHAPTER 22
neuropathy by preventing axonal degeneration. Neurology 62: 1939-1944. Olsson, Y. 1968. Topographical differences in the vascular permeability of the peripheral nervous system. Acta Neuropathol. (Berl.) 10:26-33. Rees, J. H., S. E. Soudain, N. A. Gregson, and R. A. C. Hughes. 1995. Campylobacter jejuni infection and Guillain-Barrt syndrome. N. Engl. J. Med. 333:1374-1379. Saida, K., T. Saida, M. J. Brown, and D. H. Silberberg. 1979a. In vivo demyelination induced by intraneural injection of antigalactocerebroside serum: a morphologic study. Am. J. Pathol. 95:99-116. Saida, K., T. Saida, M. J. Brown, D. H. Silberberg, and A. K. Asbury. 1978. Antiserum-mediated demyelination in vivo: a sequential study using intraneural injection of experimental allergic neuritis serum. Lab. Invest. 39:449-462. Saida, T., K. Saida, S. H. Dorfman, D. H. Silberberg, A. J. Sumner, M. C. Manning, R. P. Lisak, and M. J. Brown. 1979b. Experimental allergic neuritis induced by sensitization with galactocerebroside. Science 204:1103-1106. Sheikh, K. A., T. J. Deerinck, M. H. Ellisman, and J. W. Griffin. 1999. The distribution of ganglioside-like moieties in peripheral nerves. Brain 122449-460. Sheikh, K. A., and J. W. Griffin. 2001. Variants of the Guillain BarrC syndrome: progress toward fulfilling “Koch’s postulates.” Ann. Neurol. 49:694-696. Sheikh, K. A., G. Zhang, Y. Gong, R. L. Schnaar, and J. W. Griffin. 2004. An anti-ganglioside antibody-secreting hybridoma induces neuropathy in mice. Ann. Neurol. 56:228-239. Shu, X. M., F. C. Cai, and X. P. Zhang. 2006. Carbohydrate mimicry of Campylobacter jejuni lipooligosaccharide is critical for the induction of anti-GM1 antibody and neuropathy. Muscle Nerve 33:225-231. Susuki, K., Y. Nishimoto, M. Koga, T. Nagashima, I. Mori, K. Hirata, and N. Yuki. 2004. Various immunization protocols for an acute motor axonal neuropathy rabbit model compared. Neurosci. Lett. 368:63-67. Susuki, K., Y. Nishimoto, M. Yamada, M. Baba, S. Ueda, K. Hirata, and N. Yuki. 2003. Acute motor axonal neuropathy rabbit model: immune attack on nerve root axons. Ann. Neurol. 54: 3 83-3 88. Susuki, K., M. N. Rasband, K. Tohyama, K. Koibuchi, S. Okamoto, K. Funakoshi, K. Hirata, H. Baba, and N. Yuki. 2007. Anti-GM1 antibodies cause complement-mediated disruption of sodium channel clusters in peripheral motor nerve fibers. J. Neurosci. 27:3956-3967. Takahashi, M., M. Koga, K. Yokoyama, and N. Yuki. 2005. Epidemiology of Carnpylobacterjejuni isolated from patients with Guillain-BarrCand Fisher syndromes in Japan. J. Clin. Microbiol. 43:335-339. Takigawa, T., H. Yasuda, R. Kikkawa, Y. Shigeta, T. Saida, and H. Kitasato. 1995. Antibodies against GM1 ganglioside affect K+ and Na’ currents in isolated rat myelinated nerve fibers. Ann. Neurol. 37:436-442.
-
RABBIT MODEL OF GBS
399
Tamura, N., S. Kuwabara, S. Misawa, K. Kanai, M. Nakata, S. Sawai, M. Mori, and T. Hattori. 2007. Time course of axonal regeneration in acute motor axonal neuropathy. Muscle Nerve 35:793-795. Thomas, F. P., P. H. Adapon, G. P. Goldberg, N. Latov, and A. P. Hays. 1989. Localization of neural epitopes that bind to IgM monoclonal autoantibodies (M-proteins) from two patients with motor neuron disease. J. Neuroimmunol. 21:31-39. Thomas, F. P., W. Trojaborg, C. Nagy, M. Santoro, S. A. Sadiq, N. Latov, and A. P. Hays. 1991. Experimental autoimmune neuropathy with anti-GM1 antibodies and immunoglobulin deposits at the nodes of Ranvier. Acta Neuropathol. (Berl.) 82:378-383. Tuck, R. R., J. H. Antony, and J. G. McLeod. 1982. F-wave in experimental allergic neuritis. J. Neurol. Sci. 56:173-184. Willison, H. J., and N. Yuki. 2002. Peripheral neuropathies and anti-glycolipid antibodies. Brain 125:2591-2625. Wirguin, I., C. Briani, L. Suturkova-Milosevic, T. Fisher, P. DellaLatta, P. Chalif, and N. Latov. 1997. Induction of anti-GM1 ganglioside antibodies by Campylobacterjejuni lipopolysaccharides. J. Neuroimmunol. 78:138-142. Xiang, S. L., M. Zhong, F. C. Cai, B. Deng, and X. P. Zhang. 2006. The sialic acid residue is a crucial component of C. jejuni lipooligosaccharide ganglioside mimicry in the induction Guillain-Barri syndrome. J. Neuroimmunol. 174: 126-132. Yang, Y., S. Lacas-Gervais, D. K. Morest, M. Solimena, and M. N. Rasband. 2004. PIV spectrins are essential for membrane stability and the molecular organization of nodes of Ranvier. J. Neurosci. 24:7230-7240. Yu, Z., and V. A. Lennon. 1999. Mechanism of intravenous immune globulin therapy in antibody-mediated autoimmune diseases. N.Engl. J. Med. 340:227-228. Yuki, N. 2001. Infectious origins of, and molecular mimicry in, Guillain-Barri and Fisher syndromes. Lancet Infect. Dis. 1:2937. Yuki, N., K. Susuki, M. Koga, Y. Nishimoto, and M. Kamijo. 2002. Myelin-derived glycolipids and animal models of GuillainBarrC syndrome: reply. Ann. Neurol. 51532-533. Yuki, N., K. Susuki, M. Koga, Y. Nishimoto, M. Odaka, K. Hirata, K. Taguchi, T. Miyatake, K. Furukawa, T. Kobata, and M. Yamada. 2004. Carbohydrate mimicry between human ganglioside GM1 and Campylobacterjejuni lipooligosaccharide causes Guillain-Barrt syndrome. Proc. Natl. Acad. Sci. USA 101:1140411409. Yuki, N., T. Taki, F. Inagaki, T. Kasama, M. Takahashi, K. Saito, S. Handa, and T. Miyatake. 1993. A bacterium lipopolysaccharide that elicits Guillain-Barrt syndrome has a GM1 ganglioside-like structure. J. Exp. Med. 178:1771-1775. Yuki, N., M. Yamada, M. Koga, M. Odaka, K. Susuki, Y. Tagawa, S. Ueda, T. Kasama, A. Ohnishi, S. Hayashi, H. Takahashi, M. Kamijo, and K. Hirata. 2001. Animal model of axonal Guillain-Barrt syndrome induced by sensitization with GM1 ganglioside. Ann. Neurol. 49:712-720. Yuki, N., H. Yoshino, S. Sato, and T. Miyatake. 1990. Acute axonal polyneuropathy associated with anti-GM1 antibodies following Campylobacterenteritis. Neurology 40: 1900-1 902.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C . M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 23
Pathogenesis o f Campylobacter fetus MARTIN J. BLASER,DIANEG. NEWELL,STUARTA. THOMPSON, AND ELLENL. ZECHNER
Subspecies of C. fetus now have been designated on the basis of the typing of Berg et al. (1971). C. fetus subsp. fetus and C. fetus subsp. venerealis, although sharing many characteristics, have important biological differences and are responsible for different diseases in mammalian hosts.
Campylobacter fetus has been recognized as a significant pathogen of livestock for nearly a century. In 1909, McFaydean and Stockman cultured an unnamed vibrioid organism and identified it as the etiologic agent of abortion in sheep and cattle. In 1919, the organism was cultured by Theobald Smith from aborted bovine fetal fluids (Smith and Taylor, 1919) and was named Vibrio fetus because of its morphologic resemblance to V. cholerae. It was first recognized as a pathogen of humans in a case of infectious abortion with bacteremia in France in 1947 (Vinzent et al., 1947). As with many bacteria, the nomenclature of Vibrio fetus (Mohanty et al., 1962) changed substantially with increased study of the organism (summarized in Table 1).In 1963, Sebald and Ver6n noted that the genomic percentage of guanosine plus cytosine content of Vibrio fetus (32 to 35% G+C) was quite unlike that of other members of the genus Vibrio (47% G+C). With the additional distinction that Vibrio fetus did not ferment sugars, it was suggested that Vibrio fetus should define a novel genus, and the name Campylobacter was proposed (Sebald and VCron, 1963). Antigenic typing was added to the biotyping scheme of Bryner et al. (1962) by Berg et al. (1971), resulting in the classification of C. fetus that currently is in use (V6ron and Chatelain, 1973). Serogroup A-1 had heat-stable antigen A and the biochemical characteristics of biotype I. Similarly, groups A-sub-1 and A-2 had heat-stable antigen A, but were subtype I and type 11, respectively. Group B consisted only of biotype I1 organisms but with heat-stable antigen B. Group C organisms differed, with heat-stable antigen C, and also were distinguished from groups A and B by their growth at 42°C. C. fetus strains of serogroup C were eventually reclassified as C. jejuni and C. coli (Smibert, 1978).
C . FETUS INFECTIONS IN ANIMALS
Both C. fetus subsp. fetus and C. fetus subsp. venerealis can cause disease in cattle. It is often stated that their niches are different: C. fetus subsp. fetus is assumed to be confined to the intestinal tract, whereas C. fetus subsp. venerealis colonizes the urogenital tract. However, both subspecies can be recovered from the urogenital tract and products of abortion of cattle. Moreover, C. fetus subsp. venerealis has been recovered from the intestinal tract of a young and sickly calf (M. J. Toszeghy and D. G. Newell, unpublished observations), so the niches are apparently not exclusive. C. fetus subsp. venerealis is important because it is a statutory disease (http://www.oie.int/eng/ normes/mcode/anciens%2Ofichiers%20mcode%20 pour % 20 an Yo 2020041 en - chapitre - 2 . 3 . 2 . htm # rubrique-campylobacteriose-genitale-bovine), particularly for the purposes of the international trade of breeding bulls, semen, and embryos, and consequently it causes significant economic issues, especially for the artificial insemination industry. C. fetus subsp. venerealis infection of cattle (bovine venereal campylobacteriosis, BVC) is considered primarily a sexually transmitted disease. The reservoir for C. fetus subsp. venerealis is the penile prepuce of the bull, where the organism is maintained predominantly within epithelial crypts, although it also can be re-
Martin J. Blaser * Frederick H. King Professor of Internal Medicine, Department of Medicine, New York University School of Medicine, Veterinary Laboratories Agency, New Haw, Addlestone, Diane G. Newell 550 First Ave., OBV A-606, New York, NY 10016. Stuart A. Thompson Department of Biochemistry and Molecular Biology, Medical College Surrey KT15 3NB, United Kingdom. Ellen L. Zechner Institute of Molecular Biosciences, University of Graz, Humboltstrage 50, Aof Georgia, Augusta, GA 30912. 8010 Graz, Austria.
401
402
BLASER E T AL.
Table 1. Outcome of experimental ovine challenge with wild-type and mutant C. fetus strains Effect of challenge
Genotype Strain
Major S-layer protein expressed
sapA
recA
No. of animals
No. with positive placenta
23 D 23D:600(2) 23D:600(4) 23D:501 23D: 502
97 97 127 None None
sapA+ sapA’ sapA’ sapA sapA
WT recA recA WT recA
5 5 5 6 5
5 5 5 2 0
No. with positive feces
No. of abortions/ stillbirths 1 2 0“ 1 0
“Note that this is erroneously shown as 2 in Grogono-Thomas et al (2000)
covered from the glans penis and distal urethra (Smibert, 1978). Young bulls are relatively resistant to infection and may clear an established infection without intervention. Older bulls are highly susceptible to colonization, and once established, infection is lifelong. The increased incidence of C. fetus subsp. venerealis infection in older bulls correlates with an increase in the size and number of epithelial crypts in the prepuce, which may provide the organism with both the required microaerobic atmosphere for growth and also a sequestered site from the bull’s immune response. Because C. fetus subsp. venerealis is present in preputial fluids and semen, it is transmitted to cows during coitus. In cows and heifers, C. fetus subsp. venerealis infects the vagina, cervix, uterus, and oviducts (Smibert, 1978) in an ascending infection (Corbeil, 1999). Bacteremia is not recognized but may occur transiently. The infection is usually contained by the bovine immune response but often persists for several months. Temporary immunity to reinfection occurs after resolution of the first infection; this immunity is longer lasting in the uterus than in the vagina. Infection in cows results primarily in infertility due to failure of implantation as a result of endometritis. Over several months, the cow develops an immunoglobulin (1g)G response; this allows the cow to clear the infection from the uterus, but vaginal colonization persists. Thus, an infected cow can then become pregnant, even while being a vaginal carrier. Prolonged vaginal carriage is beneficial to C. fetus because it allows the organism to be transmitted to other animals in the herd, especially to new bulls. Cows develop secretory IgA in the vaginal mucus shortly after infection that is long lasting and diagnostic for previous exposure to C. fetus subsp. venerealis. Because BVC is typically a subclinical infection, it is not usually suspected until low calving rates are noted within a herd and losses have already occurred. Consequently, BVC has been called the “quiet profit-taker.”
The distribution of BVC is worldwide (Hum, 1996), although many countries are declared free of infection. With the introduction and widespread use of artificial insemination for bovine herds, especially in industrialized countries, the incidence of BVC has declined. However, there is little information on the carriage of C. fetus subsp. venerealis in national, naturally served cattle herds around the world, largely because investigation of cattle infertility, usually on the basis of number of repeated services, is costly and frequently inconclusive, and the tests for BVC are inefficient. Recent publications indicate that C. fetus subsp. venerealis infection in bulls remains a problem in South Africa (Pefanis et al., 1988), Australia (Hum, 1996), Nigeria (Bawa et al., 1991), Nepal (Smibert, 1978), Argentina (Eaglesome and Garcia, 1992), Jamaica (Garcia et al., 1980), and in parts of the United States (Eaglesome and Garcia, 1992). For example, in a study of beef herds in New South Wales, Australia (1989 to 1991), BVC was present in 35% of farms, and was suspected in another 11%. Twentyfive percent of herds were endemically infected, and an estimated 5% of herds acquired infections each year. The annual economic loss from BVC in New South Wales was estimated at $60 million-a significant source of loss for the industry (Hum, 1996). Nevertheless, there has been little interest in this area for many years. Interestingly, buffalo also are affected by BVC, and with the growth of buffalo meat consumption, a greater awareness of the problem may be generated. Vaccines for BVC are available in some countries, but the effectiveness of such vaccines has been questioned. The vaccination regimen is for cows or bulls, and it uses killed C. fetus subsp. venerealis cells in oil adjuvant. This generates a systemic and mucosal IgG response that can prevent infection by C. fetus subsp. venerealis and may also eradicate a previously existing infection (Corbeil, 1999). C. fetus also is recognized worldwide as a major cause of fetopathology, usually manifest as abortions, in bovines, ovines, and caprines. In the United King-
CHAPTER 23
dom from 1998 to 2005, official reports identified 232 cases of Campylobacter-associated bovine abortion (http ://www.defra.gov.uk/ corporatelvlal science/documents/science-vida04-intro.pdf). From 1997 to 2005, a total of 234 of these isolates from throughout the United Kingdom were speciated by the Veterinary Laboratories Agency. Over 60% of cases are due to C. fetus, of which 70% were due to C. fetus subsp. fetus and the remainder to C. fetus subsp. venerealis. Thus, C. fetus subsp. venerealis is not an uncommon cause of abortion. In addition, it is worth noting that the thermophilic campylobacters, C. jejuni and C. coli, also are common (causing about 20% of all cases). The remaining 10% of cases of bovine abortion were associated with C. hyointestinalis and C. sputorum. In the United Kingdom in 2004, campylobacters were the third most common cause (14%) of ovine abortion when assessed by a passive surveillance system (http://www.defra.gov.uk/corporate/vla/ science/documents/science-vida04-intro.pdf). Of the Campylobacter isolates speciated (n = 403) from 1997 to 2005, approximately 75% of cases were due to C. fetus subsp. fetus; the remainder were due to C. jejuni or C. coli. In sheep, in particular, Campylobacter-associated abortion is not usually sporadic but observed as acute outbreaks with multiple fetal losses within a flock (abortion blooms) and serious economic consequences to flock output. Such outbreaks tend to occur in 4- to 5-year cycles and are thought to reflect the development and waning of flock immunity. Because protective immunity is generated after to infection, vaccination is considered not to be economic except when used as a long-term strategy. C. fetus subsp. fetus is a common inhabitant of the ruminant gut (Atabay and Corry, 1998), and preliminary molecular typing studies that use amplified fragment-length polymorphism indicate overlapping bacteria populations from cattle and sheep (Duim et al., 2001). Thus, the route of infection causing Campylobacter-associated abortion in cattle or sheep is considered fecal-oral by contaminated food and water. Environmental contamination and spreading may also occur via wild animals and birds. However, infected abortion products contain high levels of bacteria, and in the confined conditions of breeding livestock, the aerosol route also may be important. After oral ingestion of C. fetus subsp. fetus, a transient bacteremia occurs. C. fetus subsp. fetus can be recovered from the liver, hepatic lymph nodes, gallbladders, and intestines of infected sheep; persistent biliary carriage may occur. In pregnant animals, the organism may localize to the placenta, resulting in placentitis and abortion within 20 to 25 days after inoculation
PATHOGENESIS OF C. FETUS
403
(Grogono-Thomas et al., 2000). The risk of abortion is dependent on the phase of pregnancy and is highest during the last trimester of ovine pregnancy. Bacteria can be recovered from the placenta, chorion, cotyledons, and allantoic fluid. Aborted fetuses show gross pathology of the liver, with necrotic lesions in 2 (20%) of the 10 fetal livers examined. The fetal abomasum and jejunum become red and inflamed. C. fetus subsp. fetus is recoverable from all aborted fetal tissues, including liver, jejunum, and abomasum. C. fetus strains may be isolated from reptiles but are genetically distinct (Tu et al., 2001a; Tu et al., 2005a).
C. FETUS INFECTIONS IN HUMANS C. fetus is an uncommon and opportunistic pathogen of humans, although its incidence is almost certainly underestimated and probably increasing (Blaser, 1998). Nearly all infections in humans are due to C. fetus subsp. fetus rather than C. fetus subsp. venerealis (Penner, 1988; Smibert, 1978). Several cases of C. fetus subsp. venerealis-mediated vaginosis in Sweden have been documented (Eaglesome and Garcia, 1992), paralleling bovine carriage, although this type of infection seems to be rare. The majority of patients with significant C. fetus subsp. fetus infection either are pregnant or have an underlying illness, including alcoholism, cirrhosis, neoplasm, diabetes, hematological malignancies, cardiovascular diseases, and infection with human immunodeficiency virus (Anstead et al., 2001; Rettig, 1979; Skirrow, 1994). Although diarrhea is sometimes reported as a symptom of C. fetus subsp. fetus infection, especially in previously healthy hosts, it is much less likely recognized as causing diarrhea than C. jejuni (Allos et al., 1995). Instead, most C.fetus subsp. fetus infections of humans are systemic or have a systemic component (Fig. 1). The mode of transmission of most cases of C. fetus subsp. fetus to humans is uncertain but is likely to be similar to that of livestock, by ingestion of contaminated food. Cases of diarrheal illness due to C. fetus subsp. fetus after ingestion of raw milk underscores this point (Klein et al., 1986; Taylor et al., 1979). C. fetus also can be transmitted via alternative therapies for malignancies (“nutritional therapy”), in which raw calf‘s liver is the source of the inoculum (Ginsberg et al., 1981). After colonization of the intestinal tract, a transient bacteremia can occur that may be symptomatic, but often is not. In compromised hosts, sustained bacteremia is one of the more common detectable manifestations of C. fetus disease (Bokkenheuser, 1970; Guerrant et al., 1978; Morrison et al., 1990; Ray et al., 2000; Verresen et al., 1985). Other conditions such as meningitis (Guerrant et al., 1978; Ray et al., 2000; U11-
404
BLASER ET AL.
Oral ingestion
1
Intestinal colonization ----------------* Diarrhea
1
Portal bacteremia Normal host i
Impaired host i
Transient or no Sustained bacteremia systemic bacteremia . A / \ Sepsis Secondary tissue +-------
seeding Figure 1. Model for C. fetus subsp. fetus disease of humans (Blaser, 1998). C. fetus subsp. fetus is ingested from contaminated food, followed by colonization of the intestinal tract. Bacteremia can occur but in normal hosts is limited by the immune system. In compromised hosts, the bacteremia may be prolonged due in part to bacterial virulence factors such as its surface layer, which allows secondary infection of additional anatomical sites. These may subsequently serve as a source of bacteria for sustained or renewed sepsis. From Blaser (1998).
mann et al., 1982), pericarditis (Morrison et al., 1990; Verresen et al., 1985), aortic aneurism (Marty et al., 1983), abortion, thrombophlebitis (Guerrant et al., 1978; Ponka et al., 1984; Ray et al., 2000), cellulitis (Ichiyama et al., 1998), pneumonia, pleuritis, septic arthritis, and peritonitis occur, but these are likely to be preceded by dissemination of C. fetus subsp. fetus through the bloodstream. The fact that most systemic C. fetus illness occurs in debilitated hosts underscores the importance of normal host defenses in resisting invasiveness by C. fetus beyond the gastrointestinal tract. The localization of C. fetus subsp. fetus in secondary anatomical sites can result in a sequestered focus for renewed sepsis, which can be chronic and difficult to treat (Neuzil et al., 1994); relapsing infections are increasingly being recognized (Tu et al., 2005b).
C. FETUS METABOLISM AND STRUCTURE Like other campylobacters, C. fetus are spiral, gram-negative rods, with a single polar flagellum at one or both ends of the cell. They lack alkaline phosphatase activity and do not hydrolyze DNA or sodium hippurate. The inability to hydrolyze hippurate distinguishes C. fetus from the majority of C. jejuni strains (HCbert et al., 1982). Most, but not all, strains of C. fetus are unable to grow at 42"C, and this may be dependent on the atmospheric conditions used.
They are microaerophilic, requiring less oxygen (5 to 6%) than that found in ambient air. C. fetus does not grow well anaerobically and may be missed by blood culture detection systems because it is neither an aerobe nor an anaerobe (Wang and Blaser, 1986). Oxygen toxicity may be lessened by superoxide dismutase, which was present as two bands after gel electrophoresis (Smibert, 1978). Chemically defined media are available that support C. fetus growth. Importantly, C. fetus neither utilizes nor ferments carbohydrates (Smibert, 1978). Iron (Fe) is an essential nutrient for bacteria, and they have therefore developed high-affinity uptake systems for its acquisition. Limited studies have been done on Fe uptake systems of C. fetus (Goossens et al., 1989). All 24 C. fetus strains grown under Felimiting conditions produced a 75-kDa Fe-repressed protein that cross-reacted with similar size proteins in C. jejuni, C. coli, and C. lari. Antibodies to this protein were found in convalescent sera, indicating its expression in vivo. Because in other bacteria similarly sized Fe-repressed proteins are often outer membrane receptors for extracellular Fe-sequestering proteins (siderophores), this protein may play such a role in a C. fetus Fe-uptake system. However, the 75kDa C. fetus protein did not cross-react with antibodies against the enterobactin and aerobactin receptors from Escherichia coli, and C. fetus chromosomal DNA did not hybridize with DNA probes derived from the E. coli systems. It therefore seems to represent an Fe-uptake system distinct from those in E. coli. C. fetus shows variable resistance to quinolone antibiotics. C. fetus is highly resistant to nalidixic acid, in contrast to C. jejuni. Resistance to nalidixic acid is typically determined by sequences in the DNA gyrase (GyrA) protein. Although the sequences of the C. fetus and C. jejuni GyrA proteins are known, there were no obvious amino acid differences that would explain the innate resistance of C. fetus to nalidixic acid (Taylor and Chau, 1997). Although crossresistance to ciprofloxacin is not typical for C. fetus subsp. fetus, it can occur due to a D91Y or T861 substitution in the GyrA protein (Kienesberger et al., 2007; Meier et al., 1998; Taylor and Chau, 1997). Tremblay and Gaudreau (1998) reported that 27% of C. fetus strains were resistant to tetracycline. Although the mechanism is unknown, it may be related to plasmid-borne tetracycline resistance that can be transmitted among campylobacters (Taylor et al., 1983). The genome size of C. fetus is relatively small but variable among the different strains of C. fetus examined (Salama et al., 1992). Salama and Taylor used pulsed-field gel electrophoresis to investigate differences in genome size and found the sizes of the
CHAPTER 23
C. fetus subsp. fetus, C. fetus subsp. venerealis, and C. fetus subsp. intermedius chromosomes to be 1.1, 1.3, and 1.5 Mb, respectively (Salama et al., 1992). These differences allow pulsed-field gel electrophoresis to be used in conjunction with biochemical and immunological tests in the typing of C. fetus subspecies. Another study reported the genome size of C. fetus subsp. fetus strain TK(+) to be considerably larger, at 2.0 Mb (Fujita and Amako, 1994). The reason for this apparent discrepancy is unknown. The process of horizontal gene transfer in bacteria is known to drive genetic diversity and evolution, and it certainly contributes to genome plasticity in C. fetus. The existence of different subsets of genes facilitating the different aspects of the lifestyles of C. fetus is quite probable, and similar explanations underlie a broad spectrum of distinguishing physiological and virulence properties in bacterial pathogens (Hacker et al., 2003; Maurelli, 2007; Ochman and Moran, 2001). Vehicles of this genome plasticity, the mobile genetic elements, include insertion sequences, transposons, integrons, bacteriophages, plasmids, genomic islands (e.g., pathogenicity islands), and combinations of these elements (Zechner and Bailey, 2004). Common mechanisms of bacterial horizontal gene transfer are transformation, conjugation, transduction, and transposition. Little is currently known about the molecular mechanisms promoting gene dissemination, its frequency, or the contribution to adaptive selection in C. fetus. Nonetheless, wholegenome comparisons combined with bioinformatic analyses are a promising approach to provide clear indications of evolutionary gene mobility. Comparative analysis typically reveals species, subspecies, or strain variation at the level of the mobile genetic elements present (Eppinger et al., 2004; Fouts et al., 2005; Raskin et al., 2006; Tettelin et al., 2005). The core genes, also referred to as housekeeping or backbone genes, are transferred at much lower rates and are generally viewed as highly conserved genes acquired by vertical descent. Consistent with the hypothesis that lifestyles of subspecies would parallel the existence of differential sets of genes, representational difference analysis was applied to reference strains of the C. fetus subspecies: C. fetus subsp. venerealis ATCC 19438 vs. C. fetus subsp. fetus ATCC 27374. This genomic subtractive hybridization revealed multiple genes unique or predominant for the different subspecies (G. Gorkiewicz and E. L. Zechner, unpublished data). As predicted, these included genes likely to determine differences in virulence attributes such as motility, lipopolysaccharide (LPS) production, or sensitivity to antibiotics. Importantly, the C. fetus subsp. venerealis chromosome was found to uniquely harbor a large genomic
PATHOGENESIS OF C. FETUS
405
island that was well conserved (86%) in the isolates tested (n = 64) but completely absent from C. fetus subsp. fetus isolates investigated thus far (n = 43). Several lines of evidence imply that the island was acquired laterally. It exhibits a high G+C content. The apparent site of integration is a tRNA gene adjacent to backbone genes, which are also found on the C. fetus subsp. fetus 82-40 chromosome (GenBank accession no. NC-008599). In addition, the island harbors a suite of putative mobility genes of phage, transposon, or plasmid origin, indicative of lateral dissemination via a variety of horizontal gene transfer processes. Most notably, this locus carries an operon expressing a conjugation-related secretion machinery (type IV). Thus, the system may be dedicated to translocation of protein and DNA macromolecules to targeted animal or bacterial cells, as is frequently found in gram-negative mammalian and plant pathogens (Backert and Meyer, 2006; Christie, 2004; Christie et al., 2005). Taken together, the findings argue that this large DNA fragment uniquely present on the chromosome of C. fetus subsp. venerealis represents a genomic island: either a pathogenicity island involved in C. fetus subsp. venerealis virulence, or a symbiosis island enabling C. fetus subsp. venerealis to preferentially colonize the bovine genital mucosa. The contribution of these and other subspecies-specific genes of C. fetus to functional diversity, adaptation, and hostpathogen interactions is a focus of current investigation (G. Gorkiewicz and E. L. Zechner, unpublished data). Gene inactivation and complementation is becoming easier in C. fetus, particularly regarding the more stringent requirements for genetic manipulation of C. fetus subsp. venerealis, as an expanding selection of appropriate molecular tools is developed (Kienesberger, 2007). Moreover, the mechanisms contributing to gene spread among campylobacters are becoming apparent, which in turn helps increase our understanding of the subspecies-specific adaptation and the molecular nature of C. fetus pathogenicity. Dunn et al. (1987) used two-dimensional gel electrophoresis to characterize cell envelope proteins of C. fetus. The most abundant proteins were seen at 45 to 47, 63, and 100 kDa. The 100-kDa protein was easily extractable with acidic glycine, although it was somewhat labile and tended to degrade under these conditions. It stained with periodic acid, indicating that it was a glycoprotein. The proteins that migrated at 45 to 47 kDa were the major porins. Although typically these were present in single isomeric forms, in some cases, minor isomeric forms of the 47kDa porin existed. The 63-kDa protein was the flagellar subunit. It consisted not of a single discrete entity but rather of a charge train. This suggests het-
406
BLASER ET AL.
erogeneity in the charge on flagellar subunits, which could be related to posttranscriptional modification analogous to the observations for C. coli (Alm et al., 1992).
C . FETUS GENOME The genome sequence of C. fetus subsp. fetus strain 82-40 has recently been determined. Analysis of this sequence has confirmed and extended previous observations and has also provided new insights into C. fetus metabolism, physiology, and pathogenesis. The C. fetus 82-40 genome is nearly 1.8 Mb in size, which is similar to other campylobacters (range, 1.6 to 2.1) but considerably larger than the previous estimate for C. fetus subsp. fetus (Salama et al., 1992). The difference in genome size may be due to strain differences or limitations of the previous technologies used. As stated above, C. fetus has fastidious growth requirements and does not ferment carbohydrates. Like C. jejuni (Kelly, 2 0 0 9 , C. fetus lacks the glycolytic enzyme 6-phospofructokinase, and this may explain the inability of C. fetus to use glucose (and other hexoses) as carbon sources. The pentose phospate pathway is present, but enzymes of the EntnerDoudoroff pathway (Eda and Edd) are not found. With the exception of an obvious succinate dehydrogenase (also absent in Helicobacter pylori; Kelly and Hughes, 2001), C. fetus has orthologs of all the tricarboxylic acid cycle enzymes. It is possible that the tricarboxylic acid cycle can run in a cyclic manner via the interconversion of fumarate and succinate by an alternate enzyme such as fumarate reductase. As with C. jejuni (Kelly, 200.5; Velayudhan et al., 2004), amino acids may be important carbon sources for C. fetus as the assimilatory enzymes are present, although the precise mechanisms may differ from those of C. jejuni. For example, although C. fetus has a serine transporter, it lacks the serine hydratase required for C. jejuni to grow in the presence of serine (Velayudhan et al., 2004). Orthologs of enzymes for catabolizing other critical amino acids such as aspartate, glutamate, and proline are present. C. jejuni has a highly branched electron transport chain and is capable of utilizing a variety of electron donors and terminal electron acceptors (Kelly, 200.5). The respiratory system of C. fetus 82-40 appears to be somewhat more restricted. Among the electron donors that C. fetus 82-40 is predicted to use are hydrogen, malate, and formate. NADH appears to be another important electron donor for C. fetus as a result of the presence of the complex 1 enzymes NuoE and NuoF, which are absent in both
C. jejuni and H. pylori (Kelly, 200.5). However, respiratory substrates utilized by C. jejuni but not predicted to be used by C. fetus as a result of the lack of the corresponding dehydrogenases include lactate, succinate, sulfite, and gluconate (Kelly, 200.5; Myers and Kelly, 200.5; Pajaniappan et al., in press). C. fetus has oxidative metabolism and uses oxygen as a terminal electron acceptor via a single cb-type cytochrome c oxidase; the alternative terminal cytochrome bd oxidase of C. jejuni (CydAB) is missing in C. fetus. Other putative alternative electron acceptors for C. fetus include nitrate, nitrite, fumarate, hydrogen peroxide, and trimethylamine-N-oxide/dimethylsulfoxide (Kelly, 200.5). C. fetus must be able to regulate gene expression in response to environmental conditions, but it is similar to C. jejuni in having a relatively small number of predicted regulatory proteins. C. fetus has only three sigma factors (a2’,d4, and a”); two of these are involved in flagellar regulation. C. fetus has a greater number of two-component regulatory systems than C. jejuni, with a total of 10 histidine kinase sensors and 1.5 response regulators (C. jejuni has 7 sensors and 12 response regulators; Parkhill et al., 2000). C. fetus 82-40 possesses an ortholog of both the Fur (Fe) regulator and the related PerR peroxide stress regulator found in C. jejuni (van Vliet et al., 1999), and it also encodes proteins for acquiring Fe, as described above. C. fetus appears to have uptake systems for obtaining iron as ferrous iron (feoAB system; Naikare et al., 2006), and possibly from the iron-containing compounds heme and enterochelin, but not from transferrin or lactoferrin. Both of the putative enterochelin and heme receptors are predicted to be 76 to 77 kDa, and either one could be the 7.5-kDa Fe-repressed protein reported previously (Goossens et al., 1989). Fe and Cu/Zn superoxide dismutase enzymes are encoded in the 82-40 genome and represent the two bands noted previously after gel electrophoresis (Smibert, 1978). C. fetus also appears to contain components of a system for natural transformation, including an ortholog of the C. jejuni competence-related protein CjOOll (Jeon and Zhang, 2007). This suggests that C. fetus could be naturally transformable, similar to the surface layer (S-layer)possessing Campylobacter rectus (Wang et al., 2000), although natural transformation ability of C. fetus has never been demonstrated experimentally (M. J. Blaser, unpublished observations; S. A. Thompson, unpublished observations). Regarding potential virulence mechanisms, C. fetus also shares many features with C. jejuni. C. fetus encodes homologues of the genes encoding cytolethal distending toxin. The 82-40 genome also predicts both N-linked and O-linked protein glycosylation
CHAPTER 23
systems. 0-linked glycosylation is associated with flagellar assembly and function in C. jejuni (Guerry, 2007). C. fetus is flagellated, and the presence of flagella is clearly required for the pathogenesis of Campylobacter infection (Guerry, 2007; Jagannathan and Penn, 2005). Although orthologs of many of the C. jejuni genes involved in the synthesis of 0-linked sugars are present, the C. fetus repertoire is somewhat divergent, suggesting that the sugars added to C. fetus flagella may be distinct from those that modify C. jejuni flagella. The majority of the C. jejuni N-linked general glycosylation pathway genes @gl genes; Karlyshev et al., 2005) are also found in C. fetus, although unlike in C. jejuni, the genes are arranged in multiple clusters. A surprising finding resulting from the first sequence of a C. jejuni genome was the presence of genes encoding a polysaccharide capsule (Karlyshev et al., 2005). Many of the genes of the C. jejuni capsule locus (e.g., kpsF, kpsT, kfiD, and cj1415-cj1420) also are present in C. fetus 82-40, although only weak homologies to other important C. jejuni capsular genes (e.g., kpsM) are found (Karlyshev et al., 2005). Whether these genes are involved in the expression of a novel, divergent C. fetus capsule or are involved in other aspects of C. fetus surface carbohydrate structures deserves further study. C. fetus also has orthologs of several C. jejuni proteins that have been implicated in adherence to host cells, such as the fibronectin-binding protein CadF (Konkel et al., 1997) and PEBla (Pei et al., 1998), but lacks other C. jejuni adhesins such as JlpA (Jin et al., 2001). Although C. fetus is believed to not express pili, the presence of genes encoding orthologs of the type IV pilus proteins PilT and PilD is intriguing. An ortholog of the C. jejuni invasion-related protein CiaB (Konkel et al., 1999) suggests that invasion of host cells may be an important part of the C. fetus lifestyle. Clearly, however, the major pathogenesisrelated difference of C. fetus compared with C. jejuni is the presence of the C. fetus S-layer, and this will be discussed in detail below.
C. FETUS LPS An important molecule possessed by all gramnegative bacteria is LPS, which is composed of two major portions. The interior portion of LPS is the lipid A moiety, which is widely conserved and is in effect a modified form of the outer leaflet of the outer membrane. In addition to its essential structural role in the outer membrane, lipid A is a bioreactive molecule that is responsible for the acute biological reactions leading to endotoxic shock (Moran, 1995); in pathogenic bacteria, LPS can be a major determinant
PATHOGENESIS OF C. FETUS
407
of disease outcome due to LPS-mediated interactions with the host. Although C. fetus LPS has features that are conserved for all bacterial LPS, it also has unique characteristics that contribute to the biology and virulence of C. fetus. LPS from C. fetus has extremely low biological activity, and as in other bacteria, this is due to lipid A. Moran et al. (1996) purified LPS from C. fetus subsp. venerealis (serotype A) and from C. fetus subsp. fetus (serotypes A and B), and tested their bioactivities including mitogenicity, pyrogenicity, and lethal toxicity. In all assays, C. fetus LPS preparations had significantly lower biological activities than LPS purified from Salmonella. This might have been due to the presence of longer fatty acid chains in C. fetus lipid A compared with that of Salmonella. Among C. fetus LPS preparations, LPS from a serotype A C. fetus subsp. fetus strain consistently had lower activity than those from other C. fetus serotypes. Although the exact composition of lipid A (e.g., phosphorylation) may influence biological activity, there appeared to be no correlation between the lower activity of C. fetus subsp. fetus serotype A LPS and interstrain differences in lipid A composition. The structure of the polysaccharide moiety also can have a modulating effect on LPS activity, and interstrain differences in this portion of LPS might be responsible for the varying degrees of C. fetus LPS activity. Because the major lifestyle of these organisms involves persistent colonization of hosts, LPS of low biological activity may have been selected. External to lipid A is a polysaccharide that can be subdivided into two regions, the core oligosaccharide and the 0-antigen. The core oligosaccharide often contains 2-keto-3-deoxyoctanateYwhich is present in C. fetus (Moran et al., 1994; PCrez-PCrez and Blaser, 1985). Cross-reactive antiserum against the LPS core region of C. fetus demonstrated the presence of conserved antigens among serum-sensitive C. fetus subsp. fetus strains, but these epitopes were not accessible in serum-resistant strains (PCrez-Ptrez et al., 1985). The core epitopes did not cross-react with those of C. jejuni LPS. More external on the bacterial cell surface is the 0-antigen. As with other bacteria, the 0-antigen harbors many of the epitopes to which the host immune system responds. Early studies defined three heatstable serotypes (A, B, and AB) by means of a slide agglutination test (Berg et al., 1971), and these were subsequently shown to be based on epitopes in the C. fetus 0-antigen (Perez-PCrez et al., 1986). Characterization of purified LPS from the different serotypes showed that C. fetus strains that were originally characterized as type AB had LPS profiles distinct from type A and indistinguishable from type B (PCrez-
408
BLASER ET AL.
Ptrez et al., 1986). Therefore, C. fetus is probably composed of two main serotypes (A and B), with the AB serotype being a minor variant of serotype B. In general, LPS recovered from serum-resistant C. fetus is smooth and relatively constant in size. The “ladder” effect representing variability in the number of saccharide repeat units in LPS from Enterobacteriaceae was different in LPS from C. fetus. Instead, C. fetus LPS contained few “rungs,” indicating relatively low diversity in the number of repeat units. These results indicate probable constraints on the length and variability of the 0-antigen side chains, which may be critical for its interactions with the S-layer (see below) (Moran et al., 1994; Perez-Perez and Blaser, 1985). In contrast, LPS isolated from serumsensitive C. fetus strains was rough, suggesting that longer 0-antigen side chains are associated with serum resistance (Perez-Perez and Blaser, 1985). In addition to the length and immunogenicity of the 0-antigen, there are obvious additional chemical differences among the LPS from different strains. Several different methods have been used to achieve satisfactory recovery of LPS from different strains; no single method works well for all strains, probably relating to differences in the LPS molecules themselves (Moran et al., 1994; Perez-PQez and Blaser, 1985). Furthermore, chemical differences among LPS may not define serotypes, representing heterogeneity in addition to that detected immunologically (Moran et al., 1994). In a compositional analysis, all C. fetus LPS was found to contain L-rhamnose, L-fucose, D-mannose, D-galactose, and L- and D-glycero-D-manno-heptose (Moran et al., 1994). Two strains (type B and AB) contained the unique sugar 3-O-methyl-~-rhamnose (L-acofriose, later reported to actually be D-acofriose; Senchenkova et al., 1996). All strains had amino acetylneuraminic (sialic) acid and D-galactosamine. It is unknown whether the presence of sialic acid on the LPS contributes to the serum resistance of C. fetus strains, as occurs with Neisseria gonorrhoeae. C. fetus LPS contained significant differences from that of C. jejuni. The lipid A of C. fetus contains no 2,3-diamino-2,3-dideoxy-~-glucose and therefore has a backbone different from C. jejuni LPS. In addition, the inner core of C. fetus LPS contains both L- and D-glyero-D-manno-heptose, whereas C. jejuni contains the L isomer only (Moran et al., 1994). Two main differences were found in the LPS constitutions of different serotypes of C. fetus. First, the molar ratio of rhamnose was much higher in LPS from serotype B and AB strains than from type A strains (Moran et al., 1994). Second, the unique sugar D-acofriose was found only in LPS from types B and AB and was absent from type A LPS, confirming ob-
servations about the high similarity of type B and AB strains (Perez-P&ez et al., 1986). The component of type AB LPS that cross-reacts with type A serum is currently unknown. The precise structures of the 0-antigens of type A and B C. fetus LPS were subsequently determined by lH- and 13C-nuclear magnetic resonance, and gas-liquid chromatography-mass spectrometry (Senchenkova et al., 1996, 1997). Confirming previous data, type A and type B 0-antigens were distinct. The type A 0-antigen was composed of a partially 0acetylated D-mannan chain of 10 to 12 residues of a monosaccharide unit of [-3)-a-~-Manp2Ac-(l+] (Senchenkova et al., 1997). In contrast, the type B 0antigen consisted of a D-rhamnan chain containing a disaccharide repeat of [--+3)-p-D-Rhap-(1+2)-a-DRhap-(1-1, terminated by a single residue of Dacofriose (Senchenkova et al., 1996). These chemical differences are an important basis of the differing immunological specificity of these molecules. Although two main serotypes of LPS could be identified by both biochemical and immunological techniques, there also was interstrain variability in LPS profiles within individual serotypes. Variation in mobility, staining, and banding patterns of LPS molecules in sodium dodecyl sulfate-polyacrylamide gel electrophoresis demonstrated that structural differences existed (Moran et al., 1994). LPS differences also can be exemplified by the varying partition of LPS molecules into aqueous and organic phases during hot phenol extraction of LPS from C. fetus, and can be observed both among strains and within a single strain. LPS molecules that partitioned into the aqueous phase tended to be slightly larger (Mr = 8,850 to 9,824) than those that were extracted in the phenol phase (M, = 7,270 to 8,550). In addition to the size differences, the variable extraction also reflects the nonstoichiometric presence of sugars among LPS molecules, varying levels of substitution of the sugars, and differences in hydrophobicity. Interstrain variation in the level of phosphorylation also was substantial (Moran et al., 1994). In summary, C. fetus has LPS that in many ways is typical of gram-negative bacterial LPS, containing lipid A, 2-keto-3-deoxyoctanateYand a polysaccharide 0-antigen. The lipid A portion of C. fetus LPS has low biological activity compared with members of the Enterobacteriaceae. In contrast to the rough LPS isolated from serum-sensitive C. fetus, LPS from serum-resistant strains is long and relatively constant in size, suggesting selection for the maintenance of smooth LPS. The 0-antigens of type A and type B strains are distinct and are one basis for the C. fetus serotyping scheme.
CHAPTER 23
C . FETUS SLAYER For bacterial pathogens, the interaction of the cell surface with the host often is critical for virulence. Such structures either can promote virulence or serve as targets for the host immune response in reducing virulence. For example, outer membrane proteins can mediate attachment of bacteria to host cells, mediate transfer of toxins into the extracellular space or directly into host cells, or induce blocking antibodies that prevent an effective immune response against important bacterial surface structures. The cell surface usually contains the dominant antigens to which an effective immune response is generated and can be the site of complement deposition that leads to cell lysis or phagocytosis. Similarly, the C. fetus cell surface also is critical for its virulence. However, with C. fetus, it is not the components of the outer membrane, but rather an external structure that is the prime virulence factor of this organism. This structure is the proteinaceous S-layer. Early observations on the virulence of C. fetus were as follows: (i) C. fetus was resistant to phagocytosis by mononuclear cells (Corbeil et al., 1975a, 1975b), and (ii) persistence of C. fetus in cattle was accompanied by variation of surface antigens (Schurig et al., 1973). In 1971, Myers isolated a 135-kDa protein from C. fetus broth culture supernatants (Neuzil et al., 1994). In seminal work toward the understanding of C. fetus virulence, McCoy et al. (1975) purified and characterized this superficial antigen that was responsible for both of these traits. This antigen, which was termed antigen [a], could easily be removed from the cell surface of wild-type C. fetus strain 23D by treatment with 0.2 M glycine, pH 2.2. Antigen [a] was proteinaceous and was reported to be glycosylated. When present on the cell surface, antigen [a] blocked agglutination with O-antiserum. In contrast, O-antiserum agglutinated either 23D cells from which antigen [a] had been removed, or cells of a spontaneous mutant (23B) that had lost the production of antigen [a]. Electron microscopy that used negative staining of C. fetus cells showed a distinct outer structural layer in the antigen [a]-producing 23D that was absent in the mutant 23B. This structure was designated the C. fetus “microcapsule.” The strong binding of cationized ferritin to this structure indicated that it was acidic. Importantly, in the absence of immune serum the presence of the antigen [a] “microcapsuleyyinhibited the uptake of 23D cells by macrophages. Lacking the microcapsule, phagocytosis of 23B cells was efficient. If the 23D cells were preopsonized with [a] antiserum, phagocytosis then proceeded efficiently. Therefore, McCoy et al. (1975) showed that C. fetus has an acidic outer layer (mi-
PATHOGENESIS OF C. FETUS
409
crocapsule) that covers LPS epitopes and resists phagocytosis. Antigen [a], purified from culture supernatants of strain 23D and subjected to biochemical analysis, was found to be 98 kDa in size (Winter et al., 1978), rich in glycine, alanine, and aspartic acid, and poor in cysteine, histidine, and tryptophan (Winter et al., 1978). When membrane vesicles were prepared, antigen [a] was complexed with LPS, although this association could be disrupted with EDTA. Incubation of free antigen [a] with strain 23B in the presence of 10 mM CaC1, allowed the reassembly of a microcapsule around those cells. Because of the chemical and structural similarities of the microcapsule with the Slayer of Aquaspirillum serpens (Buckmire and Murray, 1973), it was suggested that the C. fetus microcapsule formed an analogous structure. In addition to phagocytosis resistance, another explanation for the disproportionate number of systemic infections in humans caused by C. fetus subsp. fetus relative to other campylobacters (Blaser et al., 1985; Riley and Finch, 1985) was that C. fetus was more resistant to the bactericidal effects of human serum and were therefore more able to disseminate via the bloodstream. A comparison of the serum susceptibilities of C. fetus, C. coli, and C. jejuni strains subsequently showed this to be true. In a serum bactericidal assay, all C. coli and C. jejuni strains were susceptible to killing, whereas all C. fetus strains were completely resistant despite high serum concentrations, prolonged incubations, or the use of immune serum (Blaser et al., 1985, 1987). The killing of C. coli and C. jejuni was mediated by both complement and antibody; neither of these components promoted killing of C. fetus. What was the molecular basis for the ability of C. fetus to resist killing by serum? A collection of 38 C. fetus subsp. fetus strains isolated either from feces or from systemic sites was used to investigate whether certain proteins were associated with serum resistance (Blaser et al., 1987). All of the serum-resistant strains expressed cross-reactive surface proteins of approximately 100 or 125 kDa, and these were absent in serum-sensitive strains. Furthermore, spontaneous laboratory-passaged mutants of two serumresistant strains lost the production of their 100-kDa protein and had become serum sensitive. Therefore, C. fetus subsp. fetus serum resistance was associated with the expression of the 100- or 125-kDa surface proteins, presumably components of the S-layer (Blaser et al., 1987). To directly study the function of the S-layer in serum resistance, encapsulated (S+) serum-resistant C. fetus subsp. fetus and their nonencapsulated (S-) (serum sensitive) variants were examined for their
410
BLASER ET AL.
abilities to bind and be killed by components of the complement system (Blaser et al., 1988). Although a copious amount of complement factor C3 bound to S- C. fetus, little C3 was bound by the S+ bacteria (Fig. 2),and these small amounts were in a degraded (nonfunctional) form. The difference in C3 binding led to greater consumption of the components of the complement membrane attack complex (C5 and C9) by S- but not S+ cells. The low abundance of C5 and C9 on the cell surfaces of encapsulated bacteria was shown to be due to the direct inhibition of the Slayer on the binding of complement factor C3b, thereby eliminating formation of the C5 convertase and all downstream events that normally would lead to cell lysis. In other experiments, S+ C. fetus cells were found not to activate the alternative pathway of complement, explaining the low amount of C3b deposition (Washburn et al., 1991). The low binding of C3b by the S+ C. fetus strains rendered them resistant to phagocytic killing by polymorphonuclear leukocytes, which exhibited minimal uptake of these bacteria and generated a minimal chemiluminescent response (Blaser and Pei, 1993). However, if S+ bacteria were preopsonized with immune rabbit serum (either containing anticapsule antibodies or not), they were efficiently killed by polymorphonuclear leukocytes. Therefore, the presence of the C. fetus S-layer makes these bacteria resistant to the complement system, either by complement-mediated cell lysis or by phagocytic killing stimulated by cell-bound C3b. Antibody-mediated opsonophagocytosis remains effective in killing S+ bacteria.
These in vitro results were extended by using a mouse model of C. fetus infection (Pei and Blaser, 1990). Outbred HA/ICR mice challenged with S+ or S- C. fetus strains substantially differed in their rates of bacteremia and death. The S- strain was a spontaneous mutant derived from extensive laboratory passage of the S+ strain. Mice challenged intraperitoneally with the S+ C. fetus strain showed a much higher mortality than those challenged with the Sstrain. After oral inoculations, no mortality was observed for either strain; however, the S+ strain caused an immediate and high-grade bacteremia that persisted for 5 days. Bacteremia was not detected in mice administered the S- strain. Protection against bacteremia by the S+ strain was observed when mice were passively immunized with rabbit antiserum against the purified 100-kDa S-layer protein expressed by this C. fetus strain. The S-layer protein itself was not toxic to mice but could be reattached to S- C. fetus cells, regenerating increased virulence in mice. These results were not due to artifacts arising from the laboratory passage history of the S- strain because similar results were obtained with bacteria that had been made S- by removal of the S-layer by means of pronase treatment of s+cells (Blaser and Pei, 1993). Together, these studies provided direct in vivo evidence in an animal model of the conclusions drawn from in vitro studies; the S-layer is critical for inhibition of complement binding, serum and phagocytosis resistance, and for virulence. Subsequently, two groups purified and characterized the proteins that comprised the S-layer (Dubreuil et al., 1988; Pei et al., 1988). The 131-kDa S-
3
LL U \
u m
230
0
10
20
30 TIME
40
50
60
hinl
Figure 2. Inhibition of complement factor C3 binding by C. fetus. 'zSI-labeled C3 was incubated with either C. fetus 23D (S+) or 23B (S-), and the amount of bound C3 determined. The Slayer in strain 23D prevents significant C3 binding. From Blaser et al. (1988).
CHAPTER 23
layer protein (SLP) of type B strain VC119 and the 100, 127, and 149-kDa SLPs from type A strain 23D were removed from the cell surface by treatment with acidic glycine, proteinase K, or by extracting the cells in water. These proteins were similar in composition, were surface exposed, were the predominant antigens recognized by antiserum generated against whole C. fetus cells (Dubreuil et al., 1988; Pei et al., 1988), and appeared identical to antigen [a] (Winter et al., 1978). The proteins were relatively hydrophobic (37 to 39%), and like the microcapsule of McCoy, they contained large amounts of aspartic acid, threonine, glycine, leucine, valine, and alanine. Histidine and cysteine were undetected. They were heat stable and relatively acidic (isoelectric points of 6.35 for the VC119 SLP, and 4.1 to 4.3 for the 23D SLPs), and refractory to digestion by trypsin or chymotrypsin. The amino acid sequences of the amino termini of the SLPs derived from type A strains were identical to each other, but different than that of the type Bderived SLP. To determine whether the S-layer proteins were glycoproteins, the purified proteins were subjected to a variety of techniques, including Schiff staining, periodate oxidation, and lectin affinity chromatography and trifluoromethanesulfonic acid treatment. Unlike previous results (McCoy et al., 1975), glycosylation of the S-layer protein could not be detected with any of these techniques (Dubreuil et al., 1988; Pei et al., 1988). Most likely, earlier reports of glycosylation of the S-layer protein were due to contamination with minor amounts of LPS with which the S-layer proteins tend to associate. Electron microscopy provided further insight about the C. fetus S-layer and showed that variant SLPs produced by C. fetus cells give rise to S-layers with different structures (Dubreuil et al., 1988; Fujimoto et al., 1989, 1991). In strain VC119, the Slayer was arranged as a linear (or possibly tetragonal) array of 131-kDa SLP subunits with 8.75 nm centerto-center spacing that completely covered the cell surface (Dubreuil et al., 1988). In strain TK, both tetragonal and hexagonal structures were present (Fujimoto et al., 1991) (Fig. 3). Cells that expressed 98-kDa SLPs had hexagonally arrayed S-layers, with 24-nm center-to-center spacing. In contrast, those that expressed 127- or 149-kDa SLPs possessed tetragonal S-layers with 8-nm center-to-center spacing. On the basis of these results and the relative paucity of surface layers with linear symmetry, it is likely that the observation of a linear array (8.75-nm spacing) generated by the 131-kDa SLP in strain VC119 was due to an artifact of the electron microscopy shadowing technique (as suggested by Dubreuil et al.) and that the structure is in fact tetragonal (Dubreuil et al., 1988). The 98-, 127-, and 149-kDa SLPs, although
-
PATHOGENESIS OF C. FETUS
411
having unique antigenic determinants, also have immunologic epitopes in common that are detectable both by polyclonal serum (Blaser et al., 1987; Garcia et al., 1995; Pei et al., 1988; Wang et al., 1993) and monoclonal antibodies (Wang et al., 1993). The reattachment of SLPs to the C. fetus cell surface was investigated in vitro by using SLPs that had been removed from S + bacteria by extraction with water Fang et al., 1992). For cells that had been made S- by mutation or by removal of the S-layer by water extraction, SLPs could be reattached to their surfaces in a manner that was dependent on the presence of divalent cations. Neither monovalent nor trivalent cations supported reattachment. Binding of SLPs to the cell was serotype specific: SLPs purified from type A strains could be reattached only to type A cells (not type B) and vice versa. A recombinant 98-kDa SLP and truncated amino-terminal 65- and 50-kDa fragments thereof also were able to bind to S- cells. These data indicated that LPS-binding ability was contained in the amino-terminal half of the protein, and that C. fetus posttranslational modification of the SLP was not required for LPS binding. Quantitative enzyme-linked immunosorbent assay showed the copy number of SLPs to be at least lo5 per C. fetus cell. Recombinant sup-homolog encoded proteins all could bind specifically to SLP-negative cells (Tu et al., 2003b), indicating that no C. fetus specific posttranslational modification is necessary for this phenotype. Lectins were used to detect the presence of the S-layer and to differentiate between different serotypes of LPS (Fogg et al., 1990). Lectins from Bundeirueu simplicifoliu I1 (BS-11), Helix pomutiu (HPA), and wheat germ agglutinin (WGA) specifically bound to the 0-antigen side chains of type A (C. fetus) LPS, but not to LPS from type B (C. fetus) or type C (C. jejuni). However, binding to type A C. fetus did not occur in the presence of the S-layer, suggesting that the entirety of the LPS molecule is shielded by the Slayer.
CHARACTERIZATION OF sup GENES
A gene encoding a subunit of the S-layer was first cloned in 1988 and sequenced from serotype A strain 23D (Blaser and Gotschlich, 1990). This gene was designated supA, for surface array protein, type A. The supA gene was 2,802 nt in length and predictFd an acidic (PI = 4.55) 933 amino acid protein of 97 kDa. The amino terminus of the protein predicted from the DNA sequence exactly matched the aminoterminal sequence of the purified protein, and the initiating methionine was not processed. This, together
412
BLASER ET AL.
Figure 3. Electron microscopy of the C. fetus surface layer (S-layer). (A) Shown in ultrathin cross section (Dubreuil et al., 1988), the Slayer appears as a ringlike structure external to the outer membrane (arrow). (B) In freeze-etch preparations of the cell surface (Fujimoto et al., 1991), the Slayer appears as either regular tetragonal (left) or hexagonal (right) arrays. This micrograph demonstrates the ability of a single cell to express more than one type of S-layer. From (A)]ournalof Bacteriology and (B) Infection and Immunity.
with visual inspection of the DNA sequence, showed that despite its ultimate extracellular localization, SapA lacked an amino-terminal signal sequence that would direct its secretion to the cell surface. Before this, the only surface-layer protein (SLP) that lacked a signal sequence was that of Caulobacter crescentus (Fisher et al., 1988). All other SLPs had aminoterminal signal sequences, suggesting that their secretion was mediated by the general (sec-dependent) secretory pathway (Boot and Pouwels, 1996; Pugsley, 1993). Another notable characteristic of the SapA protein was its single cysteine residue, suggesting that intramolecular disulfide bonds were not required for maintaining the secondary structure of this protein. Because the cloned DNA terminated just upstream of the sapA open reading frame (ORF), no information could be predicted about promoter sequences upstream of sapA (Blaser and Gotschlich, 1990). However, a putative transcriptional terminator was located just downstream of sapA, suggesting that sapA was not cotranscribed with downstream genes.
Subsequently, an additional two SLP-encoding genes from type A strains (Dworkin et al., 1995a; Tummuru and Blaser, 1993) and two genes from type B strains (Casadkmont et al., 1998; Dworkin et al., 1995b) were sequenced. Not surprisingly, these genes had conserved and divergent characteristics, consistent with both the common functions and the antigenic differences of the encoded proteins (Fig. 4). Comparison of the three sapA alleles (sapA, sapAl, and sapA2) revealed highly conserved DNA sequences at the 5' end of each gene. The conservation began 74 bp before the ATG translational initiation codon, and continued 552 to 654 bp into the ORF (Blaser and Gotschlich, 1990; Dworkin et al., 1995a; Tu et al., 2001b; Tummuru and Blaser, 1993). Contained within the 74-bp conserved sapA upstream sequences were several features of note. First, each of the genes contains a sequence that has seven of eight bp identity with the x site (Dworkin et al., 1995a; Tummuru and Blaser, 1993), which in E. coli serves as a signal for RecBCD enzyme to initiate homolo-
CHAPTER 23
PATHOGENESIS OF C. FETUS
413
sapA1
sapA sapA4
sapm
sap* sapAS
sap43
1
II
sap.4 7
sapAp8
I!&
2kb
3kb
4kb
Figure 4. Structural features and comparison of DNA sequences of eight complete and one partial sapA homolog. The structure of each homolog is represented schematically by the aligned rectangles. Colored boxes identify shared regions of identity among the homologs. White boxes indicate nonconserved sequences, with no sequences >30 bp shared. The first 5 5 3 bp in all eight complete sapA homologs are shared. The Cf0007 ORF shared 752 bp with sapA7, but lacks the 5’ conserved region; it is referred to as sapAp8, as it is a partial homolog. From Molecular Microbiology.
gous recombination (Kowalczykowski et al., 1994). Immediately after the putative x site was a pentameric sequence (ATTTT) that is repeated three times consecutively (Dworkin et al., 1995a; Tummuru and Blaser, 1993). Finally, a 7-bp inverted repeat overlapping the ribosome binding site and terminating immediately upstream of the ATG initiation codon was found in all three genes (Blaser and Gotschlich, 1990; Dworkin et al., 1995a; Tummuru and Blaser, 1993). In contrast to the high conservation at the 5’ ends of the sapA genes, the 3’ ends of the ORFs were divergent. Consequently, downstream of the 5 ’ conserved regions, the sapA gene had only limited nucleotide identity with the sapAl and sapA2 genes (Blaser and Gotschlich, 1990; Dworkin et al., 1995a; Tummuru and Blaser, 1993), and this was reflected in significant differences in the encoded proteins (see below). Immediately downstream of each of sapA, sapAl , and sapA2 were additional conserved sequences of approximately 50 bp. In each case, these conserved sequences also contained putative transcriptional terminators. A repeated 10-nt sequence (TTTTAAATTT) was present numerous times downstream of sapA2 (Dworkin et al., 1995a) and to a
lesser extent downstream of sapA and sapAl (Blaser and Gotschlich, 1990; Tummuru and Blaser, 1993). The sapB and sapB2 genes were organized in a manner parallel to their sapA counterparts (CasadCmont et al., 1998; Dworkin et al., 1995b). Conserved sequences were found at the 5’ ends of the sapB and sapB2 genes (CasadCmont et al., 1998), although these conserved sequences were different from those at the 5’ ends of sapA homologs (Dworkin et al., 1995b). Divergence at the 3’ ends of the sapB and sapB2 genes was noted, similar to that seen at the 3’ ends of the sapA homologs. Similar to the conserved regions upstream of sapA homologs, pentameric ATTTT repeats were found upstream of the sapB and saps2 genes (CasadCmont et al., 1998; Dworkin et al., 1995b). However, the putative x sites and inverted repeats found upstream of sapA genes were absent from the corresponding locations of sapB genes (CasadCmont et al., 1998; Dworkin et al., 1995b). Several studies have addressed the genetic organization of sap genes (Dworkin et al., 1995a; Fujita and Amako, 1994; Salama et al., 1995; Tummuru and Blaser, 1993). In two separate genome maps (Fu-
Next Page 414
BLASER ET AL.
jita and Amako, 1994; Salama et al., 1995), genes encoding SLPs were found to be clustered. Hybridization studies performed during the cloning of sapA homologs also revealed the presence of clustering of multiple sapA genes (Dworkin et al., 1995a; Tummuru and Blaser, 1993). Finally, Southern hybridization was used to conduct a survey of 27 C. fetus strains for the genetic organization of SLP-encoding genes (Fujita et al., 1995). By use of a probe derived from the conserved region of sapA, all 23 of the type A strains were found to contain sapA-homologous sequences. Predictably, as a result of the divergence of the sapB DNA sequence in the region of this probe, none of the non-type A strains hybridized. In each type A strain, there were at least six copies of sapA genes. Furthermore, hybridization experiments with pulsed-field gel electrophoresis showed that in every strain the genes were located within a small region ( 5 9 0 kb) of the chromosome, suggesting that clustering of these genes is highly conserved. It now is clear that the homologs are located on a conserved sap island (see below) (Tu et al., 2003a, 2003b). The characteristics of the proteins predicted from the sapA and sapB genes characterized thus far are summarized in Table 2. These proteins range in size from 95 to 113 kDa and have isoelectric points of 4.2 to 4.6. Although they did not possess aminoterminal signal sequences, the SLPs isolated from a given serotype share nearly identical amino termini. For type A proteins, the amino-terminal 184 amino acids were more than 97% identical to each other; a similar relationship was found among the amino termini of type B proteins. However, the amino termini of type A proteins were completely different from those of type B proteins. The remainders of the type A and type B C. fetus SLPs diverged significantly within each family (Table 2). Regions of similarity and divergence in type A and type B SLPs are represented graphically in Fig. 4. At first glance, the extreme carboxy-terminal segments of each of these
proteins appeared to have little sequence conservation. However, later experimentation revealed that not only must significant carboxy-terminal sequence and/ or structural conservation exist, but that such conserved features are important for the function of these proteins (see below). With the exception of the 184 amino acids at the amino termini of SapA and SapB, these proteins were nearly identical (Fig. 4) and can therefore be considered equivalent proteins within their respective serotypes. SapA2 and SapB2 shared a similar relationship, although a region of divergence at the carboxy terminus of these proteins suggests that although closely related to SapA2, SapB2 might actually be allelic with another yet uncharacterized SapA homolog. The high similarity of the amino termini of SLPs within a serotype but divergence between serotypes suggested a role of the amino terminus in binding to homologous LPS. To test this hypothesis, deletions were made from the 3’ end of the sapA2 gene yielding progressively shorter amino-terminal fragments of the SapA2 protein (Dworkin et al., 1995a). These recombinant proteins then were tested for their abilities to attach to C. fetus cells. As expected, none of the type A proteins was able to attach to type B LPS. In contrast, SapA2 derivatives containing large carboxy terminal deletions retained the ability to bind to type A cells. These results localized the LPS binding domain to the first 189 amino acids of the type A protein. The poor immunogenicity of SapA2 peptides shorter than 189 amino acids prevented their testing for proficiency in LPS binding. Similarly, the analogous region of SapB was responsible for serospecific binding to homologous type B LPS (Dworkin et al., 1995b). Further work has shown the extension of this model of conserved and diverse feature to extend into reptile C. fetus strains as well (Fig. 5 ) . In summary, the genes encoding C. fetus SLPs are monocistronic and contain conserved DNA se-
Table 2. Biochemical properties of C. fetus S-layer proteins % Amino acid similarity to SapA or SapB
Protein
Type A SapA SapAl SapA2 Type B SapB SapB2
No. of amino acids
Mass (kDa)
Similarity to SapA at positions:
Similarity to SapB at positions:
1-184
185-end
1-184
No. of cysteine residues
Reference
185-end
933 920 1,109
97 95 112
4.55 4.36 4.22
100 98.9 99.4
100 44.2 28.1
35.8 35.8 35.8
89.3 48.0 28.3
1 1 0
Blaser and Gotschlich, 1990 Tummuru and Blaser, 1993 Dworkin et al., 1995a
936 1,112
96 113
4.40 4.24
35.8 35.8
89.3 26.4
100 98.9
100 23.6
0 0
Dworkin et al., 1995b Casadtmont et al., 1998
Previous Page CHAPTFR 23
PATHOGFNFSTS OF
c
Fm-ris
41 F I
Figure 5. Schematic representation of genes encoding type A and type B SLPs. The conserved 5 ' regions of each gene are indicated by black (type A) or green (type B) rectangles. Colors show areas of conservation between homologs; white boxes represent homolog-specific sequences. From 'l'u et al. (2004a).
quences at both the 5' and 3' ends of the genes. The presence of conserved DNA outside of the ORFs predicts that this DNA homology has functions other than simply encoding identical proteins. Among the encoded SLPs were important conserved features, as well as features that impart unique type- and proteinspecific characteristics to these molecules. Highly conserved regions at the amino termini of C. fetus SLPs are responsible for binding to LPS. The remainder of each protein has divergent amino acid sequences, which endows each protein with individual immunologic and structural characteristics.
ANTIGENIC VARIATION OF C. FETUS SLPs Early observations of experimentally infected cattle indicated that C. fetus exhibited antigenic variation of its surface antigens during infection (Corbeil et al., 1975b; Schurig et al., 1973). This was subsequently shown to be due to variations in the expression of the S-layer protein (SLP) (Dubreuil et al., 1990; Garcia et al., 1995; Tummuru and Blaser, 1993; Wang et al., 1993). Antigenic variation of surface structures is a common theme in bacterial path-
ogenesis, although the mechanisms involved differ among bacteria. Observations made during the cloning of the sapA genes were important toward understanding the mechanisms by which the expression and antigenic variation of the encoded proteins are controlled. First, all of the genes that were cloned were complete and could express a full-length product in E. coli. This indicated that antigenic variation occurred by a mechanism different from that of Neisseria gonorrhoeae pilin, which uses partial gene copies recombined into an expression locus. Second, most of the cloned genes were not situated downstream of recognizable promoters, suggesting that each was not expressed from its own promoter and required instead that each gene be brought into the context of a promoter for its expression. To understand how SLPencoding genes were expressed, the region upstream of an expressed supA gene was characterized (Tu et al., 2001a). Primer extension by means of total cellular RNA and a sapA-specific primer identified a singlc transcriptional initiation site 114 bp upstream of the sapA initiation codon, and downstream of a sequence that resembled an E. coli a 7 O promoter. This promoter was present in a single copy on the C. fetus
416
BLASER ET AL.
23D chromosome (Tummuru and Blaser, 1992). In three unrelated laboratory-passaged strains, the promoter was deleted, resulting in nonexpression of SLPs and an S- phenotype (Fujita et al., 1997; Tu et al., 2001a). This suggests both that sapA expression is controlled by a single promoter and that spontaneous deletion of this promoter is common in vitro. To investigate whether genetic changes accompanied the switch from one antigenic form of SapA protein expressed to another, two C. fetus strains were chosen (Tummuru and Blaser, 1993). One was a laboratory-passaged strain (23D-11) derived from the other, the wild-type strain 23D. 23D predominantly expressed a 97-kDa SapA protein but had switched to producing a 127-kDa protein in 23D-11. Southern hybridization using several probes encompassing the sapA locus showed that accompanying the switch in SLP production was a reciprocal exchange of DNA between the loci encoding the 97 and 127kDa proteins. This exchange resulted in the placement of the sapA gene encoding the 127-kDa protein downstream of the sapA promoter in strain 23D-11, thereby allowing its expression. In this strain, the remaining sapA genes were intact, indicating that the genetic changes that had resulted in the switch between two sapA homologs did not involve rearrangements of additional sapA genes. The sapA gene was mutagenized by the insertion of a kanamycin-resistance (aphA) cassette (Blaser et al., 1994). The mutation generated a predominant, truncated 50-kDa SapA protein, indicating that the sapA gene was downstream of the active supA promoter. The 50-kDa truncated SapA protein was not exported by C. fetus, implicating the carboxyterminal region of SapA as containing sequences necessary for its secretion (see below). However, preparations of these cells also contained minor amounts of SLPs of sizes ranging from 96 to 149 kDa. These data suggested that cells arising from a progenitor mutant that contained an interrupted sapA gene had undergone phenotypic reversion at extremely high frequency (to the expression of alternate, full-length SLPs. Selection based on serum resistance of strains expressing the full-length, but not the truncated, supA homolog confirmed the high frequency of the reversion events. The reversion events were accompanied by the reacquisition of other phenotypes associated with the S-layer-poor complement binding and bacteremia in mice. Revertants contained DNA rearrangements that had recombination end points in the conserved regions at the 5’ ends of sapA homologs, and resulted in the movement of complete alternate supA gene copies to a location downstream of the sapA promoter. Importantly, the kanamycin resistance cassette was not lost during these reversion
events, indicating that the genetic exchanges were due to reciprocal recombination (bidirectional) rather than to gene conversion (unidirectional). Similar findings were reported for C. fetus subsp. venerealis in experimental infections in cattle (Garcia et al., 1995). A C. fetus subsp. venerealis strain that expressed a 110-kDa SLP was inoculated into the vagina of a 9month-old heifer. Protein profiles of C. fetus subsp. venerealis isolates recovered weekly showed progressive switching between SLPs of sizes ranging from 110 to 160 kDa. The variation in SLP expression also was accompanied by genetic rearrangements detectable by DNA hybridization with a probe specific for the 5’ end of the sapA coding region. Therefore, both in vitro and in vivo, the antigenic variation of SLPs occurs at high frequency and is due to reciprocal recombination among sapA genes (Dworkin and Blaser, 1997a). THE sup ISLAND As predicted by pulsed-field gel electrophoresis (Dworkin et al., 1995a), all of the sap homologues are encoded in a 54-kb chromosomal region. In strain 23D, the island consists of nine ORFs, of which eight are full-length sapA homologs and one a partial homolog, as well 19 other ORFs (Tu et al., 2003a, 2003b), including the sap CDEF secretion genes (see below). The homologs are arranged in three regions with the homologs directly repeated or inverted in relation to each other. This arrangement would permit any of the homologs to be expressed downstream of the unique sapA promoter, which has been shown experimentally to occur (Tu et al., 2003b). Other sequences that are present in more than one homolog permit more complex recombination that allows chimeric forms. Examination of the sap island in strain 82-40 for which the genomic sequence has been solved (D. Fouts, unpublished observations) shows essential similarity to that of strain 23D. The use of probes and PCR primers to these homologs shows their widespread conservation among most C. fetus strains tested (Tu et al., 2004a). In relapsing human C. fetus infections, rearrangements and antigenic variation involving the sap homologs were common (Tu et al., 2005b). MOLECULAR MECHANISMS OF SLP ANTIGENIC VARIATION Questions remained regarding the expression and antigenic variation of SLPs. First, is the expression of SLPs dependent on the single promoter upstream of sapA? Several lines of evidence suggested
CHAPTER 23
that it was, including primer extension, Southern and Northern hybridization, and the lack of SLP expression in mutant 23B, in which a deletion of the known sapA promoter had occurred (Blaser et al., 1994; Tu et al., 2001a). However, the presence of additional promoters upstream of other sapA homologs could not be excluded. Second, what exactly was the nature of the recombinational events that resulted in altered expression of sapA genes? Further investigations into these questions were accomplished by the introduction of a promoterless apbA cassette into either the sapA or the sapA2 locus of wild-type strain 23D (Dworkin and Blaser, 1996). Because the cassette lacked its own promoter, kanamycin resistance resulted only when the copy of sapA containing the apbA gene was present downstream of the sapA promoter. In strains containing either of these constructs (apbA in sapA or in sapA2), when kanamycin resistance was selected, no expression of SLPs was observed. This was in contrast to previous results that used an apbA cassette containing its own promoter, when simultaneous selection of kanamycin resistance and expression of SLPs was detected (Blaser et al., 1994). This result suggested that only a single supA promoter existed, and that when it was positioned to drive kanamycin resistance, it was unavailable for the expression of SLPs. This hypothesis was tested further by placing kanamycin-resistant cells into human serum. As stated previously, survival of C. fetus in human serum requires the presence of an S-layer (Blaser et al., 1987). Therefore, exposure to serum provides a strong selection for the detection of an Slayer. When strains in which aphA was present within sapA or sapA2 had been grown on kanamycin were then subjected to serum selection, only cells in which the sapA promoter had been moved upstream of a gene encoding a native SLP would be able to survive (Fig. 6). The opposing kanamycin and serum selections then were used to isolate a series of variants that had undergone recombination events to alter the expression of sapA homologs. The original sapAapbA and sapA2-upbA strains were first selected on kanamycin as described above, and SLP expression was undetectable. However, incubation of these bacteria in serum followed by plating on medium lacking kanamycin allowed the identification of bacteria that had switched to the expression of an S-layer and therefore survive serum selection. These S+ bacteria were recovered at a frequency of lop4 and were kanamycin sensitive, as expected. Plating of these S+ cells on medium containing kanamycin allowed the recovery of kanamycin resistance (with concomitant serum sensitivity), again at a frequency of The inability to express both the supA-aphA construct (resulting
PATHOGENESIS OF C. FETUS
417
in kanamycin resistance) and another sapA homolog (resulting in an S-layer and serum resistance) verified that a single sapA promoter existed. Chromosomal DNA from these variants was subjected to Southern hybridization using probes specific for the sapA promoter or apbA gene. These experiments were used to map the locations of the sapA promoter in relation to the aphA gene and to other sapA genes, and to correlate this with the expression state of the given genes. The hybridization results predicted the model shown in Fig. 6, which indicates inversion of a DNA segment containing the promoter. All inversion events have in common a 6.2-kb segment of DNA that contains the unique outwardfacing sapA promoter flanked by sapA homologs (Fig. 6). With end points in the 600-bp conserved regions at the 5 ’ ends of each sapA homolog, this region can invert, now orienting the sapA promoter such that it initiates the expression of the alternate flanking sapA gene. This model provides a means for the alternating expression of the two sapA genes that flank the 6.2 kb invertible region. However, wild-type C. fetus strains have multiple SLP-encoding genes: strain 23D has eight (Tu et al., 2001a) and strain TK has seven (Fujita and Amako, 1994). These genes are clustered on the chromosome rather than dispersed, and this tight arrangement is conserved among C. fetus strains (Dworkin et al., 1995a), suggesting that this genetic organization has been selected for the recombination events resulting in the expression of alternate sapA homologs. To investigate the expression of sapA homologs other than sapA and sapA2, an inversion reporter system similar to that described above was constructed (Dworkin and Blaser, 1997b). A strain was isolated that contained insertions of promoterless chloramphenicol (cat) and kanamycin (apbA) resistance genes in the sapA and sapA2 genes, respectively. Selection of resistance to either antibiotic resulted in recovery of cells (at a frequency of in which the sapA promoter was upstream and driving the expression of the relevant antibiotic resistance gene. In this manner, antibiotic resistance could be used to assay simple inversion events, exactly as described above. However, if serum resistance was selected in this strain, it required that a gene that did not originally flank the invertible region be brought downstream of the sapA promoter (Fig. 7). Surprisingly, this type of inversion event occurred at a frequency similar to that of simple inversions (approximately lop4). These experiments also provided strong evidence for the utilization of extremely large inversion events such that any of the clustered sapA homologs potentially could be expressed. All of the inversion events detected in these experiments involved recombination between the 600-
418
BLASER ET AL.
-
m H
PP
N H
I
I 1
1
N
I
*
HPPCH I II I
------+
sapA2
23D
aU!nKanamvcin
+
saPA
+
* H N P P I
I
I I
Selection on serum P
H I
b
t
sapA2
23D:A2K250
+
SPA
I
Selection on kanamycin
I
I I
II
1
II
H
PP
N H
I
I 1
1
I
4. sapA2
N
HPPCH
* - - - sapA
I II I
23D:A2K350
-b 1 kb
Figure 6 . DNA inversion events in a model system using a promoterless apbA (km) cassette inserted into the wild-type sapA2 locus (top line). When the supA promoter is positioned in the proper orientation, resistance to kanamycin results in S- bacteria, at a frequency of (second line). When kanamycin-resistant cells are removed from kanamycin selection and subjected to serum selection, S+ (serum resistant), kanamycin-sensitive cells arise at a frequency of (third line). Solid arrows represent expressed genes, and broken arrows represent silent (unexpressed) genes. The stippled boxes represent the 600-bp conserved regions at the 5’ ends of supA genes, and asterisks show the positions of the embedded inverted repeats that may play a role in the inversion process. The heavy line is the 6.2-kb invertible region. From Dworkin and Blaser (1996).
bp conserved regions at the 5’ ends of the sapA homologs (Figs. 6 and 7). To investigate trans-acting factors involved in the inversion, the C. fetus recA gene was cloned and a recA mutant was constructed (Dworkin et al., 1997). RecA, a protein found in essentially all bacteria, is required for homologous recombination. As such, it might be predicted to be essential for the types of recombination events leading to sapA promoter inversion. The putative x sites upstream of sapA homologs are further suggestions of the involvement of homologous recombination. However, previously characterized inversion systems operated independently of RecA, and instead required site-specific invertases. Nevertheless, in a C. fetus recA mutant, the high-frequency sapA promoter inversion was not detectable (Dworkin et al., 1997). Thus, homologous recombination is required to be able to detect C. fetus SLP antigenic variation. Although RecA appears necessary for high-frequency
inversion, it has not been shown to be sufficient. The possibility still exists that other proteins might be required for initiating inversion events by generating site-specific DNA strand breaks. The potential roles of the putative x sites and inverted repeats remain to be shown because the sapB genes lack these features. A low-frequency recA-independent mechanism also has been detected (Ray et al., 2000). The existence of alternative mechanisms is one indication of the biological importance of supA inversion to C. fetus. An area that has not been fully investigated is the potential role of homologous recombination in generating novel sapA-encoding genes. With DNA homology both upstream and downstream of each sapA gene, there exists the possibility for either homologous recombination or gene conversion events between directly repeated genes, and not involving the sapA promoter. These types of events would not result in changes in sapA expression, but rather
CHAPTER 23
PATHOGENESIS OF C. FETUS
419
Original Genotype
/I\
Inversion Events
New Genotypes r
-
* 1kb Figure 7. Model for complex inversion events resulting in the expression of alternate sapA homologs. Simple inversion events (Fig. 6 ) can occur, as well as more complex events in which the invertible region and one or more adjacent genes invert. Black boxes and other types of shading represent the conserved 5’ regions and divergent 3’ regions of sapA genes, respectively. The bent arrow shows the location o f the unique supA promoter, which is associated with the expression of the adjacent supA homolog (straight arrow). The asterisks are sequences (x,inverted repeats) that are potentially involved in the inversion process. From Dworkin and Blaser (199713).
would create chimeric sapA homologs that would have novel antigenic determinants. Reassortment of antigenic domains occurs in this manner in gonococcal opacity proteins (Connell et al., 1988). Such an increase in the arsenal of SLP antigens would enhance the “immune avoidance” capabilities of C. fetus during infections. The high levels of conservation of the sapA homologs between strains from different parts of the world, and even between isolates from mammals and reptiles, argue against a mechanism that generates unlimited recombinants (Tu et al., 2001a, 2003b, 2004a, 2004b), but there are preliminary experimental findings that such events can occur (Tu et al., 2001b). In summary, the importance of the C. fetus Slayer is evident by the substantial segment of the small C. fetus genome that is devoted to the expression of SLPs, and by the complexity with which SLPencoding genes can be rearranged. SLPs are expressed from a family of multiple complete genes, using a single, highly active promoter. An inverting segment of DNA, ranging in size from a minimum of 6.2 kb to an undetermined maximum of perhaps 20 to 30 kb, contains an outward-facing unique sapA promoter. The inversion of this segment occurs at a frequency of and places the promoter in position to express any of a family of sapA genes. Unlike pre-
viously characterized inversion systems, the C. fetus SLP system requires RecA and homologous recombination for its high-frequency inversion events, but RecA-independent inversion occurs at much lower frequency (Grogono-Thomas et al., 2003). SECRETION OF C . FETUS SURFACELAYER PROTEINS A more recent development in understanding the biogenesis of the C. fetus S-layer relates to the secretion of the S-layer subunits before assembly at the cell surface. To understand sapA promoter inversion in greater detail, the 6.2-kb minimal invertible region from both type A strain 23D and type B strain 84107 were characterized (Thompson et al., 1998). The DNA sequences of the invertible regions from the two strains were virtually identical, underscoring the high degree of conservation among the factors controlling the biogenesis of the S-layer (Fig. 8). The invertible region of the type A strain was studied in the greatest detail, but because of the high similarity of the type B segment at both the DNA and predicted protein levels, the conclusions reached for the type A sequence are likely largely true for the type B segment as well.
420
BLASER E T AL.
1 kb
t
WPF v
v
r
111111111111111111
w-
=PAY
111111111111111111
INVERTIBLE SEGMENT Figure 8. Genetic organization of the 6.2-kb invertible region. Bold arrows indicate genes contained within the invertible region. Bent arrows represent the divergent sapA and sapCDEF promoters. Hatched lines denote the conserved 5‘ regions of the flanking sapA homologs, indicated here as s a p h and sapAy. From Thompson et al. (1998).
As stated above, the minimal type A invertible region is 6.2 kb in size. The 74 bp at either end of the type A invertible region are inverted repeats of each other. These comprise the 5’ ends of the 600bp conserved boxes that serve as the sites of homologous recombination during the inversion events that result in SLP antigenic variation (Dworkin and Blaser, 1996; Dworkin et al., 1995a; Tummuru and Blaser, 1993) (Figs. 6 and 7). Internal to the inverted repeats and at one end is the outward-facing supA promoter that allows expression of the supA homolog that is located immediately downstream (Figs. 8 and 9). Features of the 114-bp 5’ untranslated region of the supA message (inverted 7-bp repeats, repeated pentameric sequences, x site) were found as previously described (see above). Adjacent to the supA promoter, but positioned in the opposite orientation, are several potential promoter sequences that would direct the transcription of an operon of four genes (supCDEF) that occupies the remainder of the invertible region. Initial primer extension studies aimed at precisely mapping the supCDEF transcriptional start site were unable to detect this message (S. A. Thompson, unpublished data). In contrast to the abundance of cellular supA message, the steady-state level of supCDEF message apparently is extremely low. Downstream of the supCDEF promoter region begin the supCDEF genes. These genes are 1.0, 1.8, 1.3, and 1.3 kb in size, respectively, and each overlaps the adjacent gene. The predicted SapC protein is 39.7 kDa, with a PI of 9.3. The predicted SapC protein currently has no homologies in GenBank to suggest its function. A homology search of the C. jejuni genome indicates that C. jejuni does not contain a sapC homolog (Parkhill et al., 2000). Because C. jejuni does not possess an S-layer, the function of SapC may be related to expression of S-layers by Cumpylobacter species. Whether this gene is present in Slayer-producing campylobacters other than C. fetus (such as C. rectus) is currently unknown.
In contrast to the lack of SapC-homologous proteins, database searches indicated that the sapDEF genes predicted proteins that constituted the three components of a type I protein secretion apparatus. The use of a type I transporter for the secretion of C. fetus SLPs was suggested by previous characteristics of the SLPs themselves. First, none of the characterized C. fetus SLPs had an amino-terminal secretion signal, unlike most other bacterial SLPs (Blaser and Gotschlich, 1990; Boot and Pouwels, 1996; Casademont et al., 1998; Dworkin et al., 1995a, 1995b; Tummuru and Blaser, 1993). Second, SapA derivatives in which the carboxy terminus of the protein had been deleted were secretion deficient, suggesting that the carboxy terminus of the protein was required for its extracellular transport (Blaser et al., 1994). Carboxy-terminal secretion signals are typical of type I-secreted proteins. The use of type I transporters in SLP secretion has been reported for other bacterial species, including Cuulobucter crescentus and Serratiu murcescens (Awram and Smit, 1998; Kawai et al., 1998). The Cumpylobacter rectus SLP lacks an amino-terminal signal sequence and may be secreted by a similar pathway (Wang et al., 1998). On the basis of studies of type I transporters in other bacteria, the model shown in Fig. 10 can be predicted for the C. fetus SLP transporter. SapD is 64 kDa in size and by homology is an inner membrane protein and a member of the large ABC superfamily of transport proteins. These proteins use the binding and hydrolysis of ATP to energize the movement of molecules across bacterial membranes (Blight and Holland, 1990). SapE (47.9 kDa) is another inner membrane protein that belongs to the membrane fusion protein family. SapF is an outer membrane protein that may serve as the outermost component (pore) of the transport apparatus. The interaction of the C. fetus SLP secretion signal with SapD probably initiates the sequential recruitment of SapE and SapF protein to assemble the intact SLP
Zone 1
Zone 2
Zone 3
1 kb -
Figure 9 . Schematic representation and genomic organization of the sup locus in C fetus strain 23D. The adjacent ORFs that are not supA homologs (including supC, 0, E, F , and Cf0001. . .Cf0032) are indicated by shaded boxes. The PCR primers used in this study (PF, SF, TR, SR, AF-A7F, and AR-A7R) are designated by the arrows, which also denote the primer orientations. The horizontal line indicates the length of the fragment. From Molecular Microbiology.
422
BLASER E T AL.
a p n Figure 10. Model for secretion and assembly of the S-layer. A hypothetical structure of the C. fetus SLP transporter is shown, based on similarities to other type I transporters. The putative stoichiometry of the SapE and SapF proteins in the assembled transport apparatus is based on data gathered for the E. coli HlyA transporter (Thanabalu et al., 1998). Recognition of the SapA carboxy-terminal secretion signal is mediated by the SapD protein. The SapAiSapD complex initiates the sequential assembly of SapE and SapF trimers resulting in a contiguous pore through which SapA is secreted. SapA then may attach to LPS and be added to the growing S-layer.
transporter, which creates a pore bridging both inner and outer membranes (Lttofft et al., 1996; Thanabalu et al., 1998). The initial event in translocation of SapA by the SapDEF type I transporter probably is recognition of the SapA secretion signal by SapD. Unlike sec-dependent protein secretion, which utilizes amino-terminal signal sequences, type I secreted proteins typically have carboxy-terminal secretion signals (Pugsley, 1993). Inspection of the carboxyterminal 100 amino acids of several of the characterized C. fetus SLPs reveals conserved residues that may play a role in the secretion process (Thompson et al., 1998). The relative simplicity of the type I secretion apparatus facilitates its study. The C. fetus invertible region cloned into E. coli is functional and can mediate the secretion of SapA from E. coli into the culture medium (Thompson et al., 1998). This will permit study of factors important in the SapA secretion process by means of more complex genetic systems than currently are available for C. fetus. The lack of a recognizable transcriptional terminator downstream of sapF raises the question as to whether a sapA homolog correctly positioned at the end of the invertible region opposite the sapA promoter could be expressed from the sapCDEF promoter. It is possible that C. fetus transcriptional terminators do not resemble those of E. coli, and that sequences downstream of sapF in fact function to cause cessation of transcription. Similar to this, Helicobacter gylori lacks inverted repeat-type (rho-
independent) transcriptional terminators downstream of the operons predicted by the genomic sequence (Washio et al., 1998). Is the production of SLPs by C. fetus regulated? SLPs are among the most abundant proteins produced by C. fetus, and as such represent a major burden to the cell’s metabolic machinery. Furthermore, it is possible that SLPs accumulating intracellularly will exert a toxic effect on the cell. It would seem beneficial for the cell to have a mechanism to repress the accumulation of these proteins, and this may be linked to the rate of their secretion. For some proteins that are secreted by type I systems, mutations involving the transport apparatus lead to the inability to detect the nontransported protein within the cell (Thompson et al., 1998). Although not addressed experimentally, this has been assumed to be due simply to proteolytic degradation of the protein within the cell. General or specific degradation of these proteins would be one means of preventing an intracellular buildup of aberrantly nonsecreted proteins. Another possibility is that there may be a specific regulatory mechanism that prevents further synthesis of proteins once the intracellular pool of these proteins surpasses an acceptable level. An example of this type of regulation occurs with the S-layer-producing bacterium Thermus thermophilus (Fernandez-Herrero et al., 1997). The mRNA from which its SLP (SlaA) is translated contains a 127-bp 5’ untranslated region. During conditions of intracellular accumulation of SLPs,
CHAPTER 23
this 5’ untranslated region serves as a binding site for SlaA. The binding of SlaA to its own message prevents further translation of the mRNA, and the rate of SlaA synthesis is decreased. In summary, the following model of S-layer expression, variation, and biogenesis has emerged from these studies. Within both type A and type B strains, SLPs are encoded by families of sup genes. The sup genes encode serotype-specific SLPs that have highly conserved amino termini. The conserved amino termini of these proteins are responsible for the binding of the SLPs to homologous LPS, and therefore allow adherence of the S-layer to the cell surface. The remainder of each protein is divergent, and this bestows structural, antigenic, and possibly functional differences among S-layers constructed from the individual SLPs (Tu et al., 2003a, 2003b). Each sup gene is complete and contains all the information needed to encode a functional SLP. However, only one of these genes is expressed at a given time, and its expression requires that the unique sup promoter be present immediately upstream of the expressed gene. The movement of the sup promoter upstream of a sup gene usually involves high-frequency DNA inversion events mediated by homologous recombination. This recombination is strongly promoted by RecA protein and probably the conserved homologous DNA sequences found at the 5’ ends of each sup gene. Other cis-acting features (putative x sites, direct and inverted repeats, and conserved 3’ homology) also may be required. Factors that are responsible for the secretion of SLPs to the cell surface are encoded on the invertible element, and these comprise a type I protein secretion system (SapDEF) that is different than the secretion systems by which most SLPs are transported. The role of SapC is unknown at present.
ROLE OF S-LAYER PROTEINS IN OVINE ABORTION Animal models that mimic the natural disease outcomes of infection are valuable for understanding the mechanisms of pathogenesis of zoonotic organisms, such as C. fetus. Such models not only enable investigation of the disease mechanisms and holistic responses to the infection in the natural host, but also allow extrapolation to the human host. As indicated above, the pregnant ewe is highly susceptible to systemic infection with C. fetus subsp. fetus, which can result in abortion or the birth of weak lambs. To investigate the role of the S-layer proteins in ovine abortion, an in vivo model was developed that used pregnant ewes subcutaneously challenged with C. fe-
PATHOGENESIS OF C. FETUS
423
tus subsp. fetus strain 23D. Depending on the timing of the challenge, up to 90% of ewes aborted within 25 days after infection. However, the natural mutant 23B, which has a 9-kDa chromosomal deletion including the supA promoter, failed to cause abortion. The S-layer proteins expressed by C. fetus subsp. fetus strain 23D isolates recovered from the feces, placenta, and aborted fetal material of challenged ewes varied in molecular mass from 97 to 149 kDa, indicating that in vivo expression was driven by multiple supA genes. To further investigate the role of the S-layer proteins, mutants of this strain that were either unable to express or only expressed a single fixed, S-layer protein were used to challenge pregnant ewes (Grogono-Thomas et al., 2000). These mutants were produced by the insertion of an antibiotic cassette into the supA and/or recA genes, as described above (Dworkin and Blaser, 1996), to produce a series of isogenic mutants. Expression of at least one S-layer protein was essential for systemic infection resulting in ovine abortion (Table 1).However, bypassing the systemic infection phase by intraplacental injection enabled an S-layer-deficient mutant (23D:502) to cause abortion, indicating that the S-layer was involved in protecting the organism during the early infection stages of establishment of colonization or bacteremia and/or translocation to the placenta, but it does not possess the abortifactant properties. Interestingly, mutants with fixed expression of either a 97-kDa or a 127-kDa S-layer protein generated by defined mutation of the recA gene demonstrated variation in virulence, suggesting that organisms expressing a 97-kDa S-layer protein had a selective advantage over those expressing a 127-kDa S-layer protein. This hypothesis was supported by the unexpected isolation of an isolate expressing a 97kDa protein from the placenta of one ewe challenged with the mutant expressing a fixed 127-kDa S-layer protein. Subsequent investigations (Ray et al., 2000) demonstrated that this was the consequence of a RecA-independent process that enabled the inversion of the sup invertible region. The inversion event occurred at significantly lower frequency than for the RecA-dependent pathway. Without the selective pressure of this in vivo model, this observation likely would not have been made.
ROLE OF S-LAYER PROTEIN ANTIGENIC DIVERSITY IN OVINE IMMUNE RESPONSES As described previously, rearrangements in the supA locus enable variation in the expression of the S-layer proteins (Tu et al., 2003b, 2004a; Tummuru
424
BLASER ET AL.
and Blaser, 1993). Comparison of the amino acid sequences of these S-layer proteins indicates the presence of 3' variable regions. Because the S-layer is highly antigenic during infection, sapA switching allows for antigenic variation. The ovine model of infection was used to investigate the role of this antigenic variation in the outcome of infection (Grogono-Thomas et al., 2003). After subcutaneous challenge with C. fetus subsp. fetus strain 23D, antiS-layer protein antibodies were detectable in serum, milk, bile, and urine by an S-layer-protein-specific enzyme-linked immunosorbent assay. By means of an isotype-specific enzyme-linked immunosorbent assay, the serum antibody response was composed of an early increase in specific IgM followed by specific IgG1, and later responses of IgG2 and IgA. The presence of mucosal antibody responses were dependent on the route of challenge. Fecal IgA antibodies were only detected in animals challenged via the oral route; however, anti-S-layer protein IgA antibodies were detected in the bile and urine of subcutaneously infected animals. To investigate the role of antigenic diversity on the immune response, a comparison was undertaken of the kinetics of the serum IgG response of animals challenged with the wild-type strain or the isogenic mutants. The results indicated that the ability to switch S-layer protein delayed the host S-layerprotein-specific IgG immune response by approximately 1 week compared with mutants expressing a fixed S-layer protein. Thus, sapA switching provided an advantage to the pathogen by affecting the host immune response. However, epidemiological evidence shows that prior infection with C. fetus subsp. fetus protects ewes from subsequent challenge during further pregnancies, indicating that effective immune responses can be generated. To investigate this apparent paradox, a model of immune protection was established by using ewes vaccinated before mating and subsequent challenge. The results of these experiments indicated that vaccines containing S-layers, whether fixed or variable, protected the ewes from abortion. This protection was independent of the S-layer protein expressed by the vaccine strain, suggesting that common epitopes in the S-layer protein were protective. By means of a bioinformatics approach, a conserved 184-amino acid N-terminal region of the S-layer protein was identified as potentially immunogenic. This region was epitope-mapped with hyperimmune rabbit antiserum and 20-mer overlapping peptides. The amino acid regions 8 1 to 110 and 141 to 160 both reacted with the rabbit antiserum and the former region also reacted with serum IgG antibodies from the experimentally infected sheep, and could be a candidate for a future peptide-based subunit vaccine.
Thus, despite the antigenic variation afforded by the sapA gene switching, protective host immune responses can be generated against the conserved regions of the protein. Therefore, the advantage to the pathogen of such an investment in antigenic variation is not obvious. One possible explanation is that the epitopes exposed in the variable regions of the antigen are immunodominant, distracting the host immune response towards the variable regions and away from the more cryptic conserved region epitopes. This strategy appears to work for a limited period, possibly because of the restricted antigenic variation available, but may be sufficient to enable colonization to become established and bacterial growth to occur, resulting in abortion and pathogen distribution through the products of abortion. Thereafter, the immune response directed against the conserved region becomes effective and is sufficiently cross-protective to provide immunity to potential exposures during subsequent pregnancies. However, the cyclical nature of abortion blooms in sheep flocks provides evidence that such protective immunity wanes rapidly, rendering a flock susceptible to further abortions within a few years. In summary, in ovine infections, the S-layer proteins are essential virulence factors, enabling translocation of C. fetus subsp. fetus to placental tissue in the pregnant animal. The outcome of the infection in terms of effects on the fetus are dependent on the interaction between the pathogen and the host response. This model also provides a context to understand the role of S-layer proteins as virulence factors in human infections. Acknowledgments. The work in this chapter was supported by NIH (AI024145, AI043548, AIO55715, AI058284), the Medical Research Service of the Department of Veterans Affairs, the Department of Environment, Food, and Rural Affairs, Great Britain, FWF grants P18607-B12, P20479-BO5, and the Austrian Academy of Science.
REFERENCES Allos, B. M., A. J. Lastovica, and M. J. Blaser. 1995. Atypical campylobacters and related microorganisms, p. 849-865. In P. D. S. M. J. Blaser, J. I. Ravdin, H. B. Greenberg, and R. L. Guerrant (ed.), Infections of the Gastrointestinal Tract. Raven Press, New York. Alm, R. A., P. Guerry, M. E. Power, and T. J. Trust. 1992. Variation in antigenicity and molecular weight of Cumpylobacter coli VC167 flagellin in different genetic backgrounds. J. Bacteriol. 174:4230-4238. Anstead, G. M., J. H. Jorgensen, F. E. Craig, M. J. Blaser, and T. F. Patterson. 2001. Therrnophilic multidrug-resistant Cumpylobucter fetus infection with hypersplenism and histiocytic phagocytosis in a patient with acquired immunodeficiency syndrome. Clin. Infect. Dis. 32:295-296.
CHAPTER 23
Atabay, H. I., and J. E. Corry. 1998. The isolation and prevalence of campylobacters from dairy cattle using a variety of methods. J. Appl. Microbiol. 84:733-740. Awram, P., and J. Smit. 1998. The Caulobacter crescentus paracrystalline S-layer protein is secreted by an ABC transporter (type I) secretion apparatus. J. Bacteriol. 180:3062-3069. Backert, S., and T.F. Meyer. 2006. Type IV secretion systems and their effectors in bacterial pathogenesis. Cuw. Opin. Microbiol. 9:207-217. Bawa, E. K., J. 0. Adekeye, E. 0. Oyedipe, and J. U. Omoh. 1991. Prevalence of bovine campylobacteriosis in indigenous cattle of three states in Nigeria. Trop. Anim. Health Prod. 23:157-160. Berg, R. L., J. W. Jutila, and B. D. Firehammer. 1971. A revised classification of Vibrio fetus. Am. J. Vet. Res. 32:ll-22. Blaser, M. J. 1998. Campylobacter fetus-emerging infection and model system for bacterial pathogenesis at mucosal surfaces. Clin. Infect. Dis. 27:256-258. Blaser, M. J., and E. C. Gotschlich. 1990. Surface array protein of Campylobacter fetus. Cloning and gene structure. 1.Biol. Chem. 265:14529-14535. Blaser, M. J., and Z.Pei. 1993. Pathogenesis of Campylobacter fetus infections: critical role of high-molecular-weight S-layer proteins in virulence. J. Infect. Dis. 167:372-377. Blaser, M.J., P. F. Smith, J. A. Hopkins, I. Heinzer, J. H. Bryner, and W. L. Wang. 1987. Pathogenesis of Campylobacter fetus infections: serum resistance associated with high-molecularweight surface proteins. 1.Infect. Dis. 155:696-706. Blaser, M. J., P. F. Smith, and P. A. Kohler. 1985. Susceptibility of Campylobacter isolates to the bactericidal activity in human serum. 1. Infect. Dis. 151:227-235. Blaser, M.J., P. F. Smith, J. E. Repine, and K. A. Joiner. 1988. Pathogenesis of Campylobacter fetus infections. Failure of encapsulated Campylobacter fetus to bind C3b explains serum and phagocytosis resistance. J. Clin. Invest. 81:1434-1444. Blaser, M. J., E. Wang, M. K. Tummuru, R. Washburn, S. Fujimoto, and A. Labigne. 1994. High-frequency S-layer protein variation in Campylobacter fetus revealed by sapA mutagenesis. Mol. Microbiol. 14:453-462. Blight, M. A., and I. B. Holland. 1990. Structure and function of haemolysin B, P-glycoprotein and other members of a novel family of membrane translocators. Mol. Microbiol. 4:873-880. Bokkenheuser, V. 1970. Vibrio fetus infection in man. I. Ten new cases and some epidemiologic observations. Am. J. Epidemiol. 91~400-409. Boot, H. J., and P. H. Pouwels. 1996. Expression, secretion and antigenic variation of bacterial S-layer proteins. Mol. Microbiol. 2 1:1 117- 1123. Bryner, J. H., A. H. Frank, and P. A. O’Berry. 1962. Dissociation studies of vibrios from the bovine genital tract. Am. J. Vet. Res. 23:32-41. Buckmire, F. L., and R. G. Murray. 1973. Studies on the cell wall of Spirillum serpens. 11. Chemical characterization of the outer structured layer. Can. J. Microbiol. 1959-66. Casadimont, I., D. Chevrier, and J. L. Guesdon. 1998. Cloning of a sapB homologue (sapB2) encoding a putative 112-kDa Campylobacter fetus S-layer protein and its use for identification and molecular genotyping. FEMS Immunol. Med. Microbiol. 21: 269-281. Christie, P. J. 2004. Type IV secretion: the agrobacterium VirB/ D4 and related conjugation systems. Biochim. Biophys. Acta 1694~219-234. Christie, P. J., K. Atmakuri, V. Krishnamoorthy, S. Jakubowski, and E. Cascales. 2005. Biogenesis, architecture, and function of bacterial type IV secretion systems. Annu. Rev. Microbiol. 59: 451-485.
PATHOGENESIS OF C. FETUS
425
Connell, T. D., W. J. Black, T. H. Kawula, D. S. Barritt, J. A. Dempsey, K. Kverneland, Jr., A. Stephenson, B. S. Schepart, G . L. Murphy, and J. G. Cannon. 1988. Recombination among protein I1 genes of Neisseria gonowhoeae generates new coding sequences and increases structural variability in the protein I1 family. Mol. Microbiol. 2:227-236. Corbeil, L. B. 1999. Immunization and diagnosis in bovine reproductive tract infections. Adv. Vet. Med. 41:217-239. Corbeil, L. B., R. R. Corbeil, and A. J. Winter. 1975a. Bovine venereal vibriosis: activity of inflammatory cells in protective immunity. Am. 1. Vet. Res. 36(4 Pt. 1):403-406. Corbeil, L. B., G. G . D. Schurig, P. J. Bier, and A. J. Winter. 1975b. Bovine venereal vibriosis: antigenic variation of the bacterium during infection. Infect. Immun. 11:240-244. Dubreuil, J. D., M. Kostrzynska, J. W. Austin, and T. J. Trust. 1990. Antigenic differences among Campylobacter fetus S-layer proteins. J. Bacteriol. 1725035-5043. Dubreuil, J. D., S . M. Logan, S. Cubbage, D. N. Eidhin, W. D. McCubbin, C. M. Kay, T. J. Beveridge, F. G. Ferris, and T. J. Trust. 1988. Structural and biochemical analyses of a surface array protein of Campylobacter fetus. 1. Bacteriol. 170:41654173. Duim, B., P. A. R. Vandamme, A. Rigter, S. Laevens, J. R. Dijkstra, and J. A. Wagenaar. 2001. Differentiation of Campylobacter species by AFLP fingerprinting. Microbiology 147:27292737. Dunn, B. E., M. J. Blaser, and E. L. Snyder. 1987. Twodimensional gel electrophoresis and immunoblotting of Campylobacter outer membrane proteins. Infect. Immun. 55:15641572. Dworkin, J., and M. J. Blaser. 1996. Generation of Campylobacter fetus S-layer protein diversity utilizes a single promoter on an invertible DNA segment. Mol. Microbiol. 19:1241-1253. Dworkin, J., and M. J. Blaser. 1997a. Molecular mechanisms of Campylobacter fetus surface layer protein expression. Mol. Microbiol. 26:433-440. Dworkin, J., and M. J. Blaser. 1997b. Nested DNA inversion as a paradigm of programmed gene rearrangement. Proc. Natl. Acad. Sci. USA 94:985-990. Dworkin, J., 0. L. Shedd, and M. J. Blaser. 1997. Nested DNA inversion of Campylobacter fetus S-layer genes is recA dependent. 1.Bacteriol. 179:7523-7529. Dworkin, J., M. K. Tummuru, and M. J. Blaser. 1995a. A lipopolysaccharide-binding domain of the Campylobacter fetus S-layer protein resides within the conserved N terminus of a family of silent and divergent homologs. 1.Bacteriol. 177:17341741. Dworkin, J., M. K. Tummuru, and M. J. Blaser. 1995b. Segmental conservation of sapA sequences in type B Campylobacter fetus cells. J. Biol. Chem. 270:15093-15101. Eaglesome, M. D., and M. M. Garcia. 1992. Microbial agents associated with bovine genital tract infections and semen. Part I. Brucella abortus, Leptospira, Campylobacter fetus and Tritrichomonas foetus. Vet. Bull. 62:743-775. Eppinger, M.,C. Baar, G. Raddatz, D. H. Huson, and S. C. Schuster. 2004. Comparative analysis of four campylobacterales. Nat. Rev. Microbiol. 2672-8 85. Fernandez-Herrero, L. A., G . Olabarria, and J. Berenguer. 1997. Surface proteins and a novel transcription factor regulate the expression of the S-layer gene in Themzus themophilus HB8. Mol. Microbiol. 24:61-72. Fisher, J. A., J. Smit, and N. Agabian. 1988. Transcriptional analysis of the major surface array gene of Caulobacter crescentus. 1. Bacteriol. 170:4706-4713. Fogg, G. C., L. Y. Yang, E. Wang, and M. J. Blaser. 1990. Surface array proteins of Campylobacter fetus block lectin-mediated
426
BLASER ET AL.
binding to type A lipopolysaccharide. Infect. Immun. 58:27382744. Fouts, D. E., E. F. Mongodin, R. E. Mandrel], W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Fujimoto, S., A. Takade, K. Amako, and M. J. Blaser. 1991. Correlation between molecular size of the surface array protein and morphology and antigenicity of the Campylobacter fetus S layer. Infect. Immun. 59:2017-2022. Fujimoto, S., A. Umeda, A. Takade, K. Murata, and K. Amako. 1989. Hexagonal surface layer of Campylobacter fetus isolated from humans. Infect. Immun. 57:2563-2565. Fujita, M., and K. Amako. 1994. Localization of the sapA gene on a physical map of Campylobacter fetus chromosomal DNA. Arch. Microbiol. 162:375-3 80. Fujita, M., T. Moriya, S. Fujimoto, N. Hara, and K. Amako. 1997. A deletion in the supA homologue cluster is responsible for the loss of the S-layer in Campylobacter fetus strain TK. Arch. Microbiol. 167:196-20 l. Fujita, M., T. Morooka, S. Fujimoto, T. Moriya, and K. Amako. 1995. Southern blotting analyses of strains of Campylobacter fetus using the conserved region of sapA. Arch. Microbiol. 164: 444-447. Garcia, M. M., M. D. Eaglesome, C. F. Hawkins, and F. C. Alexander. 1980. campylobacteriosis in Jamaican cattle. Vet. Rec. 106~287-288. Garcia, M. M., C. L. Lutze-Wallace, A. S. Denes, M. D. Eaglesome, E. Holst, and M. J. Blaser. 1995. Protein shift and antigenic variation in the S-layer of Campylobacter fetus subsp. venerealis during bovine infection accompanied by genomic rearrangement of sapA homologs. J. Bacteriol. 177:1976-1980. Ginsberg, M. M., M. A. Thompson, C. R Peter, D. G. Ramras, and J. Chin. 1981. Cumpylobacter sepsis associated with “nutritional therapy”-California. Morb. Mortal. Wkly. Rep. 30: 294-295. Goossens, H., R. van den Abbeele, J. P. Butzler, and P. Williams. 1989. Study of iron uptake in Campylobacter fetus and C. jejuni, p. 414-415. In G. M. Ruiz-Palacios, E. Calva, and B. R. RuizPalacios (ed.), Campylobacter V. Instituto Nacional de la Nutricion, Mexico City. Grogono-Thomas, R., M. J. Blaser, M. Ahmadi, and D. G. Newell. 2003. The role of S-layer protein antigenic diversity in the immune responses of sheep experimentally challenged with Campylobacter fetus subsp. fetus. Infect. Immun. 71:147-154. Grogono-Thomas, R, J. Dworkin, M. J. Blaser, and D. G. Newell. 2000. The role of the Campylobacter fetus surface-layer protein in ovine abortion. Infect. Immun. 68:1687-1691. Guerrant, R. L., R. G. Lahita, W. C. Winn, and R. B. Roberts. 1978. Campylobacteriosis in man: pathogenic mechanisms and review of 91 bloodstream infections. Am. J. Med. 65584-592. Guerry, P. 2007. Campylobacter flagella: not just for motility. Trends Microbiol. 15:45 6-46 1. Hacker, J., U. Hentschel, and U. Dolbrindt. 2003. Prokaryotic chromosomes and disease. Science 301:790-793. Htbert, G. A., D. G. Hollis, R. E. Weaver, M. A. Lambert, M. J. Blaser, and C. W. Moss. 1982. 30 years of campylobacters: biochemical characteristics and a biotyping proposal for Cumpylobacter jejuni. J. Clin. Microbiol. 15:1065-1073. Hum, S. 1996. Bovine venereal campylobacteriosis, p. 355-358. In D. G. Newell, J. M. Ketley, and R. A. Feldman (ed.), Cam-
pylobacters, Helicobacters, and Related Organisms. Plenum Press, New York. Ichiyama, S., S. Hirai, T. Minami, Y. Nishiyama, S. Shimizu, K. Shimokata, and M. Ohta. 1998. Campylobacter fetus subspecies fetus cellulitis associated with bacteremia in debilitated hosts. Clin. Infect. Dis. 27:252-255. Jagannathan, A., and C. W. Penn. 2005. Motility, p. 331-347. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Cellular and Molecular Biology. Horizon Bioscience, Norfolk, United Kingdom. Jeon, B., and Q. Zhang. 2007. CjOOllc, a periplasmic single- and double-stranded DNA binding protein, contributes to natural transformation in Campylobacter jejuni. 1. Bacteriol. 189:73997407. Jin, S., A. Joe, J. Lynett, E. K. Hani, P. Sherman, and V. L. Chan. 2001. JlpA, a novel surface-exposed lipoprotein specific to Campylobacter jejuni, mediates adherence to host epithelial cells. Mol. Microbiol. 3 9 :1225-1236 . Karlyshev, A. V., J. M. Ketley, and B. W. Wren. 2005. The Campylobacter jejuni glycome. FEMS Microbiol. Rev. 29:377-390. Kawai, E., H. Akatsuka, A. Idei, T. Shibatani, and K. Omori. 1998. Sewatiu marcescens S-layer protein is secreted extracellularly via an ATP-binding cassette exporter, the Lip system. Mol. Microbiol. 27:941-952. Kelly, D. J. 2005. Metabolism, electron transport and bioenergetics of Campylobacter jejuni: implications for understanding life in the gut and survival in the environment, p. 275-292. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Cellular and Molecular Biology. Horizon Bioscience, Norfolk, United Kingdom. Kelly, D. J., and N. J. Hughes. 2001. The citric acid cycle and fatty acid biosynthesis, p. 135-146. In H. L. T. Mobley, G. L. Mendz, and S. L. Hazel1 (ed.), Helicobacter pylori: Physiology and Genetics. ASM Press, Washington, DC. Kienesberger, S., G . Gorkiewicz, M. M. Joainig, S. R. Scheicher, E. Leitner, and E. L. Zechner. 2007. Development of experimental genetic tools for Campylobacter fetus. Appl. Environ. Microbiol. 73:4619-4630. Klein, B. S., J. M. Vergeront, M. J. Blaser, P. Edmonds, D. J. Brenner, D. Janssen, and J. P. Davis. 1986. Campylobacter infection associated with raw milk. An outbreak of gastroenteritis due to Campylobacter jejuni and thermotolerant Campylobacter fetus subsp. fetus. JAMA 255:361-364. Konkel, M. E., S. G. Garvis, S. L. Tipton, D. E. Anderson, Jr., and W. Cieplak, Jr. 1997. Identification and molecular cloning of a gene encoding a fibronectin-binding protein (CadF) from Campylobacter jejuni. Mol. Microbiol. 24:953-963. Konkel, M. E., B. J. Kim, V. Rivera-Amill, and S. G. Garvis. 1999. Identification of proteins required for the internalization of Campylobacter jejuni into cultured mammalian cells. Adv. Exp. Med. Biol. 473:215-224. Kowalczykowski, S . C., D. A. Dixon, A. K. Eggleston, S. D. Lauder, and W. M. Rehrauer. 1994. Biochemistry of homologous recombination in Escherichia coli. Microbiol. Rev. 58:401-465. Lttofft, S., P. Delepelaire, and C. Wandersman. 1996. Protein secretion in gram-negative bacteria: assembly of the three components of ABC protein-mediated exporters is ordered and promoted by substrate binding. EMBO J. 155804-5811. Marty, A. T., T. A. Webb, K. G. Stubbs, and R. R. Penkava. 1983. Inflammatory abdominal aortic aneurysm infected by Campylobacter fetus. JAMA 249:1190-1192. Maurelli, A. T. 2007. Black holes, antivirulence genes, and gene inactivation in the evolution of bacterial pathogens. FEMS Microbiol. Lett. 267:l-8. McCoy, E. C., D. Doyle, K. Burda, L. B. Corbeil, and A. J. Winter. 1975. Superficial antigens of Cumpylobacter (Vibrio) fetus: char-
CHAPTER 23
acterization of an antiphagocytic component. Infect. Immun. 11: 5 17-525. McFaydean, J., and S. Stockman. 1909. Report of the Department Committee Appointed by the Board of Agriculture and Fisheries to Inquire into Epizootic Abortion. Vol. 3. His Majesty’s Stationery Office, London. Meier, P. A., D. P. Dooley, J. H. Jorgensen, C. C. Sanders, W. M. Huang, and J. E. Patterson. 1998. Development of quinoloneresistant Campylobacter fetus bacteremia in human immunodeficiency virus-infected patients. J. Infect. Dis. 177:95 1-954. Mohanty, S. B., G. J. Plumer, and J. E. Faber. 1962. Biochemical and colonial characteristics of some bovine vibrios. A m J. Vet. Rts. 23554-557. Moran, A. P. 1995. Structure-bioreactivity relationships of bacterial endotoxins. J. Toxicol. Toxin Rev. 14:47-83. Moran, A. P., D. T. O’Malley, T. U. Kosunen, and I. M. Helander. 1994. Biochemical characterization of Campylobacter fetus lipopolysaccharides. Infect. Immun. 62:3922-3929. Moran, A. P., D. T. O’Malley, J. Vuopio-Varkila, K. Varkila, L. Pyhala, H. Saxen, and I. M. Helander. 1996. Biological characterization of Campylobacter fetus lipopolysaccharides. FEMS Immunol. Med. Microbiol. 15:43-50. Morrison, V. A., B. K. Lloyd, J. K. Chia, and C. U. Tuazon. 1990. Cardiovascular and bacteremic manifestations of Campylobacter fetus infection: case report and review. Rev. Infect. Dis. 12:387392. Myers, J. D., and D. J. Kelly. 2005. A sulphite respiration system in the chemoheterotrophic human pathogen Campylobacter jejuni. Microbiology 151:233-242. Myers, L. L. 1971. Purification and partial characterization of a Vibrio fetus immunogen. Infect. Immun. 3562-566. Naikare, H., K. Palyada, R. Panciera, D. Marlow, and A. Stintzi. 2006. Major role for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infect. Immun. 745433-5444. Neuzil, K. M., E. Wang, D. W. Haas, and M. J. Blaser. 1994. Persistence of Campylobacter fetus bacteremia associated with absence of opsonizing antibodies. J. Clin. Microbiol. 32:17181720. Ochman, H., and N. A. Moran. 2001. Genes lost and genes found: evolution of bacterial pathogenesis and symbiosis. Science 292: 1096-1099. Pajaniappan, M., J. E. Hall, S. A. Cawthraw, D. G. Newell, E. C. Gaynor, J. A. Fields, K. M. Rathbun, W. A. Agee, C. M. Burns, S . J. Hall, D. J. Kelly, and S. A. Thompson. A temperatureregulated Campylobacter jejuni gluconate dehydrogenase is involved in respiration-dependent energy conservation and chicken colonization. Mol. Microbiol., in press. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668 Pefanis, S. M., S. Herr, C. G. Venter, L. P. Kruger, C. C. Queiroga, and L. Amaral. 1988. Trichomoniasis and campylobacteriosis in bulls in the Republic of Transkei. J. S. A f . Vet. Assoc. 59:139140. Pei, Z., and M. J. Blaser. 1990. Pathogenesis of Campylobacter fetus infections. Role of surface array proteins in virulence in a mouse model. J. Clin. Invest. 85:1036-1043. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. Z. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchtre, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni re-
PATHOGENESIS OF C. FETUS
427
duces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943. Pei, Z., R. T. D. Ellison, R. V. Lewis, and M. J. Blaser. 1988. Purification and characterization of a family of high molecular weight surface-array proteins from Campylobacter fetus. J. Biol. Chem. 263:6416-6420. Penner, J. L. 1988. The genus Campylobacter: a decade of progress. Clin. Microbiol. Rev. 1:157-172. PCrez-Ptrez, G. I., and M. J. Blaser. 1985. Lipopolysaccharide characteristics of pathogenic campylobacters. Infect. Immun. 47: 353-359. P&ez-PQez, G. I., M. J. Blaser, and J. H. Bryner. 1986. Lipopolysaccharide structures of Campylobacter fetus are related to heat-stable serogroups. Infect. Immun. 51:209-212. Ptrez-Ptrez, G. I., J. A. Hopkins, and M. J. Blaser. 1985. Antigenic heterogeneity of lipopolysaccharides from Campylobacter jejuni and Campylobacter fetus. Infect. Immun. 48528-533. Ponka, A., R. Tilvis, J. Helle, and T. U. Kosunen. 1984. Infection with Campylobacter fetus. Scand. J. Infect. Dis. 16:127-128. Pugsley, A. P. 1993. The complete general secretory pathway in gram-negative bacteria. Microbiol. Rev. 5750-108. Raskin, D. M., R. Seshadri, S. U. Pukatzki, and J. J. Mekalanos. 2006. Bacterial genomics and pathogen evolution. Cell 124: 703-714. Ray, K. C., Z.-C. Tu, R Grogono-Thomas, D. G. Newell, S. A. Thompson, and M. J. Blaser. 2000. Campylobacter fetus sap inversion occurs in the absence of RecA function. Infect. Immun. 685663-5667. Rettig, P. J. 1979. Campylobacter infections in human beings. J. Pediatr. 94: 8 55-8 64. Riley, L. W., and M. J. Finch. 1985. Results of the first year of national surveillance of Campylobacter infections in the United States. J. Infect. Dis. 151:956-959. Salama, S. M., M. M. Garcia, and D. E. Taylor. 1992. Differentiation of the subspecies of Campylobacter fetus by genomic sizing. Int. J. Syst. Bacteriol. 42:446-450. Salama, S. M., E. Newnham, N. Chang, and D. E. Taylor. 1995. Genome map of Campylobacter fetus subsp. fetus ATCC 27374. FEMS Microbiol. Lett. 132:239-245. Schurig, G. D., C. E. Hall, K. Burda, L. B. Corbeil, J. R. Duncan, and A. J. Winter. 1973. Persistent genital tract infection with Vibrio fetus intestinalis associated with serotypic alteration of the infecting strain. Am. J. Vet. Res. 34:1399-1403. Sebald, M., and M. Vtron. 1963. Teneur en bases de I’ADN et classification des vibrions. Ann. Inst. Pasteur 105:897-910. Senchenkova, S. N., A. S. Shashkov, Y. A. Knirel, J. J. McGovern, and A. P. Moran. 1997. The 0-specific polysaccharide chain of Campylobacter fetus serotype A lipopolysaccharide is a partially 0-acetylated 1,3-linked alpha-D-mannan. Eur. J. Biochem. 245: 637-64 1. Senchenkova, S. N., A. S. Shashkov, Y. A. Knirel, J. J. McGovern, and A. P. Moran. 1996. The 0-specific polysaccharide chain of Campylobacter fetus serotype B lipopolysaccharide is a Drhamnan terminated with 3-O-methyl-~-rhamnose(D-aC0friOSe). Eur. J. Biochem. 239:434-438. Skirrow, M. B. 1994. Diseases due to Campylobacter, Helicobacter and related bacteria. J. Comp. Pathol. 111:113-149. Smibert, R M. 1978. The genus Campylobacter. Annu. Rev. Microbiol. 32:673-709. Smith, T., and M. Taylor. 1919. Some morphological and biological characters of the spirilla (Vibrio fetus, n. sp.) associated with disease of the fetal membranes in cattle. J. Exp. Med. 30:299311. Taylor, D. E., and A. S. Chau. 1997. Cloning and nucleotide sequence of the gyrA gene from Campylobacter fetus subsp. fetus ATCC 27374 and characterization of ciprofloxacin- resistant
428
BLASER ET AL.
laboratory and clinical isolates. Antimicrob. Agents Chemother. 41:665-671. Taylor, D. E., R. S. Garner, and B. J. Allan. 1983. Characterization of tetracycline resistance plasmids from Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 24: 930-935. Taylor, P. R., W. M. Weinstein, and J. H. Bryner. 1979. Campylobacter fetus infection in human subjects: association with raw milk. Am. J. Med. 66:779-783. Tettelin, H., V. Masignani, M. J. Cieslewicz, C. Donati, D. Medini, N. L. Ward, S. V. Angiuoli, J. Crabtree, A. L. Jones, A. S. Durkin, R T. Deboy, T. M. Davidsen, M. Mora, M. Scarselli, I. Margarit y Ros, J. D. Peterson, C. R. Hauser, J. P. Sundaram, W. C. Nelson, R. Madupu, L. M. Brinkac, R. J. Dodson, M. J. Rosovitz, S. A. Sullivan, S. C. Daugherty, D. H. Haft, J. Selengut, M. L. Gwinn, L. Zhou, N. Zafar, H. Khouri, D. Radune, G. Dimitrov, K. Watkins, K. J. O’Connor, S. Smith, T. R. Utterback, 0. White, C. E. Rubens, G. Grandi, L. C. Madoff, D. L. Kasper, J. L. Telford, M. R. Wessels, R. Rappuoli, and C. M. Fraser. 2005. Genome analysis of multiple pathogenic isolates of Streptococcus agalactiae: implications for the microbial “pangenome.” Proc. Natl. Acad. Sci. USA 102:13950-13955. Thanabalu, T., E. Koronakis, C. Hughes, and V. Koronakis. 1998. Substrate-induced asembly of a contiguous channel for protein export from E. coli: reversible bridging of an inner-membrane translocase to an outer membrane exit pore. EMBOJ. 17:64876496. Thompson, S. A., 0. L. Shedd, K. C. Ray, M. H. Beins, J. P. Jorgensen, and M. J. Blaser. 1998. Campylobacter fetus surface layer proteins are transported by a type I secretion system. J. Bacteriol. 180:6450-6458. Tremblay, C., and C. Gaudreau. 1998. Antimicrobial susceptibility testing of 59 strains of Campylobacter fetus subsp. fetus. Antimicrob. Agents Chemother. 42:1847-1849. Tu, Z., J. Hui, and M. J. Blaser. 2004a. Conservation and diversity of sap homologues and their organization amongst Campylobacter fetus isolates. Infect. Immun. 72:1715-1724. Tu, Z., D. W. Ussery, D. T. Pride, and M. J. Blaser. 2003a. Genomic characteristics of the Campylobacter fetus sap island. Genome Lett. 2:34-40. Tu, Z., G. Zeitlin, J.-P. Gagner, T. Keo, B. Hanna, and M. J. Blaser. 2004b. Campylobacter fetus of reptile origin as a human pathogen. J. Clin. Microbiol. 42:4405-4407. Tu, Z.-C., F. E. Dewhirst, and M. J. Blaser. 2001a. Evidence that the Campylobacter fetus sap island is an ancient genomic constituent with origins before mammals and reptiles diverged. Infect. Immun. 69:2237-2244. Tu, Z. C., W. Eisner, B. N. Kreiswirth, and M. J. Blaser. 2005a. Genetic divergence of Campylobacter fetus strains of mammal and reptile origins. J. Clin. Microbiol. 43:3334-3340. Tu, 2. C., C. Gaudreau, and M. J. Blaser. 2005b. Mechanisms underlying Campylobacter fetus pathogenesis in humans: surface layer protein variation in relapsing infections.J. Infect. Dis. 191: 2082-2089. Tu, Z.-C., K. C. Ray, S. A. Thompson, and M. J. Blaser. 2001b. Campylobacter fetus uses multiple loci for DNA inversion within the 5’ conserved region of sap homologs. J. Bacteriol. 183:66546661.
Tu, 2.-C., T. M. Wassenaar, S. A. Thompson, and M. J. Blaser. 2003b. Structure and genotypic plasticity of the Campylobacter fetus sap locus. Mol. Microbiol. 48:685-698. Tummuru, M. K., and M. J. Blaser. 1992. Characterization of the Campylobacter fetus sapA promoter: evidence that the sapA promoter is deleted in spontaneous mutant strains. ]. Bacteriol. 174: 5916-5922. Tummuru, M. K., and M. J. Blaser. 1993. Rearrangement of sapA homologs with conserved and variable regions in Campylobacter fetus. Proc. Natl. Acad. Sci. USA 90:7265-7269. Ullmann, U., H. Langmaack, and C. Blasius. 1982. Campylobacteriosis in humans caused by subspecies intestinalis and fetus. Six new diseases. Infection lO(Supp1. 2):S64-S66. van Vliet, A. H., M. L. Baillon, C. W. Penn, and J. M. Ketley. 1999. Campylobacter jejuni contains two fur homologs: characterization of iron-responsive regulation of peroxide stress defense genes by the PerR repressor. J. Bacteriol. 181:6371-6376. Velayudhan, J., M. A. Jones, P. A. Barrow, and D. J. Kelly. 2004. L-Serine catabolism via an oxygen-labile L-serine dehydratase is essential for colonization of the avian gut by Campylobacter jejuni. Infect. Immun. 72:260-268. VCron, M., and R. Chatelain. 1973. Taxonomic study of the genus Campylobacter Sebald and Vtron and designation of the neotype strain for the type species Campylobacter fetus (Smith and Taylor) Sebald and Vtron. Int. J. Syst. Bacteriol. 23:122-134. Verresen, L., M. Vrolix, J. Verhaegen, and R. Lins. 1985. Campylobacter bacteraemia. Report of 2 cases and review of the literature. Acta Clin. Belg. 40:99-104. Vinzent, R., J. Dumas, and N. Picard. 1947. Septicemie grave au cours de la grossesse due a un vibrion: avortment consecutif. Bull. Acad. Natl. Med. 131:90-93. Wang, B., E. Kraig, and D. Kolodrubetz. 1998. A new member of the S-layer protein family: characterization of the crs gene from Campylobacter rectus. Infect. Immun. 66: 1521-1526. Wang, B., E. Kraig, and D. Kolodrubetz. 2000. Use of defined mutants to assess the role of the Campylobacter rectus S-layer in bacterium-epithelial cell interactions. Infect. Immun. 68: 1465-1473. Wang, E., M. M. Garcia, M. S. Blake, Z. Pei, and M. J. Blaser. 1993. Shift in S-layer protein expression responsible for antigenic variation in Campylobacter fetus. ]. Bacteriol. 175:49794984. Wang, W. L., and M. J. Blaser. 1986. Detection of pathogenic Campylobacter species in blood culture systems. J. Clin. Microbiol. 23:709-714. Washburn, R. G., N. C. Julian, E. Wang, and M. J. Blaser. 1991. Inhibition of complement by surface array proteins from Campylobacter fetus, abstr. 1328. 31st Intersci. Conf. Antimicrob. Agents Chemother., Chicago, IL. Washio, T., J. Sasayama, and M. Tomita. 1998. Analysis of complete genomes suggests that many prokaryotes do not rely on hairpin formation in transcription termination. Nucleic Acids Res. 265456-5463. Winter, A. J., E. C. McCoy, C. S. Fullmer, K. Burda, and P. J. Bier. 1978. Microcapsule of Campylobacter fetus: chemical and physical characterization. Infect. Immun. 22:963-971. Yang, L. Y., 2. H. Pei, S. Fujimoto, and M. J. Blaser. 1992. Reattachment of surface array proteins to Campylobacter fetus cells. J. Bacteriol. 174:1258-1267. Zechner, E. L., and M. J. Bailey. 2004. The horizontal gene pool. Plasmid 5 167-74.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 24
Development of a Human Vaccine DAVIDR. TRIBBLE, SHAHIDABAQAR,AND STUARTA. THOMPSON
In addition to these clinical observations, microbiologic surveillance in children who live in the developing world has documented a shift in illness-toinfection ratio occurring in children aged 2 and 5 years accompanied by an apparent development of colonization resistance and a shortened excretion period during convalescence (Taylor et al., 1993). An age-related C. jejuni-specific serology progressive increase in all isotypes was observed during first 2 years of life, followed by continued increases in immunoglobulin (Ig) A titers (possibly indicative of frequent exposure and subsequent mucosal immunity), IgG titer decline, and plateau of IgM among age group prevalence surveys. Additional evidence of immune correlates with clinical outcomes derives from studies of breast-fed Mexican infants (<6 months old) (RuizPalacios et al., 1990). Lower incidence of Campylobacter-associated diarrhea occurred in infants whose mothers had colostral Campylobacter-specific secretory IgA antibodies detected in breast milk. Demonstrations that endemic-region-resident children in their first 4 years of life have frequent Campylobacter colonization suggest that these repeated exposures could serve as immunizations against subsequent disease. The frequency and magnitude of exposure in industrialized regions is severalfold less than the experience of developing world children. Observational studies in selected populations originating in lowerincidence regions but experiencing periods of more frequent exposure provide indirect evidence of acquired immunity (Blaser et al., 1983, 1987). This situation has been observed in reduced C. jejuniassociated diarrheal attack rates and increased Campylobacter-specific antibody in chronic raw milk con-
Campylobacter species are among the most common causes of diarrheal disease worldwide. The incidence of disease varies widely, generally falling into one of three epidemiologic scenarios: (i) hyperendemic levels (40,000 per 100,000 children <5 years old) in developing regions (Coker et al., 2002), (ii) endemic levels (20 to 100 per 100,000 population), occurring most commonly as sporadic disease in young adults and infants, in developed countries (Jones et al., 2007; Samuel et al., 2004), and (iii) traveler’s diarrhea (TD) in persons from industrialized countries visiting hyperendemic regions (Blaser, 1997; Mattila, 1994; Riddle et al., 2006; Tribble et al., 2007). Studies of at-risk populations have contributed knowledge regarding Campylobacter-acquired immunity and the viability of developing an effective Campylobacter vaccine. Epidemiologic studies in developing world children have provided evidence of protective immunity through investigations of disease incidence, clinical presentation, stool microbiology, and Campylobacter-specific immunology. Children (typically <2 years old) with Campylobacter-associated illness manifest a less severe clinical course than observed in affected individuals in regions with lower incidence of campylobacteriosis (Taylor, 1992). Patients in industrialized regions who seek care for Campylobacter enteritis more commonly manifest moderate to severe symptoms and dysentery consistent with a nake or semi-immune status. The clinical outcomes do not appear to be related to regional strain differences because traveler populations visiting these areas experience patterns of clinical illness similar to Campylobacter-associated disease in industrialized countries (Tribble et al., 2007).
David R. Tribbie * Preventive Medicine & Biometrics Department, Uniformed Services University of the Health Sciences, 4301 Jones Shahida Baqar Enteric Diseases Program, Naval Medical Research Center, 503 Robert Bridge Rd., Bethesda, M D 20814-5119. Grant Ave., Silver Spring, MD 20910-7500. Stuart A. Thompson Department of Biochemistry and Molecular Biology, Medical College of Georgia 1120 15th St., Augusta, GA 30912-2100.
429
430
TRIBBLE ET AL.
sumers on dairy farms as compared with individuals with first-time raw milk exposure. Expatriate residents in Thailand were observed to have reduced prevalence rates of Campylobacter-associated diarrheal illness related to duration of residency, as follows: residency <1 year, 8 (17%) of 47, versus residency >1 year, 2 (3%) of 58 (Gaudio et al., 1996). In addition to epidemiologic evidence, experimental C. jejuni infection studies in humans documented short-term homologous strain protection from illness at 1 to 2 months after initial infection (Black et al., 1988, 1992). Experimental infection studies at the University of Maryland demonstrated infectioninduced serologic and intestinal antibody responses, higher prechallenge C. jejuni-specific (acid-extracted protein) serologic and jejunal fluid IgA levels in noninfected versus infected subjects, and increased levels of jejunal fluid IgA during rechallenge in subjects who remained well on second exposure. Homologous rechallenge provided complete protection from illness; however, colonization resistance after this second C. jejuni exposure ( l o 9 CFU challenge dose) was observed in only two (29%) of the seven subjects. Our investigative team at the U.S. Naval Medical Research Center (NMRC) and the U.S. Army Medical Research Institute of Infectious Diseases has undertaken studies reassessing the C. jejuni experimental model (D. R. Tribble, unpublished data). These studies provided confirmation of complete homologous protection from Campylobacter-associated illness as seen in the Maryland study. Subjects receiving rechallenge with a lo9 CFU infectious dose approximately 1 month after initial infection were all protected from illness, and 3 8% demonstrated colonization resistance. Observational and experimental studies provide evidence of acquired immunity developing in humans exposed to C. jejuni, lending support for vaccine development. Prevention and control strategies for enteric transmission of Campylobacter species involves a multifaceted approach that includes reducing or avoiding C. jejuni contamination across critical control points along the farm-to-table food chain and providing immunoprophylaxis targeted to at-risk populations.
TARGET POPULATIONS FOR CAMPYLOBAC TER VACCINE
Campylobacter vaccine development strategy must integrate basic science knowledge of pathogenesis and immunity, as well as an appreciation of the antigens and host responses most strongly associated with the development of protective immunity. More practical concerns include the need to optimize vac-
cine production methodology, delivery methods, and regimen selection followed by an assessment of vaccine safety, immunogenicity, and efficacy in target populations. Application of a vaccine also requires risk assessment and feasibility evaluation in at-risk populations early in the development process. Issues to consider include such areas as the multiple serotypes of C. jejuni associated with disease, protective epitopes, immune correlates of protection, duration and cross-serotypic nature of protective responses, and numerous pragmatic considerations related to the target population’s demographics and period of risk, and vaccination method. As previously stated, C. jejuni-associated disease occurs commonly in three settings: children (0 to 4 years old) in the developing world, industrialized nation population (highest rates in infants and young adults), and in naive or semi-immune travelers to developing regions. Several concerns and possible limitations exist regarding C. jejuni vaccine development for children in the developing world. The frequency and magnitude of exposure in this hyperendemic environment, 0.4 infectious episodes annually over the first 4 to 5 years of the child’s life, would require vaccination during the initial 6 months of life in order to decrease the diarrheal disease burden during this vulnerable period (Taylor, 1992). Birth cohort studies in Egypt documented the high incidence and pathogenicity of first Campylobacter infections in young children (Rao et al., 2001). The observation that prior exposure did not confer protection from subsequent symptomatic infection highlights the challenges in applying an effective vaccine strategy for this at-risk population. Additional challenges in this population include frequent and possibly high-level exposures to virulent organisms, multiple strain exposures, age-related differences in immune system development (total IgA levels are approximately 80% of adult levels by 8 years of age) (Cejka et al., 1974), coexistent non-Campylobacter-associated diarrheal disease and/ or malnutrition, and potential difficulties encountered in delivering an oral immunization to young infants. An effective vaccination program during infancy may lead to a modification of the at-risk age range by shifting age at onset of disease to older children who would likely tolerate diarrheal illness with less morbidity and mortality. A vaccine strategy may utilize repeated booster doses in a population with anticipated ongoing exposures, or the repeated exposures themselves may serve to solidify protective immunity, as observed in epidemiologic studies, with less overall risk of serious disease. The major impact of diarrheal disease in industrialized countries such as the United States relates to short-term morbidity and economic burden.
CHAPTER 24
Diarrhea-related mortality typically is limited to populations at the extremes of age, particularly the elderly. C. jejuni has been estimated to be the most common bacterial cause of diarrhea in industrialized nations, although the FoodNet surveillance system has documented a decline in the United States (Samuel et al., 2004). The incidence and the resultant morbidity, however, are not high enough to warrant broad application of a Campylobacter vaccine to the general population. Targeting subpopulations of extremely high-risk (i.e., chronically immunosuppressed patients) may be more reasonable. Naive or semi-immune travelers to Campylobacter hyperendemic regions constitute a moderate to high-risk population amenable to vaccination. This diverse population includes business, leisure, and military travelers as well as expatriates residing in developing countries. The nake or semi-immune traveler leaving their relatively low-risk environment for an area hyperendemic for bacterial enteropathogens has an approximately 40% diarrhea risk (Steffen, 1986). In most TD series in tourists and military personnel, enterotoxigenic Escherichiu coli (ETEC) predominates as the most commonly identified agent, representing between 5 and 40% of cases (Black, 1990; Riddle et al., 2006). Campylobacter jejuni ( 3 to 45%), Shigella species (2 to lo%), and nontyphoidal Salmonella species (2 to 10%) are other commonly identified causative agents, with regional and seasonal variability affecting the relative agentspecific distribution. In certain regions, such as Thailand, Campylobacter has consistently been the most common enteropathogen isolated from TD cases in deployed military populations (Petruccelli et al., 1992; Tribble et al., 2007). “Typical” TD represents a spectrum of illness from a fleeting mild diarrhea without associated symptoms or activity limitation to a serious dehydrating and/or febrile dysentery requiring hospitalization. Most commonly, TD consists of a self-limited diarrheal illness lasting 3 to 5 days (Ericsson, 1998). The mean duration of symptoms without treatment is approximately 4 days (median, 2 days). Duration of illness for C. jejuni may exceed 1 week in up to 20 to 25% of affected individuals, as seen in follow-up studies of outbreaks and sporadic disease (Kapperud et al., 1992; Millson et al., 1991). A study in Finnish travelers to Morocco demonstrated a more severe clinical illness with C. jejuni as compared with ETEC (Mattila, 1994). The Campylobacter-associated cases had increased frequency of diarrheal stools on the follow-up visits (second and third days) as well as more fevers, myalgias, abdominal cramps, and nausea or vomiting. In addition to a relatively more severe clinical course, the management of Campylobacter-associated TD is further
-
DEVELOPMENT OF A HUMAN VACCINE
431
complicated by increasing antimicrobial resistance. Since 1990, ciprofloxacin-resistant Campylobacter has been reported in numerous locations (Ruiz et al., 2007; Tribble et al., 2007). In Thailand from 1990 to the present, ciprofloxacin-resistant C. jejuni and C. coli have increased from 0 to >85% of isolates (Petruccelli et al., 1992; Tribble et al., 2007). Consideration for a Campylobacter vaccine may be in the context of a combined TD vaccine that takes into account other common bacterial enteropathogens such as ETEC and Shigella, or as a single agent vaccine targeted to travelers to high-campylobacterprevalence regions. An optimally effective TD vaccine should contain a Campylobacter component based on C. jejuni’s relative contribution as a major bacterial enteropathogen, its increasing antimicrobial resistance noted against first-line TD empiric antibiotics, and the relatively increased clinical severity of C. jejuni-associated TD as compared with other major TD enteropathogens (primarily ETEC). Surveys among travel medicine practitioners in the United States and the United Kingdom have provided estimates on the level of support and potential application for vaccines to prevent TD, and specifically, Campylobacter-only vaccines (Miller and Saunders, 2007). The survey was limited by low response rate but was able to assess providers providing care in high-volume practices. In general, there was broad acceptance of the utility of a highly effective TD and Carnpylobacter vaccine and providing an approximation of provider perceptions and estimates of potential application.
CAMPYLOBACTER VACCINE DEVELOPMENT CONSIDERATIONS Epidemiological Issues and Strain Diversity Development of a vaccine may be complicated by the tremendous antigenic diversity of the organism and the fact that the protective epitopes are not clearly defined. The most common typing schemes are the Lior system, which includes 108 serotypes, and the Penner system, which has >60 serotypes (Lior et al., 1982; Penner and Hennessy, 1980). The Penner or 0 serotyping system is based on lipopolysaccharide and lipooligosaccharide (LOS) antigens; the serodeterminant of the Lior scheme is a heatlabile antigen and was originally thought to be flagellin. Genetic analyses with site-specific flagellin mutants, however, have indicated that in most serotypes examined, flagellin is not the Lior serodeterminant (Alm et al., 1991). Animal studies have indicated that protection against intestinal colonization in at least one animal model is Lior serotype specific (Abimiku
432
TRIBBLE ET AL.
and Dolby, 1988; Pavlovskis et al., 1991). The lack of specific information on the nature of the Lior serodeterminant and the large numbers of serotypes complicate vaccine development. However, several studies indicate that a limited number of Lior serotypes predominate in different regions of the world. For example, in Thailand, approximately 75% of all isolates belong to one of five Lior serotypes. Our understanding of what constitutes a protective immune response to Campylobacter infection is also fairly rudimentary. As noted above, Campylobacter-specific secretory IgA and serum IgA antibodies correlate with protection against disease. However, all of the studies that support these conclusions used crude, complex antigens such as glycine extracts. There are few data on antigen-specific immune responses. The role of cellular immunity is also poorly defined. As previously discussed, the composite of epidemiological evidence strongly supports the development of effective protective anti-Campylobacter immunity, both disease reduction and reduced intestinal colonization, with increasing age in populations wherein the organisms are prevalent. The repeated demonstration that breast-feeding results in a reduction in symptomatic infections lends further support for specific acquired immunity being the mechanism of resistance (Georges-Courbot et al., 1987; Ruiz-Palacios et al., 1990). However, that antibodies in milk are the protective factor has not been directly demonstrated, and milk contains numerous additional substances with antibacterial capacities (Newburg et al., 2005). Epidemiological studies, at this point, have failed to identify bacterial components that are consistently associated with disease, that are targets of protective immune responses, or that are effectors of protective immunity. Case-control and cohort studies have demonstrated notable rates of carriage of Campylobacter by individuals who are free of disease (Coker et al., 2002). Explanations for isolation of campylobacters from disease-free individuals include the fact that isolated organisms lack pathogenic potential, and the fact that the host has acquired resistance to disease but not colonization. Definitive evidence is not available to confirm either of the hypotheses. Bacterial structures and capacities thought to associate with pathogenesis are present in isolates from both ill and well individuals (Sjogren et al., 1989). Increased levels of antibodies (often measured by using crude antigen preparations) relate to decreased disease; however, definitive identification of antigenic epitopes that are targets of immune protection, immunological mechanisms that confer protection, and the role of cellular immunity are lacking. The durability of these antibodies may be dependent on the endemicity of the agent. In endemic ar-
eas, the level of the acquired immunity is probably maintained as a result of repeated asymptomatic exposure to the same or cross-reactive Campylobacter serotypes. The role of mucosal antibody in protecting from Campylobacter infection is probably more important than serum levels of antigen-specific immunoglobulins. Antiflagellar IgA has been identified in human breast milk and has been shown to protect children from C. jejuni diarrhea (Ruiz-Palacios et al., 1990; Torres and Cruz, 1993). In addition, the clinical trials conducted at Naval Medical Research Center also suggest a protective role of Campylobacterspecific fecal IgA. The role of systemic or mucosal antibodies in protection is predicted from multiple epidemiological and a few volunteer studies. The relative role of human’s cellular immune responses to C. jejuni enteric disease is not known, and only limited data are available on the role of T-cell subset recognition in patients with Guillain-BarrC syndrome (GBS) (Cooper et al., 2002; Hughes et al., 1999). Campylobacterspecific T-cell activation might play important roles as helpers in antibody formation in clearance of C. jejuni and in recall of memory response on reexposure. A sustained antigen-specific, interferon-gammadependent lymphocyte proliferative response was seen in a naturally acquired C. jejuni infection (Baqar et al., 2001). Our unpublished data also support a crucial role of Th-1 type interferon-gamma-dependent cell-mediated immune response in protecting volunteers from a homologous rechallenge with C. jejuni. The composite evidence convincingly argues for the existence of acquired protective immunity to C. jejuni disease that may be achievable by vaccination with appropriately designed vaccines. Safety Concerns
A major consideration in the development of any vaccine is safety. For Campylobacter, this includes symptoms that may be caused by a vaccine in the first 24 to 72 h, as well as the potential for postvaccination sequelae a few weeks after immunization. Like Salmonella, Shigella, and Yersinia, Campylobacter enteritis has been associated with development of a reactive arthritis or arthropathy (RA), as recently reviewed (Pope et al., 2007). This review delineates the range of findings and uncertainties with this observed association in published reports. Outbreak studies have found incidence rates of RA ranging from about 0.7 to 2.6% in subjects with evidence of C. jejuni infection (Bremell et al., 1991; Hannu et al., 2004). On the basis of population survey results, the incidence of RA after Campylobacter infection in Finland has been estimated at 4.3 per 100,000 (Hannu et al.,
CHAPTER 24
2002). A genetic predisposition to acquiring a seronegative spondyloarthropathy after a bacterial enteric infection (approximate risk gradient: Yersinia spp. > Shigella spp. [S. flexneri] > nontyphoidal Salmonella > C. jejuni) has been observed in individuals with the human leukocyte antigen (HLA)-B27 (Lindsay, 1997). An overall estimated 18-fold increased risk of RA exists for HLA-B27-positive as compared with -negative individuals after one of these infections (Lindsay, 1997). The presence of HLA-B27 antigen in Campylobacter-associated RA has been variable, ranging from 14% to as high as 53% (Hannu et al., 2002). A less common but potentially more devastating complication of Campylobacter infection is GBS (Mishu and Blaser, 1993). The association between GBS and Campylobacter infection is supported by epidemiological and experimental data, discussed in chapters 13 and 22. Research supports the hypothesis that the association between C. jejuni and GBS is due to molecular mimicry between the outer LOS cores and human gangliosides (Godschalk et al., 2007; Koga et al., 2005, 2006). It is hypothesized that the Campylobacter epitopes, endogenous sialic acid (NeuNAc) incorporated into outer LOS core, stimulate the human immune system to attack similarappearing NeuNAc-containing gangliosides that are components of neural coverings throughout the human nervous system, thereby leading to the development of GBS (Ang et al., 2004). The impact of vaccination on the incidence of postinfectious sequelae is not clear. Epidemiologic studies to determine whether prior exposure to Campylobacter during asymptomatic colonization will increase or decrease the risk of RA or GBS after a subsequent infection are lacking. An effective vaccine will by definition prevent the clinical syndrome of Campylobacter enteritis but may not prevent colonization. Volunteer studies have clearly shown that colonization without disease can induce both local intestinal and serum antibody responses (Black et al., 1992), which could potentially trigger postinfectious sequelae. However, if illness and intestinal tissue damage are required, then a vaccine that prevents illness would decrease the risk of RA and GBS. If the bacterial structures involved in inducing GBS can be clearly delineated, then it would be possible to construct even live attenuated vaccines that could be provided without a risk of GBS. No bacterial factors have been identified in the pathogenesis of RA. Recent studies also suggest an increased risk of postinfectious irritable bowel syndrome after bacterial enteritis (Connor, 2005). This risk has also been observed with Campylobacter enteritis (Thornley et al., 2001). As stated above for RA and GBS, it is un-
DEVELOPMENT OF A HUMAN VACCINE
433
certain whether postinfectious irritable bowel syndrome is a result of Campylobacter exposure, and if it is, whether there would be any increased risk related to vaccination.
CANDIDATE SELECTION PROCESSES It is certain that there are surface targets (protein, carbohydrate, or glycoprotein) in addition to those tested previously that may be important components of a vaccine against Campylobacter colonization or disease. Several Campylobacter antigens have been studied to varying degrees to assess relationship to virulence and/or potential as a vaccine target. A protective immune response against such targets may result from inhibition of some important bacterial cellular function (e.g., nutrient uptake), disabling the ability of the pathogen to replicate (inhibition of colonization). Alternatively, targets may be virulence factors (e.g., toxins or adhesins), and generation of antibodies against these proteins may inhibit the pathogenic function of these molecules (inhibition of disease). There may also be conserved surface proteins whose functions may not be inhibited but that may simply serve as targets for complement deposition or for opsonizing antibodies, resulting either in complement-mediated cell lysis or opsonophagocystosis. In each of these cases, it is assumed that any of these targets would be found on or secreted from the cell surface such that they would be available on intact, living cells for attack by the host immune system. Although it is tempting to focus on a single, abundant, conserved protein that would induce a broadly protective immune response, it is possible that protective immunity would be better achieved by using a synergistic mixture of immunogenic yet more minor surface proteins, as was recently demonstrated in Neisseria meningitidis (Weynants et al., 2007). Antibodies to carbohydrates, in particular capsular polysaccharides, have proven extremely useful in vaccines (e.g., in Haemophilus infi7uenzae type b, pneumococcal, or meningococcal vaccines), and their potential usefulness in a putative C. jejuni vaccine will be discussed below. Current Knowledge of C. jejuni Outer Membrane Proteins and Secreted Proteins Only a handful of C. jejuni proteins have been experimentally localized to the outer membrane, such as the porins MOMP and Omp50 (Bolla et al., 2000; De et al., 2000; Labesse et al., 2001; Zhang et al., 2000) and cell surface adhesins PEBl (Pei and Blaser, 1993; Pei et al., 1998) and PEB3 (Linton et al.,
434
TRIBBLE ET AL.
2002; Rangarajan et al., 2007). Several of these outer membrane proteins are already known to be variable among C. jejuni strains (Caldwell et al., 1985; Harris et al., 1987; Labesse et al., 2001; Logan et al., 1989; Luneberg et al., 1998; Pawelec et al., 2000), diminishing the probability that they would be good components of a broadly cross-protective vaccine. In addition to outer membrane proteins, C. jejuni also releases a number of proteins extracellularly, and several of these have been demonstrated or hypothesized to have roles in various C. jejuni disease processes. Three such proteins are the subunits of the C. jejuni cytolethal distending toxin, CdtA, CdtB, and CdtC (Hickey et al., 2000; Lara-Tejero and Galin, 2000, 2001), which are both associated with the cell surface and secreted into the extracellular space (Hickey et al., 2000; Lara-Tejero and Galin, 2000). C. jejuni also secretes various proteins through its flagellar export apparatus (Konkel et al., 2004), including Cia proteins (e.g., CiaB) (Konkel et al., 1999) and FlaC (Song et al., 2004). Finally, some highly invasive C. jejuni strains such as 81-176 possess virulence plasmids such as pVir (Bacon et al., 2000, 2002), which encode a type IV secretion apparatus, and either components of the type IV secretion apparatus (Bacon et al., 2000, 2002) or putative effector proteins of this secretion system might also be vaccine candidates. Immune targeting of some of the extracellular proteins of C. jejuni might prevent the development of disease. Identification of Novel C. jejuni Surface Proteins Novel surface proteins may be identified by two primary means: by bioinformatic analysis of sequenced Campylobacter genomes or by direct identification of surface molecules by means of experimental methods such as proteomics. To date, neither of these methods has been explored fully for C. jejuni, although advances in these areas are beginning to be made. Bioinformatics is a powerful tool for identifying surface proteins, in particular because of the rapid increase in the availability of sequenced bacterial genomes. Analysis of 4 fully sequenced (C. jejuni strains NCTC 11168, RM1221, CG8486, and 81-176) and 13 partially sequenced Campylobacter genomes (total of nine C. jejuni and seven non-jejuni Campylobacter species) (Fouts et al., 2005; Gundogdu et al., 2007; Hofreuter et al., 2006; Parkhill et al., 2000; Poly et al., 2007) allows the possibility of not only identifying proteins that are predicted to reside in the Campylobacter cell envelope, but also of providing an initial assessment of the degree to which these proteins
are conserved across multiple C. jejuni strains or Campylobacter species. Outer membrane proteins have certain structural features that facilitate their identification, such as an N-terminal type I1 secretion signal and C-terminal phenylalanine (Struyvt et al., 1991). Despite difficulties with conventional computer programs in predicting p-barrel outer membrane protein structure, recent advances in computer algorithms such as PSORTb (http: //www.psort.org) and others have increased the reliability of predictions (Bagos et al., 2004; Berven et al., 2004; Bigelow et al., 2004; Garrow et al., 2005; Park et al., 2005; Waldispiihl et al., 2006). Thus, there is certainly substantial utility in using these programs for an initial prediction of outer membrane proteins. However, there are an increasing number of examples in bacterial pathogens of highly atypical surface proteins not predicted by bioinformatics (Pancholi and Chhatwal, 2003). For many years, it has been known that proteins with known cytoplasmic functions (enolase, glyceraldehyde-3-phophate dehydrogenase, and a surface dehydrogenase) are found on and possess pathogenesis-related properties on the cell surfaces of group A streptococci (Boel et al., 2005; Jin et al., 2005; Pancholi and Chhatwal, 2003; Pancholi and Fischetti, 1992, 1998). Other atypical bacterial surface proteins with roles in virulence include pneumococcal 6-phosphogluconate dehydrogenase (Daniely et al., 2006), Helicobacter pylori urease (Dunn and Phadnis, 1998; Marcus and Scott, 2001) and catalase (Radcliff et al., 1997; Sabarth et al., 2002), Mycoplasma pneumoniae elongation factor Tu and pyruvate dehydrogenase E1p subunit (Dallo et al., 2002), and enolase, DnaK, peroxiredoxin, and RpiL of pathogenic neisseria (Knaust et al., 2007; Spence and Clark, 2000). Likewise, a study of the cell wall subproteome of Listeria monocytogenes revealed a large number of unexpected proteins (Schaumburg et al., 2004). The presence of such proteins would not be revealed by bioinformatics, but only by direct experimentation. Therefore, the most comprehensive approach for identifying novel cell surface proteins that may be components of a Cumpylobacter vaccine will incorporate both bioinformatics and any of several experimental methodologies for directly characterizing surface proteins. Methods for Directly Detecting and Identifying Novel Outer Membrane Proteins The experimental identification of bacterial outer membrane proteins is not necessarily a straightforward task. Several commonly used purification or enrichment methods for obtaining outer membrane
CHAPTER 24
fractions, such as extraction with the detergent sarkosyl, sucrose density gradient centrifugation, and extraction with acidic glycine, have been used with varying degrees of success. However, it is important to note that each of these methods yields outer membrane protein samples that are contaminated to varying extents with proteins from other cellular compartments (cytoplasm, cytoplasmic membrane, and periplasm) (R. I. Hobb et al., personal communication), and it is important that any identification of outer membrane proteins that uses these methodologies be confirmed by independent techniques such as immunogold electron microscopy. Outer membrane proteins are also notoriously difficult to manipulate, especially in two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels, as a result of their insolubility because of their often high level of hydrophobicity. Nevertheless, outer membrane proteins can be separated in either conventional or two-dimensional gel systems. Once outer membrane protein preparations have been prepared, the proteins contained therein must be identified. Mass spectrometry for the purpose of protein identification is performed in several variations, including MALDI-TOF (matrix-assisted laser desorption/ionization-time of flight) mass spectrometry on protein samples digested with a protease (commonly trypsin), which yields peptide patterns that can be searched against databases of in silico-digested proteins (Wittmann-Liebold et al., 2006). The related MALDI-TOF allows a second dimension of separation and determination of primary acid sequence of peptide fragments (Nesvizhskii, 2007). High throughput and non-gel-based methods such as multidimensional liquid chromatography-mass spectrometry (Josic and Clifton, 2007) can be used for identification of proteins from complex samples such as bacterial outer membrane preparations. A gel-based proteomic approach for the study of C. jejuni outer membrane proteins was undertaken by Prokhorova et al. (2006). A preparation containing surface proteins of C. jejuni strain ML53 was prepared by extraction with acidic glycine, and the proteins in this extract were identified after twodimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis and mass spectrometry. A total of more than 110 proteins were identified by this approach. However, caution should be used in assigning surface localization to all of these proteins because antibody and enzyme activity studies show that glycine extraction of C. jejuni cells releases numerous proteins, not just from the surface but from every cellular compartment (Hobb et al., unpublished data). In fact, several of the proteins identified by Prokhorova et al. clearly have defined roles
DEVELOPMENT OF A HUMAN VACCINE
435
in other cellular compartments, such as translation initiation and elongation factors, ribosomal and periplasmic chaperones, cytochromes, F1 ATPase, and transcriptional regulators. Whether these constitute proteins that are atypical intracellular proteins that are truly found on the surface of intact C. jejuni cells or are simply released from within the cell during glycine extraction remains to be determined. Nevertheless, immunization of mice with some of these proteins showed a degree of protective efficacy against colonization after oral feeding of C. jejuni (Prokhorova et al., 2006). Additional studies evaluating selected proteins demonstrated significant reduction in C. jejuni colonization in a murine oral challenge model after a three-dose (2-week intervals) subcutaneous immunization schedule of the ACE393 and ACE83 candidates (Schrotz-King et al., 2007). ACE393 is reported to be a surface-localized protein, and the authors state this candidate will be investigated in clinical studies (Schrotz-King et al., 2007). There are also methods for identifying proteins that are surface exposed on intact bacteria. Because surface-exposed proteins are thought to be the most desirable vaccine targets, these approaches are the most direct for identifying novel vaccine candidates. Two such methods are labeling surface proteins by iodination and by biotinylation. Although analysis of C. jejuni surface proteins by either of these techniques has not been published, a study of the related Epsilonproteobacterium H. pylori cell surface via biotinylation was performed by Sabarth et al. (2002). Some of the 18 H. pylori surface proteins identified in this work have orthologs in C. jejuni, and it remains to be seen whether use of similar procedures holds promise for the identification of C. jejuni vaccine targets. Animal Models The most notable symptom of C. jejuni infection of humans is diarrhea, but this symptom is rarely seen when animals are challenged with this species. In humans, C. jejuni also causes fever and myalgias, and it may disseminate systemically. In animals, cytokine cascades indicative of febrile-type responses, dissemination from the gut to deep organs, and marked weight loss can follow C. jejuni challenge. Because it is unclear whether Cumpylobacter static or cidal immunity protects only from systemic manifestations or from both systemic and diarrheagenic mechanisms, it is unclear what outcomes from animal models are the best correlates of protective immunity. The possibility that the mechanisms of immunity to Cumpylobacter diarrhea are distinct from immunity to other manifestations of campylobacteriosis should be con-
436
TRIBBLE ET AL.
sidered. There is a lack of consensus on outcomes from animal models that correspond to human protective immunity to Campylobacter. In contrast, animal models are quality platforms for measurement of immunogenicity of candidate vaccines and the effect of adjuvants. The paucity of simple animal models that mimic human gastrointestinal disease has also slowed pathogenesis and vaccine development efforts. Campylobacters naturally colonize numerous animal species, but usually without symptoms. However, experimental infection of relatively young adult Macaca nemestrina with virulent C. jejuni closely mimics human disease with the same strain (Russell et al., 1989, 1990). Adult M. mulatta infected with high doses of a pathogenic strain of C. jejuni (81-176) resulted in only mild diseases in some animals, although all animals were colonized and mounted an aggressive immune response after infection (Islam et al., 2006). However, New World adult monkeys, Aotus nancymae, infected orally with the same strain of C. jejuni at similar doses resulted in infection of all animals with a 83% diarrhea rate, and all animals mounted robust anti-Campylobacter immune responses (Jones et al., 2006). At homologous challenge, protection (no diarrhea and a reduction in colonization duration) was seen (Jones et al., 2006). Data from both challenge studies (Jones et al., 2006; Russell et al., 1989) and epidemiologic studies in primate colonies indicate that prior infection provides protection against illness on rechallenge (Russell et al., 1990). These data suggest that nonhuman primates could serve as a useful model for vaccine testing; however, the expense of primate models has limited their application. Small animal models of diarrheal disease include the ferret, newborn piglets, and a surgical rabbit model (Babakhani et al., 1993; Bell and Manning, 1990; Vitovec et al., 1989; Walker et al., 1992). Both gnotobiotic and newborn colostrum-deprived piglets have been shown to develop diarrhea when challenged orally with C. jejuni (Babakhani et al., 1993; Vitovec et al., 1989). The disease affects mainly the colon and has similar histopathology to that seen in humans. However, the requirement for newborn animals makes this a logistically difficult and expensive model with limited usefulness for testing vaccine efficacy. By means of a surgical method, when removable intestinal tie adult rabbit diarrhea (RITARD) young rabbits were challenged with C. jejuni, most animals became bacteremic and developed diarrhea (Walker et al., 1992). After oral immunization of young rabbits with either live strains or killed whole-cell vaccines (CWC), a Lior-serotype-specific homologous
protection was documented. Thus, immunization with one strain of Campylobacter protects against other strains of the same Lior serogroup, but not strains of other Lior serogroups. However, surgical manipulation and the requirements of young age make this model cumbersome and less useful for vaccine development studies. Unlike most small mammals, young ferrets develop diarrhea when naturally or experimentally infected with C. jejuni via the orogastric route. Adult ferrets that have been intragastrically inoculated with at least some strains of live Campylobacter 81-176 or CG8421, a clinical isolate from Thailand (S. Baqar, unpublished data), develop enteric symptoms lasting up to 3 days, and they remain colonized for up to 8 days. The disease is characterized by mild to moderate diarrhea with mucus, fecal leukocytes, and occasionally frank blood (Bell and Manning, 1990). Ferrets that have been immunized in this manner do not develop disease when rechallenged with the same strain. However, colonization resistance does not develop until after several additional exposures. This is similar to volunteer studies where protection against disease developed before colonization resistance. These characteristics make the ferret an excellent model for studying pathogenesis. Although older than 12 weeks, animals become refractory to diarrhea, yet the model is relatively promising for studying Campylobacter candidate vaccine efficacy with protection against diarrhea used as the criterion, rather than colonization resistance. Several nondiarrheal murine models have also been used in vaccine development. Mice immunized or challenged via the orogastric route have been used to study safety and immunogenicity of oral vaccines and the induction of Campylobacter colonization resistance against oral challenge. A murine model of oral/intranasal immunization followed by intranasal challenge has been adapted from its original use in studies of Shigella immunity (Baqar et al., 1996; Mallett et al., 1993). Naive mice infected intranasally with virulent strains of Campylobacter develop pulmonary disease, accompanied by weight loss and other signs of illness, including ruffled fur, hunched back, dehydration, lethargy, and sometimes death (Baqar et al., 1996). After intranasal infection, most strains colonize the intestine, and some strains disseminate to the liver, spleen, and other organs. In addition to other immunodeficient mouse models newly reported, SCID mice with limited flora (Chang and Miller, 2006), interleukin-10-deficient C57BL/6 mice (Mansfield et al., 2007), and myeloid-differentiation-defective MyD8 8 mice (Watson et al., 2007) have shown pathological lesions similar to what is seen after infection in humans. In these mod-
CHAPTER 24
els, animals have a nonnormal immune system, although pathology at the mucosal site is observed. However, absence of diarrhea and abnormal immune systems make these models hard to use for vaccine studies. The oral/intranasal immunization followed by intranasal challenge appears to be useful in measuring the immune response and protective efficacy as measured by protection against pulmonary disease, intestinal colonization, and disseminated disease. However, correlation to protection in humans has not been established.
-
DEVELOPMENT OF A HUMAN VACCINE
437
a recA mutation (Guerry et al., 1994) would preclude reversion to wild type. However, in addition to the attenuation of the organism, more must be learned about the mechanisms by which Campylobacter is able to induce RA and GBS before a live attenuated strain would be safe. Another option related to attenuated oral vaccines is the use of another bacteria to vector Cumpylobucter antigens. Preclinical studies have been reported that use attenuated Salmonella enterica serovar Typhimurium to deliver PEBl in a murine model (Sizemore et al., 2006). A two-dose immunization schedule induced anti-PEB serologic responses; however, no protection (measured as colonization level) was observed in an oral challenge.
VACCINE APPROACHES For a vaccine to be effective against an enteric pathogen, most experts believe it must be able to stimulate intestinal immunity (Brandtzaeg, 2007; Holmgren and Czerkinsky, 2005; Holmgren et al., 1992). For this reason, the oral route of immunization has most frequently been selected to evaluate vaccines against diarrheal pathogens and was also selected in the initial evaluation of the first candidate Campylobacter vaccine (killed whole cells). Live Attenuated Vaccines Several live, attenuated oral vaccines have been shown to effectively stimulate mucosal immunity and provide excellent protection in field or volunteer challenge studies. These include vaccines attenuated by serial passage, oral polio (Sabin, 1965), chemical mutagenesis (Ty21A vaccine strain of Salmonella enterica serovar Typhi) (Black et al., 1990), or targeted deletion of genes encoding for important virulence factors (cholera toxin A subunit-deleted cholera vaccine CVD 103HgR) (Kaper et al., 1984). A genetic approach to develop a living attenuated Campylobacter vaccine would allow the inclusion of the full complement of antigens. Challenge studies have clearly shown that infection with a wild-type strain produces solid protective immunity in volunteers, so it is reasonable to expect that a live attenuated vaccine could produce similar results. The paucity of information about pathogenesis and physiology of the organism complicate this approach. In addition, the association between Campylobacter infection and the GBS makes the development of a safe, live attenuated vaccine difficult. GBS appears to be associated with specific LOS serotypes, so it may be possible to select or engineer a strain without the genetic machinery (or capacity to acquire genetic elements) to lead to molecular mimicry. Unlike other enteric pathogens, C. jejuni is naturally transformable. The inclusion of
Killed CWC Inactivated microorganisms offer several advantages as potential vaccines for mucosal immunization (Walker, 2005). Physically, they are natural microparticles, which should enhance interactions between their surface and mucosal lymphoid tissues. As vaccines, they are inexpensive to produce and possess multiple antigens that can be important for protection. Formulations can be modified to offer protection against prevalent serotypes, as is done for influenza. Presentation of multiple antigens may be particularly important for pathogens like Campylobacter for which protective antigens are not known or not available in recombinant forms. Although less appealing for parenteral administration, these wholecell preparations are generally safe for mucosal immunization. Killed whole-cell oral vaccines have been developed and marketed outside of the United States. The best studied is an oral inactivated cholera vaccine composed of heat- and formalin-killed Vibrio cholerae whole cells, of different biotypes and serotypes, plus the nontoxic B subunit (BS-B subunit purified from V. cholerae; rBS-recombinant BS) of cholera toxin with demonstrated efficacy in Bangladesh (Clemens et al., 1990), Peruvian military recruits (Sanchez et al., 1994), and Finnish travelers (Peltola et al., 1991). Although not as immunogenic as live attenuated V. cholerae strains (Ryan et al., 2006), these vaccines have proven the concept that an orally administered CWC can be safe and can provide reasonable protection against an enteric pathogen. The Enteric Diseases Program at NMRC has studied the hypothesis that a killed CWC against Campylobacter could be safe and immunogenic and can protect against disease, particularly if combined with an effective mucosal adjuvant such as E. coli heat-labile toxin (LT). Initial studies that used C. coli VC167 showed that when sonicated cells were com-
438
TRIBBLE ET AL.
bined with 25 pg of LT and provided orally to mice, the mucosal immune response was equal to live infection with the same strain (Rollwagen et al., 1993). The duration of intestinal colonization after live challenge could also be significantly shortened in mice or rabbits immunized with sonicates plus LT, but not by immunization with sonicates alone. A mixture of heat- and formalin-killed C. jejuni 8 1-176 coadministered with LT in mice demonstrated enhanced mucosal immune responses over a wide range of vaccine doses (Baqar et al., 1995a). When challenged orally with live C. jejuni 81-176, mice immunized with either lo5, lo7, or lo9 whole cells plus LT showed colonization resistance, whereas only the highest dose of CWC alone (lo9 cells) gave comparable protection. Both vaccine formulations provided similar levels of protection against systemic spread of challenge organisms. In follow-up experiments, the relative protective efficacy and duration of immunity induced by CWC or CWC-LT vaccines were compared in an oral immunization-intranasal challenge model (Baqar et al., 1996). Both formulations provided varying degrees of protection against illness and intestinal colonization for up to 4 months after the two-dose primary immunization series (14-day intervals). However, the adjuvant preparation appeared superior to formalin-inactivated whole cells alone in that immunity in this vaccination group was still evident in mice at 8 months after immunization. In vitro Tcell proliferative responses to C. jejuni antigens were also measurably enhanced by the addition of LT. Two doses of orally administered CWC vaccine with or without LT were well tolerated in rhesus monkeys (Baqar et al., 1995b). High numbers of Campylobacter-specific IgA and IgG antibody-secreting cells (ASCs) were detected in the peripheral blood of most animals after vaccination, but the IgA ASC response was significantly enhanced with coadministration of LT in a dose-dependent manner. LT also significantly enhanced the serum Campylobacter-specific IgA and IgG responses. These results suggest that both killed whole cells alone and LT adjuvant preparations are promising oral Campylobacter vaccine candidates. The addition of LT in differing animal models enhances the mucosal and serum immune response, and Thl-type cellmediated immune responses to Campylobacter antigens enhance the protective efficacy of the CWC at lower vaccine doses and prolong the duration of immunity compared with CWC alone. The disadvantage of LT is enterotoxicity. The immune modulating effects of LT are seen when CWC is combined with the LT(R192G), which has a mutation in the A subunit cleavage site that makes it less toxic in mice and
humans (D. R. Tribble et al., unpublished data; Dickinson and Clements, 1995). A monovalent, formalin-inactivated CWC made from C. jejuni 81-176 (Penner serotype 23/36; Lior serogroup 5 ) was investigated in clinical studies (D. R. Tribble et al., personal communication). This strain is invasive in cell culture (Oelschlaeger et al., 1993); it was originally isolated from the feces of a 9-year-old girl with diarrhea who became ill during a milk-borne outbreak in Minnesota (Korlath et al., 1985). The strain was used in a human volunteer study at the Center for Vaccine Development, University of Maryland (Black et al., 1988). The organism used to make the vaccine was recovered from a challenged volunteer with diarrhea (prepared by the Walter Reed Army Institute of Research, Department of Biologics Research, and Antex Biologics, Gaithersburg, MD). The cells used for vaccine preparation were grown under conditions that attempted to maximize motility, flagella expression, and ability to invade eukaryotic cells in vitro and were inactivated by 0.2% formalin. The final preparation was shown to have intact flagella by electron microscopy and to retain its ability to agglutinate in Lior 5 antisera. Clinical studies evaluated the safety and immunogenicity of two oral doses (days 0, 14) of CWC (lo5, lo', lo9, or loll vaccine particles) coadministered with LT(R192G), a mucosal adjuvant genetically modified E. coli LT (25 pg). The two-dose oral vaccination regimen of CWC (2.5 X lo9 bacterial cells plus 25 pg of LTLT(R192G)) was not sufficiently immunogenic to protect volunteers from illness in an 81-176 homologous challenge. On the basis of the lack of efficacy in the challenge study, a four-dose oral regimen (days 0 , 2 , 4 , 6 ) of CWC (10" vaccine particles) provided in combination with 25 pg of LT(R192G) (n = 15) was assessed after the efficacious serial short-interval regimen used for the licensed oral typhoid vaccine, Ty2la (Ferreccio et al., 1989; Kantele, 1991). Ferret Campylobacter infection animal model studies provided preclinical support for enhanced protection by using the shortinterval CWC regimen (Burr et al., 2005). Regimens delivering two (14-day interval) or four (days 0, 2,4, 6) oral doses of a killed C. jejuni CWC coadministered with LT(R192G) were associated with mild to moderate diarrhea in approximately 20% of vaccinees. Both regimens induced CWC-specific IgA antibody secreting cells (IgA-ASC) and in vitro interferon-gamma production; however, fecal IgA (sIgA), deemed important in protection, was only induced by the four-dose regimen (mean 13-fold increase from baseline titers). It is not known whether the enhanced immune responses would be adequate to provide protection against disease.
CHAPTER 24
Subunit Vaccines
A subunit vaccine would likely consist of a recombinant protein given alone, in combination with an adjuvant, or expressed in a carrier vaccine strain such as live attenuated Salmonella, Shigella, or Vibrio. Such a vaccine would pose significantly less risk of postinfectious sequelae than a live attenuated vaccine. Purified recombinant proteins could be delivered to the mucosa by oral or intranasal administration, or provided parenterally. Two Carnpylobacter antigens, flagellin and PEB1, have been suggested as potential subunit vaccine candidates. Another subunit candidate vaccine, ACE393 , discussed above, has entered clinical testing (Schrotz-King et al., 2007). Campylobacter vaccine development at the US. Naval Medical Research Center recently completed clinical studies of subunit Campylobacter vaccine based on the flagellin protein. Flagellin is the immunodominant antigen recognized during infection (Blaser and Duncan, 1984; Nachamkin and Hart, 1985), and development of antibodies against flagellin correlates with development of protection against disease (Martin et al., 1989). The structure of Campylobacter flagellin contains both highly conserved and highly variable regions (Power et al., 1994), in addition to glycosyl posttranslational modifications (Doig et al., 1996; Guerry et al., 1996; Szymanski et al., 1999). NMRC researchers have developed a recombinant subunit vaccine composed of truncated FlaA flagellin of C. coli VC167 fused to maltose binding protein of Escherichia coli (Lee et al., 1999). The region of the flaA gene encoding amino acids 5 to 337 was chosen because of its high conservation among different strains of Campylobacter, and because this region has been demonstrated to be immunogenic. The recombinant truncated flagellin protein (rFla-MBP) has demonstrated cross-species protection in murine models of C. jejuni pathogenesis by using intranasal rFla-MBP vaccination (Lee et al., 1999). Intranasal rFla-MBP vaccination induces higher levels of immune response in intestinal IgA and serology (IgA, IgG) than when administered orally. In addition, a ferret diarrhea model has been used to study safety, immunogenicity, and efficacy of the intranasal rFlaMBP vaccine. Intranasal immunization with a threedose regimen of rFla-MBP (100 p g ) provided an approximately 60% protective efficacy from diarrheal illness after oral challenge with 1 X lo1' CFU C. jejuni strain 81-176. A dose escalation phase I trial was conducted to evaluate safety and immunogenicity of the intranasal rFla-MBP vaccine in adult volunteers (D. R. Tribble et al., personal communication). Doses studied
DEVELOPMENT OF A HUMAN VACCINE
439
ranged from 25 to 1,000 p g with vaccinees receiving three intranasal doses (via pipette delivery) at 14-day intervals. No adjuvant was used. The vaccine was generally well tolerated, and no vaccine-associated serious adverse effects occurred. Mild transient local symptoms such as rhinorrhea were reported, with no apparent relationship between increasing vaccine dose with the exception of transient nasal burning (
440
TRIBBLE ET AL.
sequences. Continued efforts are needed to promote appropriate candidate selection, to investigate correlates of immunity, and to continue the assessment of subunit vaccines currently in development. REFERENCES Abimiku, A. G., and J. M. Dolby. 1988. Cross-protection of infant mice against intestinal colonisation by Campylobacter jejuni: importance of heat-labile serotyping (Lior) antigens. J. Med. Microbiol. 26:265-268. Alm, R. A., P. Guerry, M. E. Power, H. Lior, and T. J. Trust. 1991. Analysis of the role of flagella in the heat-labile Lior serotyping scheme of thermophilic campylobacters by mutant allele exchange. J. Clin. Microbiol. 29:2438-2445. Ang, C. W., B. C. Jacobs, and J. D. Laman. 2004. The GuillainBarre syndrome: a true case of molecular mimicry. Trends Immunol. 25:61-66. Babakhani, F. K., G. A. Bradley, and L. A. Joens. 1993. Newborn piglet model for campylobacteriosis. Infect. lmmun. 61:34663475. Bacon, D. J., R. A. Alm, D. H. Burr, L. Hu, D. J. Kopecko, C. P. Ewing, T. J. Trust, and P. Guerry. 2000. Involvement of a plasmid in virulence of Campylobacter jejuni 81-176. Infect. lmmun. 68:4384-4390. Bacon, D. J., R. A. Alm, L. Hu, T. E. Hickey, C. P. Ewing, R. A. Batchelor, T. J. Trust, and P. Guerry. 2002. DNA sequence and mutational analyses of the pVir plasmid of Campylobacter jejuni 81-176. Infect. Immun. 70:6242-6250. Bacon, D. J., C. M. Szymanski, D. H. Burr, R. P. Silver, R. A. A h , and P. Guerry. 2001. A phase-variable capsule is involved in virulence of Campylobacter jejuni 8 1-176. Mol. Microbiol. 40:769-777. Bagos, P. G., T. D. Liakopoulos, I. C. Spyropoulos, and S. J. Hamodrakas. 2004. PRED-TMBB: a web server for predicting the topology of beta-barrel outer membrane proteins. Nucleic Acids Res. 32:W400-W404. Baqar, S., L. A. Applebee, and A. L. Bourgeois. 1995a. Immunogenicity and protective efficacy of a prototype Campylobacter killed whole-cell vaccine in mice. Infect. lmmun. 63:373 1-3735. Baqar, S., L. Applebee, T. Gilliland, P. Guerry, and M. A. Monteiro. 2007a. Immunogenicity of a Campylobacter jejuni capsular polysaccharide conjugate vaccine in mice. Abstract. 107th Am. SOC.Microbiol. Baqar, S., L. Applebee, T. Gilliland, C. Lin, P. Guerry, and M. A. Monteiro. 2007b. The synthesis and efficacy in mice of a Campylobacter jejuni capsular polysaccharide. Abstract. 4th Int. Conf. Vaccines Enteric Dis. Baqar, S., A. L. Bourgeois, L. A. Applebee, A. S. Mourad, M. T. Kleinosky, Z. Mohran, and J. R. Murphy. 1996. Murine intranasal challenge model for the study of Campylobacter pathogenesis and immunity. Infect. lmmun. 64:4933-4939. Baqar, S., A. L. Bourgeois, P. J. Schultheiss, R. I. Walker, D. M. Rollins, R. L. Haberberger, and 0. R. Pavlovskis. 1995b. Safety and immunogenicity of a prototype oral whole-cell killed Campylobacter vaccine administered with a mucosal adjuvant in nonhuman primates. Vaccine 13:22-28. Baqar, S., J. Lapa, M. Nelson, C. Lin, C. Porter, C. Williams, T. Gilliland, R. Arora, J. Saunders, P. Guerry, J. Sanders, and D. Tribble. 2007c. Safety and immunogenicity of a recombinant Campylobacter flagellin vaccine in adult volunteers. In Vaccine for Enteric Diseases. Lisbon Portugal. Baqar, S., B. Rice, L. Lee, A. L. Bourgeois, A. N. El Din, D. R. Tribble, G. P. Heresi, A. S. Mourad, and J. R. Murphy. 2001. Campylobacter jejuni enteritis. Clin. Infect. Dis. 33:901-905.
Bell, J. A., and D. D. Manning. 1990. A domestic ferret model of immunity to Campylobacter jejuni-induced enteric disease. lnfect. lmmun. 58:1848-1852. Berven, F. S., K. Flikka, H. B. Jensen, and I. Eidhammer. 2004. BOMP: a program to predict integral beta-barrel outer membrane proteins encoded within genomes of gram-negative bacteria. Nucleic Acids Res. 32:W394-W399. Bigelow, H. R., D. S. Petrey, J. Liu, D. Przybylski, and B. Rost. 2004. Predicting transmembrane beta-barrels in proteomes. Nucleic Acids Res. 32:2566-2577. Black, R., D. Perlman, M. Clements, M. Levine, and M. Blaser. 1992. Human volunteer studies with Campylobacter jejuni, p. 207-215. In I. Nachamkin, M. Blaser, and L. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Black, R. E. 1990. Epidemiology of travelers’ diarrhea and relative importance of various pathogens. Rev. Infect. Dis. 12(Suppl. 1): s73-s79. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Black, R. E., M. M. Levine, C. Ferreccio, M. L. Clements, C. Lanata, J. Rooney, and R. Germanier. 1990. Efficacy of one or two doses of Ty2la Salmonella typhi vaccine in enteric-coated capsules in a controlled field trial. Chilean Typhoid Committee. Vaccine 891-84. Blaser, M. J. 1997. Epidemiologic and clinical features of Campylobacter jejuni infections. J. Infect. Dis. 176(Suppl. 2):S103S105. Blaser, M. J., and D. J. Duncan. 1984. Human serum antibody response to Campylobacter jejuni infection as measured in an enzyme-linked immunosorbent assay. Infect. lmmun. 44:292298. Blaser, M. J., D. J. Duncan, M. T. Osterholm, G. R. Istre, and W. L. Wang. 1983. Serologic study of two clusters of infection due to Campylobacter jejuni. 1.Infect. Dis. 1475320-823. Blaser, M. J., E. Sazie, and L. P. Williams, Jr. 1987. The influence of immunity on raw milk-associated Campylobacter infection. JAMA 257~43-46. Boel, G., H. Jin, and V. Pancholi. 2005. Inhibition of cell surface export of group A streptococcal anchorless surface dehydrogenase affects bacterial adherence and antiphagocytic properties. lnfect. lmmun. 73~6237-6248. Bolla, J. M., E. De, A. Dorez, and J. M. Pages. 2000. Purification, characterization and sequence analysis of Omp50, a new porin isolated from Campylobacter jejuni. Biochem. J. 352~637-643. Brandtzaeg, P. 2007. Induction of secretory immunity and memory at mucosal surfaces. Vaccine 255467-5484. Bremell, T., A. Bjelle, and A. Svedhem. 1991. Rheumatic symptoms following an outbreak of Campylobacter enteritis: a five year follow up. Ann. Rheum. Dis. 50:934-938. Burr, D. H., D. Rollins, L. H. Lee, D. L. Pattarini, S. S. Walz, J. H.Tian, J. Pace, A. L. Bourgeois, and R. I. Walker. 2005. Prevention of disease in ferrets fed an inactivated whole cell Campylobacter jejuni vaccine. Vaccine 23:43 15-4321. Caldwell, M. B., P. Guerry, E. C. Lee, J. P. Burans, and R. I. Walker. 1985. Reversible expression of flagella in Campylobacter jejuni. Infect. lmmun. 50:941-943. Cejka, J., D. W. Mood, and C. S. Kim. 1974. Immunoglobulin concentrations in sera of normal children: quantitation against an international reference preparation. Clin. Chem. 20:656659. Chang, C., and J. F. Miller. 2006. Campylobacter jejuni colonization of mice with limited enteric flora. Infect. Immun. 74: 5261-5271.
CHAPTER 24
Clemens, J. D., D. A. Sack, J. R. Harris, F. van Loon, J. Chakraborty, F. Ahmed, M. R. Rao, M. R Khan, M. Yunus, N. Huda, B. F. Stanton, B. A. Kay, S. Walter, R. Eeckels, A.-M. Svennerholm, and J. Holmgren. 1990. Field trial of oral cholera vaccines in Bangladesh: results from three-year follow-up. Lancet 335: 270-274. Cohen, D., S. Ashkenazi, M. S. Green, M. Gdalevich, G. Robin, R. Slepon, M. Yavzori, N. Orr, C. Block, I. Ashkenazi, J. Shemer, D. N. Taylor, T. L. Hale, J. C. Sadoff, D. Pavliakova, R. Schneerson, and J. B. Robbins. 1997. Double-blind vaccinecontrolled randomised efficacy trial of an investigational Shigella sonnei conjugate vaccine in young adults. Lancet 349:155-159. Coker, A. O., R. D. Isokpehi, B. N. Thomas, K. 0. Amisu, and C. L. Obi. 2002. Human campylobacteriosis in developing countries. Emerg. Infect. Dis. 8:237-244. Connor, B. A. 2005. Sequelae of traveler’s diarrhea: focus on postinfectious irritable bowel syndrome. Clin. Infect. Dis. 41(Suppl. 8 ):S577-S5 8 6. Cooper, J. C., S. Hughes, A. Ben-Smith, C. 0. Savage, and J. B. Winer. 2002. T cell recognition of a non-protein antigen preparation of Campylobacter jejuni in patients with Guillain-BarrC syndrome. J. Neurol. Neurosurg. Psychiatry 72:413-414. Dallo, S. F., T. R. Kannan, M. W. Blaylock, and J. B. Baseman. 2002. Elongation factor Tu and E l b subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol. Microbiol. 46:1041-1051. Daniely, D., M. Portnoi, M. Shagan, A. Porgador, N. Givon-Lavi, E. Ling, R. Dagan, and Y. Mizrachi Nebenzahl. 2006. Pneumococcal 6-phosphogluconate-dehydrogenase, a putative adhesin, induces protective immune response in mice. Clin. Exp. lmmunol. 144:254-263. De, E., M. Jullien, G. Labesse, J. M. Pages, G. Molle, and J. M. Bolla. 2000. MOMP (major outer membrane protein) of Campylobacter jejuni; a versatile pore-forming protein. FEBS Lett. 469:93-97. Dickinson, B. L., and J. D. Clements. 1995. Dissociation of Escherichia coli heat-labile enterotoxin adjuvanticity from ADPribosyltransferase activity. Infect. Immun. 63:1617-1623. Doig, P., N. Kinsella, P. Guerry, and T. J. Trust. 1996. Characterization of a post-translational modification of Campylobacter flagellin: identification of a sero-specific glycosyl moiety. Mol. Microbiol. 19:3 79-3 8 7. Dunn, B. E., and S. H. Phadnis. 1998. Structure, function and localization of Helicobacter pylori urease. Yale J. Biol. Med. 71: 63-73. Ericsson, C. D. 1998. Travelers’ diarrhea. Epidemiology, prevention, and self-treatment. Infect. Dis. Clin. North Am. 12:285303. Ferreccio, C., M. M. Levine, H. Rodriguez, and R. Contreras. 1989. Comparative efficacy of two, three, or four doses of TY2la live oral typhoid vaccine in enteric-coated capsules: a field trial in an endemic area. J. Infect. Dis. 159:766-769. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R J. Dodson, A. S . Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C . Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Garrow, A. G., A. Agnew, and D. R. Westhead. 2005. TMB-Hunt: an amino acid composition based method to screen proteomes for beta-barrel transmembrane proteins. BMC Bioinformatics 6: 56. Gaudio, P. A., P. Echeverria, C. W. Hoge, C . Pitarangsi, and P. Goff. 1996. Diarrhea among expatriate residents in Thailand:
DEVELOPMENT OF A HUMAN VACCINE
441
correlation between reduced Campylobacter prevalence and longer duration of stay.]. Travel Med. 3:77-79. Georges-Courbot, M. C., A. M. Beraud-Cassel, I. Gouandjika, and A. J. Georges. 1987. Prospective study of enteric Campylobacter infections in children from birth to 6 months in the Central African Republic. J. Clin. Microbiol. 25:836-839. Godschalk, P. C., M. L. Kuijf, J. Li, F. St Michael, C. W. Ang, B. C. Jacobs, M. F. Karwaski, D. Brochu, A. Moterassed, H. P. Endtz, A. van Belkum, and M. Gilbert. 2007. Structural characterization of Campylobacter jejuni lipooligosaccharide outer cores associated with Guillain-Bard and Miller Fisher syndromes. Infect. Immun. 75:1245-1254. Guerry, P., P. Doig, R. A. Alm, D. H. Burr, N. Kinsella, and T. J. Trust. 1996. Identification and characterization of genes required for post-translational modification of Campylobacter coli VC167 flagellin. Mol. Microbiol. 19:369-378. Guerry, P., P. M. Pope, D. H. Burr, J. Leifer, S. W. Joseph, and A. L. Bourgeois. 1994. Development and characterization of recA mutants of Campylobacter jejuni for inclusion in attenuated vaccines. Infect. lmmun. 62:426-432. Gundogdu, O., S. D. Bentley, M. T. Holden, J. Parkhill, N. Dorrell, and B. W. Wren. 2007. Re-annotation and re-analysis of the Campylobacter jejuni N C T C l l l 6 8 genome sequence. BMC Genomics 8:162. Hannu, T., M. Kauppi, M. Tuomala, I. Laaksonen, P. Klemets, and M. Kuusi. 2004. Reactive arthritis following an outbreak of Campylobacter jejuni infection. J. Rbeumatol. 31528-530. Hannu, T., L. Mattila, H. Rautelin, P. Pelkonen, P. Lahdenne, A. Siitonen, and M. Leirisalo-Repo. 2002. Campylobacter-triggered reactive arthritis: a population-based study. Rbeumatology (Oxford) 41:312-318. Harris, L. A., S. M. Logan, P. Guerry, and T. J. Trust. 1987. Antigenic variation of Campylobacter flagella. J. Bacteriol. 169: 5066-5071. Hickey, T. E., A. L. McVeigh, D. A. Scott, R. E. Michielutti, A. Bixby, S. Carroll, A. L. Bourgeois, and P. Guerry. 2000. Campylobacter jejuni cytolethal distending toxin mediates release of interleukin-8 from intestinal epithelial cells. Infect. lmmun. 68: 6535-6541. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galin. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. lmmun. 74:4694-4707. Holmgren, J., and C. Czerkinsky. 2005. Mucosal immunity and vaccines. Nut. Med. ll:S45-S53. Holmgren, J., C. Czerkinsky, N. Lycke, and A. M. Svennerholm. 1992. Mucosal immunity: implications for vaccine development. Immunobiology 184:157-179. Hughes, R. A., R D. Hadden, N. A. Gregson, and K. J. Smith. 1999. Pathogenesis of Guillain-Barre syndrome. J. Neuroimmunol. 100:74-97. Islam, D., M. D. Lewis, A. Srijan, L. Bodhidatta, A. Aksomboon, M. Gettayacamin, S. Baqar, D. Scott, and C. J. Mason. 2006. Establishment of a lion-human primate Campylobacter disease model for the pre-clinical evaluation of Campylobacter vaccine formulations. Vaccine 24:3762-3771. Jin, H., Y. P. Song, G. Boel, J. Kochar, and V. Pancholi. 2005. Group A streptococcal surface GAPDH, SDH, recognizes uPAW CD87 as its receptor on the human pharyngeal cell and mediates bacterial adherence to host cells. J. Mol. Biol. 350:27-41. Jones, F. R., S. Baqar, A. Gozalo, G. Nunez, N. Espinoza, S. M. Reyes, M. Salazar, R. Meza, C. K. Porter, and S. E. Walz. 2006. New World monkey Aotus nancymae as a model for Campylobacter jejuni infection and immunity. Infect. Immun. 74:790793.
442
TRIBBLE ET AL.
Jones, T. F., E. Scallan, and F. J. Angulo. 2007. FoodNet: overview of a decade of achievement. Foodborne Pathog. Dis. 4:6066. Josic, D., and J. G. Clifton. 2007. Use of monolithic supports in proteomics technology. J. Chromutogr. A 1144:2-13. Kantele, A. 1991. Immune response to prolonged intestinal exposure to antigen. Scand. Immunol. 33:225-229. Kaper, J. B., H. Lockman, M. M. Baldini, and M. M. Levine. 1984. Recombinant nontoxinogenic Vibrio cholerae strains as attenuated cholera vaccine candidates. Nature 308:655-658. Kapperud, G., J. Lassen, S . M. Ostroff, and S . Aasen. 1992. Clinical features of sporadic Campylobacter infections in Norway. Scund. J. Infect. Dis. 24:741-749. Karlyshev, A. V., J. M. Ketley, and B. W. Wren. 2005. The Cumpylobacter jejuni glycome. FEMS Microbiol. Rev. 29:377-390. Karlyshev, A. V., D. Linton, N. A. Gregson, A. J. Lastovica, and B. W. Wren. 2000. Genetic and biochemical evidence of a Campylobacter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 35529-541. Knaust, A., M. V. Weber, S . Hammerschmidt, S. Bergmann, M. Frosch, and 0. Kurzai. 2007. Cytosolic proteins contribute to surface plasminogen recruitment of Neisseria meningitidis. J. Bacteriol. 189:3246-3255. Koga, M., M. Gilbert, M. Takahashi, J. Li, S . Koike, K. Hirata, and N. Yuki. 2006. Comprehensive analysis of bacterial risk factors for the development of Guillain-Bard syndrome after Campylobacter jejuni enteritis. J. Infect. Dis. 193547-555. Koga, M., M. Takahashi, M. Masuda, K. Hirata, and N. Yuki. 2005. Campylobacter gene polymorphism as a determinant of clinical features of Guillain-BarrC syndrome. Neurology 65: 1376-1381. Konkel, M. E., B. J. Kim, V. Rivera-Amill, and S . G. Garvis. 1999. Bacterial secreted proteins are required for the internalization of Cumpylobacter jejuni into cultured mammalian cells. Mol. Microbiol. 32:69 1-70 l. Konkel, M. E., J. D. Klena, V. Rivera-Amill, M. R. Monteville, D. Biswas, B. Raphael, and J. Mickelson. 2004. Secretion of virulence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. J. Bucteriol. 186:32963303. Korlath, J. A., M. T. Osterholm, L. A. Judy, J. C. Forfang, and R. A. Robinson. 1985. A point-source outbreak of campylobacteriosis associated with consumption of raw milk. J. Infect. Dis. 152592-596. Labesse, G., E. Garnotel, S. Bonnel, C. Dumas, J. M. Pages, and J. M. Bolla. 2001. MOMP, a divergent porin from Campylobacter: cloning and primary structural characterization. Biochim. Biophys. Res. Commun. 280:380-387. Lara-Tejero, M., and J. E. Galln. 2000. A bacterial toxin that controls cell cycle progression as a deoxyribonuclease I-like protein. Science 290:354-357. Lara-Tejero, M., and J. E. Galln. 2001. CdtA, CdtB, and CdtC form a tripartite complex that is required for cytolethal distending toxin activity. Infect. Immun. 69:4358-4365. Lee, L. H., E. Burg 111, S. Baqar, A. L. Bourgeois, D. H. Burr, C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evaluation of a truncated recombinant flagellin subunit vaccine against Campylobacter jejuni. Infect. Immun. 675799-5805. Lindsay, J. A. 1997. Chronic sequelae of foodborne disease. Emerg. Infect. Dis. 3:443-452. Linton, D., E. Allan, A. V. Karlyshev, A. D. Cronshaw, and B. W. Wren. 2002. Identification of N-acetylgalactosamine-containing glycoproteins PEB3 and CgpA in Campylobacter jejuni. Mol. Microbiol. 43:497-508. Lior, H., D. L. Woodward, J. A. Edgar, L. J. Laroche, and P. Gill. 1982. Serotyping of Cumpylobacter jejuni by slide agglutination
].
based on heat-labile antigenic factors. J. Clin. Microbiol. 15: 761-768. Logan, S . M., P. Guerry, D. M. Rollins, D. H. Burr, and T. J. Trust. 1989. In vivo antigenic variation of Campylobacter flagellin. Infect. Immun. 57:2583-2585. Liineberg, E., E. Glenn-Calvo, M. Hartmann, W. Bar, and M. Frosch. 1998. The central, surface-exposed region of the flagellar hook protein FlgE of Campylobacter jejuni shows hypervariability among strains. J. Bacteriol. 180:3711-3714. Mallett, C. P., L. VanDeVerg, H. H. Collins, and T. L. Hale. 1993. Evaluation of Shigella vaccine safety and efficacy in an intranasally challenged mouse model. Vaccine 11:190-196. Mansfield, L. S., J. A. Bell, D. L. Wilson, A. J. Murphy, H. M. Elsheikha, V. A. Rathinam, B. R. Fierro, J. E. Linz, and V. B. Young. 2007. C57BL/ 6 and congenic interleukin-10-deficient mice can serve as models of Campylobacter jejuni colonization and enteritis. Infect. Immun. 75: 1099-1 115. Marcus, E. A., and D. R. Scott. 2001. Cell lysis is responsible for the appearance of extracellular urease in Helicobacter pylori. Helicobacter 6:93-99. Martin, P. M., J. Mathiot, J. Ipero, M. Kirimat, A. J. Georges, and M. C. Georges-Courbot. 1989. Immune response to Campylobucter jejuni and Campylobacter coli in a cohort of children from birth to 2 years of age. Infect. Immun. 57:2542-2546. Mattila, L. 1994. Clinical features and duration of traveler’s diarrhea in relation to its etiology. Clin. Infect. Dis. 19:728-734. Miller, N., and I. Saunders. 2007. Current perceptions of travelers’ diarrhea treatments and vaccines: results from a postal questionnaire survey and physician interviews. ]. Travel Med. 14:158167. Millson, M., M. Bokhout, J. Carlson, L. Spielberg, R. Aldis, A. Borczyk, and H. Lior. 1991. An outbreak of Cumpylobacter jejuni gastroenteritis linked to meltwater contamination of a municipal well. Can. ]. Public Health 82:27-31. Mishu, B., and M. J. Blaser. 1993. Role of infection due to Cumpylobucter jejuni in the initiation of Guillain-BarrC syndrome. Clin. Infect. Dis. 17:104-108. Nachamkin, I., and A. M. Hart. 1985. Western blot analysis of the human antibody response to Campylobacter jejuni cellular antigens during gastrointestinal infection. J. Clin. Microbiol. 21: 33-38. Nesvizhskii, A. I. 2007. Protein identification by tandem mass spectrometry and sequence database searching. Methods Mol. Biol. 367~87-119. Newburg, D. S., G. M. Ruiz-Palacios, and A. L. Morrow. 2005. Human milk glycans protect infants against enteric pathogens. Annu. Rev. Nutr. 25:37-58. Oelschlaeger, T. A., P. Guerry, and D. J. Kopecko. 1993. Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proc. Natl. Acad. Sci. USA 90:6854-6888. Pancholi, V., and G. S. Chhatwal. 2003. Housekeeping enzymes as virulence factors for pathogens. Int. J. Med. Microbiol. 293: 39 1-40 1. Pancholi, V., and V. A. Fischetti. 1998. Alpha-enolase, a novel strong plasmin(ogen) binding protein on the surface of pathogenic streptococci. J. Biol. Cbem. 273:14503-14515. Pancholi, V., and V. A. Fischetti. 1992. A major surface protein on group A streptococci is a glyceraldehyde-3-phosphatedehydrogenase with multiple binding activity. J. Exp. Med. 176: 415-426. Park, K. J., M. M. Gromiha, P. Horton, and M. Suwa. 2005. Discrimination of outer membrane proteins using support vector machines. Bioinfomzatics 21:4223-4229. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S . Hol-
CHAPTER 24
royd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. BarrC11. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pavlovskis, 0. R., D. M. Rollins, R. L. Haberberger, Jr., A. E. Green, L. Habash, S . Strocko, and R. I. Walker. 1991. Significance of flagella in colonization resistance of rabbits immunized with Campylobacter spp. Infect. Immun. 59:2259-2264. Pawelec, D. P., D. Korsak, A. K. Wyszynska, E. Rozynek, J. Popowski, and E. K. Jagusztyn-Krynicka. 2000. Genetic diversity of the Campylobacter genes coding immunodominant proteins. FEMS Microbiol. Lett. 185:43-49. Pei, Z., and M. J. Blaser. 1993. PEB1, the major cell-binding factor of Campylobacter jejuni, is a homolog of the binding component in gram-negative nutrient transport systems. J. Biol. Chem. 268: 18717-1 8725. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. Z. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchtre, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943. Peltola, H., A. Siitonen, H. Kyronseppa, I. Simula, L. Mattila, P. Oksanen, M. J. Kataja, and M. Cadoz. 1991. Prevention of travellers’ diarrhoea by oral B-subunit/whole-cell cholera vaccine. Lancet 338:1285-1289. Penner, J. L., and J. N. Hennessy. 1980. Passive hemagglutination technique for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heat-stable antigens. J. Clin. Microbiol. 12: 732-73 7. Petruccelli, B. P., G. S. Murphy, J. L. Sanchez, S. Walz, R. DeFraites, J. Gelnett, R. L. Haberberger, P. Echeverria, and D. N. Taylor. 1992. Treatment of traveler’s diarrhea with ciprofloxacin and loperamide. J. Infect. Dis. 165:557-560. Poly, F., T. Read, D. R. Tribble, S. Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. lmmun. 75:3425-3433. Pope, J. E., A. Krizova, A. X. Garg, H. Thiessen-Philbrook, and J. M. Ouimet. 2007. Campylobacter reactive arthritis: a systematic review. Semin. Arthritis Rheum. 37:48-55. Power, M. E., P. Guerry, W. D. McCubbin, C. M. Kay, and T. J. Trust. 1994. Structural and antigenic characteristics of Campylobacter coli FlaA flagellin. J. Bacteriol. 176:3303-33 13. Prokhorova, T. A., P. N. Nielsen, J. Petersen, T. Kofoed, J. S. Crawford, C. Morsczeck, A. Boysen, and P. Schrotz-King. 2006. Novel surface polypeptides of Canzpylobacterjejuni as traveller’s diarrhoea vaccine candidates discovered by proteomics. Vaccine 24: 6446-6455. Radcliff, F. J., S. L. Hazell, T. Kolesnikow, C. Doidge, and A. Lee. 1997. Catalase, a novel antigen for Helicobacter pylori vaccination. Infect. Immun. 65:4668-4674. Rangarajan, E. S., S . Bhatia, D. C. Watson, C. Munger, M. Cygler, A. Matte, and N. M. Young. 2007. Structural context for protein N-glycosylation in bacteria: the structure of PEB3, an adhesin from Campylobacter jejuni. Protein Sci. 16:990-995. Rao, M. R., A. B. Naficy, S. J. Savarino, R. Abu-Elyazeed, T. F. Wierzba, L. F. Peruski, I. Abdel-Messih, R Frenck, and J. D. Clemens. 2001. Pathogenicity and convalescent excretion of Campylobacter in rural Egyptian children. Am. 1. Epidemiol. 154:166-173. Riddle, M. S., J. W. Sanders, S. D. Putnam, and D. R. Tribble. 2006. Incidence, etiology, and impact of diarrhea among longterm travelers (U.S. military and similar populations): a systematic review. Am. J. Trop. Med. Hyg. 74:891-900. Rollwagen, F. M., N. D. Pacheco, J. D. Clements, 0. Pavlovskis, D. M. Rollins, and R. I. Walker. 1993. Killed Campylobacter
DEVELOPMENT OF A HUMAN VACCINE
443
elicits immune response and protection when administered with an oral adjuvant. Vaccine 11:1316-1320. Ruiz, J., F. Marco, I. Oliveira, J. Vila, and J. Gascon. 2007. Trends in antimicrobial resistance in Campylobacter spp. causing traveler’s diarrhea. APMIS 115:218-224. Ruiz-Palacios, G. M., J. J. Calva, L. K. Pickering, Y. Lopez-Vidal, P. Volkow, H. Pezzarossi, and M. S . West. 1990. Protection of breast-fed infants against Campylobacter diarrhea by antibodies in human milk. J. Pediatr. 116:707-713. Russell, R. G., M. J. Blaser, J. I. Sarmiento, and J. Fox. 1989. Experimental Campylobacter jejuni infection in Macaca nemestrina. Infect. lmmun. 57:1438-1444. Russell, R G., J. I. Sarmiento, J. Fox, and P. Panigrahi. 1990. Evidence of reinfection with multiple strains of Campylobacter jejuni and Campylobacter coli in Macaca nemestrina housed under hyperendemic conditions. Infect. Immun. 58:2149-2155. Ryan, E. T., S. B. Calderwood, and F. Qadri. 2006. Live attenuated oral cholera vaccines. Expert Rev. Vaccines 5:483-494. Sabarth, N., S. Lamer, U. Zimny-Amdt, P. R. Jungblut, T. F. Meyer, and D. Bumann. 2002. Identification of surface proteins of Helicobacter pylori by selective biotinylation, affinity purification, and two-dimensional gel electrophoresis. J. Biol. Chem. 277~27896-27902. Sabin, A. B. 1965. Oral poliovirus vaccine. History of its development and prospects for eradication of poliomyelitis. JAMA 194~872-876. Samuel, M. C., D. J. Vugia, S. Shallow, R. Marcus, S. Segler, T. McGivern, H. Kassenborg, K. Reilly, M. Kennedy, F. Angulo, and R. V. Tauxe. 2004. Epidemiology of sporadic Campylobacter infection in the United States and declining trend in incidence, FoodNet 1996-1999. Clin. Infect. Dis. 38(Suppl. 3): S165-S174. Sanchez, J. L., B. Vasquez, R. E. Begue, R Meza, G. Castellares, C. Cabezas, D. M. Watts, A. M. Svennerholm, J. C. Sadoff, and D. N. Taylor. 1994. Protective efficacy of oral whole-cell/ recombinant-B-subunit cholera vaccine in Peruvian military recruits. Lancet 344:1273-1276. Schaumburg, J., 0. Diekmann, P. Hagendorff, S. Bergmann, M. Rohde, S. Hammerschmidt, L. Jansch, J. Wehland, and U. Karst. 2004. The cell wall subproteome of Listeria monocytogenes. Proteomics 4:2991-3006. Schrotz-King, P., T. A. Prokhorova, P. N. Nielsen, J. S . Crawford, and C. Morsczeck. 2007. Campylobacter jejuni proteomics for new travellers’ diarrhoea vaccines. Travel Med. Infect. Dis. 5: 106-109. Sizemore, D. R., B. Warner, J. Lawrence, A. Jones, and K. P. Killeen. 2006. Live, attenuated Salmonella typhimurium vectoring Campylobacter antigens. Vaccine 24:3 793-3 803. Sjogren, E., G. Ruiz-Palacios, and B. Kaijser. 1989. Campylobacter jejuni isolations from Mexican and Swedish patients, with repeated symptomatic and/or asymptomatic diarrhoea episodes. Epidemiol. Infect. 102:47-57. Song, Y. C., S. Jin, H. Louie, D. Ng, R. Lau, Y. Zhang, R. Weerasekera, S. A1 Rashid, L. A. Ward, S. D. Der, and V. L. Chan. 2004. FlaC, a protein of Campylobacter jejuni TGH9011 (ATCC43431) secreted through the flagellar apparatus, binds epithelial cells and influences cell invasion. Mol. Microbiol. 53: 541-553. Spence, J. M., and V. L. Clark. 2000. Role of ribosomal protein L12 in gonococcal invasion of HeclB cells. Infect. Immun. 68: 5002-5 0 10. Steffen, R. 1986. Epidemiologic studies of travelers’ diarrhea, severe gastrointestinal infections, and cholera. Rev. Infect. Dis. 8(Suppl. 2):S 122-S 130. Struyvt, M., M. Moons, and J. Tommassen. 1991. Carboxyterminal phenylalanine is essential for the correct assembly of a bacterial outer membrane protein. J. Mol. Biol. 218:141-148.
444
TRIBBLE ET AL.
Szymanski, C. M., R. Yao, C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evidence for a system of general protein glycosylation in Campylobacter jejuni. Mol. Microbial. 32:1022-1030. Taylor, D. 1992. Campylobacter infections in developing countries, p. 20-30. In I. Nachamkin, M. Blaser, and L. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Taylor, D. N., D. M. Perlman, P. D. Echeverria, U. Lexomboon, and M. J. Blaser. 1993. Campylobacter immunity and quantitative excretion rates in Thai children. J. Infect. Dis. 168:754758. Thornley, J. P., D. Jenkins, K. Neal, T. Wright, J. Brough, and R. C. Spiller. 2001. Relationship of Campylobacter toxigenicity in vitro to the development of postinfectious irritable bowel syndrome. J. Infect. Dis. 184:606-609. Torres, O., and J. R. Cruz. 1993.Protection against Cumpylobacter diarrhea: role of milk IgA antibodies against bacterial surface antigens. Actu Paediutr. 82:835-838. Tribble, D.R., S. Baqar, D. A. Scott, M. L. Oplinger, F. Trespalacios, R. I. Walker, J. D. Clements, and A. L. Burgeois. 2001. Clinical development of an inactivated whole cell Cumpylobacter vaccine candidate. In Vaccines for Enteric Diseases, 2001, Tampere, Finland. Tribble, D. R., J. W. Sanders, L. W. Pang, C. Mason, C. Pitarangsi, S. Baqar, A. Armstrong, P. Hshieh, A. Fox, E. A. Maley, C. Lebron, D. J. Faix, J. V. Lawler, G. Nayak, M. Lewis, L. Bodhidatta, and D. A. Scott. 2007. Traveler’s diarrhea in Thailand: randomized, double-blind trial comparing single-dose and 3-day azithromycin-based regimens with a 3-day levofloxacin regimen. Clin. Infect. Dis. 44:338-346. Vitovec, J., B. Koudela, J. Sterba, I. Tomancova, Z. Matyas, and P.Vladik. 1989. The gnotobiotic piglet as a model for the path-
ogenesis of Campylobacter jejuni infection. Zentralbl. Bakteriol. 271:91-103. Waldispiihl, J., B. Berger, P. Clote, and J. M. Steyaert. 2006.Predicting transmembrane beta-barrels and interstrand residue interactions from sequence. Proteins 65:61-74. Walker, R. I. 2005. Considerations for development of whole cell bacterial vaccines to prevent diarrheal diseases in children in developing countries. Vaccine 23:3369-3385. Walker, R. I., D. M. Rollins, and D. H. Burr. 1992. Studies of Campylobacter infection in the adult rabbit, p. 139-147. In I. Nachamkin, M. Blaser, and L. Tompkins (ed.), Campylobacter jejuni: Current Status and Future Trends. American Society for Microbiology, Washington, DC. Watson, R. O., V. Novik, D. Hofreuter, M. Lara-Tejero, and J. E. G a l h 2007.A MyD88-deficient mouse model reveals a role for Nrampl in Campylobacter jejuni infection. Infect. Immun. 75:1994-2003. Weynants, V. E., C. M. Feron, K. K. Goraj, M. P. Bos, P. A. Denoel, V. G . Verlant, J. Tommassen, I. R. A. Peak, R. C. Judd, M. P. Jennings, and J. T. Poolman. 2007.Additive and synergistic bactericidal activity of antibodies directed against minor outer membrane proteins of Neisseria meningitidis. Infect. lmmun. 755432-5442. Wittmann-Liebold, B., H. R. Graack, and T. Pohl. 2006. Twodimensional gel electrophoresis as tool for proteomics studies in combination with protein identification by mass spectrometry. Proteomics 6:468 8-4703. Zhang, Q.,J. C. Meitzler, S . Huang, and T.Morishita. 2000. Sequence polymorphism, predicted secondary structures, and surface-exposed conformational epitopes of Campylobacter major outer membrane protein. Infect. Immun. 6856794689.
IV. GLYCOBIOLOGY
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 25
N-Linked Protein Glycosylation in Campylobacter HARALDNOTHAFT, SABAAMBER,MARKUS AEBI,AND CHRISTINE SZYMANSKI
there has been extensive work demonstrating that many bacteria do indeed add sugars through the hydroxyl groups of Ser and Thr in an 0-linkage. These studies also include characterization of the O-glycosylation pathway of flagellin in Campylobacter jejuni and Campylobacter coli, described further in chapter 26. Interestingly, C. jejuni was the first bacteria described to be capable of adding sugars to proteins through the amide group of Asn in an Nlinkage. N-linked protein glycosylation is the most common type of protein modification in eukaryotes and will be the topic of this chapter. We will demonstrate that the C. jejuni glycome is an excellent toolbox for glycobiologists to understand the fundamentals of this pathway, to develop new techniques for glycobiology, and to exploit this pathway for novel diagnostics and therapeutics. Since the discovery of the general protein glycosylation pathway in C. jejuni in 1999 (Szymanski et al., 1999) and the demonstration that the glycans are attached through an N-linkage (Wacker et al., 2002; Young et al., 2002), a surge of research activity has resulted to characterize and exploit this system. What emerged from these efforts is the finding that N-linked protein glycosylation is found in all three domains of life and that it follows a general principle in archaea, bacteria, and eukaryotes. Sequence similarity of proteins involved in these processes as well as the conserved conceptual framework strongly support the hypothesis that N-linked protein glycosylation is a homologous process in all life forms. The bacterial pathway originally identified in C. jejuni and extensively studied in the heterologous host Escherichia coli has been instrumental in defining the basic mechanisms underlying the process of N-glycosylation.
Previously, biochemists would develop tools to remove sugars from the proteins that they were examining. Today, monosaccharides represent “an alphabet of biological information similar to amino acids and nucleic acids, but with unsurpassed coding capacity.” It is recognized that glycosylation of proteins is necessary for extending protein serum halflife, protease resistance, and biological activity. Recognizing the importance of sugars in health and disease, the National Institutes of Health awarded a 5-year, $34 million “glue” grant to the Consortium for Functional Glycomics in October 2001. In the 2003 issue of MIT’s Technology Review, glycomics appeared first on the top 10 listing for emerging technologies that will change the world. Similarly, the pharmaceutical giant, Merck & Co. purchased the yeast-based recombinant glycoprotein technology of GlycoFi last year for $400 million. These investments are a result of the fact that in eukaryotes, it is believed that up to 90% of all proteins are modified with sugars. Furthermore, disruption of these sugar pathways has been implicated in multiple disease states ranging from congenital disorders of glycosylation to autoimmune disease and cancer. Although it is predicted that bacteria synthesize at least sixfold more sugar building blocks compared with their eukaryotic counterparts, it was previously believed that these single-celled organisms were incapable of modifying proteins with sugars. The reason why carbohydrate research, and particularly the characterization of protein glycosylation pathways, has lagged behind in bacteria was in part the result of the lack of reagents and sensitive techniques capable of detecting the unusual sugars that these organisms produce. However, in the past two decades,
Saba Amber and Markus Aebi Institute of Microbiology, Department of Biology, Eidgenossische Technische Hochschule (ETH) Harald Nothaft and Christine M. Szymanski Institute for Biological Sciences, National ReZurich, 8093 Zurich, Switzerland. search Council, Ottawa, Canada, K1A OR6.
447
448
NOTHAFT ET AL.
N-GLYCOSYLATION PATHWAY IN C . JEJUNI A series of 13 genes, initially termed the wla operon, encoding proteins with significant sequence similarity to enzymes involved in lipopolysaccharide (LPS) biosynthesis were cloned from C. jejuni 81116 by Fry et al. (1998). When this 16-kb DNA region was expressed in E. coli, it gave rise to an LPS that reacted with the C. jejuni Penner-typing antiserum, leading to the belief that the wla operon was involved in LPS biosynthesis. However, the role of each individual gene in LPS biosynthesis was not confirmed. Rather, the data indicated that these genes are part of a general system of protein glycosylation in Campylobacter by which numerous soluble and membrane-bound proteins are glycosylated (Szymanski et al., 1999). On the basis of these data, the genes were renamed pgl (for protein glycosylation) (Szymanski et al., 1999) folGwing t& nomenclature of Jennings et al. (1998), and their role in protein glycosylation was further characterized (Fig. 1). The proteins of the pgl operon show homology to various bacterial glycosyltransferases and sugar modification enzymes. Of special interest was PglB, a homolog of STT3, an essential and most highly conserved subunit of the oligosaccharyltransferase (OTase) complex of eukaryotes (Kelleher and Gilmore, 1994), suggesting a direct role of PglB in Nlinked protein glycosylation. STT3 orthologs contain a conserved C-terminal catalytic motif (WWDYG), of
which the second and third amino acids are essential (Wacker et al., 2002; Yan and Lennarz, 2002). This motif is present in PglB (458WWDYGY463), and the mutations W458A and D459A led to the disruption of glycosylation. Linton et al. (2002) identified two glycoproteins, PEB3 and CgpA, which were glycosylated by components of the Pgl pathway and further showed that such glycoproteins bound to the GalNAc-specific lectin, soybean agglutinin (SBA). Subsequent work by Young et al. (2002) identified 38 potential C. jejuni glycoproteins, including PEB3 and CgpA, and demonstrated that the glycan was linked to proteins through asparagines in the motif Asn-X-Ser/Thr (NX-SIT, where X is any amino acid), the same sequon used in eukaryotes for N-linked glycosylation. The structure of the N-linked glycan was determined by mass spectrometry (MS) and nuclear magnetic resonance (NMR) techniques to be a heptasaccharide with the structure GalNAc-al,4-GalNAc-al,4-[Glcp 1,3]GalNAc-a 1,4-GalNAc-a1,4-GalNAc-a173-Bac2,4diNAc-pl7N-Asn, where Bac is di-N-acetyl bacillosamine (2,4-diacetamido-2,4,6-trideoxyglucopyranose). This heptasaccharide structure is conserved throughout all C. jejuni and C. coli strains examined so far (Szymanski et al., 2003). Moreover, a mutation in pglB led to loss of glycosylation on the proteins, providing evidence that this gene is responsible for formation of the glycopeptide N-linkages. Wacker et al. (2002) established an experimental system to produce N-linked glycoproteins in E. coli
UDP
1
PglDEF
Cytoplasm
U
Periplasm
PglB
\
Glucose N-acetylglucosarnine N-acetyigalactosamine Di-N-acetyl bacillosarnine
Figure 1. Schematic pathway of N-linked protein glycosylation at the plasma membrane of Campylobacter jejuni. The assembly of a heptasaccharide takes place on the lipid bactoprenylpyrophosphate at the cytoplasmic side of the membrane by the glycosyltransferases PglC, PglA, PglJ, PglH, and PglI. UDP-bacillosamine is synthesized from UDP-GlcNAc by PglF, PglE, and PglD. The ABC-transporter PglK mediates the translocation of the lipid-linked heptasaccharide across the membrane. The oligosaccharyltransferase catalyzes the transfer of the heptasaccharide from the lipid carrier to selected asparagine residues on nascent polypeptide chains. This pathway is encoded by the pgl operon shown below, with the pglB gene encoding the oligosaccharyltransferase highlighted.
CHAPTER 25
N-LINKED PROTEIN GLYCOSYLATION
449
bled consecutively in the manner of eukaryotic 0in the presence of the pgl operon. The pgl gene cluslinked glycoproteins (Gemmill and Trimble, 1999). ter was expressed in E. coli along with a C. jejuni Consistent with the homologies, PglF and PglE glycoprotein, AcrA or Peb3, and these proteins were have recently been shown to modify the sugar shown to be glycosylated in this engineered E. coli nucleotide uridine diphosphate-N-acetylglucosamine host. This demonstrated that the pgl cluster contains (UDP-GlcNAc) to form UDP-2-acetamido-4-aminoall the genes necessary for the biosynthesis of the 2,4,6-trideoxy-a-~-glycopyranose (UDP-4-aminolipid-linked heptasaccharide and its eventual transfer sugar) (Schoenhofen et al., 2006b; Vijayakumar et al., to proteins. Additionally, it opened up avenues for 2006). PglD was also cloned, overexpressed, and puanalyzing the glycosylation machinery in a more amirified from the pgl locus of C. jejuni NCTC 11168 able host and exploiting it for glycoengineering. and identified as the acetyltransferase that modifies As shown initially by Fry and coworkers (1998), the UDP-4-amino-sugar to form UDP-N,N'-diacetyla number of pgl genes encode proteins with sequence bacillosamine, utilizing acetyl-coenzyme A as the acesimilarity to known bacterial glycosyltransferases. On tyl group donor (Olivier et al., 2006). the basis of these homologies, a hypothesis was built In C. jejuni, there are at least three distinct metfor the role of each glycosyltransferase in the Pgl hepabolic pathways that share UDP-GlcNAc as an initasaccharide biosynthesis pathway (Fig. 1). It was tial substrate: (i) UDP-GlcNAc is modified by PglF postulated that similar to LPS biosynthesis, the olito form the UDP-4-keto-sugar substrate of PglE gosaccharide is synthesized on the lipid carrier un(Schoenhofen et al., 2006b); (ii) pseudaminic acid, decaprenol pyrophosphate (Und-PP) in a stepwise the major carbohydrate modification of flagellin, is fashion, at the cytoplasmic side of the inner memsynthesized via the UDP-GlcNAc dehydratase PseB brane. Sugar nucleotides serve as activated mon(Goon et al., 2003; Schoenhofen et al., 2006b); and osaccharide donors. PglC, the first glycosyltransfer(iii) GalNAc, a major component in the biosynthesis ase, attaches a sugar phosphate to undecaprenol of glycoprotein N-linked heptasaccharide, lipooligophosphate (Fig. 1).The next glycosyltransferase PglA saccharide, and capsule (Linton et al., 2002; St then adds an al-3GalNAc to the PglC product. The Michael et al., 2002; Young et al., 2002), is syntheother glycosyltransferases, PglJ, PglH, and PglI, then sized intracellularly by the C4 epimerase Gne from sequentially add monosaccharides for assembly of the GlcNAc (Bernatchez et al., 2005). This leads to the full-length heptasaccharide on the lipid where PglI question about how the turnover is regulated. Kinetic adds the glucose branch. The lipid-linked oligoparameters of PseB and Gne as well as those for PglF saccharide (LLO) is then translocated (flipped) to clearly demonstrate that Gne is the most efficient of the outer side of the inner membrane in a proall three enzymes, with a kc,,lK, greater by two ortein-mediated fashion, possibly by the ABC transders of magnitude than those of PseB and PglF. This porter PglK. As a final step, the oligosaccharyltransis to be expected because one or more copies of ferase PglB transfers the oligosaccharide from the GalNAc/ Gal or subsequent derivatives occur more lipid to asparagine residues of target proteins in the frequently on cell surface structures (heptasaccharide, consensus sequence N-X-S/T. 2,4-Diacetamidolipooligosaccharide, capsule) than GlcNAc/Glc2,4,6-trideoxy-au-~-glucopyranose,termed N,N'related glycosides. The kc,,IK, value of PseB indicates diacetylbacillosamine (Bac2,4diNAc), is the first that this enzyme is a moderately effective dehydracarbohydrate in the glycoprotein N-linked heptasactase, more so than PglF, by a factor of 30. However, charide. In Neisseria meningitidis, the pilin is 0- pgl gene knockout studies on C. jejuni NCTC 11168 substituted at Ser63with Gal-/31,3-GaI-al,3-trideoxy- showed a significant disparity between wild-type and diacetamidohexose (Stimson et al., 1995). Three mutant strain signatures by high-resolution magic angenes that are plausible candidates for Bac synthesis gle spinning NMR analyses and whole-cell lysate are present in both the C. jejuni and N. meningitidis blotting with lectins (Kelly et al., 2006). These results loci (Jennings et al., 1998; Power et al., 2000), Cj indicate the substrate turnover rate of PseB is insufpglF (Nm pglD) a dehydratase; Cj pglE (Nm pglC), ficient to completely complement the wild-type conan aminotransferase; and Cj pglD (Nm pglB), an acesumption levels of the UDP-4-keto-sugar by the Pgl tyltransferase (annotations from Parkhill et al., 2000). pathway in the absence of PglF. Comparison of the Cj pglC is a homolog of the second domain of the kinetics data, together with the knockout study reapparently fused gene Nm pglB, which puts the first sults, supports the hypothesis of a channeling mechsugar, i.e., bacillosamine, onto a lipid carrier (Power anism where a sequential transfer of products beet al., 2000). This implies that both N- and 0-linked tween enzymes of the same biosynthetic pathway structures are preassembled on lipids before attachexists (Creuzenet, 2004). Of the two dehydratases ment to protein, rather than the latter being assemthat are compared, PseB has the lowest K,, indicating
450
NOTHAFT ET AL.
a higher affinity for substrate. Studies on PseB knockouts in C. jejuni 81-176 showed unglycosylated flagella accumulated in cells intracellularly, thus explaining a lack of motility (Goon et al., 2003). Arguably, when UDP-GlcNAc is in low abundance, the pathogen would likely channel available stores toward the function of motility over host cell interactions. The glycosylation pathway of C. jejuni was also analyzed in E. coli by expressing the pgl operon together with acceptor proteins of C. jejuni, AcrA and Peb3, in these cells (Linton et al., 2005). Individual genes of the pgl operon were disrupted by insertion of a kanamycin resistance cassette in a nonpolar manner in the respective genes. The glycoproteins in these mutant strains were analyzed by Western blotting and MS, enabling the determination of the order of glycosyltransferases acting in this pathway. PglC, the first glycosyltransferase, attaches bacillosamine phosphate to undecaprenol phosphate (Und-P) and belongs to the family of polyprenol-phosphate: Nacetyl-hexosamine 1-phosphate transferases. It was shown that a defect in the pglC locus can be complemented by wecA of E. coli (Linton et al., 2005). WecA belongs to the same family of transferases as PglC, and it plays an essential role in LPS biosynthesis, where it adds a GlcNAc-P to Und-P. In a pglC mutant strain, WecA attaches a HexNAc phosphate to Und-P, which can then be extended by the glycosyltransferases of the pgl operon. The identity of the HexNAc has not yet been determined, but it is most likely a GlcNAc, the natural substrate of WecA. This variation in the first monosaccharide suggests a relaxed specificity of the next glycosyltransferase PglA, which can then transfer a GalNAc to both bacillosamine and HexNAc, and also of the flippase PglK and the OTase PglB for the first sugar of the oligosaccharide (discussed below). These results also show that an incompletely assembled oligosaccharide is recognized and transferred to proteins in E. coli, indicating that the length of the lipid-linked oligosaccharide is not critical for PglK and PglB (Alaimo et al., 2006; Linton et al., 2005). Linton et al. could show that PglA, PglJ, and PglH act sequentially on the PglC product to synthesize the hexasaccharide (Linton et al., 2005). It was then evident that PglI acts last to add the Glc branch of the heptasaccharide. However, it could not be determined which proteins were involved in the transfer of the two terminal GalNAcs. It was suggested that either PglH adds the three terminal GalNAcs or that PglJ and PglH act alternately, adding two residues each to complete the hexasaccharide. The answer was provided by Glover et al., who carried out in vitro biochemical analysis by using chemically synthesized Und-PP-Bac and pu-
rified Pgl glycosyltransferases (Glover et al., 2005a; Weerapana et al., 2006). They confirmed the roles for PglC, PglA, PglJ, and PglI and demonstrated that PglH is a sugar polymerase, adding three a1,4linked GalNAc residues to the undecaprenylpyrophosphate-linked trisaccharide. In vitro biosynthesis of the complete heptasaccharide lipid-linked donor by the coupled action of eight enzymes (PglF, PglE, PglD, PglC, PglA, PglJ, PglH, and PglI) from the Pgl pathway in a single reaction vessel was also demonstrated (Olivier et al., 2006). Biosynthesis of the N-glycan by the pgl operon was also analyzed in its native host, Campylobacter jejuni NCTC 11168, by Kelly et al. (2006). A pglC mutation (i.e., bacillosamine transferase) in this background seems to be nonviable for the cells and could not be obtained, while the Bac biosynthesis genes could be knocked out, which led to a loss of glycosylation. Contrary to the E. coli system, incomplete oligosaccharides were not transferred effectively to protein by PglB, except the hexasaccharide in a pglI mutant strain, which could still be transferred and detected on proteins. This indicates that the bacterial glycans are transferred en bloc in C. jejuni, similar to the mechanism in eukaryotes, emphasizing the evolutionary conservation of protein N-glycosylation. It was postulated that the prokaryotic oligosaccharide is assembled onto a polyisoprenyl-pyrophosphate in a manner similar to the assembly of dolichyl-pyrophosphate linked oligosaccharide in eukaryotes. On the basis of the similarity of the bacterial N-glycosylation pathway to the LPS biosynthesis pathway, it was suggested that the polyisoprene used is undecaprenol (also known as bactoprenol), the polyisoprenol carrier in LPS biosynthesis. Undecaprenol, as the name suggests, contains 11 isoprene units, where the a-isoprene unit is unsaturated, in contrast to the a-saturated nature of the corresponding unit in the dolichols used by eukaryotes. WecA complementation of pglC suggests the polyisoprenyl carrier to be the same in N-glycosylation, as in LPS biosynthesis. Further proof was provided by the transfer of 0-antigen to proteins by PglB (Feldman et al., 2005). However, the direct evidence came from in vitro glycosylation studies where (i) PglC was able to produce undecaprenyl pyrophosphate bacillosamine (Glover et al., 2006); (ii) Und-PP-Bac, synthesized chemically, served as a substrate for PglA, the second glycosyltransferase (Weerapana et al., 2005); (iii) this undecaprenol-linked disaccharide could be further extended and elongated to form the heptasaccharide by subsequent glycosyltransferases (Glover et al., 2005b); and (iv) undecaprenyl-linked glycans were transferred to protein by PglB (Glover et al., 2 0 0 5 ~ ) .
CHAPTER 25
The 3 8 potential glycoproteins identified by Young et al. (2002) are predominantly annotated as periplasmic proteins, which suggested that glycosylation happens in the periplasm, which is functionally equivalent to the lumen of the endoplasmic reticulum in eukaryotes. C. jejuni AcrA is a periplasmic membrane-attached glycoprotein. An acrA mutant that lacks the periplasmic signal sequence localizes in the cytoplasm and is not glycosylated, but when a pelB leader sequence is cloned in front of this cytoplasmic protein, it is efficiently glycosylated (NitaLazar et al., 2005). This not only identifies the periplasm as the site of glycosylation but also shows that both membrane-bound and soluble proteins localized in the periplasm can be glycosylated by this system. These findings imply that as in the eukaryotic Nglycosylation pathway, the polyisoprenyl oligosaccharide has to be flipped (or transported) across a membrane bilayer in order to serve as a substrate for the N-glycosylation process. Non-ABC and ABC-type transporters are proposed to catalyze this transbilayer movement of LLOs, a reaction that does not occur spontaneously (Hanover and Lennarz, 1979; McCloskey and Troy, 1980; Menon, 1995). The eukaryotic and 0-antigen flippases do not have an ATPbinding cassette, but the C. jejuni flippase, PglK, is an ABC-type transporter. Mutation of the nucleotidebinding domain of PglK (formerly known as WlaB) abolishes glycosylation (Alaimo et al., 2006), and ATPase activity of purified PglK has been demonstrated in vitro. PglK has a relaxed specificity for its substrate in E. coli. Not only is it able to flip incomplete oligosaccharides, but it can also complement the 0antigen flippase wzx, a non-ABC transporter (Alaimo et al., 2006). However, in its native host, C. jejuni, a pglK mutation leads to loss of glycosylation, suggesting a more stringent control of glycosylation in C. jejuni and possibly the involvement of other factors in E. coli that might contribute to a relaxed specificity (Kelly et al., 2006). The central enzyme in the N-glycosylation pathway of C. jejuni is the oligosaccharyltransferase PglB. This transmembrane protein has low but significant sequence similarity to the most highly conserved component of the eukaryotic oligosaccharyltransferase, STT3. PglB has two very different substrates, a well-defined lipid-linked oligosaccharide and a series of different polypeptides. In contrast to the complex N-glycosylation pathway in eukaryotes, where the oligosaccharyltransferase is confronted with multiple biosynthetic intermediates of the LLO, this is not the case in the bacterial N-glycosylation system. It is therefore not surprising that the prokaryotic oligosaccharyltransferase has relaxed substrate specificity
N-LINKED PROTEIN GLYCOSYLATION
45 1
toward the lipid-linked oligosaccharide. Indeed, it was shown that lipid-linked oligosaccharides of the LPS pathway (0-antigen) can be transferred to AcrA in E. coli cells by PglB (Feldman et al., 2005). Several different 0-antigens from E. coli, Pseudomonas aeruginosa, and Shigella dysenteriae were transferred to AcrA, highlighting the relaxed oligosaccharide substrate specificity of PglB. It has been proposed that similar to eukaryotic OTases, PglB requires an acetamido group at the C-2 of the monosaccharide at the reducing end to be able to recognize or transfer it to protein (Wacker et al., 2006). On the protein side, the specificity of PglB is not so well defined, but in sharp contrast to the lipidlinked substrate of the reaction (LLO), there is increased stringency and specificity on the protein acceptor of the glycosylation reaction. PglB requires the extended consensus sequence Asp/ Glu-X-Asn-Z-Ser/ Thr (DIE-X-N-Z-S/T; X, Z # P) (Kowarik et al., 2006b) for transfer of oligosaccharides to protein. However, this sequence is essential but not sufficient for glycosylation. Not all consensus sequences were glycosylated when they were either artificially introduced or naturally present in non-C. jejuni proteins. The complete absence of proline from the -1 position suggests that the structure of the acceptor protein is important for glycosylation to occur. Although the primary sequence requirements are fulfilled in several non-C. jejuni secreted proteins, their potential glycosylation sites may be located within domains that do not allow changes in local conformation necessary for N-glycosylation by PglB. This supports the hypothesis that unfolding or flexibility is required for protein domains to be compatible with glycosylation. Also, unlike its eukaryotic counterpart, PglB was able to glycosylate folded proteins, and glycosylation is independent of the protein translocation machinery (Kowarik et al., 2006a). A model has been proposed where the conformation of the protein domain bearing the acceptor site must be compatible with an OTase-induced conformational change leading to the activated structure. Within the framework of this model, it is evident that rigid, terminally folded protein domains therefore have a lower potential to serve as glycan acceptors than thermodynamically less stable folding intermediates.
C M P Y LOBACTER N-LINKED GLYCOPROTEINS C. jejuni is predicted to encode 1,643 proteins, and among them are approximately 170 membrane, 130 periplasmic, and 40 lipoproteins (Gundogdu et al., 2007; Parkhill et al., 2000). Because protein Nglycosylation takes place in the periplasmic space, the
452
NOTHAFT ET AL.
number of potential targets to be N-glycosylated therefore includes approximately 340 proteins. In addition, it has been demonstrated that the bacterial Nglycosylation consensus sequence, D/E-X-N-Z-S/T, within a given polypeptide chain is another prerequisite for protein N-glycosylation in Campylobacter. When we perform a databank search for the presence of at least one bacterial N-glycosylation sequon, more than 150 of the 340 secreted proteins are possible candidates. This large amount of potential C. jejuni glycoproteins is unique among bacterial species and is indicative of a general role for this modification in this organism. However, there are only a few reports describing the relevance of this pathway within the native host. Even though it has been shown that NgIycosylation plays an important role within the infectious life cycle of Campylobacter (see below), there are even fewer reports describing the identification and characterization of N-linked glycoproteins, the role of this posttranslational modification on protein function or structure, and the influence of protein N-glycosylation on other cellular functions. This section summarizes the N-linked proteins identified so far (Table 1) and will provide further information on the roles for this posttranslational modification in Campylobacter. The identification of two glycoproteins in C. jejuni NCTC 11168 was described by Linton et al. in 2002. In Western blot experiments, the authors demonstrated that the lectin SBA specifically binds to terminal GalNAc residues of oligosaccharides present on multiple C. jejuni proteins. This binding was altered by mutations in the pgl operon. Subsequent SBA pulldown experiments and peptide mapping of in-gel trypsin-digested proteins by MS identified a 28-kDa protein as PEB3 (CjO289c), previously characterized as a highly immunogenic protein (Pei et al., 1991). In addition, a 32- to 34-kDa protein was identified as a putative periplasmic protein encoded by Cj167Oc. The protein was named CgpA (Campylobacter glycoprotein A). Additional glycoprzein candidateswere-identified by SBA affinity chromatography from C. jejuni glycine extracts followed by oneand two-dimensional polyacrylamide gel electrophoresis, Western blot analysis with an SBA-alkaline phosphatase conjugate, and glycopeptide identification by mass spectrometry (Kowarik et al., 2006b; Young et al., 2002) (Table 1). Among the proteins that appeared as multiple spots on two-dimensional gels were PEB3 and CgpA. In addition, the vertical pattern of spots with identical PIS displayed by CgpA and other proteins indicated the presence of multiple glycoforms with varying degrees of glycosylation that were dependent on the presence of a functional PglB protein (Young et al., 2002). CapLC-MS/MS of pu-
rified, tryptic digested PEB3 followed by MS/MS analysis showed that only 50% of native PEB3 was modified with the single heptasaccharide. By means of nano-NMR techniques, it was demonstrated that the identified heptasaccharide is common to all glycoproteins/peptides and that it is linked exclusively to an asparagine. As mentioned above, the identified glycoproteins were predominantly annotated as periplasmic proteins, suggesting that the glycosylation machinery is specific for this compartment. It has to be noted that several proteins identified from two-dimensional gels do not contain the extended glycosylation sequon necessary for bacterial N-glycosylation (Kowarik et al., 2006b) whereas modified peptides identified by MS/MS methods all contain this peptide sequence (Kowarik et al., 2006b; Young et al., 2002), indicating that a combination of several methods is necessary to finally prove the existence of this posttranslational modification (Table 1). In parallel, Wacker et al. (2002) demonstrated N-glycosylation of another C. jejuni protein. By means of a polyclonal antiserum raised against C. jejuni whole-cell extracts, several immunoreactive signals were detected in Western blot experiments that were dependent on a functional PglB protein. One immunoreactive signal was identified by matrixassisted laser desorption/ionization mass mapping as the periplasmic protein, AcrA (CjO367c). By means of AcrA-specific antibodies, the authors observed three signals in C. jejuni, while only two were recognized by the glycosylation-specific serum, indicating that AcrA is glycosylated at two sites. This was corroborated by the fact that expression of AcrA in the presence of the fully functional pgl gene cluster in E. coli led to the production of two reactive signals by using glycan-specific antiserum and SBA. MS analyses of SBA affinity purified glycopeptides obtained after trypsin digests of recombinant AcrA produced in E. coli demonstrated the presence of the heptasaccharide on two sites, Amlz3 and on AcrA and confirmed the heptasaccharide structure. Another Campylobacter protein that was recently shown to be N-linked glycosylated is CjaC/ HisJ (Cj167Oc), a ligand-binding protein that is a component of the putative ABC transport system recognizing polar amino acids and opines (Pawelec et al., 1998). Wyszynska et al. (2007) demonstrated that HisJ/CjaC of C. coli is glycosylated at two sites. Interestingly, the ortholog from C. jejuni that was expressed and glycosylated in E. coli appeared to be monoglycosylated (Nita-Lazar et al., ZOOS), whereas no glycosylation site is present in the CjaC orthologs of C. upsaliensis RM3195 and C. lari RM2100 (Wyszynska et al., 2007). It was further shown that although both glycosylation sequons have the same
CHAPTER 25
amino acid composition in C. coli, one glycosylation site was preferred over the other, and it is the “preferred” site that is missing in C. jejuni. It was therefore suggested that different N-glycosylation patterns of orthologous proteins might determine the host specificity of Campylobacter isolates.
INVESTIGATING THE ROLE OF THE N-GLYCAN O N PROTEIN STRUCTURE AND FUNCTION PEB3 To obtain more information on the potential role of N-glycosylation with regard to protein structure and function, Rangarajan et al. (2007) have determined the crystal structure of recombinant PEB3. PEB3 crystallizes as a dimer, and the obtained structural information indicates that PEB3 may function as both an adhesin and a transporter protein. However, the role of PEB3 remains obscure. A potential role to aid in interactions with eukaryotic cells has been discussed previously (Pei et al., 1991), but cell invasion and colonization studies showed that mutation in peb3 has no effect on chick colonization and INT407 invasion when compared with wild-type cells (Kakuda and DiRita, 2006). Although the structure was derived from nonglycosylated PEB3 overproduced and purified from E. coli, it was shown that the three key residues of the glycosylation sequon D/E-X-N-Z-S/T are well exposed on the surface in the folded state and therefore accessible to the oligosaccharyltransferase, PglB. Currently the authors are obtaining crystal structure information on the native PEB3 purified from C. jejuni to determine whether the presence of the heptasaccharide has any affect on protein structure. VirB 10 The VirB10 protein (Cjp3) was identified in C. jejuni 81-176 glycine extracts and appears as two specific bands in Western blot experiments, possesses affinity for SBA, and is susceptible to glycosidases specific for HexNAc (Larsen et al., 2004). The C. jejuni VirB10 protein is part of a type IV secretion system forming complex (T4SS) and homologous to the Com system for natural transformation in Helicobacter pylori (Hofreuter et al., 1998, 2001), which is located on the plasmid pVir, a nonconjugative plasmid that affects both virulence and natural competence (Bacon et al., 2000). Indeed, an insertion mutant in C. jejuni virBl0 showed reduced DNA uptake by natural transformation. VirB10 can be expressed in E. coli and become glycosylated in the presence of
N-LINKED PROTEIN GLYCOSYLATION
453
the C. jejuni pgl operon. It was shown that VirB10 is glycosylated at two sites, N32 and N97, as introduction of point mutations replacing these Asn by Ala residues led to nonglycosylated protein as detected by Western blotting with VirB10-specific antibodies. To demonstrate a potential influence on VirB10 protein function due to loss of N-linked glycosylation, the authors could show that the natural transformation defect in a virBl0 mutant can be complemented in trans by using a plasmid expressing wild-type VirBlO or an N32A substitution, but not by using a mutant expressing VirB10 with an N97A or an N32A, N97A substitution. It was therefore suggested that glycosylation of VirBlO at N97, but not at N32, is essential for wild-type levels of competence in C. jejuni 81-176. Furthermore, mutation in either pglB or pglE resulted in a more significant decrease (compared with the virBl0 mutant) in the number of transformants, and expression of pglE in trans led to partial complementation of the observed phenotype. It was therefore proposed that the competence effect was due to either a lack of VirB10 glycosylation or loss of N-glycosylation of other glycoproteins potentially involved in T4SS complex assembly or to decreased stability of nonglycosylated VirB10. The latter point was supported by the fact that VirB10 could not be detected in the glycosylation-deficient pgl mutants. Cj 1496c
A systematic approach to determine glycoprotein function was recently described by Kakuda and DiRita (2006). Null mutants in each gene known to encode a potential glycosylated protein as determined by SBA-lectin binding (Young et al., 2002) in C. jejuni were generated and examined for their ability to invade the human epithelial cell line INT407. Only mutation in Cj1496c showed an effect. An in-frame deletion mutant in Cj1496c led to approximately 20fold-reduced ability to invade cells in vitro and also led to reduced colonization of chick ceca when compared with C. jejuni wild type, an effect that could be completely complemented by expression of Cj 1496c in trans. N-linked glycan site-specific mutagenesis combined with Western blot analysis suggested that the Cj1496c protein is glycosylated at N73 and N169. However, complementation with a nonglycosylated form of Cj1496c exhibited levels of invasion and colonization equivalent to those of the parental strain, suggesting that glycans are not directly involved in the function of Cj1496c. It was further shown that the glycosylation of Cj1496c does not have a major influence on the stability of this protein. Because cell fractionation experiments
Table 1. Putative Campylobacter N-linked glycoproteins Cj no.
Gene product
No. of sequons" (N-X-S/T) Noung et al., 2002) I
Cj0081d Cj0114d CjO143c Cj0147c Cj0152cd Cj0169 Cj 0175c Cj0200cd Cj0238 Cj0289cd Cj0313d Cj033 2c Cj0334 Cj0367c
Cj0376 Cj0397cd Cj0399d Cj0414 CjO415' Cj0420 Cj0493 Cj05ll Cj0530d Cj0599d Cj0610d Cj0638c Cj0648d
SBA stainingb
pglB mutant" (Young et aI., 2002)
Extended sequon
(DIE-X-N-Z- SIT)^
Reference(s)
I
Cytochrome bd oxidase subunit Probable periplasmic protein Periplasmic solute-binding protein for ABC transport system Thioredoxin (TrxA) Putative membrane protein Superoxide dismutase (Fe; SodB) Putative iron uptake ABC transport system periplasmic iron-binding protein Putative periplasmic protein Putative integral membrane protein Major antigenic peptide (PEB3) Putative membrane protein Nucleoside diphosphate kinase (Ndk) Alkyl hydroperoxide reductase (AhpC) AcrA
1T 2S, 3T lS, 1T
Probable periplasmic protein Hypothetical protein Putative membrane protein CmeC Putative oxidoreductase sub-unit Putative periplasmic protein Translation elongation factor EF-G (FusA) Probable secreted proteinase Putative periplasmic protein Putative periplasmic protein Putative periplasmic protein Inorganic pyrophosphatase (Ppa) Hypothetical protein
lS, 1T
t
+ + +
Kowarik et al. (2006b)
-
+
Kowarik et al. (2006b)
+ + + +
Linton et al. (2002) Kowarik et al. (2006b)
-
lS, 2T 1T 5.5, 1T 2s
-
+
+ + + + +
Kowarik et al. (2006b); Nita-Lazar et al. (2005); Wacker et al. (2002) Kowarik et al. (2006b) Kowarik et al. (2006b) Kowarik et al. (2006b)
-
3T 1T 1T
+
2S, 2T
+
-
-
+ + + + +
Kowarik et al. (2006b) Kowarik et al. (2006b) Kowarik et al. (2006b) Kowarik et al. (2006b)
Cj0694 Cj0715 CjO734c
Putative periplasmic protein Transthyretin-like periplasmic protein HisJ
4S, 2T
Cj0779 CjO835c CjO843c" Cj0906c CjO944c Cj09S8cd Cj0982cd CjO998c CjlOl8c
Thioredoxin peroxidase (Tpx) Aconitate hydratase (AcnB) Putative secreted transglycosylase Putative periplasmic protein Putative periplasmic protein Putative periplasmic protein Putative periplasmic protein Putative periplasmic protein Branched-chain amino-acid ABC transport system periplasmic binding protein Probable membrane fusion component of efflux system Putative membrane protein Translation elongation factor EF-Ts (Tsf) Hypothetical protein 60-kDa chaperonin (Cpn60; GroEL) PglB Probable periplasmic protein Probable periplasmic protein Putative capsule polysaccharide export system periplasmic protein (KpsD) Putative periplasmic protein Probable bacterioferritin Paralyzed flagellum protein (PflA) Putative periplasmic protein Putative periplasmic protein Periplasmic protein (P19) Probable periplasmic protein (CgpA) VirBlO
1s
Cj1032 Cj1053cd Cjll8 l c Cj1214c Cj1221 Cj1126c Cj1345c Cj1380 Cj1444c' Cj1496c Cj1534c Cj1565c Cj1621d Cj1643 Cj1659 Cj1670cd Cjp3
lS, 2T SS, 3T 2S, 2T lS, 1T
lS, 1T lS, 2T 2T
+
-
+
+ -
+
+ + + + +
+ +
Kowarik et al. (2006b) Kowarik et al. (2006b)
+ + +
+ -
1s
*
2T 3S, 2T
+ + +
lS, 1T
+
lS, 2T SS, 2T
5S, 2T 3S, 1T
4S, 2T
-
+ -
+
Kowarik et al. (2006b)
-
+ +
Kowarik et al. (2006b)
-
+
+ + + +
Kakuda and DiRita (2006)
-
Kowarik et al. (2006b)
-
i-
+
~~
S, N-X-S sequons; T, N-X-T sequons. Reactivity with SBA in Western blot analysis of two-dimensional gels. 'Proteins with changed spot position and/or immunoreacdvity in two-dimensional gels in the pgZB mutant. CjlOl8c and Cj1643 were not isolated by SBA chromatography. dGlycopeptides observed by CapLC-MS/MS. 'Identified from one-dimensional gel. (Extended glycosylation sequon present in the amino acid sequence of the corresponding protein (http: //www.sanger.ac.uk/Projects/C-jejuni/). a
Nita-Lazar et al. (2005); Wyszynska et al. (2007)
Linton et al. (2002) Larsen et al. (2004)
456
NOTHAFT ET AL.
showed that Cj1496c was located in the periplasm, the authors suggest that the observed effect is indirect (perhaps as a result of reduced signaling or transport activity), rather than a direct one that would lead to protein nonfunction due to loss of N-glycosylation. However, the authors only examined a small subset of the 340 potential glycoprotein targets (i.e., proteins that are predicted to be secreted and have the extended glycosylation sequon).
BIOLOGICAL EFFECTS OF DISRUPTING THE N-GLYCAN PATHWAY The variety of proteins that are modified with N-linked glycans suggests that mutation in the pgl pathway will lead to multiple pleiotropic effects. In addition to the observation that N-linked glycoproteins may exist in a nonglycosylated state in vivo (i.e., PEB3) and the fact that proteins can be modified at more than one glycosylation site (i.e., AcrA), there is also the possibility of varying the glycosylation pattern of orthologous proteins among different Campylobacter species. However, the conservation of the pathway among most Epsilonproteobacteria suggests
a common role for this modification in Cumpylobacter. Adherence, Invasion, and Immune Response
The immunodominant role of the protein glycans was first demonstrated in 1999 and was the observation that led to the conclusion that a general protein glycosylation pathway was identified (Szymanski et al., 1999). pgl gene mutation resulted in reduced protein antigenicity in Western blot experiments probed with CampylobacterPenner typing sera obtained after immunization of rabbits with live whole cells of the type strains of C. jejuni (Szymanski et al., 1999) and with sera obtained after infecting human volunteers with C. jejuni 8 1-176 (Szymanski et al., 2005; Fig. 2). Furthermore, the loss of antigenicity in protein extracts could be restored by complementation in trans. In addition, adherence and invasion of the human epithelial cell line (INT407) as well as mouse colonization studies showed that a pglB and a pglE mutant adhered at 38 and 59% and invaded at 4.4 and 9.2% of the level of the isogenic wild-type strain, respectively (Szymanski et al., 2002). In addition,
Figure 2. Summary of biological effects caused by disruption of protein N-glycosylation in C. jejuni.
CHAPTER 25
N-LINKED PROTEIN GLYCOSYLATION
457
both mutant strains demonstrated a significant reduction in mouse colonization studies as early as day 7 after infection, and colonization remained significantly low through 21 days. In both assays, complementation of pglE in trans restored wild-type adherence, invasion, and colonization at all time points examined. Similarly, a pglH mutant that was shown to be deficient in its ability to glycosylate a number of proteins in C. jejuni NCTC 11168H and 81116 showed reduced attachment to and invasion of Caco2 cells in vitro, suggesting a direct or indirect biological function for N-glycosylation in C. jejuni (Karlyshev et al., 2004). SBA reactivity, adhesion, and invasion properties were fully restored after complementing the NCTC 11168H pglH mutation, demonstrating that the observed changes are attributable to mutation in this gene.
observed, no differences were observed between pglI, pglG, and wild-type colonization levels. However, mutation in pglB, pglD, pglE, and pglK resulted in complete loss of colonization. The authors therefore concluded that under the conditions studied, PglG is not involved in N-linked glycosylation, and that the glucose branch is also not important for the colonization process. These results were corroborated with the observation that mutation in pglG and pglI did not affect the N-glycoprotein pattern of whole-cell lysates probed with SBA; MS analyses of the N-linked glycan structure on purified, native PEB3; and the Nglycan patterns of whole cells by high-resolution magic angle spinning NMR. The latter technique serves as a powerful method to detect N-glycans directly from intact bacterial cells (Szymanski et al., 2003).
Chicken Colonization
EXPRESSION OF pgl GENES I N CAMPYLOBACTER JEJUNI
Several studies investigated the importance of Nlinked glycosylation in the colonization and persistence of Campylobacter in poultry, the natural host of this zoonotic pathogen (Fig. 2). A genomewide analysis performed by signature-tagged transposon mutagenesis identified 22 different genes of C. jejuni required for proper cecal colonization of chicks (Hendrixson and DiRita, 2004). Among other genes, the authors demonstrated that mutation in pglE, pglF, and pglH displayed a 100- to 1,000-fold (and up to l o 6 in individual birds) reduction in colonization in a 1-day-old chicken model (Hendrixson and DiRita, 2004). An important role for the N-glycosylation pathway in chicken colonization was also shown in other studies. As described above, a C. jejuni 81116 pglH mutant had a reduced colonization potential (8,000-fold lower at a low-dose inoculation of lo5) 5 days after infection of 1-day-old chicks (Karlyshev et al., 2004). Even at a high dose of lo7 CFU, the pglH mutant colonized four- to sixfold less efficiently than the wild-type strain. In a long-term study, mutation in pglH resulted in reduced colonization levels but did not eliminate colonization by Campylobacter when 2-week-old birds were infected and final colonization levels were determined 6 weeks after inoculation (Jones et al., 2004). The authors therefore suggested that Campylobacter glycoproteins have a nonessential but important role in colonization. A thorough mutational analysis of the pgl operon and the resulting effects on AT-linked glycosylation and chicken colonization were recently described by Kelly and coworkers (2006). The effect of pgl gene mutation on the colonization levels of 1-day-old chicks was determined in a 1-week infection model. As expected, on the basis of the glycosylation patterns
In our recent study, semiquantitative reverse transcriptase-PCR experiments showed that all pgl genes of C. jejuni NCTC 11168 are transcribed in vivo (H. Nothaft et al., personal communication). In addition, at least one polycistronic transcript spanning the complete pgl operon (including pglG) could be demonstrated, corroborating the previous observation of a common pglEFG transcript in C. jejuni 81-176 (Szymanski et al., 1999). In addition, different amounts of pgl gene transcripts were observed in C. jejuni NCTC 11168, indicating the presence of various transcriptional units within the pgl operon and leading to the conclusion that promoter structures have to be present within the pgl operon. To prove the proposed existence of promoters within the pgl operon, DNA fragments containing at least 200 bp up- and downstream of all intergenic regions (if present) were transcriptionally fused to gfp (H. Nothaft et al., personal communication) present on an E. coli-C. jejuni shuttle plasmid (Miller et al., 2000). Subsequent fluorescence microscopy studies demonstrated at least four DNA fragments with promoter activity. These fragments contained the 5 ’ upstream region of gne, pglB, pglC, and pglE and corroborated 5 ’ rapid amplification of cDNA ends studies that were done in parallel. Interestingly, pglpromoter activities were different in E. coli. The most striking difference was observed for the region upstream of pglK with high promoter activity in E. coli, and basal levels of promoter activity in C. jejuni.
AUTOREGULATION OF pgi GENES Transcriptional profiles of all available pgl mutants were compared with the wild-type strain by a
458
N O T H A F T ET AL.
C. jejuni amplicon-based DNA microarray (H. Nothaft et al., personal communication) according to the method of Carrillo et al. (2004). Expression of more than 250 genes was altered in the pgl mutants when compared with the isogenic C. jejuni wild-type strain, and the profiles also differed among the pgl mutants. This indicates that the N-glycan pathway influences multiple cellular functions in C. jejuni, consistent with the observation that disruption of this pathway results in pleiotropic biological effects (see below). Interestingly, expression of genes of the pgl operon itself showed a different expression pattern in some of the pgl mutants (Table 2). Downregulation of different sets of pgl genes was observed in the gne, pglH, pglJ, and the pglK mutants, whereas there was no influence on pgl gene expression in pglA, pglB, pglD, pglE, pglF, pglG, and pgll. The observed expression pattern further indicates the existence of transcriptional units spanning gne-pglA, pglK-pgll, pglC-pglG, and pglE-pglG, thus supporting the reverse transcriptase-PCR and g@ data. Various transcripts within the pgl operon and variations in pgl gene expression have been suggested earlier. For example, transcriptional profiling studies showed that pgl gene expression responds to extracellular iron availability and temperature shifts and also showed expression differences in the rabbit animal model when compared with in vitro growth (Palyada et al., 2004; Stintzi, 2003; Stintzi et al., 2005). It will be interesting to identify and characterize the regulatory elements (cis-acting factors) involved in pgl gene expression that would assist in understanding the role of this sophisticated pathway in more detail. It should also be noted that microarray analyses provide a broad list of changes that then need to be further confirmed through mutant screening and functional assays. INVOLVEMENT OF Pgl PROTEINS IN OTHER CELLULAR FUNCTIONS Influence on the Expression of Genes Involved in Iron Acquisition Loss of N-glycosylation leads to altered transcription of a variety of genes involved in iron acquisition and iron storage, and genes that respond to the availability and presence of iron in the environment (H. Nothaft et al., personal communication; chapter 33) (Fig. 2). We found several genes encoding proteins for iron uptake systems and iron response to be downregulated in the pgl mutants (Table 2), among them Cj0178 (a putative TonB-dependent outer membrane receptor), CjO241c (a putative iron binding protein), Cj1628-1630 (the ExbB2-ExbD2-
TonB2 iron uptake system), Cj1658-Cj1663 (encoding a putative iron uptake ABC transporter), and CjO179-CjO181 (the ExbBl-ExbDl-TonB1 iron uptake system). Interestingly, Cj1614-Cj1617 (encoding a putative hemin uptake system; Ridley et al., 2006) was found to be significantly downregulated only in the pglD mutant. In contrast, CjOO45c (putative iron-binding protein) and Cj1398 (feoB, putative ferrous iron transport protein) were upregulated in most of the pg1 mutants. Genes encoding iron uptake systems were shown to be mostly organized in operons and regulated under the control of the Fur protein (Holmes et al., 2005; Naikare et al., 2006; Palyada et al., 2004; Ridley et al., 2006; van Vliet et al., 1998; see also Parkhill et al., 2000). Our data provide evidence for an even more sophisticated and mutual interplay between both systems on the transcriptional level, as suggested previously. On the basis of microarray and real-time PCR data, it was shown that pgl gene transcript abundance is affected by iron availability (Palyada et al., 2004), resulting in reduced lectin reactivity with proteins extracted from C. jejuni grown in iron-rich medium. A tight interplay between both systems in both directions is not surprising because iron availability and N-linked protein glycosylation are both necessary for invasion, colonization, and persistence of C. jejuni. It has been shown that mutation of the regulator of the iron regulon, Fur, significantly affected the organism’s ability to colonize the gastrointestinal tract of chicks (Palyada et al., 2004). Moreover, FeoB, shown to be crucial for Fe2+ uptake, plays an important role in Cumpylobacter pathogenesis (Naikare et al., 2006) because a feoB knockout strain was significantly affected in chick colonization and invasion and in intracellular survival in human INT407 cells and porcine IPEC-1 small intestinal epithelial cells, as well as in piglet gut colonization that mimics human infection. Amino Acid Transport and Biosynthesis The transcript abundance of genes involved in amino acid catabolism and several genes encoding subunits of putative amino acid transporters showed significant expression differences in some of the pgl mutants compared with the isogenic C. jejuni wildtype strain (Table 2). One of them, CjaA (Cj0982), is downregulated at least by a factor of 3 in gne and pglKHJBAEF and by a factor of 1.5 in the pgZID mutant, demonstrating that mutation in all pgl genes except pglG has an effect on cjaA transcription. CjaA, a conserved immunodominant protein (Pawelec et al., 1998), is suggested to be the binding protein component of an ABC-type cysteine transporter system
CHAPTER 25
(Muller et al., 2005). The importance of CjaA for the in vivo survival of Carnpylobacter has recently been shown in chicken colonization studies: birds immunized with an avirulent strain of Salmonella expressing plasmid-borne cjuA showed reduced C. jejuni colonization (Wyszynska et al., 2004). Interestingly, another ABC-type amino acid transporter permease complex encoded by CjO919c, CjO92Oc, and CjO1921c (Pebla) is significantly downregulated only in the gne mutant. The observed effect can therefore be regarded as directly influenced by the gne gene product. Besides, it has been shown that Pebla (CjO921c) is responsible for glutamate and aspartate transport, two amino acids that may function as important in vivo carbon sources for this pathogen (Leon-Kempis et al., 2006). Pebla is also an important factor in host colonization because mutation in pebla led to highly reduced adherence and invasion of epithelial cells in culture (Pei et al., 1998). Because C. jejuni is asaccharolytic because it lacks an active glycolytic pathway for the use of sugars as a carbon source, the organism relies on amino acids as nutrients. Therefore, loss of chicken colonization might be due, or at least to be influenced by, reduced amino acid transport in these respective pgl mutants (Fig. 2). Flagellin Biosynthesis and 0-Linked Glycosylation The transcription of genes involved in the biosynthesis of the flagellum and the 0-linked glycosylation pathway were found to be differentially expressed in the pgl mutants (H. Nothaft et al., personal communication; Table 2). Expression of genes encoding proteins for flagellar hook and basal-body rod formation and also of both structural genes (flaA, flaB) was significantly affected. The observed changes might be because some indirect effects resulting from a potential decrease in or even loss of function of some glycoproteins involved in flagellin biosynthesis (e.g., PflA; Yao et al., 1994; Young et al., 2002). Interestingly, the expression pattern of Cj1293 (PseB) and Cj1294 (PseC) showed significant variations among the pg1 mutants. PseB and PseC were found to be less expressed in pglADE when compared with the wild type, while the transcript abundance significantly increased in pglFHIJ. Cross-talk between the proteins of the Pgl pathway and the 0-linked flagella glycosylation pathway (the pseudaminic acid pathway) was reported recently (Schirm et al., 2005; Schoenhofen et al., 2006a, 2006b). The fact that synthesis of pseudaminic acid (Pse5Ac7AcYPse) for 0linked flagellin glycosylation (Schoenhofen et al., 2006b) and synthesis of the first sugar of the N-linked heptasaccharide, bacillosamine (Young et al., 2002), both start with the same UDP-GlcNAc-activated
N-LINKED PROTEIN GLYCOSnATION
459
sugar precursor (Schoenhofen et al., 2006a) implies such cross-talk between the two pathways. Figure 2 provides a complete summary of the biological effects caused by disruption of protein N-glycosylation.
IDENTIFICATION OF FREE GLYCANS N O T ATTACHED T O PROTEIN In addition to the detection and characterization of N-linked glycoproteins and peptides, an interesting observation was recently described by Liu and coworkers (2006). A new glycomics strategy based on the combination of nonspecific proteolytic digestion and permethylation was devised that can be used for both eukaryotic and bacterial glycoproteins. Complete pronase E digestion of whole-cell extracts was used to purify Asn-glycans by using porous graphitic carbon cartridges followed by permethylation and MS. This new method was used to examine N-linked glycoproteins in total protein extracts from C. jejuni NCTC 11168H. In addition to detecting the Asnlinked bacterial heptasaccharide, an unexpected free heptasaccharide intermediate that required a functional N-glycosylation pathway was observed. Surprisingly, the free heptasaccharide comprised approximately 90% of the total heptasaccharide detected and was shown to have the same structure as the purified N-linked heptasaccharide. Current studies are aimed at determining the role of the free heptasaccharides as a possible mechanism to recycle sugars from glycoproteins because C. jejuni is incapable of sugar uptake. Alternatively, the free oligosaccharide could be the result of the hydrolysis of lipid-linked oligosaccharide catalyzed by the oligosaccharyltransferase. This futile transfer to water instead of transfer to the amide of the asparagine residue might be of relevance to the flux control of the LLO biosynthetic pathway. Importantly, the committed step in this pathway takes place at the cytoplasmic side of the plasma membrane (transfer of bacillosamine phosphate to bactoprenylphosphate by PglC), whereas the final product of the pathway is located at the periplasmic side of the plasma membrane. This does not allow for a flux control of the pathway via feedback inhibition of PglC by the lipid-linked heptasaccharide product. However, the lipid carrier bactoprenol is used in other biosynthetic pathways (such as murein biosynthesis) and has to be recycled to the cytoplasm. The simplest but also the most energy-consuming solution to the problem represents the hydrolysis of excess LLO by oligosaccharyltransferase, thereby liberating the lipid carrier for further use.
Table 2. Transcriptional profiling of pgl mutants
N-linked glycosylation
Iron acquisition and storage
Cj1123c Cj 1124c Cj1125c Cj1126c Cj1127c Cj1128c Cj1129c Cjll3Oc Cjll3lc CjOO12c CjOO45c CjO147c CjO 174c
PdD PdC PdA PdB eg4
PdI PdH
PdK gne WC
CjOO45c trxA cf@B
Cj0178
Cj0178
Cj0179 CjOl80 CjOl81 CjO241c CjO265c
exbBl exbDl tonBl CjO241c CjO265c
CjO612c Cj1224 Cj1398 Cj1534c Cj1614
Cj1224 feoB Cj1534c chuA
Cjl615
chuB
Cj1616
chuC
Cj1617
chuD
Cj1628
exbB2
Cft
HexNAc acetyltransferase Bacillosamine (Bac) transferase GalNAcal-3 transferase Oligosaccharide transferase GlcNAca1-4 transferase GlcP1-3 transferase GalNAc 011-4transferase Flippase UDP-Glc/ GlcNAc epimerase Nonheme iron protein Putative iron-binding protein Thioredoxin Putative iron-uptake ABC transport system permease protein Putative TonB-dependent outer membrane receptor Biopolymer transport protein Biopolymer transport protein TonB transport protein Putative iron-binding protein Putative cytochrome c-type hemebinding periplasmic protein Ferritin Putative iron-binding protein Ferrous iron-transport protein Putative bacterioferritin Hemin uptake system outer membrane receptor Putative hemin uptake system permease protein Putative hemin uptake system ATPbinding protein Putative hemin uptake system periplasmic hemin-binding protein Putative exbB/tolQ family transport protein
-2.5 -2.6 -2.3 -2.6 -2.9 -3.9 -3.5 -2.7 2.6 2.6 -2.5
-2.0 -2.6
-2.9 -2.5 -2.4 -2.6
-2.1 -2.2 -3.0 -2.6
-2.5 -2.6 -3.0
-3.8 -3.6 -2.6 4.5
3.6
2.6 2.8
4.6 2.9
2.6
2.1 2.5
2.6
3.7
2.0 2.2
-2.4 -2.7 -2.4
-2.3
-2.6
-2.2
-2.7
-2.4 -2.6 -2.4
-3.2 -2.6 -2.2
-3.7 -4.1 -2.8
-2.7 -3.1 -2.5
-2.1 -2.5 -2.5
-3.6 -2.2 -2.8
-2.7
-2.8 2.0
2.2
2.4 2.2
-3.0 -3.2
2.2
2.5
2.3 2.1 2.3
5.4 4.2 2.8 2.3
-2.2 2.5 4.0 2.5
2.2 2.2
2.7 2.4 2.1
2.0
-3.9 -4.6 -2.0 -3.1 -2.6
-2.9
-2.5
-2.6
-2.2
-2.7
-2.1
Amino acid transport and metabolism
Cj1629
exbD2
Cj1630 Cj1658 Cj1659 Cj1660 Cj1661
tonB2 Cj1658 p19 Cj1660 Cj1661
Cj1662 Cj1663
Cj1662 Cj1663
(211664 CjOO21c
Cj1664 CjOO21c
Cj0282c CjO289c Cj0405 Cj0891c Cj0895c
5erB peb3 aroE serA aroA
Cj0903c Cj0919c
Cj0903c Cj0919c
CjO92Oc
CjO92Oc
Cj0921c
peblA
CjO935c
Cj0935c
CjO982c
cjaA
CjlOl4c
livF
CjlOl5c
livG
CjlOl6c
livM
Putative exbD1tolR family transport protein Putative TonB transport protein Putative iron permease Periplasmic protein p19 Putative integral membrane protein Possible ABC transport system permease Putative integral membrane protein Putative ABC transport system ATPbinding protein Putative periplasmic thioredoxin Putative fumarylacetoacetate (FAA) hydrolase family protein Putative phosphoserine phosphatase Major antigenic peptide PEB3 Shikimate 5-dehydrogenase D-3-Phosphoglyceratedehydrogenase 3-Phosphoshikimate 1carboxyvinyltransferase Putative amino acid transport protein Putative ABC-type amino acid transporter permease protein Putative ABC-type amino acid transporter permease protein Aspartatel glutamate-binding ABC transporter protein Putative sodium:amino acid symporter family protein Putative amino acid transporter periplasmic solute-binding protein Branched-chain amino acid ABC transport system ATP-binding protein Branched-chain amino acid ABC transport system ATP-binding protein Branched-chain amino acid ABC transport system permease protein
-3.2
-3.6
-2.9
-3.1
-2.4
-3.2
-2.1
-3.4
-2.2
-2.0
-2.2
-2.6 -2.4
-2.8 -2.9
-2.2 -3.0
-2.4 -2.7 -2.5 -2.5
-2.6 -2.5
-2.2 -3.0
-2.2
-2.5
-2.8
-2.4 -2.2
-2.2
-2.4
-2.4 -2.3 -2.4
-2.5 -2.1 -2.1
-3.1 -3.2
-2.2 -3.4
-3.0
-2.2
-2.6 3.4
-2.4
-2.2
-2.5
-2.1
-3.0
-2.0
-2.5
-2.0
-2.8
-2.2 -2.6 -3.1
-2.4 -2.4 -2.7
-2.8 -2.3
-2.7 -2.3 -2.1
-2.1 -3.1
2.0
-2.2
-2.5
-5.0 -3.8 2.8 -3
-3.4
-2.7
-2.9
-3.8
-3.0
-2.5
-3.4
-2.7
-2.7
-2.1
-3.2
-2.4
-2.6
-3.0
-3.2
-3.1
-3.0
-3.0
-3.4
Continued on following page
Table 2. Continued
Flagellum biosynthesis and glycosylation
CjlOl7c
livH
CjlOl8c
livK
CjlOl9c
liv]
Cjl20l
metE
Cj1502c Cjl503c Cj1716c
putP putA leuD
Cj1717c
leuC
Cj1718c Cj1719c Cj1727c
leuB leuA metB
Cj0041
fliK
Cj0042
flgD
Cj0043 CjO528c Cj0547 Cj0548 Cj0549
flgE flgB BUG
fliD fliS
Branched-chain amino acid ABC transport system permease protein Branched-chain amino acid ABC transport system, periplasmic binding protein Branched-chain amino acid ABC transport system periplasmic binding protein 5-Methyltetrahydropteroyltriglutamatehomocysteine methyltransferase Putative sodium/proline symporter Putative proline dehydrogenase 3-Isopropylmalate dehydratase small subunit 3-Isopropylmalate dehydratase large subunit 3-Isopropylmalate dehydrogenase 2-Isopropylmalate synthase Putative 0-acetylhomoserine (thio1)lyase Putative flagellar hook-length control protein Putative flagellar hook assembly protein Flagellar hook protein Flagellar basal-body rod protein Flagellar protein Flagellar hook-associated protein Flagellar protein
-3.6
-3.0
-3.8
-2.8
-3.4
-2.5
-2.5
-2.4
-2.4
-3.0
-3.2
-2.5
-2.0
-2.6
-2.4
2.5
2.9
2.5
2.6
-2.4
-2.1 -2.5 2.4
3.0
2.3
4.0
2.3
-3.0 -2.3 -2.3
2.6
2.3
2.7
3.6
2.4
-2.5
2.4 2.2
2.6 2.8 2.2
2.3 2.1
3.4 4.3 3.0
2.8 2.2 2.2
-2.5 -2.7
2.2 2.2
2.3
2.4 2.2
2.3
2.6
-2.0
-2.6
2.6
2.7
3.2
3.2
2.4
-2.4
-2.1
3.7
2.3 2.2 2.5 2.2 2.8
3.1
3.8 2.7 2.5 2.6 2
3.5 2.3 3.1 2.4 2.5
-2.0
-2.6
-2.2
-2.5 -2.6 -2.3
3.7 2.5 2.3 2.5 2.2
-2.4 -2.3
CjO687c
figH
Cj0697 Cj0698 CjO72Oc CjO882c CjO887c Cj1338c Cj1339c Cj1462 Cj1463 Cj1464 Cj1466
flgG2 figG
Cj1729c Cj1293
flgE2 pseB
Cj1294
pseC
Cj1298 Cj1316c
Cj1298 pseA
Cj1319 Cj1325 Cj1328
Cj1319 Cj1325 neuC2
Cj1329
Cj1329
Cj1333 Cj1337
pseD pseE
fluC
flhA fiuD RUB fiuA fig1 fig/
figM flgK
Putative flagellar L-ring protein precursor Flagellar basal-body rod protein Flagellar basal-body rod protein Flagellin Flagellar biosynthesis protein Putative flagellin Flagellin Flagellin Flagellar P-ring protein Hypothetical protein Cj1463 Hypothetical protein Cj1464 Putative flagellar hook-associated protein Flagellar hook subunit protein UDP-GlcNAc-specific C4,6 dehydratase/ C5 epimerase C4 aminotransferase specific for PseB product Putative N-acetyltransferase Pseudaminic acid biosynthesis PseA protein Putative nucleotide sugar dehydratase Putative methyltransferase Putative UDP-N-acetylglucosamine 2-epimerase Putative sugar-phosphate nucleotide transferase PseD protein PseE protein
2.1
2.5
4.4
2.8
-2.5
-2.5
4.2
2.1 2.3 2.3
2.1 2.7
3.7 4.2
3.1 4.6
-2.1 -2.2
-2.4 -2.7 -3.5
3.4 2.6
3.1 3.1 2.8 3.0
-2.6
2.6
2.4
4.0 2.5 2.7 2.5 2.5 2.1 3.4
-3.2
2.1
4.6 3 .O 5.8 4.8 2.3 2.0 3.4
-2.2 -2.6 -2.7 -2.4
-2.1 -3.2 -3.0 -2.7
3.3
2.3
-2.7
2.4 2.8 4.7
-2.6 -2.4 -2.9 -2.1
2.8 2.2 2.2 2.7
2.3 2.4
2.4
2.3 2.4
-2.1 -2.2
-2.6 -2.2
-3.2
2.4
2.2
2.8
2.0
-3.4
-2.8
-2.5
3.0
-2.2
2.3
-2.4
2.2 -2.4
-2.2 2.3
2.3 2.3
-2.7 -2.4
-2.8
-2.1
-3.2 -2.1
~
“Equivalent amounts of Cy3- and Cy5-labeled samples were used for hybridization using Cumpylobucter O W DNA microarrays as previously described (Carrillo et al., 2004; http: //ibs-isb.nrc-cnrc.gc.ca/glycobiology/ campychips-e.htm1). Quantification, signal normalization, and data visualization were performed by Array Pro Analyzer 4.5 (Media Cybernetics). Data represent the mean of three independent biological replicates. Clustering of genes with a log, ratio of at least ? 1.0 was performed with the TIGR MultiExperiment Viewer (MEV) software package (http: //www.tigr.org/software/microarray.shtml).Numbers indicate the fold change converted from log, data. -, genes with decreased expression in the pgl mutants; +, genes with increased expression in the pgl mutants.
464
NOTHAFT ET AL.
C.jejuni 11168
C. jejuni subsp. doylei 269.97 C. coli RM2228 C. lari RM2100 C. upsaliensis RM3195
C,hominis ATCC BAA-381
C. CUNUS 525.92 C. concisus 13826 C. fetus subsp. fetus 82-40
W. succinogenes DSZM 1740
Sulfurovum sp. NBC37-1 Nitratiruptor sp. SB55-2
D. desulfuricans G20
Figure 3. Genetic organization of pgl gene orthologs in Proteobacteria. The commonality of the conserved pgl gene clusters in Delta- and Epsilonproteobacteria is shown. Orthologs of Cj1119c-1132c from C. jejuni NCTC 11168 (NC-002163) were identified by the blastp or tblastx algorithm (http://www.ncbi.nlm.nih.gov/). The arrows indicate the transcriptional orientations of the genes; gaps between pgl genes are indicated by either the number of open reading frames (orfs) or slashed lines indicating orthologs that were found elsewhere in the chromosome. The pgl genes are conserved in all Cumpylobucter species examined, including C. jejuni subsp. doylei 269.97 (NC-009707), C. coli RM2228 (NZ~AAFL00000000),C. luri RM2100 (NZAAFK00000000), C. upsaliensis RM3195 (NZ~AAFJ00000000),C. hominis ATCC BAA-381 (NC-009714), C. curvus 525.92 (NC-009715), C. concisus 13826 (NZ~AAQZOOOOOOOO),and C. fetus subsp. fetus 82-40 (NC-008599). Pgl gene orthologs were also found in the related Epsilonproteobacteria: Wolinellu succinogenes DSZM 1740 (NC-005090), Sulfurovum sp. NBC37-1 (NC-009663), Nitrutiruptor sp. SB55-2 (NC-009662), and in the Deltaproteobacterium Desulfovibrio desulfuricuns G20 (NC-007519). Genes encoding the essential oligosaccharyltransferse PglB are depicted in black. Biosynthetic Pgl enzymes (Gne, PglE (E), PglF (F), PglD (D)), glycosyltransferases (PglA (A), PglJ PglH (H), PglI (I), PglC (C)), and the flanking gene products PglG (G) and WlaA were designated according to their orthologs in C. jejuni NCTC 11168 or as glycosyltransferase if no homology to any C. jejuni NCTC 11168 Pgl glycosyltransferase was found. PglC orthologs in Sulfurovum sp. NBC37-1 and Nitrutiruptor sp. SB55-2 are indicated with a question mark because other proteins with a higher percentage identity (ID) to PglC of C. jejuni NCTC 11168 can be found in the genome of both species. Note that the PglF ortholog protein in C. jejuni subsp. doylei 269.97 is annotated as a pseudogene that might be due to a sequencing error in the unfinished genome sequencing project. Orthologs to putative ABC-type transporters that show low homologies to PglK (K) of C. jejuni NCTC 11168 are indicated by a question mark.
u),
PUTATIVE N-GLYCOSYLATION PATHWAYS IN PROTEOBACTEIUA The commonality of the pgl genes among various strains of C. jejuni and C. coli was demonstrated previously through hybridization studies comparing 81-176 with the Penner-type strains HS:1, HS:3, HS: 4, HS:19, and HS:30 (Szymanski et al., 1999). Remarkably, by means of high-resolution magic angle spinning NMR, the N-glycan structure was shown to be readily detectable from 4O-pl aliquots of intact bacterial cells (Szymanski et al., 2003). By means of
this methodology, the same heptasaccharide structure was shown to also be present in the C. jejuni HS:19 and C. coli HS:30 strains. In addition, gene clusters corresponding to the N-linked protein glycosylation pathway were shown to be present in various isolates of C. jejuni (Dorrell et al., 2001; Fouts et al., 2005; Leonard et al., 2003; Pearson et al., 2003; Taboada et al., 2004), C. lari RM2100, C. upsaliensis RM3195 (Fouts et al., 2005), C. jejuni subsp. doylei 269.97, C. coli RM2228, C. hominis ATCC BAA381, C. curvus 525.92, C. concisus 13826, and C. fetus subsp. fetus 82-40, demonstrating that this
Table 3. C. jejuni NTCT 11168 pg1 gene orthologs Protein PglG PglF PglE WlaJ PglD PglC PglA PglB PglJ PglI PglH
Putative function Unknown
Cj269.97 (%ID)
C j l l l 9 c JJD269970602 (100) (100) HexNAc Cjll2Oc JJD269970600 dehydratase (100) (97) HexNAc C j l l 2 l c JJD269970599 amino TF (100) (97) Unknown Cj1122c (100) HexNAc acetyl Cj1123c JJD269970598 TF (100) (95) BacTF Cj1124c JJD269970597 (100) (94) GalNAcal-3 Cj1125c JJD269970596 TF (100) (96) Oligosaccharyl- Cj1126c JJD269970595 transferase (100) (97) GlcNAcal-4 Cj1127c JJD269970594 TF (100) (97) Glcpl-3 TF Cj1128~ JJD269970593 (100) (98) GalNAcal-4 Cj1129c JJD269970592 TF (100) (97)
PglK
Flippase
Gne
Glc/GlcNAc epimerase Unknown
WlaA
Cj11168 (VoID)
Cc2228 C12100 Cu3195 (VoID) (%ID) (%ID)
Ch381 (%ID)
Cc525.92 (Yo ID)
Cc13826 ( O h ID)
Cf82-40 (%ID)
W~1740 NBC37-1 (%ID) (%ID)
SB155-2 (%ID)
DdG20 (VoID)
Cu0156 CHAB3810948 CCV525921201 CCC138260446 CFF82401353 Ws0616 SUN-1080 (79) (51) (58) (57) (55) (46) (42) Cc1069 C11207 Cu0155 CHAB3810949 CCV525921202 CCC138260447 CFF82401354 Ws0036 SUN-0106 (47) (59) (59) (57) (49) (85) (77) (81) (58) Cc1070 C11206 CuOl54 CHAB3810950 CCV.525921203 CCC138260448 CFF82401355 Ws0037 SUN-0105 (51) (60) (57) (89) (59) (61) (59) (80) (73)
NIL1413 (48) NIL1247 Ddes5173 (50) (35) NIL1248 Ddes5172 (52) (44)
Cc1071 (68) Cc1072 (83) Cc1073 (71) Cc1074 (81) Ccl07.5 (89) Cc1076
Ws0038 (36) Ws0039 SUN-0104 (65) (32) Ws0040 (41) Ws0043 SUN-0103 (38) (37) Ws0044 (40)
Ddes4978 (24) NIL1249 Ddes5169 (34) (49) Ddes5169 (25) NIS-1250 Ddes1870 (34) (16) Ddes1667 (31) Ddes1673 (37)
CFF82400787 Ws0052 SUN-1943 (27) (49) (28) CFF82401392 Ws0025 SUN-1933 (60) (25) (37) CFF82401398 Ws0615 (51) (50)
NIS-1102 Ddes1854 (28) (26) NIL1416 Ddes2361 (23) (43) NIS 1414
C11205 (61) C11204 (73) C11203 (62) C11202 (57) C11201 (71)
Cu0129 (69) Cu0128 (82) Cu0127 (71) Cu0126 (57) Cu0125 (80)
Cu0124 (85) (72) Cc1077 C11200 Cu0123 (84) (65) (69)
Cjll3Oc JJD269970591 Cc1078 C11199 Cu0122 (100) (97) (86) (67) (72) C j l l 3 l c JJD269970590 Cc1079 ‘211198 Cu0120 (100) (98) (93) (83) (87) Cj1132c JJD269970589 Cc1080 C11197 Cu0119 (100) (94) (81) (68) (72)
CHAB3810951 (50) CHAB3810952 (71) CHAB3810953 (49) CHAB3810954 (46) CHAB3810955 (54) CHAB3810957 (31) CHAB3810958 (40) CHAB3810956 (30) CHAB3810580 (30) CHAB3810783 (60) CHAB3810781 (40)
CCV525921210 (49) CCVS25921215 (68) CCV525921212 (48) CCV525921213 (48) CCV525921214 (51) CCV525921217 (41) CCV525921215 (39) CCVS25921216 (39) CCV525921141 (27) CCV525921224 (62) CCV525921227 (50)
CCC138260449 (52) CCC138260450 (71) CCC138260451 (47) CCC138260452 (47) CCC138260453 (48) CCC138260456 (38) CCC138260455 (42) CCC138260455 (38) CCC138260117 (27) CCC138260464 (60) CCC138260469 (52)
CFF82401380 (47) CFF82401381 (70) CFF82401382 (45) CFF82401383 (45) CFF82401384 (58)
CFF82401385 (36) CFF82401384
(38)
(50)
“ID, amino acid identity; Cj11168, Campylobacter jejuni NTCT 11168; Cj269.97, Campylobacter jejuni subsp. doylei 269.97; Cc2228; Campylobacter coli RM2228; C12100, Campylobacter lari RM2100; Cu3195, Campylobacter upsaliensis RM3195; Ch381, Campylobacter hominis ATCC BAA-381; CcS25.92, Campylobacter curvus 525.92; Cc13826, Campylobacter concisus 13826; Cf82-40, Campylobacter fetus subsp. fetus 82-40; Ws1740, Wolinella succinogenes DSZM 1740; NBC37-1, Sulfurovum sp. NBC37-1; SB155-2, Nitratiruptor sp. SB55-2; DdG20, Desulfovibrio desulfuricans G20; HexNAc, N-acetyl-hexosamine; Bac, 2,4diacetamido-2,4,6-trideoxyglucose; GalNAc, N-acety-galactosamine; Glc, glucose.
466
NOTHAFT ET AL.
pathway and potentially the bacillosamine-containing heptasaccharide are conserved among all Campylobucter species. The conservation of the pgl gene clusters in Delta- and Epsilonproteobacteria is shown in Fig. 3, which includes all Campylobacter species sequenced to date; the Epsilonproteobacteria: Wolinella succinogenes DSZM 1740, Sulfurovum sp. NBC37-1, and Nitratiruptor sp. SB55-2; and in the Deltaproteobacterium Desulfovibrio desulfuricans. In all these organisms, the critical oligosaccharyltransferase, PglB, is conserved, and the catalytic motif WWDYG (WWDWG in D. desulfuricans) is present. The closest orthologs to the membrane-associated flippase, PglK, were found in C. jejuni NCTC 11168 subsp. doylei 269.97, C. coli RM2228, C. lari RM2100, C. upsaliensis RM3195 (67 to 97% identification) and in W. succinogenes (49% identification) (Table 3). In the other species, the putative transporter (indicated by a question mark in Fig. 3 ) showed the highest homologies (27 to 30% identification) to the lipid export ABC transport protein, MsbA. Interestingly, these transporters are flanked by a putative cysteine synthase (cysK) and a putative dihydroorotate dehydrogenase (pyrD) in C. concisus 13826, C. curvus 525.9, C. fetus subsp. fetus 82-40, and C. hominis, indicating a common genomic root for these four Campylobacter species.
OUTLOOK AND APPLICATIONS Glycoproteins and the associated glycan structures are ubiquitous biomolecules involved in many biological processes ranging from cell recognition to disease pathogenesis. In contrast to polypeptides and nucleic acids, glycoproteins have escaped biotechnological applications so far. This is primarily because of the complexity of the biosynthetic pathways in eukaryotic cells. However, advances in analytical methods and genome sequencing have demonstrated the presence of both 0-linked and N-linked protein glycosylation pathways in bacterial systems, which is leading to a new era of homogeneous glycoprotein engineering that can be used in diverse applications such as disease diagnosis and intervention. To achieve this goal, a thorough knowledge of bacterial glycosylation machineries and sophisticated methods for glycan analyses will be required. C. jejuni provides researchers with an excellent model system because this organism has both well-characterized 0-linked and N-linked protein glycosylation systems. In addition, the C. jejuni N-linked pathway can be functionally transferred into Escherichiu coli to modify recombinant proteins. Furthermore, the sole Campylobacter oligosaccharyltransferase required for pro-
tein modification, PglB, has been shown to have relaxed specificity for its sugar donors. These observations open up the possibility for a new era of protein glycoengineering in bacteria such as E. coli. However, in order to take complete advantage of converting E. coli into an efficient glycofactory, it is important to fully understand the process in the native host. Recent studies are demonstrating the complexities of N-glycosylation in C. jejuni and identifying new techniques that will enable optimization of the recombinant glycofactories. Furthermore, additional promiscuous glycosyltransferases and biosynthetic enzymes are being identified that expand the glycoengineering toolbox, so that in the not-so-distant future, humanized therapeutic glycoproteins and bacterial glycoconjugates can be manufactured at an industrial scale. In addition, the ability of C. jejuni to produce large quantities of free glycan, which again could potentially be manipulated to form complex novel glycans, could be further exploited in chemoenzymatic approaches to produce large quantities of oligosaccharides that are synthetically challenging, that are expensive, and that would otherwise generate large amounts of chemical waste. In parallel, development of sensitive techniques such as the detection of lipid-linked intermediates (chapter 29) not only allows us to examine the C. jejuni pathway in greater detail, but also assists in optimizing the recombinant glycofactories and has great potential in further characterizing human congenital disorders of glycosylation and in identifying new disorders affecting glycan assembly at the lipid level. We are indeed in a new era of bacterial glycobiology, which has the possibility of making a significant impact in changing the future. Acknowledgments. We thank Tom Devecserci for help with figures and Bruno Lacelle for the graphics in Fig. 2. This work has been supported by the NRC Genomics and Health Initiative to C.S. and by grants from the Swiss National Science Foundation to M.A.
REFERENCES Alaimo, C., I. Catrein, L. Morf, C. L. Marolda, N. Callewaert, M. A. Valvano, M. F. Feldman, and M. Aebi. 2006. Two distinct but interchangeable mechanisms for flipping of lipid-linked oligosaccharides. EMBO J. 25:967-976. Bacon, D. J., R. A. Alm, D. H. Burr, L. Hu, D. J. Kopecko, C. P. Ewing, T. J. Trust, and P. Guerry. 2000. Involvement of a plasmid in virulence of Campylobacter jejuni 8 1-176. Infect. Immun. 68 :43 84-4390. Bernatchez, S., C. M. Szymanski, N. Ishiyama, J. Li, H. C. Jarrell, P. C. Lau, A. M. Berghuis, N. M. Young, and W. W. Wakarchuk. 2005. A single bifunctional UDP-GlcNAc/ Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni. J. Biol. Cbem. 280:4792-4802. Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R. Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K.
CHAPTER 25
Amoako, S . Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Genome-wide expression analyses of Campylobacter jejuni NC TC lll68 reveals coordinate regulation of motility and virulence by flu.J. Biol. Chem. 279:20327-203 38. Creuzenet, C. 2004. Characterization of CJ1293, a new UDPGlcNAc C6 dehydratase from Campylobacter jejuni. FEBS Lett. 559:136-140. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. A1 Ghusein, B. G. Barrell, J. Parkhill, N. G. Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Campylobacter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11~1706-1715. Feldman, M. F., M. Wacker, M. Hernandez, P. G . Hitchen, C. L. Marolda, M. Kowarik, H. R. Morris, A. Dell, M. A. Valvano, and M. Aebi. 2005. Engineering N-linked protein glycosylation with diverse 0 antigen lipopolysaccharide structures in Escherichia coli. Proc. Natl. Acad. Sci. USA 102:3016-3021. Fouts, D. E., E. F. Mongodin, R E. Mandrell, W. G . Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:l-14. Fry, B. N., V. Korolik, J. A. ten Brinke, M. T. Pennings, R. Zalm, B. J. Teunis, P. J. Coloe, and B. A. van der Zeijst. 1998. The lipopolysaccharide biosynthesis locus of Campylobacter jejuni 81116. Microbiology 144(Pt. 8):2049-2061. Gabius, H.-J., and F. Sinowatz. 1998. Preface. Actu Anat. 161(14):6. Gemmill, T. R. and R. B. Trimble. 1999. Overview of N- and 0linked oligosaccharide structures found in various yeast species. Biochim. Biophys. Acta 1426:227-237. Glover, K. J., E. Weerapana, M. M. Chen, and B. Imperiali. 2006. Direct biochemical evidence for the utilization of UDP-bacillosamine by PglC, an essential glycosyl-1-phosphate transferase in the Campylobacter jejuni N-linked glycosylation pathway. Biochemistry 455343-5350. Glover, K. J., E. Weerapana, and B. Imperiali. 2005a. Glycosyl donor biosynthesis for prokaryotic N-linked glycosylation: in vitro assembly of an undecaprenylpyrophosphate-linkedheptasaccharide. Proc. Natl. Acad. Sci. USA 102:14255-14259. Glover, K. J., E. Weerapana, and B. Imperiali. 2005b. In vitro assembly of the undecaprenylpyrophosphate-linked heptasaccharide for prokaryotic N-linked glycosylation. Proc. Natl. Acad. Sci. USA 102:14255-14259. Glover, K. J., E. Weerapana, S. Numao, and B. Imperiali. 2005c. Chemoenzymatic synthesis of glycopeptides with PglB, a bacterial oligosaccharyl transferase from Campylobacter jejuni. Chem. Biol. 12:13 11-1315. Goon, S., J. F. Kelly, S. M. Logan, C. P. Ewing, and P. Guerry. 2003. Pseudaminic acid, the major modification on Campylobacter flagellin, is synthesized via the Cj1293 gene. Mol. Microbiol. 50:659-671. Gundogdu, O., S . D. Bentley, M. T. Holden, J. Parkhill, N. Dorrell, and B. W. Wren. 2007. Re-annotation and re-analysis of the Campylobacter jejuni NC TC lll68 genome sequence. BMC Genomics 8:162. Hanover, J. A., and W. J. Lennarz. 1979. The topological orientation of N,N'-diacetylchitobiosylpyrophosphoryldolicholin artificial and natural membranes. J. Biol. Chem. 254:9237-9246. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484.
N-LINKED PROTEIN GLYCOSYLATION
467
Hofreuter, D., S . Odenbreit, and R. Haas. 2001. Natural transformation competence in Helicobacter pylori is mediated by the basic components of a type IV secretion system. Mol. Microbiol. 4 1:379-39 1. Hofreuter, D., S . Odenbreit, G. Henke, and R. Haas. 1998. Natural competence for DNA transformation in Helicobacter pylori: identification and genetic characterization of the comB locus. Mol. Microbiol. 28: 1027-103 8. Holmes, K., F. Mulholland, B. M. Pearson, C. Pin, J. McNichollKennedy, J. M. Ketley, and J. M. Wells. 2005. Campylobacter jejuni gene expression in response to iron limitation and the role of Fur. Microbiology 151:243-257. Jennings, M. P., M. Virji, D. Evans, V. Foster, Y. N. Srikhanta, L. Steeghs, L. P. van der, and E. R. Moxon. 1998. Identification of a novel gene involved in pilin glycosylation in Neisseria meningitidis. Mol. Microbiol. 29:975-984. Jones, M. A., K. L. Marston, C. A. Woodall, D. J. Maskell, D. Linton, A. V. Karlyshev, N. Dorrell, B. W. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni NCTClll68 to high-level colonization of the avian gastrointestinal tract. Infect. lmmun. 72:3769-3776. Kakuda, T., and V. J. DiRita. 2006. Cj1496c encodes a Campylobacter jejuni glycoprotein that influences invasion of human epithelial cells and colonization of the chick gastrointestinal tract. Infect. Immun. 74:4715-4723. Karlyshev, A. V., P. Everest, D. Linton, S . Cawthraw, D. G. Newell, and B. W. Wren. 2004. The Campylobacter jejuni general glycosylation system is important for attachment to human epithelial cells and in the colonization of chicks. Microbiology 150: 1957-1964. Kelleher, D. J., and R. Gilmore. 1994. The Saccharomyces cereuisiae oligosaccharyltransferase is a protein complex composed of Wbplp, Swplp, and four additional polypeptides. J. Bid. Chem. 269:12908-12917. Kelly, J., H. Jarrell, L. Millar, L. Tessier, L. M. Fiori, P. C. Lau, B. Allan, and C. M. Szymanski. 2006. Biosynthesis of the Nlinked glycan in Campylobacter jejuni and addition onto protein through block transfer. J. Bacteriol. 188:2427-2434. Kowarik, M., S . Numao, M. F. Feldman, B. L. Schulz, N. Callewaert, E. Kiermaier, I. Catrein, and M. Aebi. 2006a. N-linked glycosylation of folded proteins by the bacterial oligosaccharyltransferase. Science 3 14: 1148-1 150. Kowarik, M., N. M. Young, S. Numao, B. L. Schulz, I. Hug, N. Callewaert, D. C. Mills, D. C. Watson, M. Hernandez, J. F. Kelly, M. Wacker, and M. Aebi. 2006b. Definition of the bacterial N-glycosylation site consensus sequence. EMBO J. 25: 1957-1966. Larsen, J. C., C. Szymanski, and P. Guerry. 2004. N-linked protein glycosylation is required for full competence in Campylobacter jejuni 81-176.1. Bacteriol. 186:6508-6514. Leon-Kempis, M. R., E. Guccione, F. Mulholland, M. P. Williamson, and D. J. Kelly. 2006. The Campylobacter jejuni PEBla adhesin is an aspartate/glutamate-binding protein of an ABC transporter essential for microaerobic growth on dicarboxylic amino acids. Mol. Microbiol. 60:1262-1275. Leonard, E. E., T. Takata, M. J. Blaser, S. Falkow, L. S . Tompkins, and E. C. Gaynor. 2003. Use of an open-reading framespecific Campylobacter jejuni DNA microarray as a new genotyping tool for studying epidemiologically related isolates. J. Infect. Dis. 187:691-694. Linton, D., E. Allan, A. V. Karlyshev, A. D. Cronshaw, and B. W. Wren. 2002. Identification of N-acetylgalactosamine-containing glycoproteins PEB3 and CgpA in Campylobacter jejuni. Mol. Microbiol. 43 :497-5 08. Linton, D., N. Dorrell, P. G. Hitchen, S . Amber, A. V. Karlyshev, H. R. Morris, A. Dell, M. A. Valvano, M. Aebi, and B. W.
468
NOTHAFT ET AL.
Wren. 2005. Functional analysis of the Campylobacter jejuni Nlinked protein glycosylation pathway. Mol. Microbiol. 55:16951703. Liu, X., D. J. McNally, H. Nothaft, C. M. Szymanski, J. R. Brisson, and J. Li. 2006. Mass spectrometry-based glycomics strategy for exploring N-linked glycosylation in eukaryotes and bacteria. Anal. Chem. 78:6081-6087. McCloskey, M. A., and F. A. Troy. 1980. Paramagnetic isoprenoid carrier lipids. 2. Dispersion and dynamics in lipid membranes. Biochemistry 19:2061-2066. Menon, A. K. 1995. Flippases. Trends Cell Biol. 5:355-360. Miller, W. G., A. H. Bates, S. T. Horn, M. T. Brand], M. R. Wachtel, and R. E. Mandrell. 2000. Detection on surfaces and in Caco-2 cells of Campylobacter jejuni cells transformed with new gfp, yfp, and cfp marker plasmids. Appl. Environ. Microbiol. 66~5426-5436. Muller, A., G. H. Thomas, R. Horler, J. A. Brannigan, E. Blagova, V. M. Levdikov, M. J. Fogg, K. S. Wilson, and A. J. Wilkinson. 2005. An ATP-binding cassette-type cysteine transporter in Campylobacter jejuni inferred from the structure of an extracytoplasmic solute receptor protein. Mol. Microbiol. 57:143155. Naikare, H., K. Palyada, R. Panciera, D. Marlow, and A. Stintzi. 2006. Major role for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infect. Immun. 745433-5444. Nita-Lazar, M., M. Wacker, B. Schegg, S. Amber, and M. Aebi. 2005. The N-X-SIT consensus sequence is required but not sufficient for bacterial N-linked protein glycosylation. Glycobiology 15:361-367. Olivier, N. B., M. M. Chen, J. R. Behr, and B. Imperiali. 2006. In vitro biosynthesis of UDP-N,N'-diacetylbacillosamine by enzymes of the Campylobacter jejuni general protein glycosylation system. Biochemistry 45 :13659-13669. Palyada, K., D. Threadgill, and A. Stintzi. 2004. Iron acquisition and regulation in Campylobacter jejuni. J. Bacteriol. 186:47144729. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pawelec, D., J. Jakubowska-Mroz, and E. K. Jagusztyn-Krynicka. 1998. Campylobacter jejuni 72Dz/92 cjaC gene coding 28 kDa immunopositive protein, a homolog of the solute-binding components of the ABC transport system. Lett. Appl. Microbiol. 26: 69-76. Pearson, B. M., C. Pin, J. Wright, K. I'Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS Lett. 554~224-230. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. Z. Huang, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the pebZA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. Immun. 66:938-943. Pei, Z. H., R. T. Ellison, 111, and M. J. Blaser. 1991. Identification, purification, and characterization of major antigenic proteins of Campylobacter jejuni. J. Biol. Chem. 266:16363-16369. Power, P. M., L. F. Roddam, M. Dieckelmann, Y. N. Srikhanta, Y. C. Tan, A. W. Berrington, and M. P. Jennings. 2000. Genetic characterization of pilin glycosylation in Neisseria meningitidis. Microbiology 146:967-979.
Rangarajan, E. S., S. Bhatia, D. C. Watson, C. Munger, M. Cygler, A. Matte, and N. M. Young. 2007. Structural context for protein N-glycosylation in bacteria: the structure of PEB3, an adhesin from Campylobacter jejuni. Protein Sci. 16:990-995. Ridley, K. A., J. D. Rock, Y. Li, and J. M. Ketley. 2006. Heme utilization in Campylobacter jejuni. 1. Bacteriol. 188:78627875. Schirm, M., I. C. Schoenhofen, S. M. Logan, K. C. Waldron, and P. Thibault. 2005. Identification of unusual bacterial glycosylation by tandem mass spectrometry analyses of intact proteins. Anal. Chem. 77:7774-7782. Schoenhofen, I. C., D. J. McNally, J. R. Brisson, and S. M. Logan. 2006a. Elucidation of the CMP-pseudaminic acid pathway in Helicobacter pylori: synthesis from UDP-N-acetylglucosamineby a single enzymatic reaction. Glycobiology 16:8C-l4C. Schoenhofen, I. C., D. J. McNally, E. Vinogradov, D. Whitfield, N. M. Young, S. Dick, W. W. Wakarchuk, J. R. Brisson, and S. M. Logan. 2006b. Functional characterization of dehydratasei aminotransferase pairs from Helicobacter and Campylobacter: enzymes distinguishing the pseudaminic acid and bacillosamine biosynthetic pathways. J. Biol. Chem. 281:723-732. St Michael, F., C. M. Szymanski, J. Li, K. H. Chan, N. H. Khieu, S. Larocque, W. W. Wakarchuk, J. R. Brisson, and M. A. Monteiro. 2002. The structures of the lipooligosaccharide and capsule polysaccharide of Campylobacter jejuni genome sequenced strain NCTC 11168. Eur. J. Biochem. 2695119-5136. Stimson, E., M. Virji, K. Makepeace, A. Dell, H. R. Morris, G. Payne, J. R. Saunders, M. P. Jennings, S. Barker, M. Panico, I. Blench, and R. E. Moxon. 1995. Meningococcal pilin: a glycoprotein substituted with digalactosyl 2,4-diacetamido-2,4,6trideoxyhexose. Mol. Microbiol. 17:1201-1214. Stintzi, A. 2003. Gene expression profile of Campylobacter jejuni in response to growth temperature variation. J. Bacteriol. 185: 2009-2016. Stintzi, A., D. Marlow, K. Palyada, H. Naikare, R. Panciera, L. Whitworth, and C. Clarke. 2005. Use of genome-wide expression profiling and mutagenesis to study the intestinal lifestyle of Campylobacter jejuni. Infect. Immun. 73:1797-1810. Szymanski, C. M., D. H. Burr, and P. Guerry. 2002. Campylobacter protein glycosylation affects host cell interactions. Infect. lmmun. 70:2242-2244. Szymanski, C. M., S. Goon, B. Allan, and P. Guerry. 2005. Protein glycosylation in Campylobacter, p. 259-273. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norwich, United Kingdom. Szymanski, C. M., F. S. Michael, H. C. Jarrell, J. Li, M. Gilbert, S. Larocque, E. Vinogradov, and J. R Brisson. 2003. Detection of conserved N-linked glycans and phase-variable lipooligosaccharides and capsules from Campylobacter cells by mass spectrometry and high resolution magic angle spinning NMR spectroscopy. ]. Biol. Chem. 278:24509-24520. Szymanski, C. M., R. Yao, C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evidence for a system of general protein glycosylation in Campylobacter jejuni. Mol. Microbiol. 32:1022-103 0. Taboada, E. N., R. R. Acedillo, C. D. Carrillo, W. A. Findlay, D. T. Medeiros, 0. L. Mykytczuk, M. J. Roberts, C. A. Valencia, J. M. Farber, and J. H. Nash. 2004. Large-scale comparative genomics meta-analysis of Campylobacter jejuni isolates reveals low level of genome plasticity. 1. Clin. Microbiol. 42:45664576. van Vliet, A. H., K. G. Wooldridge, and J. M. Ketley. 1998. Ironresponsive gene regulation in a Campylobacter jejuni fur mutant. J. Bacteriol. 1805291-5298. Vijayakumar, S., A. Merkx-Jacques, D. B. Ratnayake, I. Gryski, R. K. Obhi, S. Houle, C. M. Dozois, and C. Creuzenet. 2006. Cjl121c, a novel UDP-4-keto-6-deoxy-GlcNAc C-4 aminotrans-
CHAPTER 25
ferase essential for protein glycosylation and virulence in Campylobacter jejuni. J. Biol. Chem. 281:27733-27743. Wacker, M., M. F. Feldman, N. Callewaert, M. Kowarik, B. R. Clarke, N. L. Pohl, M. Hernandez, E. D. Vines, M. A. Valvano, C. Whitfield, and M. Aebi. 2006. Substrate specificity of bacterial oligosaccharyltransferase suggests a common transfer mechanism for the bacterial and eukaryotic systems. Proc. Nutl. Acud. Sci. USA 103:7088-7093. Wacker, M., D. Linton, P. G. Hitchen, M. Nita-Lazar, S. M. Haslam, S. J. North, M. Panico, H. R Morris, A. Dell, B. W. Wren, and M. Aebi. 2002. N-linked glycosylation in Campylo6acter jejuni and its functional transfer into E. coli. Science 298:17901793. Weerapana, E., K. J. Glover, M. M. Chen, and B. Imperiali. 2005. Investigating bacterial N-linked glycosylation: synthesis and glycosyl acceptor activity of the undecaprenyl pyrophosphatelinked bacillosamine. J. Am. Chem. SOC.127:13766-13767. Weerapana, E., and B. Imperiali. 2006. Asparagine-linked protein glycosylation: from eukaryotic to prokaryotic systems. Glycobiology 16:9 l R- 10l R. Wyszynska, A., A. Raczko, M. Lis, and E. K. Jagusztyn-Krynicka. 2004. Oral immunization of chickens with avirulent Salmonella
N-LINKED PROTEIN GLYCOSYLATION
469
vaccine strain carrying C. jejuni 72Dz/92 cjaA gene elicits specific humoral immune response associated with protection against challenge with wild-type Campylobacter. Vaccine 22: 1379-1 3 89. Wyszynska, A., K. Tomczyk, and E. K. Jagusztyn-Krynicka. 2007. Comparison of the localization and post-translational modification of Cumpylo6ucter coli CjaC and its homolog from Cumpylobucter jejuni, Cj0734c/HisJ. Acta Biochim. Pol. 54:143150. Yan, Q., and W. J. Lennarz. 2002. Studies on the function of oligosaccharyl transferase subunits. Stt3p is directly involved in the glycosylation process. J. Biol. Chem. 277:47692-47700. Yao, R., D. H. Burr, P. Doig, T. J. Trust, H. Niu, and P. Guerry. 1994. Isolation of motile and non-motile insertional mutants of Campylobacter jejuni: the role of motility in adherence and invasion of eukaryotic cells. Mol. Microbiol. 14:883-893. Young, N. M., J. R. Brisson, J. Kelly, D. C. Watson, L. Tessier, P. H. Lanthier, H. C. Jarrell, N. Cadotte, F. St Michael, E. Aberg, and C. M. Szymanski. 2002. Structure of the N-linked glycan present on multiple glycoproteins in the gram-negative bacterium, Campylobacter jejuni. J. Biol. Chem. 277:4253042539.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C . M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 26
0-Linked Flagellar Glycosylation in Campylobacter SUSAN
M. LOGAN,IAN
c. SCHOENHOFEN,AND PATRICIA GUERRY
Campylobacter and related organisms are actively motile by means of a polar flagella, and this motility, which is critical to virulence, allows them to colonize the mucous lining of the gastrointestinal tract in both animal and human hosts (Black et al., 1988; Caldwell et al., 1985; Hendrixson and DiRita, 2004). Unlike the extensively characterized flagellar systems of Escherichia coli and Salmonella enterica serovar Typhimurium where the flagellar filament is composed of a single structural protein (simple flagella), the flagella produced by Campylobacter are complex in nature, being composed of two flagellin structural proteins, FlaA and FlaB (Guerry et al., 1990, 1991). It is only relatively recently that the complex flagella from a number of organisms, including Campylobacter, have been characterized and in many cases shown to be glycosylated (Logan, 2006). Protein glycosylation has long been recognized as an important posttranslational modification in eukaryotic systems and one that imparts unique and diverse biological functions to the respective proteins. Although there is a considerable gap in our knowledge on the process of 0-linked glycosylation in prokaryotes as a result of the significant glycan diversity among prokaryotes, the 0-linked flagellar glycosylation system of Campylobacter has received considerable attention and is one of the more detailed prokaryotic systems studied to date.
gion of the primary amino acid sequence were modified in this analysis (Logan et al., 1989). Initial evidence for glycosylation as the posttranslational modification was provided by using periodate oxidation and lectin affinity studies with the purified protein (Doig et al., 1996). By means of mild periodate and biotin hydrazide labeling, flagellin from C. coli, C. jejuni, and C. fetus were shown to be glycosylated. Reactivity of each of these flagellins with the lectin LFA (Limax fldvus agglutinin) that recognizes sialic acid suggested that terminal sialic acid may be a component of the novel flagellar glycan. A definitive structural characterization of the flagellin structural protein was completed and revealed that FlaA was one of the most heavily 0-glycosylated prokaryotic proteins characterized to date (Thibault et al., 2001, Logan et al., 2002). Electrospray mass spectrometry analysis on purified flagellin proteins from C. jejuni 81-176, NCTC 11168, and OH4384 and C. coli VC167 gave broad heterogeneous molecular mass envelopes ranging from 64,500 to 65,400 Daywith a few discrete glycoform peaks for each flagellin. Because the predicted mass of the FlaA protein is 59,240 Day this 6,000-Da excess represents 10% of total glycoprotein molecular mass. To more precisely determine the location of glycosylation, the flagellin protein from C. jejuni 81-176 was digested with trypsin and peptides analyzed by liquid chromatography-electrospray (LC-ES) mass spectrometry (MS) and MS-MS with or without alkaline hydrolysis (Thibault et al., 2001). A total of 19 sites of 0-linked glycosylation were identified on FlaA, and the sites were shown to reside in the central region of the primary amino acid sequence (Fig. 1). The FlaB protein is glycosylated in a similar manner (J. F. Kelly and C. M. Szymanski, unpublished observations). Mapping of glycosylation sites of C. coli VC167 flagellin also confirmed a conservation in lo-
STRUCTURAL CHARACTERIZATION OF FLAGELLIN GLYCOPROTEINS The first evidence for posttranslational modification of Campylobacter flagellin came from C. coli VC167 flagellin by direct chemical analysis of purified flagellar peptides. It was shown that a number of serine residues that were localized to the central re-
Susan M. Logan and Ian C. Schoenhofen * Institute for Biological Sciences, National Research Council of Canada, Ottawa, Canada O N K1A OR6. Patricia Guerry Naval Medical Research Center, Silver Spring, MD 20910.
471
LOGAN ET AL.
472
D2-D3
A
B
(
CD 1 3
GFRINTNVAA LNAKANSDLN AKSLDASLSR LSSGLRINSA ADDASGMAIA
50
DSLRSQANTL GQAISNGNDA LGILQTADKA MDEQLKILDT IKTKATQAAQ
100
DGQSLKTRTM LQADINKLME ELDNIANTTS FNGKQLLSGN FTNQEFQIGA
150
SSNQTVKATI GATQSSKIGV TRFETGAQSF TSGVVGLTIK NYNGIEDFKF
200
DNWI~TSVG TGLGALAEEI NKSADKTGVR ATYDVKTTGV YAIKEGTTSQ
250
DFAINGVTIG KIEYKDGDGN GSLISAINAV KDTTGVQASK DENGKLVLTS
300
ADGRGIKITG DIGVGSGILA NQKENYGRLS LVKNDGRDIN I&TNL~AIG
350
MGTTDMISQS SVSLRESKGQ ISATNADAMG FNSYKGGGKF VFzQNV$Srg
400
AF TLKGAMAVMD IAETAITNLD
500
QIRADIGSIQ NQVTSTINNI TVTQVNVKAA ESQIRDVDFA SESANYSKAN
QFAALKTTAA NTTDETA
550
[ ILAQSGSYAM AQANSSQQNV LRLLQ
19 modification sites Figure 1. Localization of flagellar glycosylation sites to surface-exposed regions of each monomer within the assembled filament. (A) Structure of FliC from S. enterica serovar Typhimurium with major domains labeled as previously described (Samatey et al., 2000; Yonekura et al., 2003, 2005). The Protein Data Bank accession code is 1UCU and is displayed in Protein Explorer. (B) Assignment map of C. jejuni flagellin 81-174. The primary amino acid sequence of FlaA is shown, and modified residues are highlighted. The sequences corresponding to NDO, ND1, D2, D3, CD1, and CDO domains from the FliC structure are indicated for FlaA on the left side.
calization to the central region of the monomer (Logan et al., 2002). This region of the flagellin protein can now be modeled by using the flagellin structures obtained from S. enterica serovar Typhimurium via cryo-electron microscopy. In comparing sites of 0linked glycosylation to this model, the sites of glycosylation would be found in the D2 and D3 domains of the folded protein (Fig. 1). Of significance is the fact that these domains form a substantial projection at the filament surface in the assembled flagellar filament (Samatey et al., 2000; Yonekura et al., 2003, 2005). The presence of 19 sites of glycosylation on these notable surface projections of each flagellin monomer within the assembled Campylobacter flagellar filament highlights the potential for biological interaction via these O-linked glycans. It is also consistent with earlier work that demonstrated at least some of the glycans were surface exposed in the flagellar filament (Power et al., 1994).
GLYCAN STRUCTURE DETERMINATION Both traditional bottom-up MS analysis involving proteolytic cleavage of the flagellin protein with separation and analysis of enzymatic products by either capillary electrophoresis-nanoelectrospray-MS or nano-LC-MS as well as a more recent top-down approach that uses intact flagellin glycoprotein and low-energy collisional dissociation have been used to successfully characterize posttranslational modifications of C. jejuni 81-176 and C. coli VC167 flagellin
glycoproteins (Logan et al., 2002; Schirm et al., 2005; Thibault et al., 2001). For C. jejuni 81-176 flagellin, initial LC-ES MS analysis, with front-end collision-induced dissociation conditions (orifice voltage, 120 V), revealed three distinct glycan oxonium ions (mlz 317, mlz 316, and mlz 409) that were present on a number of tryptic peptides and corresponded to glycans of residue mass 316, 315, and 408 Da. However, in contrast to earlier lectin work, no evidence for modification of the flagellin protein with sialic acid was found (residue mass, 291 Da). Secondary fragmentation patterns of the mlz 317 oxonium ion indicated that the glycosyl moiety was a diamino sugar containing an acid group with two N-acetyl functionalities and the empirical formula C,,H,OO,N, (Mr 316.126). The other glycan moieties observed were shown to be structurally related to this 316-Da residue. For the 315-Da residue, prominent neutral losses of NH, and CH,CH(NH (NH,) were consistent with substitution of an N-acetyl group with an acetamidino group (l-Da loss of mass), while the 408-Da residue fragmentation pattern was consistent with substitution of the two N-acetyl groups with two N-2,3-dihydroxyprorionylgroups. The more recent top-down analysis of this flagellin glycoprotein revealed the presence of minor amounts of an N-acetyl glutamine substituent on the 3 15-Da acetamidino moiety and an O-acetyl derivative of the 316-Da residue (Schirm et al., 2005). MS analysis of flagellar tryptic peptides from C. coli VC167 revealed that this flagellin was also modified with an identical 316-Da residue as well as a
CHAPTER 26
novel 3 15-Da acetamidino variant (Logan et al., 2002). Secondary fragmentation patterns revealed that this 3 15-Da residue was distinct in structure to that observed from C. jejuni 81-176. Additionally, this analysis revealed minor amounts of two other unique glycan moieties of mass 431 and 432 Day which by fragmentation analysis appear to be a substitution of a hydroxyl group of the 315- and 316Da residues with a deoxypentose. Top-down analysis of C. coli VC167 flagellin revealed the presence of an additional glycan moiety of mass 329 Da (McNally et al., 2007). Nuclear magnetic resonance (NMR) spectroscopy experiments with either purified glycopeptide or nucleotide-activated sugars purified from the metabolome have been used to definitively determine the structures of the major sugars decorating Campylobacter flagellin (McNally et al., 2006a, 2007; So0 et al., 2004; Thibault et al., 2001). The development of metabolomic technologies to purify nucleotideactivated sugars (chapter 29) has contributed substantially to the successful elucidation of the structures of the novel glycan derivatives. The use of purified flagellar glycopeptides for NMR structural analysis can present significant challenges in terms of sensitivity and resolution, and these novel metabolomic approaches have provided an alternative route for structural elucidation of novel glycan moieties. All sugars found on C. jejuni 81-176 or C. coli VC167 flagella belong to the novel class of sialic acid-like sugars, which are 5,7-diacetamido-3,5,7,9tetradeoxy-nonulosonate derivatives (Knirel et al., 2003). This class of sialic acid-like sugars is unique to microorganisms and has been found in many bacterial species as constituents of cell-surface glycoconjugates such as lipopolysaccharide, capsular polysaccharide, pili, and flagella (Castric et al., 2001; KISSet al., 2001; Knirel et al., 1984; Thibault et al., 2001). The predominant modifications of residue mass 316 and 315 Da found on the flagellin of C. je-
0-LINKED FLAGELLAR GLYCOSYLATION
juni 8 1-176 are 5,7-diacetamido-3,5,7,9-tetradeoxyL-glycero-a-L-manno-nonulosonicacid (pseudaminic acid, PseSAc7Ac) and 5-acetamido-7-acetamidino3,5,7,9-tetradeoxy-~-glycero-a-~-manno-nonulosonic acid (PseSAc7Am). In contrast, the flagellin of C. coli VC167, while decorated with Pse5Ac7AcYis also glycosylated with two structurally distinct nonulosonate sugars that are the acetamidino and N-methyl acetimidoyl derivatives of legionaminic acid (5,7diacetamido-D-glycero-P-D-galacto-nonulosonic acid) Leg5Am7Ac and LegSArnNMe7Ac. Table 1 provides a summary of glycan masses observed on C. jejuni 81-176 and C. coli VC167 flagellin by MS. The precise structural configuration of the major glycan moieties of C. jejuni 81-176 and C. coli VC167 flagellin were elucidated by NMR and are presented in Fig. 2. The combination of MS and NMR analysis was critical to revealing the subtle yet significant diversity in glycan structures present on the flagellin proteins from these two strains. A second related mucosal pathogen Helicobacter pylori has also been shown to produce flagellins that are glycosylated with PseSAc7Ac. In a similar manner to Campylobacter, both the FlaA and FlaB structural proteins are glycosylated within the central D2-D3 domain at 7 and 10 sites, respectively (Schirm et al., 2003). In contrast to Campylobacter, only a single nonulosonate sugar, Pse5Ac7AcYwas found to modify the flagellin proteins of H. pylori. This restricted diversity in flagellar glycan composition is reflected in the flagellar glycan biosynthetic gene content of H. pylori when compared with Campylobacter spp. Uosenhans et al., 2002; Schirm et al., 2003). The presence of a sheath structure covering the flagellar filament of H. pylori may reduce the need or advantage of diversity in flagellar glycan structure in this organism (Fox, 2002). However, as with Campylobacter, the process of glycosylation is still essential for flagellar assembly.
Table 1. Cumpylobucter 0-linked flagellar glycans Strain
C. jejuni 81-176
C. coli VC167
473
Flagellar glycan residue mass (Da)
Neutral glycan mass (Da)
Structural assignment
316 3 15 358 486 316 3 15 329 43 1 432
334 333 376 504 334 333 347 449 450
PseSAc7Ac" PseSAc7Am" PseSAc7AcOAc Pse5Ac7AcO-GlnAc PseSAc7Ac" LegSAm7Ac" LegSAmNMe7Ac a PseSAm7AcO-deoxypentose Pse5Ac7AcO-deoxypentose
"Structures elucidated by NMR analysis of glycopeptide or purified nucleotide activated sugars.
474
LOGAN ET AL.
Sialic acid +
OH
"
H
B @->cH3
R,=
HC ,
NAc OH
Legionaminic acid
I.,
I
H
R,=
H,C
12
I
H
(€-form NAmNMe)
(Z-form NAmNMe)
Figure 2. Structures of glycans found on Cumpylobucter flagellins compared with that of sialic acid. (A) Sialic acid or 5acetamido-3,5-dideoxy-~-g~ycero-~-~-gu~uc~o-nonu~osonic acid (Neu5Ac). (B) Pseudaminic acid or 5,7-diacetamido-3,5,7,9tetradeoxy-L-glycero-a-L-munno-nonulosonic acid and derivatives. (C) Legionaminic acid or 5,7-diacetamido-3,5,7,9tetradeoxy-D-glycero-p-D-guhcto-nonulosonic acid and derivatives. Known R groups are shown for both (B) and (C), illustrating the diversity of functional groups observed for each. Sialic acid and legionaminic acid exhibit the same D-glyceroD-gulucto absolute configuration. Confirmation of (Y or p linkage of B and C to flagellin has yet to be determined.
FLAGELLAR GLYCOSYLATION LOCUS Comparative genomic analyses of Cumpylobucter isolates has revealed that the flagellar locus displays considerable genetic variability (Dorrell et al., 2001; Kim et al., 2002; Leonard et al., 2003). The flagellar glycosylation locus of C. jejuni NCTC 11168 comprises 50 genes, while the locus in C. jejuni 81-176 is much simpler, having half the number of genes and lacking the ptm genes responsible for legionaminic acid biosynthesis (Guerry et al., 2006; Parkhill et al., 2000). Whole-genome comparisons of 18 strains identified the flagellar locus as one of seven hypervariable plasticity regions, and a comparative phylogenomic study revealed that a cluster of six genes from this locus (Cj1321 to Cj1326) were characteristic of a livestock clade and was suggestive of a role for a distinct flagellar glycan synthesized by these genes that is necessary for colonization of livestock (Champion et al., 2005; Pearson et al., 2003).
Preliminary annotation of the locus made on the basis of homology studies identified approximately 50% of the genes in C. jejuni NCTC 11168 to be involved in carbohydrate biosynthesis, and a number of these displayed significant homology to sialic acid biosynthetic genes. Earlier studies had provided evidence for a role of a number of genes from this locus in the glycosylation process. Site-directed mutants of sialic acid biosynthetic gene homologs Cj13 11 (neuA) and Cj1317 (neuB3) in strains 81-176 and NCTC 11168 were shown to abolish motility of cells, and no flagellin protein was produced (Linton et al., 2000; Thibault et al., 2001). In contrast, Cj1332 (ptmA) and Cj133 1 (ptmB) site-directed mutants from C. coli VC167 and Cj1316 (pseA) in C. jejuni 81-176 did not effect motility but did result in an altered flagellin isoelectric footprint and loss of immunoreactivity indicative of a change in glycosylation profile (Guerry et al., 1996; Thibault et al., 2001). The change in glycosylation profile for the C. jejuni
CHAPTER 26
81-176 pseA mutant was shown to be due to loss of Pse5Ac7AmYproviding evidence that pseA (Cj1316) is responsible for the production of PseSAc7Am. Flagellin isolated from ptmA and ptmB mutants of C. coli VC167 lacked LegSAm7Ac. Subsequent mutational analyses identified additional genes from the ptm locus of VC167 as being involved in the production of LegSAm7Ac (neuB2lptmClCj1327, ptmDl Cj1328, ptmEICj 1329 and ptmFICjl330) (Logan et al., 2002; McNally et al., 2007). Recognition of Cj1293 (pseB) and Cj1294 (pseC) from this locus as homologs of genes from the glycosylation locus of Caulobacter crescentus and of HP0840 (pseB) and HP0366 (pseC) from H. pylori prompted studies to explore the role these genes played in flagellar glycosylation in C. jejuni. As expected, a pseB mutant of C. jejuni 81-176 was nonmotile and nonflagellated and could be complemented with the H. pylori pseB gene (Goon et al., 2003). PseB expression was shown to be under the control of a d4promoter, which is the promoter that controls basal body and hook gene expression in the C. jejuni flagellar hierarchy (chapter 30) (Goon et al., 2003; Hendrixson and DiRita, 2003; Jagannathan et al., 2001). Thus, a gene involved in the synthesis of PseSAc7Ac is regulated with the middle flagellar genes. The only other glycosylation gene that appears to be transcriptionally regulated with the other flagellar genes is pseA, which is regulated by a a2' promoter, which controls flaA and other late flagellar and genes. The reason why pseB is controlled by aS4 pseA is controlled by a2' remains to be determined. The flagellar glycosylation locus in many strains of C. jejuni, including NCTC 11168, contain families of related genes, some of which appear to play a role in motility (Golden and Acheson, 2002; Guerry et al., 2006; Karlyshev et al., 2002). The first family of related genes in NCTC 11168 includes Cj1318, Cj1333, Cj1334, Cj1335, Cj1337, Cj1340, and Cj1341. These homologs in NCTC 11168 share between 52 and 87% identity at the protein level, and many of the members have homopolymeric tracts, suggesting phase variation (Karlyshev et al., 2002). Metabolomics studies with both a Cj1337 (pseE) mutant in 81-176 and a mutant in the homolog HP0114 of H. pylori 1061 demonstrated that these genes had no effect on accumulation of nucleotide-activated monosaccharides, although assembly of a flagella filament was abolished in both cases, suggesting that they may function as sugar-protein glycosyltransferases (McNally et al., 2006a; Schirm et al., 2003). In contrast, mutation of Cj1333 (pseD) in 81-176, although it did not affect motility or production of the cytidine monophosphate (CMP)-activated sugars CMP-PseSAc7Ac and CMP-Pse5Ac7AmYresulted
0-LINKED FLAGELLAR GLYCOSYLATION
475
in flagella that were no longer glycosylated with Pse5Ac7AmYindicating that the product of this gene may be involved in the attachment of this sugar to the flagellin protein (Guerry et al., 2006; McNally et al., 2006a). At the very minimum, these results point to a role of this family in flagellar glycosylation at a later stage in the assembly process, rather than glycan biosynthesis per se. A recent proteomewide protein interaction map has revealed that the PseE protein of C. jejuni NCTC 11168 interacts directly with both the FlaA and FlaB structural proteins (Parrish et al., 2007). The identification and characterization of the glycosyltransferase or -transferases responsible for the addition of these novel monosaccharides to the flagellin protein have yet to be made, although it seems likely that this family of related proteins may indeed be responsible for the transfer of Pse and Leg sugar derivatives that are found on Cumpylobacter flagellin. It is of significance that the genomes of two other bacterial species, Aeromonas spp. and Helicobacter spp., which have been shown to decorate their flagellins with Pse derivatives, also contain homologs of this gene family (Schirm et al., 2005). The second family of related genes in the flagellar glycosylation locus include Cjl30.5, Cj1306, Cj1310, and Cj1342 as well as Cj0618, which lies at a distant locus. The function of these genes is currently unknown, although signature-tagged transposon mutagenesis studies identified a Cj0618 mutant of 81-176 that was attenuated for colonization of the chick ceca and was nonmotile (Hendrixson and DiRita, 2004). Mutation of Cjl305 and Cj1342 had no effect on motility of 81-176, however, and strain 81-176 lacks homologs of Cj1306 and Cj1310 (Guerry et al., 2006).
GLYCAN BIOSYNTHETIC PATHWAY ELUCIDATION Metabolomics studies have provided considerable insight into the roles played by genes from the flagellar glycosylation locus. In combination with in vitro characterization of recombinantly expressed gene products, specific functional assignments for a significant number of glycan biosynthetic genes from this locus have been made. Previously annotated as sialic acid biosynthetic genes or glycan biosynthetic genes, a number of genes from this locus have been shown to be involved in the production of the nonulosonate sugar Pse (Chou et al., 2005; Creuzenet, 2004; Guerry et al., 2006; McNally et al., 2006a; Obhi and Creuzenet, 2005). These studies have led to the detailed elucidation of the PseSAc7Ac pathway using recombinant enzymes from both C. jejuni and H. pylori (Schoenhofen et al., 2006b, 2 0 0 6 ~ )By . the
LOGAN ET AL.
476
nine-carbon nonulosonate, PseSAc7Ac. CMP activasequential addition of biosynthetic enzymes with tion of PseSAc7Ac is then performed by PseF. Finally, their expected cofactors and cosubstrates, the entire the crystal structures of the first two biosynthetic pathway starting from UDP-GlcNAc was determined enzymes of this pathway from H. pylori have been and proceeds as indicated in Fig. 3 . PseB performs a elucidated, providing many insights into their enzyC-4,6 dehydration and C5 epimerization of UDPGlcNAc to form UDP-2 acetamido-2,6-dideoxy-p-~- matic mechanisms (Ishiyama et al., 2006; Schoenhofen et al., 2006a). arubino-hexos-4-ulose. C-4 aminotransfer of this The dehydratase PseB and aminotransferase arabino-hexos-4-ulose intermediate by PseC proPseC show homology to the dehydratase PglF and duces UDP-4-amino-4,6-dideoxy-P-~-AltNAc, which aminotransferase PglE, respectively, from the Pgl Nis then N-acetylated by PseH to form UDP-2,4linked glycan biosynthetic pathway (chapter 25). Aldiacetamido-2,4,6-trideoxy-~-~-altropyranose. The though the Pse pathway enzymes produce UDP-4actions of the UDP-sugar hydrolase PseC result in an intermediate of the release of 2,4-diacetamido-2,4,6-trideoxy-~-altro- amino-4,6-dideoxy-P-~-AltNAc, the PseSAc7Ac pathway, PglF and E produce UDPpyranose, which is subsequently condensed with the 4-amino-4,6-dideoxy-a-~-GlcNAc, a precursor used three-carbon molecule pyruvate by PseI, forming the
HO
I I1 PseH M -UDP
AcNH NAc
H
The CMP-Pse Biosynthetic Pathway
IV
VII Figure 3. CMP-Pse pathway of Campylo6acter jejuni. The enzymes and biosynthetic intermediates of the CMP-pseudaminic acid pathway, in order, are PseB, NAD(P)-dependent dehydratase/epimerase; PseC, PLP-dependent aminotransferase; PseH, N-acetyltransferase; PseG, UDP-sugar hydrolase; PseI, pseudaminic acid synthase; PseF, CMP-pseudaminic acid synthetase; and (I) UDP-GlcNAc; (11) UDP-2-acetamido-2,6-dideoxy-~-~-u~u6~~o-hexos-4-ulose; (111) UDP-4-amino-4,6-dideoxy-p-~AltNAc; (IV) UDP-2,4-diacetamido-2,4,6-trideoxy-~-~-altropyranose; (V) 2,4-diacetamido-2,4,6-trideoxy-~-altropyranose; (VI) pseudaminic acid; (VII) CMP-pseudaminic acid. Pyranose rings are shown as their predominant chair conformation in solution determined from nuclear Overhauser effects (NOES) and JH,H coupling constants. Schematic representations of PseB and PseC structures from H. pylori are shown (Ishiyama et al., 2006; Schoenhofen et al., 2006a).
CHAPTER 26
in the production of 2,4-diacetamido-2,4,6-trideoxya-D-glucopyranose, a component of the Pgl N-linked glycan (Olivier et al., 2006; Young et al., 2002). Functional characterization of PseB has revealed that cross talk between the O-linked pathway and Nlinked pathway can occur as a consequence of the C5 epimerase function of the PseB enzyme. Accumulation of the initial urubino-hexulose product in vitro leads to a shift in reaction equilibrium to produce the xylo-hexulose intermediate of the Pgl pathway (McNally et al., 2006b; Schoenhofen et al., 2 0 0 6 ~ ) . The observation of this pathway cross talk has recently been verified in vivo by analysis of pathway mutants (McNally et al., 2006a; Guerry et al., 2007). Accumulation of UDP-2,4-diacetamido-2,4,6-trideOX~-~-D-G a ~Pgl C ,glycan precursor, in the metabolome of a pseC mutant provided the first in vivo evidence of this cross talk. It has also been shown that a pglF mutant of C. jejuni 81-176, which can no longer produce Pgl glycan, was partially suppressed by a subsequent pseC mutation. The pseC mutation blocks the ability of cells to produce PseSAc7Ac and results in an accumulation of the arabino-hexulose product within the cell. Consequently, this product is converted by the PseB enzyme to the required xylohexulose intermediate of the Pgl pathway, thus restoring Pgl glycan production and the corresponding phenotypes of lectin reactivity and transformability. Finally, metabolomics studies have revealed that the ptm genes are responsible for the biosynthesis of the structurally distinct yet related nonulosonate sugar, legionaminic acid, and the LegSAm7Ac and LegSAmNMe7Ac derivatives, which were found on the flagellin of C. coli VC167 (Logan et al., 2002; McNally et al., 2007) (chapter 29). Although the precise functional characterization of each of the Leg pathway enzymes has yet to be completed, metabolomics studies have provided evidence that PtmH is the methyltransferase producing LegSAmNMe7Ac from CMP-Leg5Am7AcYwhereas PtmG is required to produce CMP-LegSAm7Ac and PtmB is responsible for activation of Leg5Ac7Ac with CMP (McNally et al., 2007).
MECHANISM OF GLYCOSYLATION O-glycosylation of Campylobucter flagellins, unlike the Pgl N-glycosylation system, appears to be a specific process, with only the flagellin structural proteins of Campylobucter being identified to date as glycosylated with the novel nonulosonate sugars. The process is integral to the flagellar assembly process in Campylobacter, and prevention of glycosylation leads to an inability to assemble the flagellar filament and
O-LINKED FLAGELLAR GLYCOSYLATION
477
cells are no longer motile. Unlike other prokaryotic glycosylation systems, which occur in association with the cell membrane in the periplasmic space (Aas et al., 2007; Chaban et al., 2006; Smedley et al., 2005; Szymanski and Wren, 2005), flagellin monomers are secreted through the flagellar apparatus and are never found in the periplasm. Therefore, glycosylation must take place either in the cytoplasm, possibly in close association with the flagellar machinery or in the sequestered basal body environment. Unlike the A?-linked Pgl glycosylation process, no consensus sequence for the O-linked flagellar glycosylation process has been identified. The localization of the majority of sites to the central region of the flagellin monomer may reflect the state of folding of the protein when it comes into contact with the specific glycosyltransferases or as it exists in the cytoplasm when associated with the flagellar chaperone, FliS, before assembly in the filament. The precise mechanistic basis for glycosylation of particular serines and threonines within the primary sequence remains to be established.
BIOLOGICAL ROLE OF FLAGELLAR GLYCANS The glycans on flagellin appear to play complex roles in the biology of Campylobacter. The requirement that flagellin be glycosylated for filament assembly to proceed underscores the biological significance of glycosylation. The localization of the glycan attachment sites to a region of the flagellin monomer that, when folded and assembled into the flagellar filament, is exposed on the surface points to significant potential for biological interactions. Evidence exists that indicates that the glycans play a key role in virulence through surface-associated interactions of these glycan moieties. Autoagglutination (AAG) is a recognized marker of virulence for a number of bacterial pathogens and has been shown to be associated with flagellar expression and glycosylation in C. jejuni (Golden and Acheson, 2002; Misawa and Blaser, 2000). Investigations into the role that flagellar glycans play in this AAG phenotype have been presented and indicate that the composition of the glycans present on the flagellar filament is critical to the process (Guerry et al., 2006). C. jejuni 81-176 cells producing flagella that lacked PseAm (pseA and pseD mutants) failed to autoagglutinate when compared with the parent strain, as did C. coli VC167 ptmD and pseB mutant strains, which could only glycosylate their flagella with Pse or Leg derivatives, respectively. Thus, AAG requires the presence of both types of glycan moiety, which suggests that the AAG phenotype may occur as a consequence of interactions be-
478
LOGAN ET AL.
tween glycans on distinct flagellar filaments. AAG in strain 81-176 is also affected by loss of capsule and changes in lipooligosaccharide composition, suggesting that interactions between glycans on flagellin and other surface structures also contribute to AAG (Guerry et al., 2006). The role of the glycans in hostcell interactions have been studied in a 81-176 pseA mutant that decorates its flagella only with PseSAc7Ac. This AAG mutant showed minor defects in adherence and invasion of intestinal epithelial cells in a 2-h incubation. However, when the period of interaction with the epithelial cells was extended, the pattern of binding of wild-type cells labeled with green fluorescent protein and that of a pseA mutant were markedly different. The mutant adhered as single cells, while the wild-type strain formed microcolonies composed of bacteria, which appeared to be interacting via flagellar bridges, as shown in Fig. 4. Finally, the role of glycan composition on flagellar filaments was examined in vivo. By means of a ferret diarrheal disease model, the pseA mutant of 8 1-176, which was still fully motile, was clearly attenuated in its ability to cause disease when compared with wildtype cells. The ability to form microcolonies may also be a first step in biofilm formation. Although biofilms
have not been demonstrated in vivo for C. jejuni, there are multiple reports of biofilm formation on inert surfaces, and evidence exists that flagella are critical to this process (Joshua et al., 2006; Kalmokoff et al., 2006) Flagellin is recognized as a major stimulus of human epithelial cells where the protein is recognized by TLRS receptors (Hayashi et al., 2001). Flagellins from the Epsilonproteobacteria, including both Campylobacter and Helicobacter spp., are not recognized by TLRS receptors. The lack of TLRS stimulation in these organisms has been shown to be due to unique amino acid sequences in the D1 domain, and glycosylation of flagellin does not play a role in this process (Andersen-Nissen et al., 2005; Lee et al., 2003). Flagellin is, however, an immunodominant antigen during infection, and the role of glycans in conferring serospecificity has been demonstrated (Alm et al., 1992). Immunodominance of glycans during experimental infection in animals and human volunteers has also been shown, and it has been suggested these glycans may mask other primary amino acid epitopes (Lee et al., 1999). Highly specific immune responses have been shown to result from subtle changes in structure of lipopolysaccharide sugars, and the distinct changes in configuration of the Pse and Leg derivatives, in addition to acetamidino and N-methyl functionalities, may also contribute to a unique antigenic response during infection (Luneberg et al., 2000). The space filing models of Pse5Ac7Ac7 Leg5Am7Ac7and LegSAmNMe7Ac presented in Fig. 5 show the structural diversity of each of these molecules and emphasize how antibodies directed toward any one of the structures may not recognize other derivatives. Alternatively, the structural diversity observed in glycan moieties may provide distinct specificities for binding to host ligands. It remains to be established whether the flagellar livestock clade locus (Cj1321 to Cj1326) is responsible for the production of a novel glycan of particular relevance for survival within the chicken or livestock gut (Champion et al., 2005).
SUMMARY AND FUTURE DIRECTIONS
Figure 4. Adherence pattern of green fluorescent protein (GPF)tagged C. jejuni 81-176 cells to INT407 cells. A microcolony of C. jejuni 81-176 tagged with GFP on INT407 cells in culture after 18-h incubation (Guerry et al., 2006). The image shows complex interactions of bacteria on the intestinal epithelial cells, and some of the bacteria are visible as apparent chains. Phase microscopy of the same field confirmed that the monolayer remained intact.
Significant progress has been made in defining at the molecular level the structural nature of the novel sialic acid-like nonulosonate sugars found to be decorating the flagellar filaments of Campylobacter, and it is clear that these types of studies will be integral to future work exploring the role of novel glycan moieties in biological interactions. In addition, both metabolomics approaches and detailed biochemical characterization of recombinantly expressed proteins
CHAPTER 26
Pse5Ac7Ac
5Ac
LegSAm7Ac
7Ac
0-LINKED FLAGELLAR GLYCOSYLATION
479
LegSAmNMe7Ac
7Ac
Figure 5. Molecular models of PseSAc7Ac, LegSAm7Ac, and LegSAmNMe7Ac. The exocyclic chain (C7 to C9) and pendant groups at C5 and C7 are displayed as reduced van der Waals spheres to show the extent of their structural diversities. Coordinates are available at http: //ibs-isb.nrc-cnrc.gc.ca/facilities/NMR/molecularmodeling-e.html. From McNally et al. (2007).
have led to a precise functional assignment for a substantial number of genes from the flagellar glycosylation locus. These two approaches have also been instrumental in the biochemical elucidation of novel glycan biosynthetic pathways. Diversity in glycan composition both on the flagellin proteins from an individual strain as well as among isolates has been shown to confer unique biological properties. In strain 81-176, the composition of glycans on the flagellar filament has been shown to play a role in virulence. Future work will likely be directed toward defining the precise roles of the remaining genes in the flagellar glycosylation locus of 8 1-176 and NCTC 11168, as well as a more comprehensive analysis of Campylobacter flagellar glycan biosynthetic loci in other strains. Only at this stage will the true flagellar glycan biosynthetic capacity of Campylobacter be known. The potential of utilizing these novel sugars and derivatives to explore additional biological roles after directed in vitro enzymatic synthesis is now possible. The utility of the biosynthetic enzymes for synthesis of CMP-Pse was shown recently in a single reaction combining six enzymes and UDP-GlcNAc. It remains to be established whether the structural similarity of the Campylobacter flagellar nonulosonate sugars to sialic acid confers on Campylobacter cells the ability to interact with specific eukaryotic ligands, which may facilitate as yet undiscovered novel hostpathogen interactions. REFERENCES Aas, F. E., A. Vik, J. Vedde, M. Koomey, and W. Egge-Jacobsen. 2007. Neisseriu gonowboeue 0-linked pilin glycosylation: functional analyses define both the biosynthetic pathway and glycan structure. Mol. Microbiol. 65:607-624. A h , R. A., P. Guerry, M. E. Power, and T. J. Trust. 1992. Variation in antigenicity and molecular weight of Cumpylobucter coli VC167 flagellin in different genetic backgrounds. J. Bucteriol. 174:4230-4238.
Andersen-Nissen, E., K. D. Smith, K. L. Strobe, S. L. Barrett, B. T. Cookson, S. M. Logan, and A. Aderem. 2005. Evasion of Tolllike receptor 5 by flagellated bacteria. Proc. Nutl. Acud. Sci. USA 102~9247-9252. Black, R E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Cumpylobucter jejuni infection in humans. J. Infect. Dis. 157:472-479. Caldwell, M. B., P. Guerry, E. C. Lee, J. P. Burans, and R. I. Walker. 1985. Reversible expression of flagella in Cumpylobucter jejuni. Infect. Immun. 50:941-943. Castric, P., F. J. Cassels, and R. W. Carlson. 2001. Structural characterization of the Pseudomonus uerugino~a1244 pilin glycan.J. Biol. Chem. 276:26479-26485. Chaban, B., S. Voisin, J. Kelly, S. M. Logan, and K. F. Jarrell. 2006. Identification of genes involved in the biosynthesis and attachment of Methanococcus uoltue N-linked glycans: insight into N-linked glycosylation pathways in Archaea. Mol. Microbiol. 61:259-268. Champion, 0. L., M. W. Gaunt, 0. Gundogdu, A. Elmi, A. A. Witney, J. Hinds, N. Dorrell, and B. W. Wren. 2005. Comparative phylogenomics of the food-borne pathogen Campylobucter jejuni reveals genetic markers predictive of infection source. Proc. Natl. Acad. Sci. USA 102:16043-16048. Chou, W. K., S. Dick, W. W. Wakarchuk, and M. E. Tanner. 2005. Identification and characterization of NeuB3 from Cumpylobucter jejuni as a pseudaminic acid synthase. J. Biol. Chem. 43:35922-35928. Creuzenet, C. 2004. Characterization of CJ1293, a new UDPGlcNAc C6 dehydratase from Cumpylobucter jejuni. FEBS Lett. 559:136-140. Doig, P., N. Kinsella, P. Guerry, and T. J. Trust. 1996. Characterization of a post-translational modification of Cumpylobucter flagellin: identification of a sero-specific glycosyl moiety. Mol. Microbiol. 19:3 79-3 87. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. A1 Ghusein, B. G. Barrell, J. Parkhill, N. G. Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Cumpylobucter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11~1706-1715. Fox, J. G. 2002. The non-H. pylori helicobacters: their expanding role in gastrointestinal and systemic diseases. Gut 50:273-283. Golden, N. J., and D. W. Acheson. 2002. Identification of motility and autoagglutination Cumpylobucter jejuni mutants by random transposon mutagenesis. Infect. Immun. 70:1761-1771. Goon, S., J. F. Kelly, S. M. Logan, C. P. Ewing, and P. Guerry. 2003. Pseudaminic acid, the major modification on Cumpylo-
480
LOGAN ET AL.
bacter flagellin, is synthesized via the Cj1293 gene. Mol. Microsonic acids in bacterial glycopolymers: chemistry and biochemistry. Adv. Carbohydr. Chem. Biochem. 58:371-417. biol. 50:659-671. Knirel, Y. A,, E. V. Vinogradov, V. L. L'vov, N. A. Kocharova, Guerry, P., R. A. A h , M. E. Power, S. M. Logan, and T. J. Trust. A. S. Shashkov, B. A. Dmitriev, and N. K. Kochetkov. 1984. 1991. Role of two flagellin genes in Campylobacter motility. J. Sialic acids of a new type from the lipopolysaccharides of PseuBacteriol. 173:4757-4764. domonas aeruginosa and Shigella boydii. Carbohydr. Res. 133: Guerry, P., P. Doig, R. A. A h , D. H. Burr, N. Kinsella, and T. J. C5-C8. Trust. 1996. Identification and characterization of genes reLee, L. H., E. Burg, 111, S. Baqar, A. L. Bourgeois, D. H. Burr, quired for post-translational modification of Campylobacter coli C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evaluation of a VC167 flagellin. Mol. Microbiol. 19:369-378. truncated recombinant flagellin subunit vaccine against CampyGuerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Patlobacter jejuni. Infect. Immun. 675799-5805. tarini, G. Majam, P. Thibault, and s. M. Logan. 2006. Changes Lee, S. K., A. Stack, E. Katzowitsch, S. I. Aizawa, S. Suerbaum, in flagellin glycosylation affect Campylobacter autoagglutination and C. Josenhans. 2003. Helicobacter pylori flagellins have very and virulence. Mol. Microbiol. 60:299-311. low intrinsic activity to stimulate human gastric epithelial cells Guerry, P., C. P. Ewing, I. C. Schoenhofen, and S. M. Logan. via TLR.5. Microbes Infect. 5:1345-1356. 2007. Protein glycosylation in Campylobacter jejuni: partial supLeonard, E. E., T. Takata, M. J. Blaser, S. Falkow, L. S. Tomppression of pglF by mutation of pseC. /. Bacteriol. 189:6731kins, and E. C. Gaynor. 2003. Use of an open-reading frame6733. specific Campylobacter jejnni DNA microarray as a new genoGuerry, P., S. M. Logan, S. Thornton, and T. J. Trust. 1990. Getyping tool for studying epidemiologically related isolates. J. nomic organization and expression of Campylobacter flagellin Infect. Dis. 187:691-694. genes. /. Bacteriol. 172:1853-1860. Linton, D., A. V. Karlyshev, P. G. Hitchen, H. R. Morris, A. Dell, Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R N. A. Gregson, and B. W. Wren. 2000. Multiple N-acetyl neuGoodlett, J. K. Eng, S. Akira, D. M. Underhill, and A. Aderem. raminic acid synthetase (neuB) genes in Campylobacter jejuni: 2001. The innate immune response to bacterial flagellin is meidentification and characterization of the gene involved in siadiated by Toll-like receptor 5 . Nature 410:1099-1103. lylation of lipo-oligosaccharide. Mol. Microbiol. 35:1120-1134. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of CamLogan, S. M. 2006. Flagellar glycosylation-a new component of pylobacter jejuni genes involved in commensal colonization of the motility repertoire? Microbiology 152: 1249-1262. the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Logan, S. M., J. F. Kelly, P. Thibault, C. P. Ewing, and P. Guerry. Hendrixson, D. R., and V. J. DiRita. 2003. Transcription of 2002. Structural heterogeneity of carbohydrate modifications afsigma54-dependent but not sigma28-dependent flagellar genes fects serospecificity of Campylobacter flagellins. Mol. Microbiol. in Campylobacter jejuni is associated with formation of the fla46:587-597. gellar secretory apparatus. Mol. Microbiol. 50:687-702. Logan, S. M., T. J. Trust, and P. Guerry. 1989. Evidence for postIshiyama, N., C. Creuzenet, W. L. Miller, M. Demendi, E. M. translational modification and gene duplication of CampylobacAnderson, G. Harauz, J. S. Lam, and A. M. Berghuis. 2006. ter flagellin.]. Bacteriol. 171:3031-3038. Structural studies of FlaAl from Helicobacter pylori reveal the Luneberg, E., N. Zetzmann, D. Alber, Y. A. Knirel, 0. Kooistra, mechanism for inverting 4,6-dehydratase activity.]. Biol. Chem. U. Zahringer, and M. Frosch. 2000. Cloning and functional 281:24489-24495. characterization of a 30 kb gene locus required for lipopolysacJagannathan, A., C. Constantinidou, and C. W. Penn. 2001. Roles charide biosynthesis in Legionella pneumophila. lnt. 1.Med. Miof rpoN, fliA, and flgR in expression of flagella in Campylobacter crobiol. 290:37-49. jejuni. 1.Bacteriol. 183:2937-2942. McNally, D. J., A. J. Aubry, J. P. Hui, N. H. Khieu, D. Whitfield, Josenhans, C., L. Vossebein, S. Friedrich, and S. Suerbaum. 2002. C. P. Ewing, I?. Guerry, J. R. Brisson, S. M. Logan, and E. C. The neuA/flmD gene cluster of Helicobacter pylori is involved Soo. 2007. Targeted metabolomics analysis of Campylobacter in flagellar biosynthesis and flagellin glycosylation. FEMS Microcoli VC167 reveals legionaminic acid derivatives as novel flagelbiol. Lett. 210:165-172. lar glycans. J. Biol. Chem. 282: 14463-14475. Joshua, G. W., C. Guthrie-Irons, A. Karlyshev, and B. Wren. McNally, D. J., J. P. Hui, A. J. Aubry, K. K. Mui, P. Guerry, J. R. 2006. Biofilm formation in Campylobacter jejuni. Microbiology Brisson, S. M. Logan, and E. C. Soo. 2006a. Functional char152:3 87-3 96. acterization of the flagellar glycosylation locus in Campylobacter Kalmokoff, M., P. Lanthier, T. L. Tremblay, M. Foss, P. C. Lau, jejuni 81-176 using a focused metabolomics approach. /. Biol. G. Sanders, J. Austin, J. Kelly, and C. M. Szymanski. 2006. Chem. 281: 18489-18498. Proteomic analysis of Campylobacter jejuni 11168 biofilms reMcNally, D.J., I. C. Schoenhofen, E. F. Mulrooney, D. M. Whitveals a role for the motility complex in biofilm formation. ]. field, E. Vinogradov, J. s. Lam, s. M. Logan, and J. R Brisson. Bacteriol. 188:4312-4320. 2006b. Identification of labile UDP-ketosugars in Helicobacter Karlyshev, A. V., D. Linton, N. A. Gregson, and B. W. Wren. pylori, Campylobacter jejuni and Pseudomonas aeruginosa: key 2002. A novel paralogous gene family involved in phase-variable metabolites used to make glycan virulence factors. Chembioflagella-mediated motility in Campylobacter jejuni. Microbiology chemistry 7: 1865-1 868. 148:473-480. Misawa, N., and M. J. Blaser. 2000. Detection and characterizaKim, C. C., E. A. Joyce, K. Chan, and S. Falkow. 2002. Improved tion of autoagglutination activity by Campylobacter jejuni. Inanalytical methods for microarray-based genome-composition fect. Immun. 68 :6168-6 175. analysis. Genome Biol. 3:RESEARCHOOG.S. Obhi, R. K., and C. Creuzenet. 2005. Biochemical characterization Kiss, E., A. Kereszt, F. Barta, S. Stephens, B. L. Reuhs, A. Konof the Campylobacter jejuni Cj1294, a novel UDP-4-keto-6dorosi, and P. Putnoky. 2001. The rkp-3 gene region of Sinordeoxy-GlcNAc aminotransferase that generates UDP-4-aminohizobium meliloti Rm41 contains strain-specific genes that de4,6-dideoxy-GalNAc. 1.Biol. Chem. 280:20902-20908. termine K antigen structure. Mol. Plant Microbe Interact. 14: Olivier, N. B., M. M. Chen, J. R. Behr, and B. Imperiali. 2006. 1395-1403. In vitro biosynthesis of UDP-N,N'-diacetylbacillosamineby enKnirel, Y. A., A. S. Shashkov, Y. E. Tsvetkov, P. E. Jansson, and zymes of the Campylobacter jejuni general protein glycosylation U. Zahringer. 2003. 5,7-Diamino-3,5,7,9-tetradeoxynon-2-ulo- system. Biochemistry 45:13659-13669.
CHAPTER 26
Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Parrish, J. R., J. Yu, G. Liu, J. A. Hines, J. E. Chan, B. A. Mangiola, H. Zhang, s. Pacifico, F. Fotouhi, V. J. DiRita, T. Ideker, P. Andrews, and R. L. Finley, Jr. 2007. A proteome-wide protein interaction map for Campylobacter jejuni. Genome Biol. 8: R130. Pearson, B. M., C. Pin, J. Wright, K. I'Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS Lett. 554:224-230. Power, M. E., P. Guerry, W. D. McCubbin, C. M. Kay, and T. J. Trust. 1994. Structural and antigenic characteristics of Campylobacter coli FlaA flagellin.]. Bacteriol. 176:3303-3313. Samatey, F. A., K. Imada, F. Vonderviszt, Y. Shirakihara, and K. Namba. 2000. Crystallization of the F41 fragment of flagellin and data collection from extremely thin crystals. J. Struct. Biol. 132~106-111. Schirm, M., I. C. Schoenhofen, S. M. Logan, K. C. Waldron, and P. Thibault. 2005. Identification of unusual bacterial glycosylation by tandem mass spectrometry analyses of intact proteins. Anal. Cbem. 77:7774-7782. Schirm, M., E. C. Soo, A. J. Aubry, J. Austin, P. Thibault, and S. M. Logan. 2003. Structural, genetic and functional characterization of the flagellin glycosylation process in Helicobacter pyLori. Mol. Microbiol. 48 :1579-1592. Schoenhofen, I. C., V. V. Lunin, J. P. Julien, Y. Li, E. Ajamian, A. Matte, M. Cygler, J. R. Brisson, A. Aubry, S. M. Logan, S. Bhatia, W. W. Wakarchuk, and N. M. Young. 2006a. Structural and functional characterization of PseC, an aminotransferase involved in the biosynthesis of pseudaminic acid, an essential fla-
0-LINKED FLAGELLAR GLYCOSYLATION
481
gellar modification in Helicobacter pylori. 1.Biol. Chem. 281: 8907-8916. Schoenhofen, I. C., D. J. McNally, J. R. Brisson, and S. M. Logan. 2006b. Elucidation of the CMP-pseudaminic acid pathway in Helicobacter pylori: synthesis from UDP-N-acetylglucosamineby a single enzymatic reaction. Glycobiology 16:8C-l4C. Schoenhofen, I. C., D. J. McNally, E. Vinogradov, D. Whitfield, M. Young, S. Dick, W. W. Wakarchuk, J.-R. Brisson, and S. M. Logan. 2006c. Functional characterisation of dehydratase/aminotransferase pairs from Helicobacter and Campylobacter: enzymes distinguishing the pseudaminic acid and bacillosamine biosynthetic pathways.]. Biol. Chem. 281:723-732. Smedley,J. G., 111, E. Jewell, J. Roguskie, J. Horzempa, A. Syboldt, D. B. Stolz, and P. Castric. 2005. Influence of pilin glycosylation on Pseudomonas aeruginosa 1244 pilus function. Infect. lmmun. 73~7922-7931. Soo, E. C., A. J. Aubry, S. M. Logan, P. Guerry, J. F. Kelly, N. M. Young, and P. Thibault. 2004. Selective detection and identification of sugar nucleotides by CE-electrospray-MS and its application to bacterial metabolomics. Anal. Chem. 76:619-626. Szymanski, C. M., and B. W. Wren. 2005. Protein glycosylation in bacterial mucosal pathogens. Nut. Rev. Microbiol. 3:225-237. Thibault, P., S. M. Logan, J. F. Kelly, J. R Brisson, C. P. Ewing, T. J. Trust, and P. Guerry. 2001. Identification of the carbohydrate moieties and glycosylation motifs in Campylobacter jejuni flagellin.]. Biol. Chem. 276:34862-34870. Yonekura, K., S. Maki-Yonekura, and K. Namba. 2005. Building the atomic model for the bacterial flagellar filament by electron cryomicroscopy and image analysis. Structure 13:407-412. Yonekura, K., S. Maki-Yonekura, and K. Namba. 2003. Complete atomic model of the bacterial flagellar filament by electron cryomicroscopy. Nature 424:643-650. Young, N. M., J. R. Brisson, J. Kelly, D. C. Watson, L. Tessier, P. H. Lanthier, H. C. Jarrell, N. Cadotte, F. St Michael, E. Aberg, and C . M. Szymanski. 2002. Structure of the N-linked glycan present on multiple glycoproteins in the gram-negative bacterium, Campylobacter jejuni. 1. Biol. Chem. 277:4253042539.
Campylobacter, 3rd ed.
Edited by I. Nacharnkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 27
Campylobacter jejuni Lipooligosaccharides: Structures and Biosynthesis MICHELGILBERT, CRAIGT. PARKER,AND ANTHONY P. M o w
Many pathogenic bacteria have cell-surface glycoconjugates such as capsules, glycosylated flagellins, glycosylated surface-layer proteins, and, in gramnegative bacteria, lipopolysaccharides (LPS) or lipooligosaccharides (LOS). The variability of cell surface poly- or oligosaccharides plays a major role in virulence in mucosal pathogens (Moran et al., 1 9 9 6 ~ ) . Carnpylobacter jejuni, as an intestinal mucosal colonizer and an important causal agent of acute gastroenteritis in humans, has been shown to have variable expression of cell surface carbohydrates (Guerry et al., 2002; Karlyshev et al., 2005b; Moran et al., 2000; Prendergast et al., 2004). In general, LPSs and LOSS are families of phosphorylated lipoglycans and glycolipids, respectively, that are considered toxic with potent immunomodulating and immune-stimulating properties. They are found anchored in the outer membrane of gramnegative bacteria, and hence they are commonly referred to as endotoxins (Raetz and Whitfield, 2002; Rietschel et al., 1994). High-molecular-weight (M,) LPS is composed of the 0-polysaccharide chain, composed of a polymer of repeating oligosaccharide units, linked to the core oligosaccharide of up to 10 to 15 sugar residues, and anchored by the lipid A, in the outer membrane. On the other hand, low-M, LOS differs from LPS by lacking an 0-polysaccharide chain and exhibits greater structural diversity in the outer core than seen in LPS. Like certain other bacterial pathogens of mucosae, C. jejuni produces LOS (Moran et al., 1996c) that displays considerable interstrain variation in the structure of its outer core (Moran and Penner, 1999; Moran et al., 2000; Prendergast and Moran, 2000). Consistent with this phenotypic variation, microarray analysis, PCR-based
gene probing and sequencing studies have shown that there is extensive variation in the gene content of the locus responsible for the biosynthesis of C. jejuni LOS (Dorrell et al., 2001; Parker et al., 2005, 2006; Taboada et al., 2004). In early studies, C. jejuni was considered to produce low-M, LOS and high-M, molecules that were deduced to be LPS (Preston and Penner, 1987), both of which were inferred to contribute to reactions in the Penner heat-stable (HS) antigen serotyping scheme (Penner and Aspinall, 1997). The exact basis of the serospecificity within the scheme remained unclear and was a confusing issue for a number of years (Moran and Penner, 1999). Although a number of high-M, polysaccharide molecules independent of LPS had been described (Aspinall et al., 1995a; Hanniffy et al., 1999), the subsequent identification of a cluster of genes in C. jejuni with significant sequence homology to capsular polysaccharide (CPS) (kps) genes led to the discovery of the production of a serodominant polysaccharide capsule by C. jejuni (Karlyshev et al., 2000). The presently accepted paradigm is that in general C. jejuni produces low-Mr LOS and high-M, CPS (chapter 28), though other polysaccharide-related molecules and exopolysaccharides are also produced by certain C. jejuni strains (Corcoran and Moran, 2007; Kilcoyne et al., 2006; Muldoon et al., 2002). The different moieties of C. jejuni LOS, i.e., the core and lipid A regions, have been the subject of intensive investigation. The lipid A component of C. jejuni LOS has been structurally characterized (Moran et al., 1991b) and represents the first so-called mixed lipid A-containing D-glucosamine (GlcN) and 2,3-diamino-2,3-dideoxy-~-glucose
Michel Gilbert Institute for Biological Sciences, National Research Council Canada, Ottawa, Ontario, Canada K1A OR6. Craig T. Parker United States Department of Agriculture, Agricultural Research Service, Produce Safety and Microbiology Research, Albany, California, CA 94710. Anthony P. Moran Department of Microbiology, National University of Ireland Galway, University Road, Galway, Ireland, and Institute for Glycomics, Gold Coast Campus, Griffith University, Queensland, Australia.
483
484
GILBERT ET AL.
(GlcN3N) of any gram-negative bacterial species for which the complete structure has been established. Despite the occurrence of GlcN3N and GlcN in the lipid A backbone (Moran et al., 1991a, 1991b), in contrast to the GlcN disaccharide backbone of enterobacterial lipid A (Raetz and Whitfield, 2002; Raetz et al., 2007; Rietschel et al., 1990, 1994), the architecture of the molecule resembles that of Escherichia coli with similar phosphorylation and acylation patterns, but whose fine detail differs (Moran, 1997). Thus, C. jejuni lipid A antigenically resembles classical enterobacterial lipid A but has slightly lower, yet comparable, endotoxic activity in biological test systems to those of enterobacterial species (Moran, 1995). Biophysical measurements have shown that C. jejuni LOS and lipid A exhibit higher phase-transition temperatures than those of Salmonella preparations, and thus the former have lower fluidity at 37°C (Moran, 1995). Potentially, the lower fluidity may influence the biological activities of C. jejuni LOS, but acyl chain characteristics in this molecule and the partial replacement of GlcN with GlcN3N may also influence the supramolecular structure of C. jejuni lipid A, thereby affecting endotoxic activities (Schromm et al., 2000). Moreover, emphasizing the uniqueness of lipid A structure to each bacterial species, and in contrast to C. jejuni, C. fetus LPS and lipid A contain only GlcN and have a different fatty acid composition (Moran et al., 1994, 1996a) that influences their biological activities (Moran et al., 1996b). Significant interest in the structure and biosynthesis of the core oligosaccharide of C. jejuni has stemmed from the potential involvement of ganglioside-like mimicking structures in the outer core of C. jejuni LOSs in inducing the neurological disorders Guillain-BarrC syndrome (GBS) and Miller Fisher syndrome (MFS) (Komagamine and Yuki, 2006; Prendergast and Moran, 2000). It is now generally accepted that these paralytic disorders are induced by an autoimmune response to ganglioside-like structures present in the core regions of LOSs produced by certain C. jejuni strains (Godschalk et al., 2004; Moran and Prendergast, 2001; Yuki et al., 1993) involving the induction of cross-reactive antineuronal antibodies (Moran et al., 2005; Prendergast and Moran, 2000), although the antiglycolipid T-cell response may also contribute (De Libero et al., 2005). Furthermore, sialylation of the LOS core, which is associated with ganglioside mimicry, affects immunogenicity and serum resistance of C. jejuni (Guerry et al., 2000). The ganglioside mimicry is further complicated by phase variation of the ganglioside mimic in a given C. jejuni strain (Guerry et al., 2002), and evidence from human volunteer studies has shown that this variation can occur in vivo (Prendergast
et al., 2004). Nevertheless, the implications of this phase variation for C. jejuni pathogenesis have not been unequivocally established, and examination of their importance remains an area of intensive investigation. The present chapter reviews the structure, biosynthesis, and genetic determination of C. jejuni LOSs, with particular reference to the core oligosaccharide region. In particular, the genetic bases for variation of the outer core are discussed, including gene content variation between LOS loci and mechanisms generating LOS core region variation. Although this review predominantly addresses the biosynthesis of the variable LOS outer core of C. jejuni, the biosynthesis and genetic determination of the inner core lipid A region is also examined.
STRUCTURES OF LOS CORES Since the initial discovery of N-acetylneuraminic acid (NeuSAc), commonly known as sialic acid, in C. jejuni LOS (Moran et al., 1991a), subsequent structural studies have shown that the structures of the outer core of many C. jejuni strains, including those associated with GBS and MFS, mimic the saccharide portion of human gangliosides (Moran et al., 1996c, 2000; Prendergast and Moran, 2000). As shown in Fig. lA, extensive structural analyses have shown that the C. jejuni HS:19 serostrain (i.e., serotype reference strain) expresses a heterogeneous outer core composed of a mixture of GMla and GDla ganglioside mimics (Aspinall et al., 1994). Other studies have reported that LOSs of C. jejuni HS:19 isolates from GBS patients and enteritis patients contain terminal tetra- and pentasaccharide moieties identical to those of GMla and GDla gangliosides, respectively (Moran and O’Malley, 1995; Yuki et al., 1993). In addition to this molecular mimicry, mimicry of GTla and GD3 gangliosides by terminal hexasaccharides and trisaccharides, respectively, was observed in the LOS outer cores of two serotype HS:19 isolates (C. jejuni OH4384 and OH4382) from Japanese siblings with GBS (Aspinall et al., 1994) (Fig. 1B and C). The outer core from a C. jejuni HS:41 GBS isolate was found to have a tetrasaccharide structure consistent with GMla mimicry (Prendergast et al., 1998), and additional sialylation to produce a GDla mimic has also been observed in the C. jejuni serostrain HS:41 (Szymanski et al., 2003). Mimicry of GD3 ganglioside over a trisaccharide was demonstrated in the outer core of C. jejuni PG836, a MFS isolate of serotype HS:10 (Nam Shin et al., 1998) (Fig. lD), and although the same form of mimicry occurs in LOS of another MFS isolate, a C. jejuni HS:23 strain,
CHAPTER 27
A.
PEtn I 6 Gal-p(1,3)-GalNAc-p(1,4)-Gal-p(1,3)-Hep-a(l,3)-Hep-a(l,5)-Kdo 3 3 4
I 2 ?[ Neu5AcaI
I 2
485
PEtn I 6 GalNAc-p(1,4)-Gal-p(1,3)-Gal-P(1 ,3)-Hep-a(l,3)-Hep-a(l,S)-Kdo 3 2 2 4
E.
I
I
2 Neu5Aca
1 Glcp
NeuSAca
C. TETUNI LIPOOLIGOSACCHARIDES
I
I
I
1 Gala
1 GlcP
I Glcp
F. B.
P I (617) Gal-p(1,3)-GalNAc-p(l,4)-Gal-p(1,3)-Hep-a(1 ,J)-Hep-a(1,5)-KdO 3 3 4
I 2 Neu5Aca
I
PPEtn I 6
2[ [Gal-p(I ,3)-]GalNAc-p(1,4)]]-Gal-p(1,3)-Gal-p(1,3)-Hep-a(l ,I)-Hep-a(l ,5)-Kdo 3 2 2 4
I
I
2 Neu5Aca
2 NeuSAca
1 Glcp
I
I
I
1 Gala
1 Glcp
1 GlcP
8
G.
I
PEtn I 6 GalNAc-p(1,4)-Gal-~(1,4)-Glc-p(l,2).Hep-a(1,3)-Hep-a(l,5)-Kdo 3 4
2 NeuSAca
PEtn I (617) Gal-p(I ,J)-Hep-a(l ,J)-Hep-a(I ,5)-Kdo 3 4
C.
I
I 1 Glcp
2 Neu5Aca
I
I
2
1 Glcp
Neu5Aca
H. PEtn I 6
8
I
f[GalNAc-p(I ,4)-]Gal-p(1,4)-Glc-B( 1,2)-Hep-a(l,3)-Hep-a(l,S)-Kdo 3 4
2 Neu5Aca
I 2 Neu5Aca
D.
P/PEtn I 7 Gal-p(I ,3)-GalNAc-p(1,4)-Gal-p(1,3)-Hep-a(l,3)-Hep-a(l,5)Xdo 3 2 4
I 2 NeuSAca
I
I
1 Glcp
1 Glcp
8
I 2 NeuSAca
I 1 Glcp
AND PEtn
I 6 +[Gal-p(I ,J)-]GalNAc-f%(I ,4)-Gal-P(l,4)-Glc-P(1,2)-Hep-~~l ,3)-He~-a(la-Kdo 3 4
I 2 NeuSAca
I 1 Glcp
8
I 2 Neu5Aca
Figure 1. C. jejuni LOS outer core structures exhibiting molecular mimicry of gangliosides. Core structures of (A) serostrains HS:19 and HS:4 (GMla and GDla mimics), (B) isolate OH4384 (GTla mimic), ( C ) isolate OH4382 (GD3 mimic), (D) serostrain HS:10 and isolate PG836 (GD3 and GDlc mimics), (E) serostrain HS:1 (GM2 mimic), (F) serostrain HS:2 (GM3 and NCTC 11168 (also HS:2 serotype but outer core mimics GMla and mimic, i.e., without Gal-/3(1,3)-GalNAc-/3(1,4)) GM2 due to phase-variable terminal Gal), (G) serostrains HS:23 and HS:36 (GM2 mimic), and (H) phase-variable strain 81176 (GM2 and GM3 and mimics predominantly, but also GDlb and GD2). All sugars are in the pyranosidic form and are D-enantiOmerS. Gal, galactose; GalNAc, N-acetyl-galactosamine; Glc, glucose; Hep, L-glycero-D-manno-heptose;Kdo, 2-keto3-deoxyoctulosonic acid; NeuSAc, N-acetylneuraminic acid; P, phosphate; PEtn, phosphoethanolamine.
microheterogeneity is also present (Aspinall et al., 1998). This leads to mimicry of GM3-, GD2-, and GD3-like ganglioside structures. To date, the core oligosaccharide structures of eight C. jejuni serostrains (HS:l, HS:2, HS:3, HS:4, HS:10, HS:19, HS:23, HS:36, and HS:41), five C. jejuni HS:19 GBS isolates, one C. jejuni HS:41 GBS isolate, and two C. jejuni MFS isolates (serotypes HS: 10 and HS:23) have been characterized (Aspinall et al., 1993b, 1993c, 1994, 1995a; Moran and
O’Malley, 1995; Nam Shin et al., 1998; Szymanski et al., 2003; Yuki et al., 1993). In the LOS outer core of C. jejuni serostrain HS: 1, the extent of mimicry of GM2 ganglioside is limited to a terminal trisaccharide (Aspinall et al., 1993c) (Fig. lE), whereas in LOS of serostrain HS:2 (ATCC 43430), structural homology with gangliosides is limited to that of a disaccharide NeuSAc-a(2,3)-Gal (Aspinall et al., 1993b) (Fig. lF), which is found in a range of gangliosides including GM4, GDla, GTlb, and GM3 (Moran, 1997; Pren-
486
GILBERT ET AL.
dergast and Moran, 2000). Interestingly, the LOS of an aerotolerant form of the latter strain expresses the same core except for the absence of terminal sialylation (Hanniffy et al., 2001). On the other hand, the initial genome-sequenced strain C. jejuni NCTC 11168 (HS:2) exhibits GM2 and GMla mimicry over tri- and tetrasaccharides, respectively (Oldfield et al., 2002; Prendergast and Moran, 2000; St Michael et al., 2002) (Fig. 1F). Thus, core LOS heterogeneity in expression of ganglioside mimics within C. jejuni strains of the same serotype is not simply limited to HS:19 strains. The LOS of C. jejuni serostrain HS:4 has a core oligosaccharide with terminal regions that mimic GMla (
2A). For this reason, HS:3 LOS has been used as a negative control in serological studies examining C. jejuni cross-reactive antibodies in GBS patients (Moran et al., 200.5; Prendergast et al., 1999, 2001). Unusually, this LOS core contains a residue of 3amino-3,6-dideoxy-~-glucose (also known as quinovosamine, Qui3N) and no P-D-Gal residues able to accept NeuSAc, although GalNAc and GlcNAc are present. The sugar Qui3N is found in the core of a number of strains of Campylobacter coli (Beer et al., 1986), but in particular in LOS of C. coli HS:30 (Fig. 2B), where two Qui3N residues in the outer core carry two acyl groups (Aspinall et al., 1993a). The structures of the LOS cores of two C. lari strains, ATCC 35221 and PC 637, have been determined (Fig. 2C and D), and although containing Glc, GalNAc, and GlcNAc in their outer cores, they show no similarity to those of other C. jejuni strains (Aspinall et al., 199Sb, 199Sc). However, the outer core structures of the C. lari strains differ significantly, perhaps accounting for their recognition as distinct serotypes, but the inner cores also differ significantly from those of C. jejuni strains (Moran et al., 2000). In contrast to these other Campylobacter spp., C. fetus strains of different serotypes produce high111, LPSs (Moran et al., 1994), resembling those of enterobacterial LPSs (Raetz and Whitfield, 2002; Rietschel et al., 1990, 1994), and whose O-polysaccharide chains are required for interaction and anchoring of the virulence-associated S-layer to the bacterium (Yang et al., 1992). The structure of the 0-polysaccharides of C. fetus serotypes A, B, and AB have been established (Senchenkova et al., 1996, 1997), but only compositional data are available on the LPS core of C. fetus strains, which is sialylated (Moran et al., 1994). On the other hand, the presence of D,D-Hep in addition to L,D-Hep in the composition suggests the occurrence of a differing inner core region to that of C. jejuni where only L,D-Hep is found.
GENETIC VARIATION IN LOS BIOSYNTHETIC LOCUS The variability of the C. jejuni LOS structures is likely to have arisen as a consequence of selection for antigenic diversity imposed by different host immune mechanisms. Variation in LOS structures is due to diversity of the constituent sugars (number of carbon atoms, ring form, isomeric form, anomeric configuration, etc.), the derivatization of the sugars with noncarbohydrate moieties, and the linkages between the individual sugars (or monosaccharides). The formation of these linkages is determined by the glyco-
CHAPTER 27
A.
c. r E r m r LIPOOLIGOSACCHAEUDES
487
P I 7 [P-tIGalNAc-a( 1,4)-GalNAc-P(1,3)-Gal-a(1,3)-Hep-a(1,3)-Hep-a(1,5)-Kdo 3 2 4
I 1 Qui3NAcP
I
I
1 GlcNAcP
1 GlcP
B. Qui3Nacyl-~(1,2)-Qui3Nacyl-~(1,3)-GlcNAc-~(1,3)-Gal-~(1,3)-Hep-a(1,3)-Hep-a(1,5)-Kd0 2 2 4
I
I
I
1 Gala
1 Gala
1 GlcP 3
I 1 Hepa
C.
Glcp 1
PEtn
I
I
4 4 GlcNAc-P(l ,3)-GalNAc-a(l,3)-Gal-a(l ,3)-Hep-a(1,3)-Hep-a(1,5)-Kdo- a(l,5)-Kdo 2 2
I
I
1 Glca
1 Gala GlcP 1
D.
I 4 Glc-a(l,3)-GalNAc-P(1,3)-GalNAc-P(1,3)-Gal-P(1,3)-Hep-a(l,3)-Hep-a(1,fi)-Kdo 2 2 2
I
I
I
1 Gala
I Gala
1 Gala
Figure 2. Curnpylobacter LOS outer core structures that are not mimicking gangliosides. C. jejuni serostrain HS:3 (A), C. coli serostrain HS:30 (B), C. lari ATCC 35221 (C), and C. lari PC 637 (D). All sugars are in the pyranosidic form and are Denantiomers. Gal, galactose; GalNAc, N-acetyl-galactosamine; Glc, glucose; GlcNAc, N-acety-glucosamine; Hep, L-glyceroD-manno-heptose; Kdo, 2-keto-3-deoxyoctulosonicacid; P, phosphate; PEtn, phosphoethanolamine; QuiSNAc, N-acetylated 3-amino-3,6-dideoxygh1cosamine (quinovosamine); Qui3Nacy1, quinovosamine N-acylated with 3-hydroxybutanol or 3hydroxy-2,3-dimethyl-5-oxoprolyl chains.
syltransferases and other transferases encoded in the LOS biosynthesis locus. Also, sugar availability for incorporation into LOS is often determined by enzymes from the LOS biosynthesis locus that are involved in synthesis of sugar intermediates such as cytidine monophosphate-5-N-acetylneuraminicacid (CMPNeu5Ac). Comparisons of the various C. jejuni LOS biosynthesis loci indicate that a variety of genetic mechanisms are responsible for the synthesis of different structures (Gilbert et al., 2005). Notably, the most obvious genetic differences are differences in gene content and organization. Bioinformatics analysis of the complete genome sequence of C. jejuni NCTC 11168 identified a LOS
biosynthesis gene cluster that spans from Cjll3 1 (galE) to C j l l 5 2 (rfao or gmhD), encoding gene products involved in both inner and outer core LOS biosynthesis (Parkhill et al., 2000). This gene cluster is organized with inner core biosynthesis genes located at both ends of the cluster. At one end are galE, waaC, and htrB (WUUM)encoding uridine diphosphate- (UDP-)GlcNAc/Glc-4-epimerase (Bernatchez et al., ZOOS), heptosyltransferase I, and a Kdo-dependent lipid A acyltransferase, respectively. At the other end of the cluster are w a d , gmhA, gmhD encoding heptosyltransferase 11, sedoheptulose-7-phosphate isomerase, and ADP-L-glyceroD-rnanno-heptose- (L,D-Hep-) 6-epimeraseY respec-
488
GILBERT ET AL.
can be grouped together on the basis of gene content and organization: classes A, ByC, My and R (Fig. 3); classes D, F, I, J, K, N, Q, and S (Fig. 4); classes E, H, 0, and P (Fig. 5 ) ; and classes G and L (Fig. 4). It should be noted that another region in the LOS cluster that shows variability in gene content is found between waaF (Cj1148) and gmhA (Cj1149), where one or two genes are inserted in some strains (Fouts et al., 2005; Kanipes et al., 2007; Parker et al., 2006). Despite this, we will focus the discussion on the hypervariable region between htrB and waaF. The significant interest in the LOS biosynthesis gene cluster stems from the potential involvement of the ganglioside-like structures of C. jejuni LOSS in inducing GBS and MFS (Prendergast and Moran, 2000). As mentioned above, these ganglioside-like LOS structures contain sialic acid. A group of five LOS biosynthesis locus classes (A, By C, M, and R; Fig. 3) possess the necessary genes to synthesize sialylated LOS, in particular genes encoding a sialyltransferase (at-II), a sialic acid synthase (neuB), an N-acetylglucosamine-6-phosphate 2-epimerase (neuC), and a CMP-Neu5Ac synthetase (neuA). Of these particular LOS locus classes, the class C locus appears more distantly related to the others on the basis of sequence comparison. When the Cst-I1 variants of classes A, B, My and R are aligned together, the level of protein sequence identity was >77%
tively. In between are genes involved in biosynthesis of the LOS outer core, including those encoding glycosyltransferases and those involved in biosynthesis of sialic acid. Comparative genomic analyses by whole-genome microarray identified this outer core biosynthesis region as an intraspecies hypervariable region within C. jejuni (Dorrell et al., 2001; Leonard et al., 2003, 2004; Parker et al., 2006; Pearson et al., 2003; Taboada et al., 2004). Moreover, the identification of the LOS biosynthesis locus from strain NCTC 11168 facilitated the cloning of LOS biosynthesis genes as well as the loci from other C. jejuni strains possessing genes that were highly variable and often nonhomologous (Gilbert et al., 2000, 2002; Guerry et al., 2002; Parker et al., 2005). The sequences of the LOS biosynthesis genes between Cj1133 (waaC) and Cj1148 (wuaF) are available for over 70 C. jejuni strains (Gilbert et al., 2005; Godschalk et al., 2007; Parker et al., 2005; C. T. Parker et al., unpublished data). The 70 loci can be grouped into 19 classes (A to S; Figs. 3 to 5 ) on the basis of the gene content and organization showing extensive variability in length (from 5 to 16 outer core genes). There are a total of 50 distinct genes found in one or more of the LOS locus classes of C. jejuni. Some genes are unique to one class, whereas some other genes are present in more than one class but not in all. Indeed, several of the LOS locus classes
“A” class l a 2a
3a
4a
5a
6a
7a
8a
9a
10a I l a 12a 13a
cgtA
cgf5 csfli neu5 neuC neuA
5bI
6b 7b
cgfAl
cgt5 csfll neu5 neuC cgfAN neuA
mmm-.
hW5
waaF
“B”class Ib 2b
3b
4b
8b
9b
5bII 10b l l b 12b 13b
.+
----a
“C”class Ic 2c
3c
4c
14c 15c
6c
7c
8c
9c
5c11Oc
16c 12c 13c
....,
mm-m.
cgt5 csflll neu5 neuC cgtAIl/neuA
“M” class lm2m
19m 3m
17m
7m
8m
9m 10mIlm 12m 13m
cstli neu5 neuC neuA
“R” class
----.I r 2r
3r
4r
5r cgtA
6r
7r
cgt5cstll
8r
9r
10r 16r 12r 13r
.--+
neuB neuC neuA
Figure 3. C. jejuni LOS classes that possess a sialyltransferase gene. Classes A, B, C, M, and R (C. T. Parker et al., unpublished data).
CHAPTER 27
“D” class Id
Zd
3d 17d
18d
19d
.,
20d 18d 12d 13d
htrB
If
---,
2f
3f
18f
19f
12f 13f
waaF
“J” class
“I”class 171
I81
19i
20i
40i
-
,.-
411 42i 43i 44i
45i 121 13i
1
441 45j 121 13j
w,
- 1 - 1
Gene cassette from HS:41 caosule locus
Gene cassette from HS41 capsule locus
“S” class
“K” class I k 2k
.,
ZOf 16f
htrB
waaF
3i
489
“F” class
---I
11 2i
C. TE7UNI LIPOOLIGOSACCHARIDES
3k 17k
1s 2s 18k
19k
3s
18s
19s
20s 40s
41s 42s 43s 44s
,.
.,
45s 12s 13s
49k50k 12k 13k
Gene cassette from HS 41 capsule iocus
“N” class I n 2n
3n 17n
18n
38n
- ,.
Class G-like gene
“0”class
--,
I g 2g
---,
3g 35g
3%
379 169 38g 12g 139 waaF
hfrB
“L” class I1 21
31
351
361
371
471 481 161 121 131
Figure 4. C. jejuni LOS classes related to LOS class D, class F, and class G (C. T. Parker et al., unpublished data).
(C. T. Parker et al., unpublished data). The pairwise alignments between Cst-I11 of the class C and each variant of Cst-I1 gave 53% protein sequence identity on average. It should be noted that strains possessing locus classes A, By and C are commonly isolated, with these classes constituting >70% of >lo0 strains examined in one study (Parker et al., 2005). Also, the majority of GBS-related strains have the LOS locus class A, while for the majority of MFS-related strains examined, it is class B (Godschalk et al., 2004,2007; Koga et al., 2006; Parker et al., 2005). Several genomes of non-jejuni Cumpylobucter species have recently been sequenced. There is no evidence of a LOS core biosynthesis locus in C. hominis ATCC BAA-381 (GenBank accession no. CP000776), at least on the basis of the absence of WaaC and WaaF homologs. LOS core biosynthesis loci with an organization somewhat related to C. jejuni are found in C. coli RM2228 (GenBank accession no. AAFLO1000002), C. luri RM2100 (GenBank accession no. AAFI
(GenBank accession no. AAFJ01000006), C. curvus 525.92 (GenBank accession no. NC-009715), C. concisus 13826 (GenBank accession no. CP000792), and C. fetus 82-40 (GenBank accession no. NC008599). Thus, at either end of these LOS loci are the heptosyltransferase genes, wauC and waaF, and between are genes involved in biosynthesis of the LOS outer core, including those encoding glycosyltransferases (Fouts et al., 2005; JCVI sequencing). In particular, the LOS locus of C. coli RM2228 is quite reminiscent of C. jejuni LOS class F, sharing many of the same gene products on the basis of amino acid similarity, and it is possible that other LOS loci of C. coli are similar to certain C. jejuni LOS locus classes. Indeed, the presence of the sugar Qui3N in the core of a number of strains of C. coli (Beer et al., 1986) and in a C. jejuni LOS class H suggests that a class resembling E, H, 0, or P may be found in C. coli. Moreover, microarray analysis of C. coli and C. luri shows that there is extensive variation in the gene content of the LOS locus in these non-jejuni Cum-
GILBERT ET AL.
490
r
pylobacter species (C. T. Parker et al., unpublished data), and this may extend to the other Campylobacter species. It should also be noted that within the LOS loci of C. lari RM2100 and C. upsaliensis RM3 195, there are insertions of clusters of contiguous genes that have homologs in NCTC 11168 that are unrelated to LOS biosynthesis, including several genes involved in flagellar modification in C. lari RM2100 (open reading frames [ORFs] CLAO453, CLAO454, CLA0457, CLA0458, CLAO459, and CLAO460 in GenBank accession no. AAFK01000004). It is unclear what role this genomic reorganization plays in the biosynthesis of LOS in these organisms, or whether these recombination events are common within these species.
RECOMBINATION AND LATERAL EXCHANGE BETWEEN LOCI OF LOS CLASSES
I I
"
I I I
I
I
Sequence analyses suggest that the LOS loci of C. jejuni strains are hot spots for genetic exchange and can lead to the creation of mosaic LOS loci. It has been observed that the introduction of a complete class of LOS biosynthesis loci can occur between strains by horizontal transfer (Gilbert et al., 2004; Phongsisay et al., 2006). Evidence suggests that these events involve recombination between homologous regions that flank the LOS biosynthesis locus, e.g., by exchanging LOS locus class C with class A as observed in strain G B l l (Gilbert et al., 2004). There is also evidence of recombination occurring within the LOS locus to create mosaics of different classes. Two examples are the LOS classes M and R that may have arisen by recombination between classes A and D loci (Figs. 3 and 4) and classes A and C loci (Fig. 3 ) , respectively (C. T. Parker et al., unpublished data). Moreover, considering the similarity between LOS classes D and F, there has been some recombination event that resulted in the replacement of orf3 and orfl7 of class D with orf3 of class F, or vice versa. Comparison of the LOS class A, class By and class C DNA sequences allowed the proposal of a sequence of evolutionary steps that has resulted in class C strains arising from a class A strain (Gilbert et al., 2005). The fundamental difference between LOS locus classes A and B is the presence of a second cgtA (cgtA-II) between neuC and neuA in the latter class (Gilbert et al., 2005). The cgtA-II is likely to be the result of a duplication of cgtA, although horizontal transfer cannot be ruled out. In any case, the DNA sequence of cgtA-II is well conserved (99% identity) among the nine GenBank entries available on August 20, 2007. The insertion of cgtA-II between neuC and neuA is precise because it has resulted in the addition
CHAPTER 27
of a single base upstream of its start codon and in no base addition after its stop codon (i.e., the TGA stop codon of cgtA-11 overlaps with the ATG start codon of neuA). The class B locus appears to be an evolutionary intermediate between classes A and C, with at least three more recombination events necessary to generate a class C locus. The first would be the in-frame fusion of @A-11 and neuA that is distinct in the class C locus. The second would be the insertion of orf14c and orfl Sc encoding two glycosyltransferases with the deletion of cgtA-I. The third event involves the concomitant insertion of orfl6c and the deletion of orfl 16. It is certainly possible that other evolutionary intermediates exist with different combinations of inserted or deleted genes and with cgtA and neuA either as separate genes or as an in-frame fusion. In particular, LOS class R may be an intermediate, possessing the insertion of orfl6d with the deletion of orfl l a . In addition, there are important evolutionary events common to both class A and B sequences that lead to significant structural differences between strains of the same LOS class. For instance, the sequence identity for CgtA (encoding a P-1,4-Nacetylgalactosaminyltransferase, P-1,4-GalNAcT) and CgtB (~-1,3-galactosyltransferase, P-lY3-GalT)can be as low as 73 and 57%, respectively, when comparing variants from the same class (Gilbert et al., 2002). Comparisons of classes A and B sequences have also shown the transfer of a 1,200-bp cassette that spans the 3' ends of both cgtA (cgtA-I in class B) and cgtB. Because cgtA and cgtB are translated in opposite (but converging) orientations, the divergence of this DNA region results in a large number of amino acid substitutions in the C terminus of both the plY4-GalNAcTand the /3-1,3-GalT. The C-terminal regions of the P-lY4-GalNAcT and the p-lY3-GalT are responsible for their transferase specificity for either sialylated or nonsialylated acceptors (Bernatchez et al., 2007; Gilbert et al., 2002). Consequently, the transfer of this DNA cassette between strains will have an impact on the structure of the outer core being synthesized. For instance, ATCC 43438 (HS: 10) has a class A locus while MF6 has a class B locus, but they both have CgtA (CgtA-I for MF6) and CgtB variants that are specific for nonsialylated acceptors. Similarly, ATCC 43446 (HS:19) has a class A locus and ATCC 43456 (HS:36) has a class B locus, but this cassette is nearly identical (99% DNA sequence identity) between these two strains. ATCC 43446 (HS:19) has CgtA and CgtB variants specific for sialylated acceptors, while ATCC 43456 (HS:36) has a CgtB variant with the same specificity but an inactive CgtA-I as a result of the deletion of a base. A group of LOS classes sharing genes similar to those required for the biosynthesis of 6-deoxysugarsY
C. 7E7UNI LIPOOLIGOSACCHARIDES
491
i.e., rmlA and rmlB homologs, demonstrate clearly that novel classes of LOS loci arise from various insertion and deletion events occurring within particular LOS loci. Including the mnZA and rmlB homologs, these related LOS locus classes (classes E, H, 0, and P; Fig. 5 ) have at least 14 distinct outer core LOS genes in common. When compared with class E, the class 0 LOS locus has a deletion of orf28, while the class P LOS locus has an insertion of a putative butyrltransferase gene (orf39) with the concomitant deletion of the 5' region of 0rf26. Both of these changes exist in the LOS loci of classes H and 0, and the class P LOS locus appears to be intermediate between classes E and H. Moreover, the slightly higher sequence identity between LOS locus classes P and H from the region htrB to orf39 and the slightly higher sequence identity between classes 0 and H in the orf27 to waaV LOS region suggest that class H arose by homologous recombination between class 0 and class P loci (Parker et al., 2005; C. T. Parker et al., unpublished data). There is also evidence of the creation of mosaic LOS classes through the acquisition of CPS biosynthesis locus genes (C. T. Parker et al., unpublished data). Like the LOS biosynthesis locus, the C. jejuni CPS biosynthesis locus contains genes encoding glycosyltransferases and other transferases for the synthesis of this polysaccharide (Karlyshev et al., 2005a). There are also genes in the CPS locus that are involved in synthesis of sugar intermediates used in the synthesis of the polysaccharide. Comparisons of classes D and F LOS loci with classes I, J, and S (Fig. 4) have demonstrated the insertion of a cassette of five contiguous CPS biosynthesis locus genes from C. jejuni HS:41 strains (C. T. Parker et al., unpublished data). It has yet to be determined whether these genes affect LOS structures, and it is possible that the gene products have a role in the synthesis of capsular structures or other putative polysaccharide molecules. Indeed, it has previously been observed that both LOS and capsular gmhA genes (sedoheptulose7-phosphate isomerase) can be involved in the heptose pathways of LOS and CPS biosynthesis (Karlyshev et al., 2005a). Until both LOS and CPS structures are determined for these strains, it will remain unclear whether these gene products play a role in these pathways. The mechanisms of recombination that leads to many of the insertion and deletion events at the LOS loci are not known, and our ability to identify LOS classes relies on the evolutionary stability of the locus class. Aside from the complete replacement of a LOS locus through recombination at homologous flanking genes, there is often a lack of homology between two classes where insertions and insertion/ deletions oc-
492
GILBERT E T AL.
cur. Despite this barrier, new classes may be produced by recombination between the large number of A and T polynucleotides in these LOS loci. Indeed, the G+C content of the LOS loci (22 to 28%) is slightly lower than the G+C content for the rest of the genome (30.5%) (Parkhill et al., 2000). It is likely that this sort of nonspecific mechanism would result in many losses of function insertion/deletion events, and the resulting nonfunctional LOS classes would be lost by selection and hence not identified. It is also quite possible that recombination events introduce incompatible gene, e.g., glycosyltransferases or modification enzymes that do not recognize the existing LOS structures. These new LOS loci certainly substantiate the need for further studies of the LOS biosynthesis region and resulting LOS structures.
LOS BIOSYNTHESIS PATHWAYS Lipid A and Inner Core Biosynthesis
C. jejuni has a so-called mixed lipid A moiety composed of a hybrid backbone of a p(l’-.6)-linked GlcN3N-GlcN disaccharide carrying phosphate groups at positions 1 and 4’ and substituted with six fatty acids (both primary and secondary) in an asymmetrical distribution on the disaccharide backbone (Moran et al., 1991b). The synthesis of the Kdo,lipid A subunit in gram-negative bacteria has been intensively studied (Raetz and Whitfield, 2002; Raetz et al., 2007). Biosynthesis of the Kdo,-lipid A is a constitutive pathway, whereas the modification of the Kdo,-lipid A tends to be induced or repressed by growth conditions. Raetz et al. (2007) reported that the nine enzymes involved in the biosynthesis of the Kdo,-lipid A subunit and the single-copy genes encoding them (lpxA, lpxB, lpxC, lpxD, lpxH, IpxK, IpxL, IpxM, and kdtA) are conserved in most gram-negative bacteria, except for 1pxH (UDP-2,3diacylglucosamine hydrolase), which is absent in about one-third of the gram-negative genomes, including C. jejuni (an alternative pyrophosphatase is likely to be present but has not been identified to date). Seven of the other eight genes have homologs in C. jejuni NCTC 11168 and Rh41221, the only exception being lpxM (msbB), which encodes a myristoyl-acyl carrier protein-dependent acyltransferase (Raetz et al., 2007). Indeed, C. jejuni HtrB (Cj1134, ORF 2) shows low protein identity with both E. coli LpxL (23%) and LpxM (19%), which are responsible for the incorporation of laurate (C12) and myristate (C14), respectively. Because the two secondary acyl chains are mostly palmitate (CJ (Moran et al., 1991b), it is possible that HtrB can
transfer to both positions. Phongsisay et al. (2007) have shown that the C. jejuni htrB can partially complement a Salmonella htrB mutant, but they could not obtain a viable htrB mutant in C. jejuni, which suggests that it is an essential gene. The synthesis of the so-called mixed lipid A moiety in C. jejuni requires the conversion of UDPGlcNAc to UDP-GlcNAc3N to produce a hybrid backbone with a p( 1’-+6)-linkedGlcN3N-GlcN disaccharide (Moran et al., 1991b). The conversion of UDP-GlcNAc to UDP-GlcNAc3N has been studied in Acidithiobacillus ferrooxidans (Sweet et al., 2004) and involves GnnA (a dehydrogenase) and GnnB (a transaminase). Cj0504c and Cj0505c from C. jejuni NCTC 11168 have low homology with GnnA and GnnB, respectively, so these putative ORFs might play a similar role in C. jejuni lipid A biosynthesis. Further variation of the lipid A unit is due to an increased degree of phosphorylation with phosphoethanolamine at the 1 and/or 4’ phosphate groups (Szymanski et al., 2003). In E. coli, the phosphoethanolamine are added at these positions by EptA (Trent and Raetz, 2002). There is an EptA homolog (Cj0256, 40% identity) in C. jejuni NCTC 11168, but it has not been studied to date. The biosynthesis pathway of ADP-L,D-Hep from ~-sedoheptulose-7-phosphatehas been proposed by Karlyshev et al. (2005a, 2005b) (chapter 28). ADPL,D-Hep is the donor used by the heptosyltransferase WaaC (Cj1133, ORF 1) that transfers HepI to Kdo (Kanipes et al., 2006; Klena et al., 1998), and heptosyltransferase WaaF (Cjl148, ORF 13) that transfers HepII to HepI (Oldfield et al., 2002). The inner core can be further modified by the addition of phosphate or phosphoethanolamine to the 6 or 7 position of HepI (Moran et al., 2000). Two genes (lpt3 and lpt6) encoding phosphoethanolamine transferases specific for the inner core HepII have been identified in Neisseria meningitidis (Wright et al., 2004). However, Lpt6 has no significant homologs in C. jejuni, while Lpt3 has low homology (23% protein sequence identity) with Cj0256 that seems more likely to be responsible for the transfer of phosphoethanolamine to the lipid A unit (see above). Consequently the genes responsible for the addition of phosphate or phosphoethanolamine to HepI in C. jejuni remain elusive.
LOS Outer Core Biosynthetic Pathway The elucidation of a biosynthetic pathway involves identifying the genes responsible for each step, and it is also desirable to assign a function to each gene present in a locus, although it is possible that a specific gene might be involved in a distinct pathway.
C. TETUNI LIPOOLIGOSACCHARIDES
CHAPTER 27
assays performed with synthetic fluorescent acceptors have shown that CgtA (ORF 5a), CgtB (ORF 6a), and Cst-I1 (ORF 7a) are a P-1,4-GalNAcTYa P-lY3-GalT, and a bifunctional a-2,3-/2,8-sialyltransferase (Gilbert et al., 2000). We propose that Cst-I1 is adding the single sialic acid (i.e., NeuSAc) on the inner Gal residue, and because Cst-I1 is a bifunctional a-2,3-/ 2,8-sialyltransferaseYthat it also adds both sialic acid residues to the terminal Gal residue. If it is responsible for adding the single a-2,3-sialic acid on the inner Gal residue, then it is curious that it does not also add a second a-2,8-sialic at this position. It is established that CgtA from C. jejuni OH4384 cannot transfer a p-ly4-GalNAc residue to a disialylated acceptor (Gilbert et al., 2002), which means that the transfer of a second Neu5Ac on the inner Gal residue in OH4384 would result in a truncated structure (i.e., a GD3 mimic) such as the one observed in C. jejuni OH4382 (Fig. 1C) where CgtA is inactive because of a frameshift mutation. It is possible that a small amount of the GD3 mimic is actually present in C. jejuni OH4384 but was undetected because the GTla mimic is the main outer core structure that is synthesized. Three of the remaining ORFs are involved in the synthesis of CMP-NeuSAc, the donor used by Cst-11. ORF 8a is homologous with NeuB, a sialic acid synthase (Sundaram et al., 2004); OW 9a is homologous with NeuC (UDP-GlcNAc-2-epimerase/N-acetylmannosamine synthase); and ORF 10a is homologous with NeuA that is found in various organisms and that is responsible for the synthesis of CMPNeuSAc, from CTP and NeuSAc. As mentioned above, ORF 2a is homologous with HtrB, which has been proposed to transfer the two secondary acyl
Gilbert et al. (2005) proposed a model for C. jejuni OH4384 possessing a class A locus. Here, we update the pathway for LOS biosynthesis that has been proposed for C. jejuni OH4384 (Fig. 6). Functions can be assigned to 1 2 of the 13 ORFs present in the LOS biosynthesis locus of OH4384 (Fig. 6), the only exception being ORF 12a. ORF 12a is present in all 19 classes (Figs. 3 to 5 ) , and according to sequence homology, it is a putative glycosyltransferase. Its function has not been determined yet, and it is possible that it plays a role in specific classes only. Of the remaining 12 ORFs, 7 encode glycosyltransferases whose functions have been deduced either by homology with well-characterized glycosyltransferases, knockout constructs, or functional characterization of the purified recombinant glycosyltransferases. ORFs l a and 13a (waaC and waaF) show homology with heptosyltransferases I and 11, respectively, and they are proposed to add the first and second heptosyl residues (HepI and HepII) in strain OH4384. ORF 3a (Cj113.5) is homologous to LgtF from Haemophilus ducreyi that encodes a P-1,4-glucosyltransferase (P-lY4-GlcT)capable of transferring glucose to heptose (Filiatrault et al., 2000), and we propose that ORF 3a performs a similar role in C. jejuni OH4384. ORF 4a (Cj1136) is homologous with many bacterial /3-lY3-GalTsand is a good candidate for transference of the p-lY3-Galresidue to HepII. Nevertheless, none of these ORF 4a homologs has been shown experimentally to perform exactly this function, and this assignment has to be considered speculative. There is substantially more experimental evidence to support the proposed pathway for the addition of the other outer core residues. Biochemical
P cj1136
waaF
I
waaC
6V) Gal-p(1,3)-GalNAc-fi(1,4)-Gal-p(1,3)-Hep-a(l,3)-Hep-a(1,5)3 3 4 cjf135 I cst-I1 I cst-ll I 2 2 I NeuSAca NeuSAca Glcp 8
I
cst-11
2 Neu5Aca HepTase-l
Lipid A acylTase GlcTase
....... l a ]
2a
waaC htrB
>I
3a
cj1135
CMP-NeuAc synthetase
Sialyltransferase GalTase GalNAcTaseGalTase
> I4a)I cj1136
5a
cgtA
>(
493
6a
I
1
7a
lNeuAc
>I
cgtB cst-I1
8a
neuB
>I
1
SOAT
HepTase-ll ORF
J
neuC neuA orfll orf12a waaF
1 kb
Figure 6. Biosynthesis of the LOS core in C. jejuni OH4384. Adapted from Gilbert et al. (2005).
494
GILBERT ET AL.
chains in lipid A. Finally, ORF l l a has been demonstrated to be a sialylated 0-acetyltranferase (Houliston et al., 2006). However, the ORF l l a variant present in C. jejuni OH4384 is inactive because of a Gly75Asp mutation. When active, this enzyme catalyzes the transfer of 0-acetyl groups onto oligosaccharide-bound sialic acid, with a high specificity for terminal a2,8-linked residues. The modification is directed to C9 and not C7, as is considered to occur more commonly in other organisms. Mechanisms Generating LOS Core Region Variation In addition to gene content differences, C. jejuni uses four genetic mechanisms to affect glycosyltransferase activity and hence vary the LOS outer core structures: (i) phase variation due to homopolymeric tracts, (ii) gene inactivation by the deletion or insertion of single or multiple bases, but without phase variation, (iii) amino acid substitution resulting in an inactive variant of the glycosyltransferase, and (iv) single or multiple mutations leading to variant glycosyltransferases with different acceptor specificities (Gilbert et al., 2002, 2005). These mechanisms are observed mostly in the four genes encoding enzymes involved in the synthesis of the distal outer core (cgtA, cgtB, cst-II/III, and o r f l l ; Table 1). So far, only the mechanism involving premature translational stop due to non-phase-variable deletion or insertion of a base has been observed for the two glycosyltransferases involved in the extension of the inner core (Cj1135 and Cj1136; Table 1). Phase variation is a mechanism resulting in highfrequency on-off switching of a gene (Levinson and Gutman, 1987). Homopolymeric tracts are the simplest motifs of the short sequence repeats that are associated with phase-variation. They are subject to a slipped-strand mispairing mechanism during replication, which results in a mixture of translationally in-frame or out-of-frame variants in microorganisms where phase-variation occurs. Phase-variable homopolymeric G-tracts have been found in at least nine C. jejuni LOS biosynthesis genes (Gilbert et al., 2005), and there is one example of a phase-variable homopolymeric A-tract, in the cst-I1 gene of C. jejuni MF6 (GenBank accession no. AY422196). Interestingly, many genes can be present as either phasevariable or non-phase-variable forms (e.g., cst-II), while other genes have been observed only as phase variable (e.g., cgtA-II) or only as non-phase variable (e.g., C j l l 3 5 and Cj1136). Examples of homologous glycosyltransferases having different acceptor specificities are found
among the three glycosyltransferases transferring the distal sugar residues of the sialylated outer cores, namely CgtA, CgtB, and Cst-II/III. Variants of the p1,4-GalNAcT (CgtA, CgtA-I and CgtA-11) only share 34% conserved residues, although they are clearly homologous (Gilbert et al., 2002). Biochemical assays that used synthetic fluorescent oligosaccharides showed that CgtA of C. jejuni ATCC 43438 (HS:10) was specific for a nonsialylated acceptor; the version from OH4384 was specific for a monosialylated acceptor, whereas those from C. jejuni ATCC 43456 (HS:36) and NCTC 11168 were capable of using both monosialylated and disialylated acceptors (Gilbert et al., 2002). A similar study has been performed with CgtB variants that share 47% conserved residues (Bernatchez et al., 2007), and CgtB of strain ATCC 43438 (HS:10) was found to be specific for a nonsialylated acceptor, whereas the versions from C. jejuni OH4384 and NCTC 11168 were specific for monosialylated acceptors. The acceptor specificities of CgtA and CgtB observed with the model compounds correlate with the natural acceptors, i.e., whether the inner Gal residue is sialylated or not. The Cst-I1 sialyltransferase is present either as a monofunctional a(2,3)- or bifunctional a(2,3/ 8)variant, which correlates with the presence of NeuSAc-a(2,3)-Gal and NeuAc-a(2,8)-Neu5Aca(2,3)-Gal in the LOS outer cores. The crystal structure of an active truncated version of Cst-I1 from C. jejuni OH4384 has been reported and shown the importance of a histidine (a general base) and two tyrosine residues (responsible for coordination of the phosphate leaving groups) in the catalytic mechanism (Chiu et al., 2004). The 15 Cst-I1 variants available to date in GenBank share 77% identical residues. Site-directed mutagenesis was used to study the two closely related Cst-I1 proteins from strains OH43 84 and ATCC 43446 (HS:19) that share 97% identity (Gilbert et al., 2002). Asn5 1 was found to be the critical residue for the bifunctional activity of Cst-I1 from C. jejuni OH4384. However, a Cst-I1 variant with Asn5 1, and that has only a-2,3-sialyltransferase activity, has been observed in C. jejuni GB26 (Godschalk et al., 2007). Cst-I1 from GB26 has diverged significantly from the other Cst-I1 sequences, and it is possible that one (or several) amino acid substitution has inactivated the a-2,8-sialyltransferase activity in that variant. The six distinct entries of Cst-I11 (from different strains) are 100% identical. There is actually an inactive variant of Cst-I11 (from C. jejuni GB1, GenBank accession no. EF066651) resulting from the deletion of five bases that causes a premature translation stop; otherwise, the protein sequence would be 100% identical to the other Cst-I11 entries.
CHAPTER 27
c. r E r m r LIPOOLIGOSACCHARIDES
495
Table 1. Variants of glycosyltransferasesand sialate 0-acetyltransferase (SOAT) involved in the biosynthesis of LOS outer cores Genes Cj1135
Cj1136
cgtAl cgtA-11cgtA-11
Function
cst-Ill cst-Ill
orfl1
Examples of strains
One domain (transfers Glc-p1,4 to HepI) One domain with premature translational stop (inactive) Two domains (transfers Glc-P1,4 to HepI and Glc-p1,2 to HepII) Two domains with premature translational stop in second domain (transfers Glc-p1,4 to HepI) ~-1,3-Galactosyltransferase (to HepII) Full length (transfers Gal-p1,3 to HepII) Premature translational stop (single-base deletion)
ATCC 43431 LI087
p- 1,4-N-acetyl-
OH4382 MF6 (CgtA-1) ATCC 43438
p-Glucosyltransferase
galactosaminyltransferase (to Gal)
cgtB
Variants
~-1,3-Galactosyltransferase(to GalNAc)
a-2,3- or a-2,3/ 8-sialyltransferase
Sialate 0-acetyltransferase (specific for a-2,8-linked NeuAc)
Premature translational stop (single-base deletion) The acceptor is a nonsialylated Gal The acceptor is a nonsialylated Gal (phasevariable G tract) The acceptor is a monosialylated Gal The acceptor can be a mono- or disialylated Gal and cgtA is fused with neuA The acceptor can be a mono- or disialylated Gal (phase-variable G-tract) Inactive due to Cys92Tyr mutation Premature translational stop (single-base deletion) The acceptor GalNAc is attached to a nonsialylated Gal The acceptor GalNAc is attached to a monosialylated Gal The acceptor GalNAc is attached to a monosialylated Gal (phase-variable G tract) Monofunctional a-2,3- to Gal ((3-11) Monofunctional a-2,3- to Gal ((3-111) Bifunctional a-2,3- to Gal and a-2,8- to NeuAc Bifunctional a-2,3- to Gal and a-2,8- to NeuAc (phase-variable G tract) Bifunctional a-2,3- to Gal and a-2,8- to NeuAc (phase-variable A tract) Premature translational stop (5-base deletion) Full-length non-phase variable Phase-variable G tract Premature translational stop due to G476A mutation Premature translational stop due to base deletion Inactive version due to Gly75Asp mutation Inactive version due to Gly91Glu mutation
Alternative Extensions from the HepII Residue in Sialylated Outer Cores The HepII-a(1,3)-HepI disaccharide in the inner core is substituted with either only one Glc residue on HepI or with two Glc residues, the second one being on HepII (Fig. 1).The Glc residues are transferred by ORF 3 (Cj1135), which can be present either as a single domain enzyme transferring only the first Glc residue or as a two-domain enzyme that can transfer both Glc residues (Gilbert et al., 2005). Some variants of the two-domain ORF 3 have a frameshift mutation in the second domain and can only transfer the first Glc. The presence of one or two Glc residues on the HepII-a(1,3)-HepI disaccharide is the first
NCTC 11168 OH4384 OH4384 81-176
OH4384 NCTC 11168 ATCC 43456 (CgtA-11) ATCC 43430 81-176 ATCC 43438 OH4384 NCTC 11168 ATCC 43446 NCTC 11168 OH4384 CCUG 10954 MF6 GB 1 HB93-13 ATCC 43438 MF7 81-176 OH4384 ATCC 43456
branching point leading to the various sialylated outer cores (Fig. 7). Most sialylated LOS outer cores are extended from HepII after the addition of a p1,3linked Gal residue, but there are a few exceptions that are extended from the p172-linked Glc residue on the HepII after addition of a plY4-linkedGal residue (Fig. 7). These strains seem to have an inactive ORF 4 (Cj1136) and an additional GalT that can use the plJ-linked Glc residue as an acceptor. The best studied example is that of C. jejuni 81-176 (Guerry et al., 2002, Kanipes et al., 2007). The annotation of the LOS locus of C. jejuni 81-176 (GenBank accession no. AY862985) suggests a complete ORF for the Cj 1136 homolog that overlaps with the previous ORF (Cjll35) by 10 bp. Nevertheless, we propose
496
GILBERT E T AL.
WaaF
WaaC
HepII-a( 1,3)-HepI-a(1,s)-
\
/
One-domain Cj1135 or second domain inactive
Two-domainCj1135
HepII-a( 1,3)-HepI-a(1,S)4
I
I
I GlcP
1
GlcP Cj1136
1
Gal+( 1,3)-He I1 a(l,3)-HepI-a(1,5)4
2I
I
1 GlcP
1
GlcP
Gal+( 1,3)-Gal-P(1,3)-HepII-a(1,3)-HepI-a(1,5)2 2 4 I I I
i
Gala
I I GlcP
Inactive Cj1136 and presence of Galp-(I p)-Tase
Gal-P( 1,3)-HepII-a(l,3)-HepI-a(1,5)4 Continued in Figure 9
HepII-a( 1,3)-HepI-a(1,5) 2 4
i
GlcP
I
HepII-a(l,3)-HepI-a( 1,5)2 4
I
1 GlCP
1 GlcP
4
I
1
GlcP
Continued in Figure 10
i
GlcP
Continued in Figure 8
Figure 7. Alternative extensions of the LOS outer core from HepII.
that the real start codon is 4 bp downstream of the stop codon of Cjll35. This alternative start codon would generate an N-terminal protein sequence that aligns perfectly with the Cj1136 homologs that are known to be active. However, a single base deletion after 64 bp results in a truncated (31 amino acid) Cj1136 in C. jejuni 81-176, which is consistent with the absence of Gal/31,3 extension from HepII in that strain. Indeed, Kanipes et al. (2007) have reported that the inactivation of Cj1136 in C. jejuni 81-176 had no effect on the LOS structure. Their data suggests that a putative glycosyltransferase located between waaF and gmhA (GenBank accession no. AY423899) is responsible for the addition of the plY4-linkedGal residue to the plY2-linkedGlc.
most extended class C structure is a mixture of GMlal GM2 mimics that result from phase variation of the CgtB (Cj1139c) /3-1,3-GalT (Linton et al., 2000; St Michael et al., 2002). A truncated structure mimicking GM2 in ATCC 43429 (HS:1 serostrain) is due to a phased-off CgtB (Cj1139c), and a structure mimicking GM3 in ATCC 43430 (HS:2 serostrain) is due to a Cys92Tyr mutation that inactivates CgtA in that strain (Gilbert e t al., 2002; Fig. 8). C. jejuni GB1 has a 5-bp deletion in Cst-I11 (Cj1140), which results in a LOS outer core without Neu5Ac (Godschalk et al., 2007).
Biosynthesis of Class C Core Structures
Many class A structures have a single Glc residue attached to the inner core on HepI (Fig. 1A to C; Fig. 9). These strains all have a two-domain OW 3c with a frameshift mutation that inactivates the second domain and prevents the addition of a second Glc residue. The absence of a Glc residue on HepII seems to allow the sialylation of the Gal residue attached to the HepII. By comparison, class A strains with a fulllength two-domain ORF 3 have two Glc residues on the inner core (Fig. 1D and 10) and no Neu5Ac on the inner Gal residue. Depending on whether CgtA (/3-lY4-GalNAcT)is active or not, two series of struc-
Karlyshev et al. (2005b) proposed a model for the biosynthesis of the LOS outer core in C. jejuni NCTC 11168 possessing a class C locus. As mentioned above, the two Glc residues are transferred by ORF 3c (Cj1135). On the basis of sequence similarities, the transfer of the Gal-P(1,3) to HepII and subsequent attachment of a p-1,3 Gal and a-1,2 Gal to Gal-/3(1,3)-HepII were proposed to be mediated by ORF 4c (Cj1136), ORF 15c (Cj1138), and ORF 14c (Cjl137), respectively (Karlyshev et al., 2005b). The
Biosynthesis of Class A Core Structures with a Sialylated Inner Gal
c. rEruNr LIPOOLIGOSACCHARIDES
CHAPTER 27
497
Gal-P(1,3)-G 1-p(1,3)-HepII-a( 1,3)-HepI-a(1,5)-
P
1 Gala
2
4
I
I GlcP
I
I
GBI (inactive Cst-Ill)
i
GlcP
cst-Ill (Cj1140)
Gal-P(1,3)-G 1 P( 1,3)-HepII-a( 1,3)-HepI-a(1,5)3 $2 4
I
2 Neu5Aca
I
1 Gala
I
1 GlcP
I
I
Serostrain HS:2 (inactive CgtA)
1 GlcP
(Cjll43)
cg*
GalNAc-P(1,4)-Gal-P(1,3)-Gal-P(1,3)-HepII-a( 1,3)-HepI-a(1,5)2 3 4
I
2 Neu5Aca
?1
1
Gala
GlcP
I
I
I
(Cj1139) CgtB
+/-[Gal+( 1,3)-]GalNAc-P(1,4)-Gal-P(1,3)-Gal-P(1,3)-HepII-a( 1,3)-HepI-a(1,5)3 2 2
I
2 Neu5Aca
I
1 Gala
Serostrain HS:I (inactive CgtB)
1 GlcP
I
1 ~ l ~ - j
41
NCTC 11168 (phase-variable CgtB)
GlcP
Figure 8. Biosynthesis of class C core structures.
tures can be synthesized once the inner Gal is sialylated. Some class A strains have a 1-bp deletion that causes a premature translation stop of CgtA and result in truncated LOS outer cores. The presence of a mono- or bifunctional Cst-I1 will then direct the synthesis of either a GM3 or GD3 mimic, respectively. The presence of an active CgtA will direct the synthesis of a GM2 mimic. Strain GB23 has an inactive CgtB (P-lY3-GalT)resulting from a frameshift mutation that prevents further elongation (Godschalk et al., 2007). Strains with an active CgtB will add a terminal Gal residue that can be sialylated with one (GDla mimic) or two (GTla mimic) NeuSAc residues, depending on the presence of a mono- or bifunctional Cst-11, respectively. Incomplete sialylation can also be observed, as is the case for ATCC 43446 (HS:19 serostrain), which expresses a mixture of GMla and GDla ganglioside mimics (Aspinall et al., 1994) and in eight GBS-associated strains (Godschalk et al., 2007). Biosynthesis of Classes A and B Core Structures with a Nonsialylated Inner Gal All of the currently known class B LOS loci have a full-length two-domain ORF 3 (Cjll35). As mentioned above, some class A strains also have a full-
length two-domain ORF 3. A full-length ORF 3 will direct the addition of two Glc residues to the inner core (Figs. 1D and 10). The addition of the second Glc residue (to HepII) seems to prevent the sialylation of the inner Gal residue. These strains all have CgtA and CgtB variants that are specific for nonsialylated acceptors, which are necessary to extend the nonsialylated inner Gal. A class B strain (GBS) with an inactive CgtB (due to a frameshift mutation) expresses a GA2 mimic (Godschalk et al., 2007).A class A strain (GB27) with an inactive Cst-I1 (due to a frameshift mutation) expresses a GA1 mimic (Godschalk et al., 2007). The presence of a mono- or a bifunctional Cst-I1 will direct the synthesis of either a G M l b or a GDlc mimic, respectively (Fig. 10). It is interesting to note that identical structures can be synthesized by either class A or class B strains, depending on the variants of ORF 3, CgtA, CgtB, and Cst-I1 that are present. incompatible^' combination of glycosyltransferase variants could theoretically exist in nature. For instance, the combination of a fulllength two-domain ORF 3 and variants of CgtA/ CgtB specific for sialylated acceptors would prevent the extension of the LOS outer core beyond the inner Gal residue. Such a combination of alleles has not been observed so far, and they might have been selected against in natural reservoirs.
498
GILBERT ET AL.
Gal+( 1,3)-HepII-a( 1,3)-HepI-a(1,s)4 I 1
I
Mono- or bi-functional Cst-ll
Gal+( 1,3)-HepII-a( 1,3)-HepI-a(1 ,S)3 4
1
I
1 GlcP
NeuSAca
J
Inactive CgtA
G 1 p( 1,3)-HepII-a( 1,3)-HepI-a(1,5)4
3-
4
GC175
NeuSAca
1
GalNAc-P(1,4)-Gal-P( 1,3)-HepII-a( 1,3)-HepI-a(I ,S)4
I
GB23 (Inactive CgtB)
1 GlcP
Bi-functional Cst-ll
4i
I
2 NeuSAca 8
I
2
I
1 GlcP
NeusFtB
Gal-P(1,3)-GaINAc-P( 1,4)-GaI-p( 1,3)-HepII-a( 1,3)-HepI-a(1,5)3 4
Gal-P(1,3)-HepII-a(l,3)-HepI-a( 1,5)-
3
9 I
2 NeuSAca
GlcP
1
I
1 GlcP
Mono- or bi-functional Cst-ll
OH4382
Gal+( 1,3)-GalNAc-P(1,4)-Gal-P( 1,3)-HepII-a( 1,3)-HepI-a(1,S)3 4
NeuSAca
Serostrain HS:19
+/-[NeuSAca] 2
I
2
NeuSAca
I
1 GlcP
1
Bi-functional Cst-ll
Gal-p( 1,3)-GalNAc-P(1,4)-Gal-P( 1,3)-HepII-a(1,3)-Hepl-a( 1,S)3 4 3
I
I
2 NeuSAca
NeuSAca 2 I
I
1 GlcP
OH4384
2 Neu5Aca Figure 9. Biosynthesis of class A core structures with a sialylated inner Gal.
Biosynthesis of Class B Core Structures Extended from Glc-P(1,2)-HepII As explained above, some strains extend their LOS outer core from Glc-P(1,2)-HepII rather than from Gal-P(1,3)-HepII. All these strains have a class B locus and seem to have a frameshift mutation in ORF 4 (Cj1136) that inactivates the transfer of the Gal-P(1,3) residue to HepII. These strains also have a P-lY4-GalTthat transfers a Gal residue to GlcP( 1,2)-HepII, creating an acceptor for the sialyltransferase (Kanipes et al., 2007). CCUG 10954, a strain with an inactive Cst-I1 (due to a frameshift mutation), expresses a GA3 (lactosyl) mimic (Szymanski et al., 2003; Fig. 11).The presence of an active sialyltransferase will result in the synthesis of either a GM3 or GD3 mimic, depending on whether a mono- or a bifunctional Cst-I1 variant is present. All the characterized strains of this group have a phase-variable CgtAI1 (P-lY4-GalNAcT)that can add a GalNAc residue to either a mono- or disialylated acceptors. The ratio of LOS outer core with a terminal GalNAc is pre-
sumed to depend on the status (on or off) of the homopolymeric G-tract in CgtA-11. A mix of GM3 and GM2 mimics is observed when a monofunctional Cst-I1 is present. A mix of GD3, GD2, and GM2 mimics is observed when a bifunctional Cst-I1 is present. The monosialylated GM2 mimic would be due to incomplete sialylation, possibly as a result of the addition of the GalNAc before the second NeuSAc is added because Cst-I1 cannot act after the GalNAc has been added. O-Acetylation of Terminal a-2,8 NeuSAc
C. jejuni strains with classes A, B, and M carry gene encoding a sialate O-acetyltransferase (SOAT or ORF ll), which targets LOS-bound NeuSAc, with a high specificity for terminal ( ~ 2 ~ 8 linked residues (Houliston et al., 2006). The SOAT gene, which was recombinantly expressed in E. coli, is soluble and homologous with members of the NodL-LacA-CysE family of O-acetyltransferases. The a
CHAPTER 27
-
C . TETUNI LIPOOLIGOSACCHARIDES
499
Gal-P(1,3)-HepII-a( 1,3)-HepI-a( 1,5)2 4
1
1
Glci
GlcP
I
1
CgtA
GalNAc-P( 1,4)-Gal-P( 1,3)-HepII-a( 1,3)-HepI-a( 1,5)2 4
1 GlcP
I
Glee
J
Inactive CgtB
tI -
GalNAc-P( 1,4)-Gal-P(1,3)-HepII-a(1,3)-HeI a( 1,s)GB5
2 I Glcb
1
GlcP
Gal-p( 1,3)-GalNAc-P(1,4)-Gal-P( 1,3)-HepII-a( 1,3)-HepI-a( 1,s)-
4 I
2 I
1 GlcP
1 GlcP
J
Inactive Cst-ll
1
Mono- or bi- functional Cst-ll
1
Bi-functional Cst-ll
Gal+( 1,3)-GalNAc-P(1,4)-Gal-P(I ,3)-HepII-a( I ,3)-HepI-a( 1,5)Gal-P( 1,3)-GalNAc-P(1,4)-Gal-P(I ,3)-HepII-a( 1,3)-HepI-a( 1,5)4 2 2 4 I I I I GB27 I 1 1 GB26 1 Neu5Aca GlcP GlcP GlcP GlcP
4
Gal-P( 1,3)-GalNAc-P(1,4)-Gal-P( 1,3)-HepII-a( 1,3)-HepI-a( 1,5)2 4
i
Neu5Aca 8 I 2 NeuSAca
I
1
Serostrain HS:IO
GlcP
I
1 GlcP
Figure 10. Biosynthesis of classes A and B core structures with a nonsialylated inner Gal.
SOAT gene was found to exist as variants that can be either active or inactive as a result of frameshift mutations or amino acid substitutions (Table 1).Furthermore, the active variants can be either phase variable or non-phase variable. LOS-bound acetylatedNeuSAc requires the presence of a bifunctional Cst-I1 because the SOAT can only transfer to a NeuSAc residue that is a2,S-linked to a subterminal NeuSAc. By means of a mass spectrometry technique that allows the analysis of intact LOS, a survey of C. jejuni strains with various combinations of Cst-I1 and SOAT variants was performed to confirm that the detection of LOS-bound O-acetylated Neu5Ac correlated with the presence of a bifunctional Cst-I1 and of an active SOAT (Dzieciatkowska et al., 2007). Figure 12 summarizes the different NeuSAc O-acetylation patterns that can be observed, depending on the variants of Cst-I1 and SOAT that are present. The possibility to O-acetylate the terminal NeuSAc in the LOS outer core further extends the ability of C. jejuni to vary the cell-surface epitopes presented to the host. Although a role as a virulence factor has not been shown for the SOAT, the large number of alleles suggests a strong pressure to modulate this modification.
Several bacteria O-acetylate capsule bound NeuSAc, which resulted in altered immunogenicity and susceptibility to glycosidases (Deszo et al., 2005; Lewis et al., 2004)
CONCLUSIONS AND FUTURE OUTLOOK Although structural variation between strains can reflect gene content variation in the LOS biosynthesis loci, and recombinant and lateral gene transfer between loci can occur, four other mechanisms that C. jejuni uses to vary its LOS outer core have been demonstrated: (i) phase variation because of homopolymeric tracts, (ii) gene inactivation by the deletion or insertion of a single or multiple bases (without phase variation), (iii) single mutation leading to the inactivation of a glycosyltransferase, and (iv) single or multiple mutations leading to glycosyltransferases with different acceptor specificities. These four mechanisms have resulted in glycosyltransferases with potential to phase-vary their expression or modulate their specificity. The genetic bases for the variation of LOS outer cores in C. jejuni provides a good ex-
500
GILBERT ET AL.
Gal+( 1,4)-Glc-P( 1,2)-HepII- a(1,3)-HepI- a( 1,5)4
4
CCUG 10954 (Inactive Cst-ll)
GIcP
1
Mono- or bi-functional Cst-ll
Gal-P( 1,4)-Glc-P(1,2)-HepII-a( 1,3)-HepI-u( 1,5)-
3
41
4
Neu5Aca
GlcP
J
Phase-variable CgtA
Bi-functional Cst-ll
+I-[GalNAc-P( 1,4)-]Gal-P(1,4)-Glc-P( 1,2)-HepII-a( 1,3)-HepI-a(1,5)-
3
4 NeuSAca
Serostrain HS:36
4 I
Gal$( 1,4)-Glc-P( 1,2)-HepII-a( 1,3)-Hepl-a( 1,5)3 4
I
1
1 GlcP
1 GlcP
NeuSAca 8 I
2
NeuSAca
1
Phase-variable CgtA
+/-[GalNAc-P( 1,4)-]Gal-P( 1,4)-Glc-P( 1,a)-HepII-a( 1,3)-HepI-a( 1,5)4 3
I 2
Neu5Aca 8
I 2
I
MF7
1 GlcP
NeuSAca
Figure 11. Biosynthesis of class B core structures extended from Glc-P(1,2)-HepII.
ample of various adaptive evolutionary strategies used by a mucosal pathogen to modulate the structure of a cell-surface carbohydrate in order to better survive in a host. Thus, the ability to generate a variety of LOS outer core structures represents a mechanism that should help C. jejuni to survive better in hosts by presenting variable cell-surface antigens. Nonetheless, it is difficult to demonstrate experimentally that a large repertoire of LOS outer core structures is essential for the survival or maintenance of this bacterial species. It can be presumed that this diversity confers selective advantages in adapting to different niches and in colonization of various hosts (or individuals). Medini et al. (2005) introduced the concept of the microbial pan-genome to describe the global gene repertoire of a bacterial species. Some species have a so-called open pan-genome that includes a core of genes present in all strains, and a number of dispensable genes whose number keeps increasing as more strains are sequenced. This concept can be applied to the repertoire of genes encoding the LOS outer core structures that are expressed by C. jejuni, leading to its definition as a pan-glycome. Moreover, C. jejuni can be interpreted as possessing an open
pan-glycome whereby new LOS outer core structures are likely to be observed as more strains are studied. This is supported by the observation of new LOS locus classes that are generated by gene duplication and lateral gene exchanges between C. jejuni strains. The diversification of the glycosyltransferases into variants that have modulated activity and specificity further expands the number of structures that can be synthesized with a set of genes. The acquisition of genes from other species will further increase the Campylobacter pan-genome and result in a dynamic and open pan-glycome. The complete pan-glycome of C. jejuni would need to include the other cellsurface glycans, namely, the general protein N-linked glycan, the 0-linked flagellin modification, and CPSs. The variety of CPS structures and biosynthesis loci observed in C. jejuni (Karlyshev et al., 2005b) suggests that it would contribute as much as the LOS repertoire to an open pan-glycome. Acknowledgments. Investigations in our laboratories o n LOS are supported by the Human Frontier Science Program, grant RGP 38/ 2003 (to M.G.), by the United States Department of Agriculture, Agricultural Research Service CRIS project 5325-42000-045 (to C.T.P.), and by the Higher Education Authority PRTL-3 program of the National Development Plan and a European Union Marie
CHAPTER 2 7
C. 7E7UNI LIPOOLIGOSACCHARIDES
501
Gal-P(1,3)-
/ \
Mono-functional Cst-ll
Bi-functional Cst-ll
Gal+( 1,3)3
I
I
2
NeuSAca The SOAT cannot transfer to
NeuAca-(2,3)-Gal
I
Gal-P( 1,3)3
Gal+( 1,3)-
Gal-p( 1,3)3
3
I
NeuSAca 8
2 P N e u S A c a 8
2 NeuSAca
2 NeuSAca
Inactive SOAT
2
I
I
SOAT
I
I
Phase-variable
Gal-P( 1,3)3
I
2 NeuSAca
2
Gal+( 1,3)3
I
2
NeuSAca 8
I
NeuSAca 8
I
2 NeuSAca 9
I
+I-Acetyl
2 NeuSAca 9
I
Acetyl Figure 12. Alternative structures resulting from different combinations of Cst-I1 and SOAT variants.
Curie program grant 2005-029774 (to A.P.M.). We thank Warren Wakarchuk for critical reading of the manuscript.
REFERENCES Aspinall, G. O., C . M. Lynch, H. Pang, R. T. Shaver, and A. P. Moran. 1995a. Chemical structures of the core region of Campylobacter jejuni 0 : 3 lipopolysaccharide and an associated polysaccharide. Eur. 1.Biochem. 231570-578. Aspinall, G. O., A. Mainkar, M. M. Prendergast, B. Jacobs, and A. P. Moran. 1998. Lipopolysaccharides from a neuropathyassociated infection from Campylobacter jejuni serotype 0:23, p. 93-96. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter, and Related Organisms. Institute of Child Health and University of Cape Town, Cape Town, South Africa. Aspinall, G. O., A. G. McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1994. Lipopolysaccharides of Campylobacter jejuni serotype 0:19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-BarrC syndrome. Biochemistry 33:241-249. Aspinall, G. O., A. G. McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1993a. Lipopolysaccharide of Campylobacter coli serotype 0:30. Fractionation and structure of liberated core oligosaccharide. J. Biol. Chem. 268:6263-6268. Aspinall, G. O., A. G. McDonald, T. S. Raju, H. Pang, L. A. Kurjanczyk, J. L. Penner, and A. P. Moran. 1993b. Chemical structure of the core region of Campylobacter jejuni serotype 0: 2 lipopolysaccharide. Eur. J. Biochem. 213:1029-1037. Aspinall, G. O., A. G. McDonald, T. S. Raju, H. Pang, A. P. Moran, and J. L. Penner. 1993c. Chemical structures of the core
regions of Campylobacter jejuni serotypes 0:1,0:4, 0:23, and 0 : 3 6 lipopolysaccharides. Eur. J. Biochem. 216:880. Aspinall, G. O., M. A. Monteiro, and H. Pang. 1995b. Lipooligosaccharide of the Campylobacter lari type strain ATCC 35221. Structure of the liberated oligosaccharide and an associated extracellular polysaccharide. Carbohydr. Res. 279:245264. Aspinall, G. O., M. A. Monteiro, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1995c. Lipo-oligosaccharide of Campylobacter lari strain PC 637. Structure of the liberated oligosaccharide and an associated extracellular polysaccharide. Carbohydr. Res. 279: 227-244. Beer, W., M. Adam, and G. Seltmann. 1986. Monosaccharide composition of lipopolysaccharides from Campylobacter jejuni and Campylobacter coli. 1.Basic Microbiol. 26:201-204. Bernatchez, S., M. Gilbert, M.-C. Blanchard, M.-F. Karwaski, J. Li, S. DeFrees, and W. W. Wakarchuk. 2007. Variants of the p1,3-galactosyltransferase CgtB from the bacterium Campylobacter jejuni have distinct acceptor specificities. Glycobiology 17: 1333-1343. Bernatchez, S., C. M. Szymanski, N. Ishiyama, J. Li, H. C. Jarrell, P. C. Lau, A. M. Berghuis, N. M. Young, and W. W. Wakarchuk. 2005. A single bifunctional UDP-GlcNAc/Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni. J. Biol. Chem. 280:4792-4802. Chiu, C . P., A. G. Watts, L. L. Lairson, M. Gilbert, D. Lim, W. W. Wakarchuk, S. G. Withers, and N. C. Strynadka. 2004. Structural analysis of the sialyltransferase CstII from Campylobacter jejuni in complex with a substrate analog. Nut. Stmct. Mol. Biol. 11:163-170. Corcoran, A. T., and A. P. Moran. 2007. Influence of growth conditions on diverse polysaccharide production by Campylobacter jejuni. FEMS Immunol. Med. Microbiol. 49:124-132.
502
GILBERT ET AL.
De Libero, G., A. P. Moran, H. J. Gober, E. Rossy, A. Shamshiev, 0. Chelnokova, Z. Mazorra, S. Vendetti, A. Sacchi, M. M. Prendergast, S. Sansano, A. Tonevitsky, R. Landmann, and L. Mori. 2005. Bacterial infections promote T cell recognition of selfglycolipids. Immunity 22:763-772. Deszo, E. L., S. M. Steenbergen, D. I. Freedberg, and E. R. Vimr. 2005. Escherichiu coli K1 polysialic acid 0-acetyltransferase gene, neuO, and the mechanism of capsule form variation involving a mobile contingency locus. Proc. Nutl. Acud. Sci. USA 1025564-5569. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. Al-Ghusein, B. G. Barrell, J. Parkhill, N. G. Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Campylobucter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11:1706-1715. Dzieciatkowska, M., D. Brochu, A. V. Belkum, A. P. Heikema, N. Yuki, R. S. Houliston, J. C. Richards, M. Gilbert, and J. Li. 2007. Mass spectrometric analysis of intact lipooligosaccharide: direct evidence for 0-acetylated sialic acids and discovery of 0linked glycine expressed by Campylobacter jejuni. Biochemistry 46~14704-14714. Filiatrault, M. J., B. W. Gibson, B. Schilling, S. Sun, R. S. Munson, Jr., and A. A. Campagnari. 2000. Construction and characterization of Haemophibs ducreyi lipooligosaccharide (LOS) mutants defective in expression of heptosyltransferase I11 and betal,4-glucosyltransferase: identification of LOS glycoforms containing lactosamine repeats. Infect. Immun. 68:3352-3361. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Cumpylobacter species. PLoS Biol. 3:e15. Gilbert, M., J. R. Brisson, M. F. Karwaski, J. Michniewicz, A. M. Cunningham, Y. Wu, N. M. Young, and W. W. Wakarchuk. 2000. Biosynthesis of ganglioside mimics in Cumpylobacter jejuni OH4384. Identification of the glycosyltransferasegenes, enzymatic synthesis of model compounds, and characterization of nanomole amounts by 600-mhz (1)h and (13)c NMR analysis. J. Biol. Chem. 275:3896-3906. Gilbert, M., M. F. Karwaski, S. Bernatchez, N. M. Young, E. Taboada, J. Michniewicz, A. M. Cunningham, and w. w. Wakarchuk. 2002. The genetic bases for the variation in the lipooligosaccharide of the mucosal pathogen, Cumpylobacter jejuni. Biosynthesis of sialylated ganglioside mimics in the core oligosaccharide. J. Biol. Chem. 277:327-337. Gilbert, M., P. C. Godschalk, M. F. Karwaski, C. W. Ang, B. A. van, J. Li, W. W. Wakarchuk, and H. P. Endtz. 2004. Evidence for acquisition of the lipooligosaccharide biosynthesis locus in Cumpylobacter jejuni GB11, a strain isolated from a patient with Guillain-Barrt syndrome, by horizontal exchange. Infect. Immun. 72:1162-1165. Gilbert, M., P. C. R. Godschalk, C. T. Parker, H. P. Endtz, and W. W. Wakarchuk. 2005. Genetic basis for the variation in the lipooligosaccharide outer core of Cumpylobacter jejuni and possible association of glycosyltransferasegenes with post-infectious neuropathies, p. 219-248. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Godschalk, P. C., A. P. Heikema, M. Gilbert, T. Komagamine, C. W. Ang, J. Glerum, D. Brochu, J. Li, N. Yuki, B. C. Jacobs, B. A. van, and H. P. Endtz. 2004. The crucial role of Cumpy-
Zobacter jejuni genes in anti-ganglioside antibody induction in Guillain-Barri syndrome. J. Clin. Investig. 114:1659-1665. Godschalk, P. C., M. L. Kuijf, J. Li, M. F. St, C. W. Ang, B. C. Jacobs, M. F. Karwaski, D. Brochu, A. Moterassed, H. P. Endtz, B. A. van, and M. Gilbert. 2007. Structural characterization of Campylobacter jejuni lipooligosaccharide outer cores associated with Guillain-Barrt and Miller Fisher syndromes. Infect. Immun. 7.5~1245-1254. Guerry, P., C. P. Ewing, T. E. Hickey, M. M. Prendergast, and A. P. Moran. 2000. Sialylation of lipooligosaccharide cores affects immunogenicity and serum resistance of Campylobacter jejuni. Infect. Immun. 68:6656-6662. Guerry, P., C. M. Szyrnanski, M. M. Prendergast, T. E. Hickey, C. P. Ewing, D. L. Pattarini, and A. P. Moran. 2002. Phase variation of Campylobacter jejuni 8 1-176 lipooligosaccharideaffects ganglioside mimicry and invasiveness in vitro. Infect. Immun. 70:787-793. Hanniffy, 0. M., A. S. Shashkov, A. P. Moran, M. M. Prendergast, S. N. Senchenkova, Y. A. Knirel, and A. V. Savage. 1999. Chemical structure of a polysaccharide from Campylobacter jejuni 176.83 (serotype 0:41) containing only furanose sugars. Carbohydr. Res. 319:124-132. Hanniffy, 0. M., A. S. Shashkov, A. P. Moran, S. N. Senchenkova, and A. V. Savage. 2001. Chemical structure of the core oligosaccharide of aerotolerant Campylobacter jejuni 0 : 2 lipopolysaccharide. Carbohydr. Res. 330:223-229. Houliston, R S., H. P. Endtz, N. Yuki, J. Li, H. C. Jarrell, M. Koga, B. A. van, M. F. Karwaski, W. W. Wakarchuk, and M. Gilbert. 2006. Identification of a sialate 0-acetyltransferase from Campylobacter jejuni: demonstration of direct transfer to the C9 position of terminal alpha-2, 8-linked sialic acid.J. Biol. Chem. 281:11480-11486. Kanipes, M. I., E. Papp-Szabo, P. Guerry, and M. A. Monteiro. 2006. Mutation of waaC, encoding heptosyltransferase I in Campylobucter jejuni 81-176, affects the structure of both lipooligosaccharide and capsular carbohydrate. J. Bacteriol. 188: 3273-3279. Kanipes, M. I., D. Rockabrand, X. Tan, M. A. Monteiro, and P. Guerry. 2007. Biological and structural characterization of the Campylobacter jejuni 8 1-176 galactosyltransferasemutant, abstr. D-020. In Abstracts of the American Society for Microbiology 107th Annual General Meeting. ASM Press, Washington, DC. Karlyshev, A. V., 0. L. Champion, C. Churcher, J. R. Brisson, H. C. Jarrell, M. Gilbert, D. Brochu, M. F. St, J. Li, W. W. Wakarchuk, I. Goodhead, M. Sanders, K. Stevens, B. White, J. Parkhill, B. W. Wren, and C. M. Szymanski. 2005a. Analysis of Campylobucter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol. Microbiol. 55:90-103. Karlyshev, A. V., J. M. Ketley, and B. W. Wren. 2005b. The Campylobacter jejuni glycome. FEMS Microbiol. Rev. 29:377-390. Karlyshev, A. V., D. Linton, N. A. Gregson, A. J. Lastovica, and B. W. Wren. 2000. Genetic and biochemical evidence of a Campylobucter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 3.5529-541. Kilcoyne, M., A. P. Moran, A. S. Shashkov, S. N. Senchenkova, J. A. Ferris, A. T. Corcoran, and A. V. Savage. 2006. Molecular origin of two polysaccharides of Campylobacter jejuni 81116. FEMS Microbiol. Lett. 263:214-222. Klena, J. D., S. A. Gray, and M. E. Konkel. 1998. Cloning, sequencing, and characterization of the lipopolysaccharide biosynthetic enzyme heptosyltransferase I gene (wauC) from Campylobacter jejuni and Campylobacter coli. Gene 222: 177-185. Koga, M., M. Gilbert, M. Takahashi, J. Li, S. Koike, K. Hirata, and N. Yuki. 2006. Comprehensive analysis of bacterial risk fac-
CHAPTER 27
tors for the development of Guillain-Barrt syndrome after Campylobacter jejuni enteritis. J. Infect. Dis. 193547-555. Komagamine, T., and N. Yuki. 2006. Ganglioside mimicry as a cause of Guillain-Barre syndrome. CNS Neurol. Disord. Drug Targets 5 :3 9 1-400. Leonard, E. E., T. Takata, M. J. Blaser, S. Falkow, L. S. Tompkins, and E. C. Gaynor. 2003. Use of an open-reading framespecific Campylobacter jejuni DNA microarray as a new genotyping tool for studying epidemiologically related isolates. J. Infect. Dis. 187:691-694. Leonard, E. E., L. S. Tompkins, S. Falkow, and I. Nachamkin. 2004. Comparison of Campylobacter jejuni isolates implicated in Guillain-Barrt syndrome and strains that cause enteritis by a DNA microarray. Infect. Immun. 72:1199-1203. Levinson, G., and G. A. Gutman. 1987. Slipped-strand mispairing: a major mechanism for DNA sequence evolution. Mol. Biol. Evol. 4:203-221. Lewis, A. L., V. Nizet, and A. Varki. 2004. Discovery and characterization of sialic acid 0-acetylation in group B Streptococcus. Proc. Natl. Acad. Sci. USA 101:11123-11128. Linton, D., M. Gilbert, P. G. Hitchen, A. Dell, H. R. Morris, W. W. Wakarchuk, N. A. Gregson, and B. W. Wren. 2000. Phase variation of a beta-1,3 galactosyltransferase involved in generation of the ganglioside GM1-like lipo-oligosaccharide of Campylobacter jejuni. Mol. Microbiol. 37501-514. Medini, D., C. Donati, H. Tettelin, V. Masignani, and R. Rappuoli. 2005. The microbial pan-genome. Curr. Opin. Genet. Dev. 15589-594. Moran, A. P. 1995. Biological and serological characterization of Campylobacter jejuni lipopolysaccharides with deviating core and lipid A structures. FEMS Immunol. Med. Microbiol. 11:121130. Moran, A. P. 1997. Structure and conserved characteristics of Campylobacter jejuni lipopolysaccharides. J. Infect. Dis. 176(S~ppl.2):S115-S121. Moran, A. P., H. Annuk, and M. M. Prendergast. 2005. Antibodies induced by ganglioside-mimicking Campylobacter jejuni lipooligosaccharides recognise epitopes at the nodes of Ranvier. J. Neuroimmunol. 165:179-185. Moran, A. P., I. M. Helander, B. Lindner, and D. T. O'Malley. 1996a. Structural analysis of lipid A from Campylobacter fetus smooth-form LPS. J. Endotoxin Res. 3(Suppl. 1):34. Moran, A. P., and D. T. O'Malley. 1995. Potential role of lipopolysaccharides of Campylobacter jejuni in the development of Guillain-Barrt syndrome. J. Endotoxin Res. 2:233-235. Moran, A. P., D. T. O'Malley, T. U. Kosunen, and I. M. Helander. 1994. Biochemical characterization of Campylobacter fetus lipopolysaccharides. Infect. Immun. 62:3922-3929. Moran, A. P., D. T. O'Malley, J. Vuopio-Varkila, K. Varkila, L. Pyhala, H. Saxen, and I. M. Helander. 199613. Biological characterization of Campylobacter fetus lipopolysaccharides. FEMS Immunol. Med. Microbiol. 15:43-50. Moran, A. P., and J. L. Penner. 1999. Serotyping of Campylobacter jejuni based on heat-stable antigens: relevance, molecular basis and implications in pathogenesis. J. Appl. Microbiol. 86:361377. Moran, A. P., J. L. Penner, and G. 0. Aspinall. 2000. Campylobacter lipopolysaccharides, p. 241-257. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Moran, A. P., and M. M. Prendergast. 2001. Molecular mimicry in Campylobacter jejuni and Helicobacter pylori lipopolysaccharides: contribution of gastrointestinal infections to autoimmunity. J. Autoimmun. 16:241-256. Moran, A. P., M. M. Prendergast, and B. J. Appelmelk. 1996c. Molecular mimicry of host structures by bacterial lipopolysac-
c. r E r m n LIPOOLIGOSACCHARIDES
503
charides and its contribution to disease. FEMS Immunol. Med. Microbiol. 16:105-1 15. Moran, A. P., E. T. Rietschel, T. U. Kosunen, and U. Zahringer. 1991a. Chemical characterization of Campylobacter jejuni lipopolysaccharides containing N-acetylneuraminic acid and 2,3diamino-2,3-dideoxy-~-glucose. J. Bacteriol. 173:618-626. Moran, A. P., U. Zahringer, U. Seydel, D. Scholz, P. Stutz, and E. T. Rietschel. 1991b. Structural analysis of the lipid A component of Campylobacter jejuni CCUG 10936 (serotype 0:2) lipopolysaccharide. Description of a lipid A containing a hybrid backbone of 2-amino-2-deoxy-D-glucose and 2,3-diamino-2,3dideoxy-D-glucose. Eur. J. Biochem. 198:459-469. Muldoon, J., A. S. Shashkov, A. P. Moran, J. A. Ferris, S. N. Senchenkova, and A. V. Savage. 2002. Structures of two polysaccharides of Campylobacter jejuni 8 1116. Carbohydr. Res. 337:2223-2229. Nam Shin, J. E., S. Ackloo, A. S. Mainkar, M. A. Monteiro, H. Pang, J. L. Penner, and G. 0. Aspinall. 1998. Lipooligosaccharides of Campylobacter jejuni serotype 0:10. Structures of core oligosaccharide regions from a bacterial isolate from a patient with the Miller-Fisher syndrome and from the serotype reference strain. Carbohydr. Res. 305:223-232. Oldfield, N. J., A. P. Moran, L. A. Millar, M. M. Prendergast, and J. M. Ketley. 2002. Characterization of the Campylobacterjejuni heptosyltransferase I1 gene, waaF, provides genetic evidence that extracellular polysaccharide is lipid A core independent. J. Bacteriol. 184:2100-2107. Parker, C. T., S. T. Horn, M. Gilbert, W. G. Miller, D. L. Woodward, and R. E. Mandrell. 2005. Comparison of Campylobacter jejuni lipooligosaccharide biosynthesis loci from a variety of sources. J. Clin. Microbiol. 43:2771-2781. Parker, C. T., B. Quinones, W. G. Miller, S. T. Horn, and R. E. Mandrell. 2006. Comparative genomic analysis of Campylobacter jejuni strains reveals diversity due to genomic elements similar to those present in C. jejuni strain RM1221. J. Clin. Microbiol. 44:4125-4135. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pearson, B. M., C. Pin, J. Wright, K. I'Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS Lett. 554:224-230. Penner, J. L., and G. 0. Aspinall. 1997. Diversity of lipopolysaccharide structures in Campylobacter jejuni. J. Infect. Dis. 176(S~ppl.2):S 135-S 138. Phongsisay, V., V. N. Perera, and B. N. Fry. 2006. Exchange of lipooligosaccharide synthesis genes creates potential GuillainBarre syndrome-inducible strains of Campylobacter jejuni. Infect. Immun. 74:1368-1372. Phongsisay, V., V. N. Perera, and B. N. Fry. 2007. Expression of the htrB gene is essential for responsiveness of Salmonella typhimurium and Campylobacter jejuni to harsh environments. Microbiology 153:254-262. Prendergast, M. M., T. U. Kosunen, and A. P. Moran. 2001. Development of an immunoassay for rapid detection of ganglioside GM(1) mimicry in Campylobacter jejuni strains. J. Clin. Microbiol. 39:1494-1500. Prendergast, M. M., A. J. Lastovica, and A. P. Moran. 1998. Lipopolysaccharides from Campylobacter jejuni 0 : 4 1 strains associated with Guillain-Barrt syndrome exhibit mimicry of GM1 ganglioside. Infect. Immun. 66:3649-3655.
504
GILBERT ET AL.
Prendergast, M. M., and A. P. Moran. 2000. Lipopolysaccharides in the development of the Guillain-Barre syndrome and Miller Fisher syndrome forms of acute inflammatory peripheral neuropathies. J. Endotox. Res. 6:341-359. Prendergast, M. M., D. R. Tribble, S. Baqar, D. A. Scott, J. A. Ferris, R. I. Walker, and A. P. Moran. 2004. In vivo phase variation and serologic response to lipooligosaccharide of Campylobacter jejuni in experimental human infection. Infect. Immun. 72~916-922. Prendergast, M. M., H. J. Willison, and A. P. Moran. 1999. Human monoclonal immunoglobulin M antibodies to ganglioside GM1 show diverse cross-reactivities with lipopolysaccharides of Campylobacter jejuni strains associated with Guillain-Barre syndrome. Infect. Immun. 67:3698-3701. Preston, M. A., and J. L. Penner. 1987. Structural and antigenic properties of lipopolysaccharides from serotype reference strains of Campylobacter jejuni. Infect. Immun. 55:1806-18 12. Raetz, C. R., C. M. Reynolds, M. S. Trent, and R. E. Bishop. 2007. Lipid A modification systems in gram-negative bacteria. Ann. Rev. Biochem. 76:295-329. Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Ann. Rev. Biochem. 71:635-700. Rietschel, E. T., L. Brade, 0. Holst, V. A. Kulschin, B. Lindner, A. P. Moran, U. F. Schade, U. Zahringer, and H. Brade. 1990. Molecular structure of bacterial endotoxin in relation to bioactivity, p. 15-32. In A. Nowotny, J. J. Spitzer, and E. J. Ziegler (ed.), Cellular and Molecular Aspects of Endotoxin Reactions, vol. 1. Elsevier Science, Amsterdam. Rietschel, E. T., T. Kirikae, F. U. Schade, U. Mamat, G. Schmidt, H. Loppnow, A. J. Ulmer, U. Zahringer, U. Seydel, F. Di Padova, M. Schreier, and H. Brade. 1994. Bacterial endotoxin: molecular relationships of structure to activity and function. FASEB J. 8:217-225. Schromm, A. B., K. Brandenburg, H. Loppnow, A. P. Moran, M. H. Koch, E. T. Rietschel, and U. Seydel. 2000. Biological activities of lipopolysaccharides are determined by the shape of their lipid A portion. Eur. J. Biochem. 267:2008-2013. Senchenkova, S. N., A. S. Shashkov, Y. A. Knirel, J. J. McGovern, and A. P. Moran. 1997. The 0-specific polysaccharide chain of Campylobacter fetus serotype A lipopolysaccharide is a partially 0-acetylated 1,3-linked alpha-D-mannan. Eur. J. Biochem. 245: 637-64 1. Senchenkova, S. N., A. S. Shashkov, Y. A. Knirel, J. J. McGovern, and A. P. Moran. 1996. The 0-specific polysaccharide chain of
Campylobacter fetus serotype B lipopolysaccharide is a Drhamnan terminated with 3-0-methyl-~-rhamnose(D-aCOfriOSe). Eur. J. Biochem. 239:434-438. St Michael, F., C. M. Szymanski, J. Li, K. H. Chan, N. H. Khieu, S. Larocque, W. W. Wakarchuk, J. R. Brisson, and M. A. Monteiro. 2002. The structures of the lipooligosaccharide and capsule polysaccharide of Carnpylobacter jejuni genome sequenced strain NCTC 11168. Eur. J. Biochem. 2695119-5136. Sundaram, A. K., L. Pitts, K. Muhammad, J. Wu, M. Betenbaugh, R. W. Woodard, and W. F. Vann. 2004. Characterization of Nacetylneuraminic acid synthase isoenzyme 1 from Campylobacter jejuni. Biochem. J. 383:83-89. Sweet, C. R., A. A. Ribeiro, and C . R. Raetz. 2004. Oxidation and transamination of the 3"-position of UDP-~-acetylglucosamine by enzymes from Acidithiobacillus ferrooxidans. Role in the formation of lipid a molecules with four amide-linked acyl chains. J. Biol. Chem. 279:25400-25410. Szymanski, C. M., F. S. Michael, H. C. Jarrell, J. Li, M. Gilbert, S. Larocque, E. Vinogradov, and J. R. Brisson. 2003. Detection of conserved N-linked glycans and phase-variable lipooligosaccharides and capsules from Campylobacter cells by mass spectrometry and high resolution magic angle spinning NMR spectroscopy. J. Biol. Chem. 278:24509-24520. Taboada, E. N., R. R. Acedillo, C. D. Carrillo, W. A. Findlay, D. T. Medeiros, 0. L. Mykytczuk, M. J. Roberts, C. A. Valencia, J. M. Farber, and J. H. Nasb. 2004. Large-scale comparative genomics meta-analysis of Campylobacter jejuni isolates reveals low level of genome plasticity. J. Clin. Microbiol. 42:45664576. Trent, M. S., and C. R. H. Raetz. 2002. Cloning of EptA, the lipidA phosphoethanolamine transferase associated with polyxin resistance. J. Endotoxin Res. 8:158. Wright, J. C., D. W. Hood, G. A. Randle, K. Makepeace, A. D. Cox, J. Li, R. Chalmers, J. C . Richards, and E. R. Moxon. 2004. lpt6, a gene required for addition of phosphoethanolamine to inner-core lipopolysaccharide of Nezsseria meningitidis and Haemophilus influenzae. J. Bacteriol. 186: 6970-6982. Yang, L. Y., Z. H. Pei, S . Fujimoto, and M. J. Blaser. 1992. Reattachment of surface array proteins to Campylobacter fetus cells. J. Bacteriol. 174: 1258-1267. Yuki, N., T. Taki, F. Inagaki, T. Kasama, M. Takahashi, K. Saito, S . Handa, and T. Miyatake. 1993. A bacterium lipopolysaccharide that elicits Guillain-BarrC syndrome has a GM1 ganglioside-like structure. J. Exp. Med. 178:1771-1775.
Campylobacter, 3rd ed. Edited by 1. Nacharnkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 28
Campylobacter jejuni Capsular Polysaccharide ANDREYV. KARLYSHEV, BRENDAN W. WREN,AND ANTHONYP. M o m
coli. Investigations of Campylobacter CPSs have been assisted by a wealth of information on CPSs from capsules in other bacteria, especially in Escherichia coli, and therefore, a comparative overview of CPSs found in other bacteria is included. Despite significant progress in this area of research since 2000, when both genetic and biochemical evidence of the existence of C. jejuni CPSs was reported (Karlyshev et al., 2OOO), there still remain many unanswered questions that will be discussed in the last section of the chapter, along with future avenues for research on this important cell surface structure.
Many bacteria contain cell surface structures called capsules, which are thought to play a protective role against adverse factors that bacteria may encounter in the environment or, in case of commensal and pathogenic strains, when counteracting the host. However, the biological role of capsules is diverse and varies among different bacteria. Most bacterial capsules are of a polysaccharide (PS) nature, although capsules made of polypeptides, e.g., in Yersiniu pestis (Karlyshev et al., 1992), Bacillus anthrucis (Ezzell and Welkos, 1999), Francisella tulurensis (Su et al., 2007), and the S-layer of Campylobacter fetus (Blaser et al., 1988), are also known. There are various forms of capsular polysaccharides (CPS) (Taylor and Roberts, 2005). In general, these molecules consist of a PS chain, made of a variable number of repeating units containing one or more sugar residues and a lipid anchor, that is involved in the attachment of these molecules to the bacterial cell surface (Roberts, 1996). The nature of the sugar residues, their variability, and their number in the repeating units, as well as the many ways these residues can be linked to each other and the presence of various moieties and modifications result in a huge variety of CPS structures among different strains and different bacterial species (Whitfield and Roberts, 1999). In addition, a single bacterial strain can produce CPSs of diverse structures (Roberts, 1996; Whitfield and Roberts, 1999). The diversity of CPS structures is considered the driving force for evading host immune responses, and this variation is attributed to multiple genetic mechanisms, including rearrangements and phase variation (Roberts, 1996). In this chapter, the biochemistry, genetics, and biological roles of Campylobacter CPSs are reviewed. This chapter is limited to consideration of PS capsules that have been detected in Campylobacter jejuni and other closely related species such as Campylobacter
BIOLOGICAL ROLES OF BACTERIAL CPSs
CPSs may be involved in various functions, e.g., protecting bacteria from adverse environmental conditions (e.g., by increasing resistance to desiccation) and playing a role in pathogenesis (Roberts, 1996; Taylor and Roberts, 2005). These molecules are different from the other cell surface PS-related molecules in that they are usually loosely associated with the cell surface and may be released under mild conditions (Whitfield and Roberts, 1999). In contrast to lipopolysaccharides (LPSs) and lipooligosaccharides (LOSS),which are anchored to the outer membrane by lipid A (Moran, 1995a; Moran et al., 1991), group II/III CPSs are attached to the bacterial cell surface via a phospholipid moiety, as exemplified by those of E. coli (Schmidt and Jann, 1982), and are susceptible to the action of phospholipases. CPSs and exopolysaccharides (EPSs) are required for biofilm formation, which contributes to virulence of bacteria infecting the gastrointestinal tract (Moran and Annuk, 2003) and leads to increased resistance to antibacterial drugs (Moran and Ljungh, 2003; Otto, 2006). In some cases, the presence of a capsule
Andrey V. Karlyshev and Brendan W. Wren London School of Hygiene & Tropical Medicine, Keppel St., London, WClE 7HT, Anthony P. Moran Department of Microbiology, National University of Ireland, Galway, Ireland, and Institute United Kingdom. for Glycomics, Gold Coast Campus, Griffith University, Queensland 4222, Australia.
505
506
KARLYSHEV ET AL.
may prevent bacterial aggregation and biofilm formation (Davey and Duncan, 2006), presumably by masking bacterial protein adhesins involved in surface recognition. Similarly, in certain instances, the capsule may mask other bacterial adhesins required for interaction with host cells (Moran and Ljungh, 2003). The CPS of Streptococcus pneumoniue is required for colonization and invasive disease (Magee and Yother, 2001). On the other hand, a significant reduction of CPS production (e.g., by as much as 80%) does not have a dramatic effect on the ability of these bacteria to colonize. It may be beneficial for bacteria at some stages of infection to produce less CPS to allow a greater exposure of cell surface adhesins. Complete elimination of the s. pneumoniae capsule, however, has a dramatic effect on bacterial ability to colonize (Magee and Yother, 2001). Because of their highly hydrophilic nature, the PS chains of bacterial CPSs and EPSs have a propensity to retain water molecules (Moran and Ljungh, 2003), and by enhancing surface fluidity, they assist bacterial motility. This is dramatically illustrated in the uropathogen Proteus mirabilis, where the CPS is required for motility and the main constituent is designated Cmf after colony migration factor (Rahman et al., 1999). Also, bacterial capsules and CPSs can influence the immune response. These molecules can interfere with complement activity and cause evasion of the host immune response by mimicking host cell antigens (reviewed in Sussman, 1997). For example, the E. coli K1 CPS, composed of a(2,8)-linked polysialic acid, is identical to the terminal carbohydrate region of the neonatal neural cell adhesion molecule nCAM, and the repeat unit of E. coli K5 CPS is identical to that of the first polymeric intermediate of heparin (Jann and Jann, 1987; Roberts, 1996). Because tolerance exists in the immune system against these molecules, antibodies are not or are only inefficiently produced against these bacterial CPS, although tolerance is greatest against the K5 CPS because the molecular mimic is found at all stages of the host’s life, whereas that of the K1 capsule is found only in the neonate (Jann and Jann, 1987). Furthermore, as a result of their surface location and presence during initial stages of infection, CPSs can protect bacteria from the innate immune response. The Neisseria meningitidis capsule is required for increased resistance to cationic antimicrobial peptides, which was assessed by increased sensitivity of noncapsulated mutants to defensins, cathelicidins, and protegrins (Spinosa et al., 2007). The role of some bacterial CPSs in abscess development has also been reported (Tzianabos and Kasper, 2002). Pneumococcal CPSs, especially those carrying negatively charged
residues, were found to be involved in bacterial interaction within the mucous layer (Nelson et al., 2007). CPS produced by E. coli is also required for survival within Acanthumoeba (Jung et al., 2007). This is an important finding that extends the range of functions of bacterial capsules and has prompted relevant studies on the survival of C. jejuni in protozoa (Axelsson-Olsson et al., 2005; Snelling et al., 2005b). To summarize, CPSs can play important roles in multiple stages of the life cycle of bacteria.
TYPES, GENETICS, AND BIOSYNTHESIS OF CPSs According to current classification, CPSs can be divided into four groups. Differentiation between groups is not based on antigenic or structural differences of the PS chains, but on such features as the mechanisms of biosynthesis, assembly, genetic regulation, and sequence similarity (Whitfield, 2006; Whitfield and Roberts, 1999). The CPSs found in C. jejuni resemble both group I1 and group I11 enterobacterial capsules, which are characterized by typical organization of the respective gene clusters containing a major internal biosynthetic region (region 2) flanked by two groups of genes involved in CPS transport and assembly (kps genes). The CPSs belonging to these groups contain a phospholipid moiety attached to a repeating unit either with or without a relatively labile linkage involving the sugar 3-deoxy-~-manno-oct-2-ulosonic acid (Kdo) linkage (Roberts, 1996; Whitfield, 2006). In contrast to group 111, group I1 capsules are characterized by thermoregulated expression and the increased level of CMP-Kdo synthetase, which is encoded by kpsU gene located in the capsule gene cluster (Whitfield, 2006; Whitfield and Roberts, 1999). One remarkable feature distinguishing the two groups of CPSs is the organization and relative order of the genes in the flanking (kps) regions of the respective gene clusters. Typically, the gene order is kpsFEDUCS...El4 for group I1 and kpsDMTE ...CS for group I11 clusters (Whitfield, 2006) (Fig. 1). Organization of the kps genes in the reported c. jejuni gene clusters, e.g., those involved in the biosynthesis of heat-stable (HS) antigens HS:2 (Parkhill et al., 2000), HS:4 (Poly et al., 2007), HS:23/36 (Karlyshev et al., 2005), and HS:53 (Parker et al., 2006), is kpsMTEDF ...CS, somewhat a cross between group I1 and I11 gene clusters (Fig. 1) but actually different from both, thus making it difficult to definitively assign it to either group. The genetic organization is more similar to group I11 capsules because of a lack of ther-
c. r E r m u CAPSULAR POLYSACCHARIDE so7
CHAPTER 2s
c.ieiuni >> I
>I
-b
- -b
kpsM kpsT kpsE kpsD kpsF
-b
-b
kpsC
kpsS
Figure 1. Organization of gene clusters involved in the biosynthesis of enterobacterial groups I1 and I11 and Cumpylobucter jejuni CPSs. Open block arrows, kpsMIT genes encoding ABC transporter; solid block arrows, other genes involved in CPS transport and assembly; thin arrow, direction of transcription of the genes in the internal biosynthetic region (region 2 genes according to Whitfield, 2006).
moregulation (Stintzi and Whitworth, 2003) and the absence of the kpsU gene encoding CPM-Kdo synthetase, but similar to group I1 in that it contains a kpsF gene and the biosynthetic genes in region 2 are collinear with kpsM and kpsT. Although the assortment of CPSs structures depends on the internal region 2 of a gene cluster (Fig. l), which is highly variable between the strains belonging to different serotypes, the flanking genes involved in CPS transport and assembly are highly conserved, indicating a common mechanism of capsule biosynthesis and formation. The mechanism of C. jejuni CPS transport and assembly can be inferred from the model deduced for E. coli CPSs (Whitfield, 2006). According to this model, after assembly in the cytoplasm the polymer is translocated through cell membranes with the help of KpsM (transmembrane protein) and KpsT (nucleotide binding protein), which form an ABC transporter system. Mutation of kpsT results in intracellular accumulation of the PS, which is also observed in some other kps mutants. The roles of the other kps genes in transport are less clear and sometimes controversial. For example, orthologues of both KpsF and KpsU are involved in the biosynthesis of activated Kdo (i.e., CMP-Kdo). However, because not all group I1 capsules contain Kdo as a linkage between the PS component and lipid moiety, it was suggested that KpsF and KpsU may play an additional role independent of CPM-Kdo biosynthesis, which is supported by the presence of other genes performing similar functions (e.g., genes similar to kpsU) and located elsewhere on the chromosome. Although KpsC and KpsS proteins are predicted to be cytoplasmic and are known to be involved in CPS translocation as a result of intracellular accumulation of the PS in the respective mutants, their exact function remains unknown. The initial suggestion of a potential role of these gene products in lipid attachment during biosynthesis of the E. coli K5 CPS could not be confirmed by results of similar
studies on the K1 capsule (Whitfield, 2006). The KpsD and kpsE gene products are involved in the translocation of the polymer through the outer membrane as a result of the accumulation of the PS in the periplasm in kpsD and kpsDE mutants.
CAMPYLOBACTER CPS BIOCHEMISTRY AND GENETICS Background of Discovery of CPS Production in Campylobacters
Campylobacter serotyping based on HS antigens by means of passive hemagglutination was first described by Penner and Hennessy in 1980 (Penner and Hennessy, 1980) and has been adopted worldwide as the Penner serotyping scheme, the gold standard for differentiating C. jejuni (reviewed in Moran and Penner, 1999). The system has subsequently undergone further development, with as many as 66 different antisera being used for Campylobacter typing (both C. jejuni and C. coli) by 2001 (McKay et al., 2001). The fact that typing antisera were detecting specific variations in CPS structure was unknown at that time. Although it was found to be of PS or glycolipid origin, the exact nature of the Penner serotyping antigen remained an unclear and confusing issue for several years (Moran and Penner, 1999). Generally, C. jejuni was considered to produce two different types of glycolipids: low-molecular-weight (M,) LOS and high-M, molecules deduced to be LPS. Despite some immunoblotting evidence that the high-M, molecules of C. jejuni were in part involved in serospecificity, it was LOS that was considered to be the main contributory molecules, although in certain serotypes, both the high-M, molecules and LOS were suggested to contribute to serological reactions (Mills et al., 1985; Preston and Penner, 1987). Additional contribution of the high-M, molecules was supported
508
KARLYSHEV ET AL.
by the fact that some strains with identical LOS structures sometimes belonged to different serotypes (e.g., serostrains HS:4 and HS:19) (Moran and Penner, 1999). In contrast to C. jejuni, it was deduced from immunoblotting studies that the HS antigenic specificities of C. coli are contained within the high-M, molecules, not LOS (Mandatori and Penner, 1989). By 1997, these high-M, molecules were detected in about half the strains analyzed; 16 of 38 C. jejuni strains that were tested revealed such molecules by Western blot testing (Penner and Aspinall, 1997). Remarkably, no high-M, molecules were found in strains belonging to serotypes HS:1 and HS:2, for which a complete PS structure is now known (Karlyshev et al., 2005; McNally et al., 2005). Failure to detect such molecules in these early studies has been attributed to their particular chemical properties, presence in a low amount, and/or technical problems (e.g., the use of nitrocellulose instead of polyvinylidene diflouride membranes in blotting experiments). Thus, the nature of the molecules contributing to serotyping remained unclear until the discovery of C. jejuni CPS, and with some exceptions, the high-M, molecules were considered to be a derivative of LOS, i.e., LPS in nature, rather than being a completely separate macromolecular structure (reviewed in Moran et al., 2000). One of the few exceptions was a high-M, PS from serostrain HS:3 found to be independent of LOS (Aspinall et al., 1995). A second exception was a neutral high-M, PS from a HS:41 clinical isolate (Hanniffy et al., 1999). However, the latter studies, although unable to establish the origin of these PSs, provided invaluable information on the chemical structure of the high-M, molecules that later were found to be CPSs in origin. The first indirect evidence of CPS produced by C. jejuni was reported in 1996 (Chart et al., 1996b). It was suggested that a PS that was loosely associated with the bacterial cell surface and easily extractable by a simple incubation of bacterial suspension in saline at 50 to 70°C was of capsular origin rather than LPS. Unfortunately, there were no follow-up studies for several years until those resulting from publication of the first genome sequence of C. jejuni (strain NCTC 11168, HS:2) in 2000 (Parkhill et al., 2000). Attempts to detect a Campylobacter capsule by electron microscopy were unsuccessful at that time. The reason for this did not become apparent until subsequent studies demonstrated that ruthenium red, a dye traditionally used for detection of capsules in bacteria, was unsuitable for CPS stabilization and hence visualization by electron microscopy (Karlyshev et al., 2001). An interpretation that HS serospecificity linked to LOS alone, rather than serospecificity derived
from CPS, led to some confusing conclusions regarding an association of Guillain-BarrC syndrome (GBS) with distinct serotypes. It is now generally accepted that GBS is induced by an autoimmune response to ganglioside-like structures present in core regions of some LOSs produced by C. jejuni (Komagamine and Yuki, 2006; Prendergast and Moran, 2000). The earlier findings of an association of GBS with serotype HS:19 in Japan (Kuroki et al., 1993) and serotype HS:41 in South Africa (Lastovica et al., 1997) could be interpreted as misleading because strains of identical serotypes were able to produce LOSs of different structures (Aspinall et al., 1994b; Nachamkin et al., 2001). Instead, the finding of a link between a serotype and GBS is likely to be due to the clonal nature of the isolates studied. Indeed, association of a serotype with GBS reported in one country could not be confirmed in studies with another strain of the same serotype isolated in another country (Wassenaar et al., 2000), although it should be noted that host factors in addition to ganglioside mimicry also can play a role in the development of this autoimmune syndrome (Prendergast and Moran, 2000). Evidence of CPS from the First Genome Sequence of C. jejuni The genetic evidence of a potential to produce CPSs came to light from the analysis of genome sequence data from C. jejuni NCTC l l 168 (Parkhill et al., 2000) when putative gene products with sequence similarity to those involved in CPS formation in other species were identified. This prompted construction and phenotypic analysis of the respective mutants (Karlyshev et al., 2000). The first one to be studied was the kpsM mutant of strain NCTC 11168. The KpsM ortholog of E. coli and other bacteria was known to be involved in CPS transport and assembly, forming a complex with ATP-binding KpsT protein (Roberts, 1996). Various phenotypic assays, including Penner serotyping and Western blot testing, were used. Surprisingly, the kpsM mutant of this strain became untypeable. The loss of serotype was also confirmed with kpsM, kpsC, and kpsS mutants of a number of other C. jejuni strains, suggesting that the molecule involved in serotyping was of CPS origin. The results were further confirmed by the inability to detect high-M, PS (CPS) in kpsM mutants in Western blotting experiments with Penner antisera. The observation that there were no detectable changes in the low-M, LOS molecules of mutants suggested that the CPS and LOS are unrelated molecules. This conclusion was supported by demonstrating that mutations in the LOS locus, particularly mutations affecting the inner core region that would influence attachment of
CHAPTER 2s
an O-chain to the LOS core, had no effect on the high-M, PS production, thereby emphasizing its CPS nature (Oldfield et al., 2002). Although the discovery of the CPS capsule was acknowledged in a review article published as early as 2002 (Nachamkin, 2002), a number of publications still mistakenly refer to the high-M, PS as LPS by using “0” instead of a more appropriate “HS” designation of the serotype (Chatzipanagiotou et al., 2003; Yun et al., 2004). The latter designation had been introduced in 1999 because of the apparent molecular complexity of the serotyping system not reflecting classical LPS serospecificity (Moran and Penner, 1999). As a result of the discovery of the serospecific contribution of CPS, this is now the recommended nomenclature. Comparative Genomics Comparative genomics analysis of different Campylobacter strains revealed that structural diversity of CPS structures is associated with variation in region 2 (Fig. 1) within the cps gene clusters (Karlyshev et al., 2005). Although the exact mechanism leading to such variation remains unknown, it was suggested that it was a result of both horizontal gene transfer and recombination events (deletions, insertions, and duplications) within the gene clusters. Originally, by similarity to cps gene clusters found in other bacteria, it was thought that the genes present in the capsule gene cluster are sufficient for capsule formation (Roberts, 1996). Subsequent studies indicated that genes outside this locus may also be involved (Karlyshev et al., 2005). In addition, a single
HS:2
c. i E r m r CAPSULAR POLYSACCHARIDE so9
bifunctional gene gne was found to be required for both CPS and LOS biosynthesis, and even glycoprotein synthesis (Bernatchez et al., 2005). In strain NCTC 11168, the gne gene product supplies Nacetyl-D-galactosamine (GalNAc) for synthesis of the CPS by converting UDP-N-acetyl-D-glucosamine (-GlcNAc) to UDP-GalNAc as a result of its 4epimerase activity. The gne gene product was previously considered to possess solely UDP-glucose-4epimerase activity and was therefore annotated as galE (Fry et al., 2000) by analogy to the UDPglucose-4-epimerase of enterobacterial LPS. The internal biosynthetic regions in the cps gene clusters of C. jejuni G1 (HS:2) and RM1221 (HS:53) reveal no similarity, while in other cases, the similarity was usually limited to the genes involved in the biosynthesis of particular sugar residues or modifications (Karlyshev et al., 2005). A typical example is a cassette of genes involved in heptose (Hep) biosynthesis and transfer (Fig. 2). The CPSs of HS:2, HS: 23/36, HS:41, and HS:53 serotypes all contain various forms of Hep residues concomitant with the presence of signature genes hddC, gmhA2, and hddA related to Hep biosynthesis (Karlyshev et al., 2005). In addition, gene clusters of HS:23/36, HS:41, and HS:53, but not HS:2 serotypes, contain a linked gene dmhA encoding a putative Hep dehydratase, which is correlated with the presence of a 6-deoxy form of Hep in these strains. Despite the high level of variation in gene order and sequences detected in other parts of these gene clusters, the hddC, gmhA2, hddA, and dmhA genes, if present, are almost identical in nucleotide sequence and are always present as adjacent genes, suggesting that they may have recently been acquired from a common organism by a horizontal gene transfer mechanism.
--hddC
gmhA2
Deoxy form of Heptose
-
hddA
<=
HS:41
1111
HS:53
111
hddC
hddC
gmhA2
gmhA2
HS41.09 hddA
hddA
<=
dmhA
dmhA
+ +
Figure 2 . Gene cassettes involved in Hep biosynthesis and correlation between gene dmhA (open arrows) and the presence of a deoxy form of Hep in the C. jejuni CPS repeating unit. Strains RM1221 (HS53) (Gilbert et al., 2007; Parker et al., 2006), 81-176 (HS:23/36) (Karlyshev et al., ZOOS), and CG8486 (HS:4) (Poly et al., 2007). Solid arrows, other Hep-related genes; white arrows, other genes.
510
KARLYSHEV ET AL.
Structural Analyses Early reports describing chemical structures of high-M, PSs, assumed to be 0-chains of LPS in the precapsule era, were in fact dealing with CPSs (Aspinall et al., 1992, 1994a, 1995; Hanniffy et al., 1999). A recently developed method that uses magic angle spinning nuclear magnetic resonance spectroscopy coupled with nuclear magnetic resonance spectrometry, which does not require the purification and isolation of PSs, is sensitive and efficient at monitoring changes in CPS from defined mutants and complemented strains, and can also be used for gene expression studies (Szymanski et al., 2003). In particular, the method has allowed the detection of an unusual phosphoramidate modification of CPSs of serotypes HS:1 (McNally et al., 2005), HS:19 (McNally et al., 2006), and HS:2 (Szymanski et al., 2003). The method was subsequently applied to the detection of this modification in a large number of other Campylobacter strains and allowed the genetic analysis of phosphoramidate modification of CPS in strain NCTC 11168 (McNally et al., 2007). Importantly, the absence or presence of these modifying groups affects the stainability of the PSs with silver and/ or its interaction with homologous antiserum, which may in part explain their previous lack of detection in some serotypes. The C. jejuni CPS structures associated with particular serotypes are shown in Fig. 3 . The repeating unit of HS:1 CPS consists of just one sugar residue (galactose) linked via glycerolphosphate bridges (McNally et al., 2005) (Fig. 3). Despite a relatively simple structure, the repeating unit is decorated with two fructofuranose branches, which are in turn substituted by an O-methylphosphoramidate group (MeOPN). Decoration of the CPS repeating unit by MeOPN has also been described in C. jejuni 81-176 (HS:23/36) (Kanipes et al., 2006). In addition to MeOPN, the disaccharide repeating unit of HS: 19 CPS is also decorated by sorbofuranose (McNally et al., 2006) (Fig. 3). Structural variation of this PS is attributed both to these groups and variable methylation of the phosphoramidate group itself. A distinct feature of HS:2 CPS (Fig. 3 ) is the presence of glucuronic acid residues amidated by aminoglycerol and a decoration of the GlcA residue with an 0-methyl-heptose side chain (Szymanski et al., 2003). Early studies had reported the occurrence of uronic acid in phenol-water extracts (Moran, 1995b), but this finding was overlooked in subsequent structural studies. Interestingly, the HS:2 capsule is still produced even after inactivation of the Hep biosynthesis pathway in cj1428 mutant, indicating that heptose is not required for capsule assembly
(Michael et al., 2002). According to more recent studies, GlcA in HS:2 CPS has either NGro or ethanolamine substitution at position 5, the Hep branch has 0-methyl groups at both the 3 and 6 positions, and MeOPN at position 4 (Fig. 3) (McNally et al., 2007).
CPS Stabilization with Alcian Blue
It has been observed that bacterial capsules are fragile and are often difficult to detect by electron microscopy without special measures to stabilize their structure. Because capsules are amorphous structures composed of >90% water, the procedure for preparation of samples for electron microscopic examination, which involves sample desiccation, leads to the collapse of these structures on the bacterial surface. To overcome this, one approach is to use specific antibodies or lectins to cross-link the individual CPS molecules and give structural stability during the microscopic preparation. For example, such an approach has been successfully used for the visualization of a capsule produced by Streptococcus suis (Charland et al., 1998). An alternative method is to use a chemical compound with specific affinity for CPSs. One of such compounds is Alcian blue dye (AB). The use of AB for stabilization of bacterial capsules has been known since 1971 (Shea, 1971). AB is a positively charged molecule with high affinity to negatively charged biopolymers such as PSs (Fig. 4A). It was found that the Neisseria gonorrhoeae capsule, for example, can easily be removed by simple washing steps unless stabilized with AB (Hendley et al., 1981).Such treatment significantly improves the consistency and reliability of capsule detection by electron microscopy. Furthermore, AB treatment before silver staining was found to be essential for the detection of subnanogram quantities of proteoglycans and glycosaminoglycans in electrophoresis gels (Moller et al., 1993). The method was later applied to staining bacterial PSs. For example, CPS (also known as colony migration factor) of P. mirabilis could not be detected without AB treatment (Rahman et al., 1999). Staining with the AB technique has been successfully used in the detection of various forms of C. jejuni CPSs directly in electrophoresis gels (Fig. 4B). In particular, it allowed the detection of a lipid-free form of CPS of strain G1 (HS:2) as a product of enzymatic digestion (Karlyshev and Wren, 2001). The mechanism of phospholipase action on C. jejuni CPS became clear after determination of the structure of the phospholilpid anchor (Corcoran et al., 2006) (Fig. 5). The anchor is composed of 1,2-dipalmitoyl3-phosphoglycerol to which the PS component of
C. 7E7UNI CAPSULAR POLYSACCHARIDE
CHAPTER 28
HS:1
[-P+4)-&D-Galp-(l+ 2 3
3)-Gro-(l+
In
/T\
2
P- D-Fruf-3f {MeOPN} HS:2
{MeOPN}
+
4 {3,6-O-Me}-D-glycero-&L-glc-Hepp 1
+ 3
[ +2)-P-~-Ribf-(1~5)-j3-~-GaljNAc-(l34)-aD-GlcpA6-(1 +]n
5
%
{MeOPN } HS:3
{NGro/NEtn}
[3 3)-L-glycero-a-D-ido-Hepp-(l+4)-&~-Galp-(l+ In
HS:6
HS: 19
[+~)-P-D-G~~~A~NG~o-(~+~)-P-D-G~~~NAc-( 1%
;
2 a-L-Sorf HS :23/36
$
{MeOPN}
[+3)-P-D-GlcpNAc-(1+3)-a-D-Galp-(l+ 2)-6d-a-D-aZtro-Hepp-(l+In
-6d-3Me-a-D-aZtro-Hepp-(l+In -D-gZycero-a-D-aZtro-Hepp-(l +In -3Me-D-gZycero-a-D-aZtro-Hepp-(l+In HS:53
a-Xlu 2
+ 2
[-P+ 3)-6d-P-D-manno-Hepp-(l~3)-6dda-DDmm~no-Hepp-(l ~3)-6d-a-D-manno-Hepp-(l +In
$ 2 P-XlU Figure 3. CPS structures of C. jejuni. HS:1 (McNally et al., 2005), HS:2, (Karlyshev et al., 2005; McNally et al., 2007), HS: 3 (Aspinall et al., 1995), HS:6, (Muldoon et al., 2002), HS:19 (McNally et al., 2006), HS:23/36 (Aspinall et al., 1992), and HS:53 (Gilbert et al., 2007). Some side chain groups may be absent as a result of structural variation. Fru, fructose, Hep, heptose, Gal, galactose; Glc, glucose; GlcA, glucuronic acid; Gro, glycerol; Man, mannose; Rib, ribose; NGro, aminoglycerol; NEtn, ethanolamine; MeOPN, 0-methyl phosphoramidate. Furanose and pyranose configurations of sugars are denoted by letters f and p , respectively.
511
5 12
KARLYSHEV ET AL.
A
1
B
2
3
4
CPS
* * CPS-PL + LOS
CH3
Figure 4. (A) Structure of Alcian blue dye (http://en.wikipedia.org/wiki/Image:AlcianBlue.png#file).(B) Electrophoretic detection of a lipid-free form of CPS from strain G1(HS:2) released after hydrolysis with a mixture of phospholipases D and A2 (Karlyshev and Wren, 2001); lane 1, native CPS; lane 2, lipid-free form of CPS; lane 3, LPS of S. enterica serovar Typhimurium; lane 4, protein size markers. CPS, capsular polysaccharide; CPS-PL, lipid-free form of CPS; LOS, lipooligosaccharide.
CPS is linked through a phosphate group as in some CPSs of E. coli (Schmidt and Jann, 1982). Strong affinity of CPS of strain G1 (HS:2) to AB appears to be due to the presence of negatively charged phosphate groups in the repeating unit (McNally et al., 2005). Consistent with this, AB stained acidic but not neutral PS of C. jejuni 81116
(Muldoon et al., 2002), and more recently, AB has been successfully used for investigation of CPS production in strain 81116 (Bachtiar et al., 2007). By means of AB staining, it was observed that CPS lacking the phospholipid moiety could be accumulated in the culture medium in a form of EPS, although the mechanism of its release from the cell surface remains unknown (Karlyshev and Wren, 2001). In addition, AB staining allowed stabilization and detection of the C. jejuni capsule by electron microscopy (Fig. 6) (Karlyshev et al., 2001).
BIOLOGICAL FUNCTIONS OF CAMPYLOBACTER CPS Attenuation of kps Mutants and In Vitro Models of Infection
Figure 5 . Mechanism of release of a lipid-free form of CPS by phospholipases. The structure of the CPS lipid anchor determined for strain 81-176 (Corcoran et al., 2006) with predicted cleavage sites by phospholipases A2 and D (according to Ghannoum et al., 2000). X, polysaccharide moiety.
The CPS does seem to be an important virulence factor because kpsM mutant of C. jejuni 81-176 was attenuated in ferret diarrheal disease model (Bacon et al., 2001). The finding is supported by other studies that used in vitro tissue cell lines as a model of infection. In an early study, C. jejuni LPS was found to
CHAPTER 28
Figure 6. Detection of capsule from C. jejuni G1 (HS:2) by electron microscopy (Karlyshev et al., 2001).
play a role in adhesion to INT-407 cells (McSweegan and Walker, 1986). However, because no distinction between LPS and CPS was possible at that time, it is unclear what molecules were responsible for such binding. Subsequent investigations showed that the KpsM mutant of C. jejuni 81-176 was 2-fold less adhesive and 10-fold less invasive compared with the wild-type strain when an INT-407 cell invasion model was used (Bacon et al., 2001). Despite partial restoration of adhesive properties after complementation, such marginal reduction in adhesion cannot fully provide evidence of involvement of CPS in adhesion. Another mutant affecting CPS biosynthesis, a knockout in KpsE gene of C. jejuni 81116, has also been studied (Bachtiar et al., 2007). Although the KpsE knockout mutant was found to have significantly reduced adhesive ability and showed a twofold reduction in invasion of INT-407 cells, the data were not supported by complementation of defective mutants, and thus it cannot be excluded that the observed changes are attributable to concomitant spontaneous mutation or mutations elsewhere on the chromosome. Conversely, noncapsulated KpsM mutants of other strains of C. jejuni revealed increased attachment ability to Caco-2 cells (A. V. Karlyshev et al., unpublished data). The reduced attachment in capsulated strains may be attributed to a masking effect of capsule on cell surface adhesins. Taken together, the data indicate that CPS may be essential for survival after invasion, although its role in adhesion is less certain. Immunogenicity and Role in Colonization
A possibility to detect various forms of CPS by means of Penner antisera (Karlyshev et al., 2000; Kilcoyne et al., 2006) indicates a strong humoral immune response against these molecules in rabbits. In
C. TETUNI CAPSULAR POLYSACCHARIDE
513
contrast, an inadequate immune response to PSs in chickens was reported, and this could be one of the reasons for the ability of C. jejuni to persist in chickens as a commensal organism (Jeurissen et al., 1998). Chart and coworkers (1996a) demonstrated expression of an “intermediate chain LPS” of C. jejuni NCTC 12501 (HS:2) in chickens. It is plausible that the PS induced was in fact CPS, and its induction in chickens could shed some light on a possible role of these molecules in chicken colonization. More recently, some changes in the in vivo level of expression of cps-related genes were also detected when rabbits were used as a model of infection (Stintzi et al., 2005). The requirement of CPS production for colonization has been investigated by CPS knockout mutants. In one study, a noncapsulated KpsE mutant of C. jejuni 81116 did colonize chickens at 10 days, but the efficiency of colonization was much reduced when compared with the wild-type strain, with cecum and colon contents showing 30 and 100 times lesser bacterial numbers, respectively (Bachtiar et al., 2007). In another study, a noncapsulated kpsC transposon mutant of strain 81-176 produced over six orders of magnitude lower bacterial count than the wild-type strain (Grant et al., 2005). Taken together, these data indicate that CPS does play a role in colonization of chickens. On the other hand, the transposon mutagenesis technique allowed the detection of 29 mutants representing 22 genes involved in chicken colonization (Hendrixson and DiRita, 2004), although none of them were related to CPS biosynthesis. This result suggests that CPS plays little or no role in chicken colonization, although this would be a premature conclusion because the screening, as the investigators admitted, was not exhaustive. Even such an important adhesin-encoding gene as cudF (Konkel et al., 2005), whose product is involved in fibronectin binding by C. jejuni (Konkel et al., 2005; Kuusela et al., 1989), was not detected in this test, although the role of flagella and motility in prolonged colonization was confirmed (Hendrixson and DiRita, 2004). Furthermore, transcriptional profiling identified 59 genes differentially expressed during chicken colonization, none of which was involved in CPS production (Woodall et al., 2005). Because CPS is an amorphous and enveloping structure on the bacterial surface, as visualized in electron microscopy studies (Karlyshev et al., 2001), its production could mask many of the cell surface structures, including adhesins interacting with host cell receptors. However, because of the high water content of capsules, receptor-ligand interactions can still sometimes occur, as evidenced by the recognition
5 14
KAFUYSHEV ET AL.
of bacterial surface molecules by antibodies in encapsulated strains (Jann and Jann, 1987). A number of C. jejuni adhesins have been identified and found to play a role in infection and colonization (Snelling et al., 2005a), including a recently discovered outer membrane transporter, protein CapA (Ashgar et al., 2007). The data on the role of bacterial adhesins in chicken colonization are contradictory because on the one hand, the CPS could prevent adhesins from interacting with host cell receptors, but on the other hand, noncapsulated bacteria exhibit a reduced ability to colonize chickens (Bachtiar et al., 2007; Grant et al., 2005). The contradiction could be resolved by assuming that capsules and adhesins are differentially regulated and are expressed at different stages of infection. CPS may be required and induced at the initial stage of bacterial interaction with the mucous layer and then be downregulated to allow interaction of cell surface protein adhesins with host receptors. Chicken mucus was found to be remarkably different from intestinal mucus of human origin in that it appears to prevent C. jejuni interaction with epithelial cell surfaces (Byrne et al., 2007), which may explain the difference in behavior of C. jejuni as a commensal organism in birds (in particular chickens) and highly pathogenic for other animals, with a low infectious dose for humans. Bacterial Cell Adhesion and Surface Recognition Biofilms have been implicated as a source of contamination by C. jejuni of poultry house water systems (Trachoo et al., 2002). Although in other bacteria CPSs and EPSs were shown to be components of extracellular matrix in the biofilms (Moran and Ljungh, 2003; Yi et al., 2004), the role of C. jejuni CPS in biofilm formation remains unclear. Remarkably, a noncapsulated kpsM mutant of C. jejuni NCTC 11168 retained its ability to form a biofilm at liquid-air interface (Joshua et al., 2006), and no cpsrelated gene products specific to biofilms were found in comparative proteomic study of planktonic and biofilm-grown bacteria (Kalmokoff et al., 2006). On the other hand, the occurrence of PS in C. jejuni biofilms has been demonstrated by selective staining (Corcoran and Moran, 2007), and it cannot be ruled out that PS production independent of CPS may play a role in their formation. Thus, the nature of the biofilm matrix remains unclear, although PSs other than CPS may be involved. More recently, it was shown that in certain conditions, C. jejuni can form a different (classical) form of biofilm (Reeser et al., 2007) that may require the biosynthesis of CPS. Not surprisingly, as a surface molecular structure, it has been found that C. jejuni capsule may play
a role as a bacteriophage receptor (Coward et al., 2006). Nonetheless, this is a significant finding indicating that CPS, along with other bacteriophage receptors, may be involved in a transduction mechanism of DNA exchange and points to a possible genetic mechanism of generating CPS variability resulting from interaction with bacteriophages that are found in abundance in the avian gut. Such a mechanism warrants thorough investigation.
ADDITIONAL PSs Under certain conditions, some C. jejuni strains may produce high-MI PSs different from CPS (Corcoran and Moran, 2007). In C. jejuni 81116, a second structurally characterized PS was found to be neutral, in contrast to the acidic CPS of the strain, and did not interact with homologous HS:6 typing antiserum (Kilcoyne et al., 2006; Muldoon et al., 2002). The neutral PS was deduced to be of LPS origin because LPS-associated core sugar signals were present in nuclear magnetic resonance analysis of this purified PS; reaction occurred in the Limulus amoebocyte lysate assay despite the proven absence of LOS contamination. The HS: 7 reactivity associated with the neutral PS was not detected in a waaF mutant affecting the core of LOS or LPS, thereby indicating an LPS origin of this PS (Kilcoyne et al., 2006; A. P. Moran et al., unpublished data). The simultaneous production of neutral and acidic PSs as an O-chain of LPS and CPS, respectively, would therefore resemble that of some E. coli strains (Jann and Jann, 1987). In another study, an additional PS in C. jejuni 81176 was found to be an a(l,4)-linked glucan that is independent of LOS and serodominant CPS, although it was inferred to be capsular in nature (PappSzabo et al., 2005). Moreover, a calcofluor whitereactive PS that is independent of CPS and produced during specific stress responses has been reported in strain 81-176 (McLennan et al., 2005) but has not been structurally characterized. Although the structures of some of these additional molecules have been determined, no data on the genetic origin of these additional PSs are available to date. It has been noted that in liquid cultures, CPS can detach from bacterial cell surface in a phospholipidfree form (Corcoran et al., 2006; Karlyshev and Wren, 2001). This is an intriguing finding because a detached CPS PS may also play an additional role in virulence and colonization. Because such a form of CPS can be generated on hydrolysis of the purified CPS with commercially available phospholipases (Karlyshev et al., 2000), a similar enzyme was deduced to be involved in the release of the Campy-
CHAPTER 28
lobacter CPS. Surprisingly, a phospholipase pldAknockout mutant still produced a lipid-free form of the CPS (Istivan et al., 2004). The biological role of the PldA remains unclear. Further studies are required for the detailed analysis of the biological function of this form of CPS and the mechanism of its formation. Production of CPSs of different structures by the same bacterial cells has also been documented for other bacteria. One remarkable example is that of the intestinal commensal bacterium Bacteroides fragilis, which has developed a strategy of circumventing the host immune response by producing as many as eight different CPSs, each involving individual differentially regulated gene clusters (Krinos et al., 2001). Although the C. jejuni genome of strain NCTC 11168 contains just one cps gene cluster (Parkhill et al., 2000), these bacteria have adapted a different mechanism of genetic variation through genetic rearrangements and phase variation. However, this does not explain the genetic origin of additional PSs because of the lack of evidence of any genes or gene clusters that would be responsible for their biosynthesis, although one cannot exclude involvement of some hypothetical genes with no assigned function. Further studies are required to determine the genetic origin of these other PS molecules of C. jejuni.
INFLUENCE OF GROWTH CONDITIONS O N CPS PRODUCTION AND STRUCTURE Although CPS production by C. jejuni can occur in liquid and on solid media (Corcoran and Moran, 2007; Karlyshev and Wren, 2001), the PS structure produced in a number of strains has been shown to be growth phase dependent, irrespective of growth at 3 7 or 42°C (Corcoran and Moran, 2007). This structural variation in CPS potentially reflects phasevariable capsule expression that can influence virulence and antigenicity (Bacon et al., 2001). On the other hand, expression of group I1 enterobacterial CPSs is temperature dependent. For example, the K5 capsule of E. coli belonging to this group is only expressed at temperatures above 20°C, and the expression is regulated at the transcriptional level (Rowe et al., 2000). However, the use of microarray-based gene expression analysis detected no such changes in expression of kps genes of C. jejuni at low (up to 4°C) and high (42°C) temperatures (Stintzi, 2003; Stintzi and Whitworth, 2003) compared with standard growth conditions (37°C). Nevertheless, it remains to be determined whether gene expression affecting CPS production is dependent on environmental or physicochemical fac-
C. TETUNI CAPSULAR POLYSACCHARIDE
515
tors other than temperature both in vivo and in vitro. Also, the influence of environmental conditions on production of additional PSs, independent of serodominant CPS, requires investigation because expression of these PSs has been reported to be influenced by laboratory growth conditions (McLennan et al., 2005; Corcoran and Moran, 2007). In particular, production of an additional PS by C. jejuni NCTC 11168 has been observed when these bacteria were grown in liquid but not on solid media (Corcoran and Moran, 2007). Expression of the PS was growth phase independent but temperature dependent, being expressed at 37°C but not 42°C. In another study, expression of a CPS-independent PS in C. jejuni 81176 on solid media was enhanced after prolonged microaerobic incubation and even more dramatically enhanced at 42°C in an anaerobic environment (McLennan et al., 2005). Moreover, this PS production was shown to be linked to the spoT-mediated bacterial stress response. However, the identity of these additional PSs is unclear, and their role in the stress response and pathogenesis of C. jejuni requires investigation.
CONCLUSIONS AND FUTURE OUTLOOK Further studies are required to investigate the genetic stability of Penner serotypes and the possibility of using CPS-related genes as genetic markers for diagnostic and epidemiological studies on C. jejuni. Moreover, the roles of specific CPS structures in bacteria-host interactions and gastroenteritis development need further study. Conversely, as virulence factors, CPSs or their partial structures could form important components of conjugate vaccines or in mixture be included in subunit vaccines, and hence the possibility of using these PSs in C. jejuni vaccines should be explored. Mechanisms of Genetic Rearrangements within cps Gene Clusters and Stability of Penner Serotypes One of the mechanisms of genetic diversity in
kps gene clusters is attributed to acquisition of exogenous material via horizontal gene transfer. This is not surprising taking into account the fact that C. jejuni bacteria are naturally competent and exchange of genetic information between Campylobacter strains in vivo was confirmed experimentally (Boer et al., 2002). Unless maintained on an autonomously replicating compatible plasmid, the newly acquired genes have to integrate into chromosomal DNA via homologous recombination. Nevertheless, in some cases, integration may also occur via an illegitimate
5 16
KARLYSHEV ET AL.
recombination event that does not require any homologous sequences (Richardson and Park, 1997). The mechanism of such recombination is unknown, but similar events may lead to the diversity of the capsule gene clusters. Despite the identified inter- and intrastrain structural diversity of CPSs, a relative stability of Penner serotypes under laboratory growth conditions seems puzzling. No changes of Penner serotype after subculturing 50 times in vitro and during colonization of mice for up to 26 days were observed (Nielsen et al., 2001). This finding of serotype stability contradicts other research demonstrating high heterogeneity in clonal isolates of strain NCTC 11168 producing CPSs with different affinity to Penner 2 antiserum (Szymanski et al., 2003). The variability was found to be correlated with modifications in this PS. For example, the presence of ethanolamine modification increased interaction of the CPS with antiserum in Western blotting experiments, while other modifications had an opposite effect (Szymanski et al., 2003). Moreover, CPS structure and biosynthesis has been found to be regulated by in vitro conditions (Corcoran and Moran, 2007). In addition, the kpsMdependent capsule was found to undergo phase variation at a high frequency (Bacon et al., 2001) in laboratory conditions, and a change in C. jejuni serotype profile in a patient infected with this pathogen has been reported (Mills et al., 1992). Therefore, without selective pressure (e.g., when maintained in vitro), C. jejuni strains appear able to produce a vast variety of CPS molecules. Not all of these alterations would necessarily result in serotype changes, and the dominant serotype present in the mixture may be masking other more dramatic changes. By producing a certain variety of CPSs, bacteria may have a selective advantage when interacting with the host and its immune system. Additional studies are necessary to clarify the origin and mechanisms of variation of the C. jejuni CPS structures and the role of such variability in pathogenesis. Use of cps Genes as Genetic Markers for Diagnosis and Epidemiology Studies Because of the high conservation of the C. jejuni kps genes involved in CPS transport and assembly (Karlyshev et al., 2005), they could be suitable genetic markers in epidemiological studies. The genomes of a number of Campylobacter spp. are being sequenced, including C. jejuni (various strains), C. coli, C. lari, C. upsaliensis, C. doylei, C. concisus, C. curvus, C. fetus, and C. hominis (Fouts et al., 2005; http://msc.tigr.org/campy/index.shtml).Scrutiny of partially sequenced genomes has allowed identifica-
tion of kpsM orthologs in C. coli, C. lari, and C. upsuliensis (A. V. Karlyshev, unpublished data), indicating that these species are closely related and similar to C. jejuni in their potential to synthesize capsules. Differentiation of CPS-producing strains could be based on the probes corresponding to the genes involved in particular biosynthetic pathways, e.g., the mobile yet highly conserved genes involved in Hep biosynthesis. Role of Specific CPS Structures in Bacterium-Host Interaction and Gastroenteritis The presence of Hep residues in some Campylobacter CPSs, conservation, and possible genetic interstrain exchange of the genes involved in encoding the biosynthesis and transport of these residues, especially genes hddC, gmhA2, and hddA (Karlyshev et al., 2005) (Fig. 2), may be attributed to bacterial virulence. In HS:2 CPS, the Hep is present as a branch residue, while in HS:23/36, HS:41, and HS:53 CPSs, it is one of main backbone sugar residues forming a repeating unit (Fig. 3). The contributions of Hep residues and modification groups to the biological properties of C. jejuni, such as interaction with host cell receptors, immune system interaction, and contribution to increased survival under adverse conditions in the environment, warrant further investigation. One of the most remarkable findings is the presence of unusual phosphoramidate modifications in a number of C. jejuni CPSs (Fig. 3) (Karlyshev et al., 2005; McNally et al., 2005, 2006, 2007), which may be involved in changes in biological activity of the CPS that could contribute to virulence. The presence of a nonmethylated phosphoramidate group would result in the formation of a so-called zwitterinonic PS (i.e., PS containing both positively and negatively charged groups in the repeating unit) found to be responsible for abscess formation during infection caused by B. fragilis (Tzianabos and Kasper, 2002). Such PSs are able to modulate host immune response via interaction with CD4+ T cells. It would also be interesting to study a role of phosphoramidate modification of CPS in colonization and interaction with the host immune system. These studies would assist in the development of specific compounds to inhibit the production of CPS in chickens, thus potentially eliminating the pathogen from the livestock. Another emerging area of research is the study of Cumpylobacter survival in protozoa. It was found that C. jejuni is protected from environmental stress conditions within Acanthamoeba castellanii and Tetruhymena pyriformis (Snelling et al., 2005b), as well as Acunthumoeba polyphuga (Axelsson-Olsson et al., 2005), although the mechanism of such survival re-
CHAPTER 28
mains unknown. It is likely that CPS can play a role in adaptation of bacteria to environmental conditions, including survival in protozoa, and hence, the regulation of CPS biosynthesis under different conditions and the role of variable CPS structures in survival within protozoa is worthy of further investigation.
CPS and Vaccine Development Unfortunately, no efficient vaccine against C. jejuni infection in poultry and/or humans is available to date (chapter 24). Such a vaccine would be useful for elimination of C. jejuni from livestock as well as prophylaxis for traveler’s diarrhea, which is often caused by Campylobacter species (Scott, 1997). Initial attempts to develop a vaccine against C. jejuni infection, particularly in chickens, have not been very successful (Ziprin et al., 2002). Only limited protection against C. jejuni infection using a ferret model was achieved when using a killed bacterial preparation (Burr et al., 2005). Initial studies in the development of a human vaccine have focused on the use of heat- and formalin-attenuated whole-cell vaccines (Prendergast et al., 2004; Walker, 2005). Phase I trials have been completed, and protection against rechallenge with the same strain was demonstrated (Black et al., 1988; Prendergast et al., 2004; Walker, 2005). Nevertheless, one drawback associated with the routine use of killed or attenuated C. jejuni in vaccines is a risk for the development of autoimmune complications, such as GBS and Miller Fisher syndrome (Moran, 2007; Prokhorova et al., 2006) as a result of mimicry of host gangliosidies by bacterial LOSS and induction of cross-reactive antineuronal antibodies (Moran and Prendergast, 2001; Moran et al., 2005). Furthermore, because C. jejuni LOS can undergo phase variation both in vitro (Guerry et al., 2002) and in vivo (Prendergast et al., 2004), which thereby influences the expression of ganglioside mimicry, the choice of a strain for use in a whole-cell vaccine but also in challenge studies is complicated (Moran, 2007). Because of these complications, other C. jejuni vaccine strategies that use a flagellin-hybrid protein (Khoury and Meinersmann, 1995) and a truncated flagellin subunit vaccine (Lee et al., 1999) have been explored in animal models and have met with limited success. In another study, a live attenuated Salmonella strain vectoring the PEBl antigen of C. jejuni, which is implicated in colonization ability (Pei et al., 1998), was evaluated as a vaccine against Campylobacter infection, but no protection was observed (Sizemore et al., 2006). An avirulent Salmonella enterica serovar strain expressing highly conserved C.
C. 7ETUNI CAPSULAR POLYSACCHARIDE
517
jejuni genes cjaA (cjO982c), cjaC (cjO734c), and cjaD (cjOll3) were used for immunization of chickens. The strain expressing CjaA was found to be able to induce serum IgG and mucosal IgA antibody responses in chickens and was able to protect chicken from challenge with heterologous wild-type C. jejuni strain (Wyszynska et al., 2004). In another study, a live attenuated Salmonella strain producing PEBl antigen of C. jejuni was evaluated as a vaccine against Campylobacter infection (Sizemore et al., 2006), but no protection was observed. More recently, CE BioSciences has undertaken a systematic, proteomicbased study to identify cell surface proteins as potential targets for vaccine development (Prokhorova et al., 2006). Reduction in colonization of mice was achieved by immunization with selected cell surface proteins (Prokhorova et al., 2006; Schrotz-King et al., 2007), thus demonstrating that the subunit vaccine approach with suitable antigens could prove successful. However, the subunit approach does not preclude the use of PS antigens or CPS structures as vaccine components instead of proteins or in combination with selected proteins. CPSs have been proven to be excellent human vaccine components to protect against Haemophilus influenzae, S. pneumoniae, and some N.meningitidis species (Lee et al., 2001). CPS may be considered a component for the development of a subunit vaccine against Campylobacter infection. Thus, a further direction for subunit vaccine development would be investigation of the protective properties of CPSs against C. jejuni infection using different animal models. Despite significant structural variation, it may be possible to identify common motifs of CPS structures that are specifically linked to virulence and protection. Some common structures that have already been identified from previous studies are heptosyl and phosphoramide residues (Kanipes et al., 2006; McNally et al., 2005, 2006; Szymanski et al., 2003). As a result of similarity between C. jejuni CPS structures, a set of a few CPSs could prove efficient in protection against a large number of different C. jejuni strains, and this hypothesis requires urgent testing. Hence, a combination or synthetic modification of PS antigens might prove crossprotective against antigenically diverse C. jejuni strains. Although PS and carbohydrate-related antigens in general possess low immunogenicity, this can be overcome by their combination in carbohydrateprotein conjugates. The conjugate vaccine approach would have the further advantage of inclusion of protective protein and carbohydrate moieties in a potential vaccine. The use of CPS structures as vaccine candidates is also attractive: they are inexpensive to produce because they can be extracted easily from the
5 18
KARLYSHEV ET AL.
bacterial cell surface and are readily purified from contaminating LOS and proteins (Corcoran et al., 2006; Karlyshev and Wren, 2001). Overall, consideration should be given to the potential inclusion of CPS-based antigens in Campylobacter vaccines and the advantages that could be attained. Acknowledgments. Investigations on CPSs in our laboratories are supported by the Biotechnology and Biological Sciences Research Council, Leverhulme Trust, and Wellcome Trust (to A.V.K and B.W.W.) and by the Higher Education Authority PRTL-3 program of the National Development Plan and a European Union Marie Curie program grant 2005-029774 (to A.P.M.). REFERENCES Ashgar, S. S., N. J. Oldfield, K. G. Wooldridge, M. A. Jones, G. J. Irving, D. P. Turner, and D. A. Ala’Aldeen. 2007. CapA, an autotransporter protein of Campylobacter jejuni, mediates association with human epithelial cells and colonization of the chicken gut. J. Bacteriol. 189:1856-1865. Aspinall, G. O., C. M. Lynch, H. Pang, R. T. Shaver, and A. P. Moran. 1995. Chemical structures of the core region of Campylobacter jejuni 0 : 3 lipopolysaccharide and an associated polysaccharide. Eur. J. Biochem. 231570478. Aspinall, G. O., A. G. McDonald, and H. Pang. 1994a. Lipopolysaccharides of Campylobacter jejuni serotype 0:19: structures of 0 antigen chains from the serostrain and two bacterial isolates from patients with the Guillain-Barrt syndrome. Biochemistry 33:250-255. Aspinall, G . O., A. G. McDonald, and H. Pang. 1992. Structures of the 0 chains from lipopolysaccharides of Campylobacter jejuni serotypes 0:23 and 0:36. Carbohydr. Res. 231:13-30. Aspinall, G. O., A. G. McDonald, H. Pang, L. A. Kurjanczyk, and J. L. Penner. 1994b. Lipopolysaccharides of Campylobacter jejuni serotype 0:19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barrt syndrome. Biochemistry 33:241-249. Axelsson-Olsson, D., J. Waldenstrom, T. Broman, B. Olsen, and M. Holmberg. 2005. Protozoan Acanthamoeba polyphaga as a potential reservoir for Campylobacter jejuni. Appl. Environ. Microbiol. 71:987-992. Bachtiar, B. M., P. J. Coloe, and B. N. Fry. 2007. Knockout mutagenesis of the kpsE gene of Campylobacter jejuni 81116 and its involvement in bacterium-host interactions. FEMS Immunol. Med. Microbiol. 49:149-154. Bacon, D. J., C. M. Szymanski, D. H. Burr, R. P. Silver, R. A. Alm, and P. Guerry. 2001. A phase-variable capsule is involved in virulence of Campylobacter jejuni 81-176. Mol. Microbiol. 40:769-777. Bernatchez, S., C. M. Szymanski, N. Ishiyama, J. Li, H. C. Jarrell, P. C. Lau, A. M. Berghuis, N. M. Young, and W. W. Wakarchuk. 2005. A single bifunctional UDP-GlcNAc/ Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni. 1.Biol. Chem. 280:4792-4802. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Blaser, M. J., P. F. Smith, J. E. Repine, and K. A. Joiner. 1988. Pathogenesis of Campylobacter fetus infections. Failure of encapsulated Campylobacter fetus to bind C3b explains serum and phagocytosis resistance. J. Clin. Invest. 81:1434-1444. Boer, P., J. A. Wagenaar, R P. Achterberg, J. P. Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:35 1-359.
Burr, D. H., D. Rollins, L. H. Lee, D. L. Pattarini, S. S. Walz, J. H. Tian, J. L. Pace, A. L. Bourgeois, and R. I. Walker. 2005. Prevention of disease in ferrets fed an inactivated whole cell Campylobacter jejuni vaccine. Vaccine 23:43 15-4321. Byrne, C. M., M. Clyne, and B. Bourke. 2007. Campylobacter jejuni adhere to and invade chicken intestinal epithelial cells in vitro. Microbiology 153561-569. Charland, N., J. Harel, M. Kobisch, S. Lacasse, and M. Gottschalk. 1998. Streptococcus suis serotype 2 mutants deficient in capsular expression. Microbiology 144:325-332. Chart, H., D. Conway, J. A. Frost, and B. Rowe. 1996a. Outer membrane characteristics of Campylobacter jejuni grown in chickens. FEMS Microbiol. Lett. 145:469-472. Chart, H., J. A. Frost, A. Oza, R. Thwaites, S. Gillanders, and B. Rowe. 1996b. Heat-stable serotyping antigens expressed by strains of Campylobacter jejuni are probably capsular and not long-chain lipopolysaccharide. J. Appl. Bacteriol. 81:635-640. Chatzipanagiotou, S., K. Kilidireas, G . Trimis, C. Nicolaou, M. Anagnostouli, C. Athanassaki, A. Giannoulia, N. Legakis, and S. Youroukos. 2003. Campylobacter jejuni 0:19 serotypeassociated Guillain-Barrt syndrome in a child: the first case reported from Greece. Clin. Microbiol. Infect. 9:69-72. Corcoran, A. T., H. Annuk, and A. P. Moran. 2006. The structure of the lipid anchor of Campylobacter jejuni polysaccharide. FEMS Microbiol. Lett. 257:228-235. Corcoran, A. T., and A. P. Moran. 2007. Influence of growth conditions on diverse polysaccharide production by Campylobacter jejuni. FEMS Immunol. Med. Microbiol. 49: 124-132. Coward, C., A. J. Grant, C. Swift, J. Philp, R. Towler, M. Heydarian, J. A. Frost, and D. J. Maskell. 2006. Phase-variable surface structures are required for infection of Campylobacter jejuni by bacteriophages. Appl. Environ. Microbiol. 72:4638-4647. Davey, M. E., and M. J. Duncan. 2006. Enhanced biofilm formation and loss of capsule synthesis: deletion of a putative glycosyltransferase in Porphyromonas gingivalis. J. Bacteriol. 188: 55 10-5523. Ezzell, J. W., and S. L. Welkos. 1999. The capsule of Bacillus anthracis: a review. J. Appl. Microbiol. 87:250-255. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Fry, B. N., S. Feng, Y. Y. Chen, D. G. Newell, P. J. Coloe, and V. Korolik. 2000. The galE gene of Campylobacter jejuni is involved in lipopolysaccharide synthesis and virulence. Infect. Immun. 68:2594--2601. Ghannoum, M. A. 2000. Potential role of phospholipases in virulence and fungal pathogenesis. Clin. Microbiol. Rev. 13: 122143. Gilbert, M., R. E. Mandrell, C. T. Parker, J. Li, and E. Vinogradov. 2007. Structural analysis of the capsular polysaccharide from Campylobacter jejuni RM1221. Chembiochemistry 8:625631. Grant, A. J., C. Coward, M. A. Jones, C. A. Woodall, P. A. Barrow, and D. J. Maskell. 2005. Signature-tagged transposon mutagenesis studies demonstrate the dynamic nature of cecal colonization of 2-week-old chickens by Campylobacter jejuni. Appl. Environ. Microbiol. 71:8031-8041. Guerry, P., C. M. Szymanski, M. M. Prendergast, T. E. Hickey, C. P. Ewing, D. L. Pattarini, and A. P. Moran. 2002. Phase variation of Campylobacter jejuni 8 1-176 lipooligosaccharide af-
CHAPTER 28
fects ganglioside mimicry and invasiveness in vitro. Infect. Immun. 70:787-793. Hanniffy, 0. M., A. S. Shashkov, A. P. Moran, M. M. Prendergast, S. N. Senchenkova, Y. A. Knirel, and A. S. Savage. 1999. Chemical structure of a polysaccharide from Campylobacter jejuni 176.83 (serotype 0:41) containing only furanose sugars. Carbohydr. Res. 3 19: 124- 132. Hendley, J. O., K. R. Powell, N. L. Salomonsky, and R. R Rodewald. 1981. Electron microscopy of the gonococcal capsule. ]. Infect. Dis. 143:796-802. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Istivan, T. S., P. J. Coloe, B. N. Fry, P. Ward, and S. C. Smith. 2004. Characterization of a haemolytic phospholipase A2 activity in clinical isolates of Campylobacter concisus.J. Med. Microbiol. 53~483-493. Jam, K., and B. Jann 1987 Polysaccharide antigens of Escherichia coli. Rev. Infect. Dis. 9(Suppl. 5):S517-S526. Jeurissen, S. H., E. M. Janse, N. van Rooijen, and E. Claassen. 1998. Inadequate anti-polysaccharide antibody responses in the chicken. Immunobiology 198:385-395. Joshua, G. W., C. Guthrie-Irons, A. V. Karlyshev, and B. W. Wren. 2006. Biofilm formation in Campylobacter jejuni. Microbiology 152:387-396. Jung, S. Y., A. Matin, K. S. Kim, and N. A. Khan. 2007. The capsule plays an important role in Escherichia coli K 1 interactions with Acanthamoeba. Int. 1.Parasitol. 37:417-423. Kalmokoff, M., P. Lanthier, T. L. Tremblay, M. Foss, P. C. Lau, G. Sanders, J. Austin, J. Kelly, and C. M. Szymanski. 2006. Proteomic analysis of Campylobacter jejuni 11168 biofilms reveals a role for the motility complex in biofilm formation.]. Bacteriol. 188:4312-4320. Kanipes, M. I., E. Papp-Szabo, P. Guerry, and M. A. Monteiro. 2006. Mutation of waaC, encoding heptosyltransferase I in Campylobacter jejuni 81-176, affects the structure of both lipopolysachharide and capsular carbohydrate. J. Bacteriol. 188: 3273-3279. Karlyshev, A. V., 0. L. Champion, C. Churcher, J. R. Brisson, H. C. Jarrell, M. Gilbert, D. Brochu, F. St Michael, J. Li, W. W. Wakarchuk, I. Goodhead, M. Sanders, K. Stevens, B. White, J. Parkhill, B. W. Wren, and C. M. Szymanski. 2005. Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol. Microbiol. 55:90-103. Karlyshev, A. V., E. E. Galyov, 0. Smirnov, A. P. Guzayev, V. M. Abramov, and V. I?. Zav’yalov. 1992. A new gene of the f 2 operon of Y. pestis involved in the capsule biogenesis. FEBS Lett. 297:77-80. Karlyshev, A. V., D. Linton, N. A. Gregson, A. J. Lastovica, and B. W. Wren. 2000. Genetic and biochemical evidence of a Campylobacter jejuni capsular polysaccharide that accounts for Penner serotype specificity. Mol. Microbiol. 35529-541. Karlyshev, A. V., M. V. McCrossan, and B. W. Wren. 2001. Demonstration of polysaccharide capsule in Campylobacter jejuni using electron microscopy. Infect. Immun. 695921-5924. Karlyshev, A. V., and B. W. Wren. 2001. Detection and initial characterization of novel capsular polysaccharide among diverse Campylobacter jejuni strains using Alcian blue dye. J. Clin. Microbiol. 39:279-284. Khoury, C. A., and R. J. Meinersmann. 1995. A genetic hybrid of the Campylobacter jejuni flaA gene with LT-B of Escherichia coli and assessment of the efficacy of the hybrid protein as an oral chicken vaccine. Avian Res. 39:812-820. Kilcoyne, M., A. P. Moran, A. S. Shashkov, S. N. Senchenkova, J. A. Ferris, A. T. Corcoran, and A. V. Savage. 2006. Molecular
c. rEruNI CAPSULAR POLYSACCHAFUDE
519
origin of two polysaccharides of Campylobacter jejuni 8 1116. FEMS Microbiol. Lett. 263:214-222. Komagamine, T., and N. Yuki. 2006. Ganglioside mimicry as a cause of Guillain-Barrt syndrome. CNS Neurol. Disord. Drug Targets 5:3 9 1-400. Konkel, M. E., J. E. Christensen, A. M. Keech, M. R. Monteville, J. D. Klena, and S . G. Garvis. 2005. Identification of a fibronectin-binding domain within the Campylobacter jejuni CadF protein. Mol. Microbiol. 57:1022-1035. Krinos, C. M., M. J. Coyne, K. G. Weinacht, A. 0. Tzianabos, D. L. Kasper, and L. E. Comstock. 2001. Extensive surface diversity of a commensal microorganism by multiple DNA inversions. Nature 414555-558. Kuroki, S., T. Saida, M. Nukina, T. Haruta, M. Yoshioka, Y. Kobayashi, and H. Nakanishi. 1993. Campylobacter jejuni strains from patients with Guillain-BarrC syndrome belong mostly to Penner serogroup 19 and contain P-N-acetylglucosamine residues. Ann. Neurol. 33:243-247. Kuusela, P., A. P. Moran, T. Vartio, and T. U. Kosunen. 1989. Interaction of Campylobacter jejuni with extracellular matrix components. Biochim. Biophys. Acta 993:297-300. Lastovica, A. J., E. A. Goddard, and A. C. Argent. 1997. GuillainBarre syndrome in South Africa associated with Campylobacter jejuni 0:41 strains.]. Infect. Dis. 176(Suppl 2):S139-S143. Lee, C. J., L. H. Lee, C. S. Lu, and A. Wu. 2001. Bacterial polysaccharides as vaccines-immunity and chemical characterization. Adv. Exp. Med. Biol. 491:453-471. Lee, L. H., E. Burg, S. Baqar, A. L. Bourgeois, D. H. Burr, C. P. Ewing, T. J. Trust, and P. Guerry. 1999. Evaluation of a truncated recombinant flagellin subunit vaccine against Campylobacter jejuni. Infect. Immun. 675799-5805. Magee, A. D., and J. Yother. 2001. Requirement for capsule in colonization by Streptococcus pneumoniae. Infect. Immun. 69: 3755-3761. Mandatori, R,and J. L. Penner. 1989. Structural and antigenic properties of Campylobacter coli lipopolysaccharides.Infect. Immun. 57:3506-35 11. McKay, D., J. Fletcher, P. Cooper, and F. M. Thomson-Carter. 2001. Comparison of two methods for serotyping Campylobacter spp.]. Clin. Microbiol. 39:1917-1921. McLennan, M., D. Ringoir, H. Jarrell, C. Szymanski, and E. Gaynor. 2005. Characterization of a Campylobacter jejuni surface moiety that cross-reacts with calcofluor white: implications for surface carbohydrates, stress response, and pathogenesis, p. 87. In V. Korolik, A. Lee, H. Mitchell, G. Mendz, B. Fry, and P. Coloe (ed.), Abstracts of the 13th International Workshop on Campylobacter, Helicobacter and Related Organisms. Griffith University Publications, Gold Coast, Australia. McNally, D. J., H. C. Jarrell, N. H. Khieu, J. Li, E. Vinogradov, D. M. Whitfield, C. M. Szymanski, and J. R. Brisson. 2006. The HS: 19 serostrain of Campylobacter jejuni has a hyaluronic acidtype capsular polysaccharide with a nonstoichiometric sorbose branch and 0-methyl phosphoramidate group. FEBS 1. 273: 3975-3989. McNally, D. J., H. C. Jarrell, J. Li, N. H. Khieu, E. Vinogradov, C. M. Szymanski, and J. R. Brisson. 2005. The HS:1 serostrain of Campylobacter jejuni has a complex teichoic acid-like capsular polysaccharide with nonstoichiometric fructofuranose branches and 0-methyl phosphoramidate groups. FEBS 1.272: 4407-4422. McNally, D. J., M. P. Lamoureux, A. V. Karlyshev, L. M. Fiori, J. Li, G. Thacker, R. A. Coleman, N. H. Khieu, B. W. Wren, J. R. Brisson, H. C. Jarrell, and C. M. Szymanski. 2007. Commonality and biosynthesis of the 0-methyl phosphoramidate capsule modification in Campylobacter jejuni. 1. Biol. Chem. 282:28566-28576.
520
KARLYSHEV ET AL.
McSweegan, E., and R. I. Walker. 1986. Identification and characterization of two Campylobacter jejuni adhesins for cellular and mucous substrates. Infect. lmmun. 53:141-148. Michael, F. S., C. M. Szymanski, J. Li, K. H. Chan, N. H. Khieu, S . Larocque, W. W. Wakarchuk, J. R. Brisson, and M. A. Monteiro. 2002. The structures of the lipooligosaccharide and capsule polysaccharide of Campylobacter jejuni genome sequenced strain NCTC 11168. Eur. J. Biochem. 2695119-5136. Mills, S. D., W. C . Bradbury, and J. L. Penner. 1985. Basis for serological heterogeneity of thermostable antigens of Campylobacter jejuni. Infect. lmmun. 50:284-291. Mills, S. D., B. Kuzniar, B. Shames, L. A. Kurjanczyk, and J. L. Penner. 1992. Variation of the 0 antigen of Campylobacter jejuni in vivo. J. Med. Microbiol. 36:215-219. Moller, H. J., D. Heinegard, and J. H. Poulsen. 1993. Combined Alcian blue and silver staining of subnanogram quantities of proteoglycans and glycosaminoglycans in sodium dodecyl sulfatepolyacrylamide gels. Anal. Biochem. 209:169-175. Moran, A. P. 1995a. Structure-bioactivity relationships of bacterial endotoxins. J. Toxicol. Toxin Rev. 14:47-83. Moran, A. P. 1995b. Biological and serological characterization of Campylobacter jejuni lipopolysaccharides with deviating core and lipid A structures. FEMS lmmunol. Med. Microbiol.l1:121130. Moran, A. P. 2007. Vaccinating humans against Campylobacter, p. 75. In Abstracts of the International Association for Food Protection 94th Annual Meeting (Orlando, Florida). International Association for Food Protection, Des Moines, IA. Moran, A. P., and H. Annuk. 2003. Recent advances in understanding biofilms of mucosae. Rev. Environ. Sci. Biotechnol. 2: 121-140. Moran, A. P., H. Annuk, and M. M. Prendergast. 2005. Antibodies induced by ganglioside-mimicking Campylobacter jejuni lipooligosaccharides recognise epitopes at the nodes of Ranvier. J. Neuroimmunol. 165: 179-1 85. Moran, A. P., T. U. Kosunen, and M. M. Prendergast. 2001. Biochemical and serological analysis of the heat-stable antigens involved in serotyping Campylobacter jejunilcoli. Int. J. Med. Microbiol. 291 (Suppl. 31):72. Moran, A. P., and A. Ljungh. 2003. Physico-chemical properties of extracellular polymeric substances, p. 91-112. In P. Lens, A. P. Moran, T. O’Mahony, P. Stoodley, and V. O’Flaherty (ed.), Biofilms in Medicine, lndustry and Environmental Biotechnology: Characteristics, Analysis and Control. IWA Publishing, London, United Kingdom. Moran, A. P., and J. L. Penner. 1999. Serotyping of Campylobacter jejuni based on heat-stable antigens: relevance, molecular basis and implications in pathogenesis. J. Appl. Microbiol. 86:361377. Moran, A. P., J. L. Penner, and G. 0. Aspinall. 2000. Campylobacter lipopolysaccharides, p. 241-257. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Moran, A. P., and M. M. Prendergast. 2001. Molecular mimicry in Campylobacter jejuni and Helicobacter pylori lipopolysaccharides: contribution of gastrointestinal infections to autoimmunity. J. Autoimmun. 16:241-256. Moran, A. P., U. Zahringer, U. Seydel, D. Ssholz, P. Stiitz, and E. T. Rietschel. 1991. Structural analysis of the lipid A component of Campylobacter jejuni CCUG 10936 (serotype 0 : 2 ) lipopolysaccharide. Description of a lipid A containing a hybrid backbone of 2-amino-2-deoxy-~-g~ucose and 2,3-diamino-2,3dideoxy-D-glucose. Eur. J. Biochem. 198:459-469. Muldoon, J., A. S. Shashkov, A. P. Moran, J. A. Ferris, S. N. Senchenkova, and A. V. Savage. 2002. Structures of two poly-
saccharides of Campylobacter jejuni 8 1116. Carbohydr. Res. 337~2223-2229. Nachamkin, I. 2002. Chronic effects of Campylobacter infection. Microbes Infect. 4:399-403. Nachamkin, I., J. Engberg, M. Gutacker, R. J. Meinersman, C. Y. Li, P. Arzate, E. Teeple, V. Fussing, T. W. Ho, A. K. Asbury, J. W. Griffin, G. M. McKhann, and J. C. Piffaretti. 2001. Molecular population genetic analysis of Campylobacter jejuni HS: 19 associated with Guillain-BarrC syndrome and gastroenteritis. J. Infect. Dis. 184:221-226. Nelson, A. L., A. M. Roche, J. M. Gould, K. Chim, A. J. Ratner, and J. N. Weiser. 2007. Capsule enhances pneumococcal colonization by limiting mucus-mediated clearance. Infect. lmmun. 75:83-90. Nielsen, E. M., J. Engberg, and V. Fussing. 2001. Genotypic and serotypic stability of Campylobacter jejuni strains during in vitro and in vivo passage. Int. J. Med. Microbiol. 291:379-385. Oldfield, N. J., A. P. Moran, L. A. Millar, M. M. Prendergast, and J. M. Ketley. 2002. Characterization of the Campylobacterjejuni heptosyltransferase I1 gene, waaF, provides genetic evidence that extracellular polysaccharide is lipid A core independent. J. Bacteriol. 184: 2 100-2 107. Otto, M. 2006. Bacterial evasion of antimicrobial peptides by biofilm formation. Cum. Top. Microbiol. lmmunol. 306:251-258. Papp-Szabo, E., M. I. Kanipes, P. Guerry, and M. A. Mouteiro. 2005. Cell-surface a-glucan in Campylobacter jejuni 8 1-176. Carbohydr. Res. 340:2218-2221. Parker, C. T., B. Quinones, W. G. Miller, S. T. Horn, and R. E. Mandrel]. 2006. Comparative genomic analysis of Campylobacter jejuni strains reveals diversity due to genomic elements similar to those present in C. jejuni strain RM1221. J. Clin. Microbiol. 44:4125-4135. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pei, Z., C. Burucoa, B. Grignon, S. Baqar, X. Z. Hunag, D. J. Kopecko, A. L. Bourgeois, J. L. Fauchere, and M. J. Blaser. 1998. Mutation in the peblA locus of Campylobacter jejuni reduces interactions with epithelial cells and intestinal colonization of mice. Infect. lmmun. 66:938-943. Penner, J. L., and G. 0. Aspinall. 1997. Diversity of lipopolysaccharide structures in Campylobacter jejuni. J. Infect. Dis. 176(S~ppl.2):S135-S138. Penner, J. L., and J. N. Hennessy. 1980. Passive hemagglutination technique for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heat-stable antigens. J. Clin. Microbiol. 12: 732-737. Poly, F., T. Read, D. R. Tribble, S. Baqar, M. Lorenzo, and P. Guerry. 2007. Genome sequence of a clinical isolate of Campylobacter jejuni from Thailand. Infect. lmmun. 75:3425-3433. Prendergast, M. M., and A. P. Moran. 2000. Lipopolysaccharides in the development of the Guillain-Barrt syndrome and Miller Fisher syndome forms of acute inflammatory peripheral neuropathies. J. Endotoxin Res. 6:341-359. Prendergast, M. M., D. R. Tribble, S. Baqar, D. S. Scott, J. A. Ferris, R. I. Walker, and A. P. Moran. 2004. In vivo phase variation and serological response to lipooligosaccharide of Campylobacter jejuni in experimental human infection. Infect. lmmun. 72:9 16-922. Preston, M. A., and J. L. Penner. 1987. Structural and tigneic properties of lipopolysaccharides from serotype reference strains of Campylobacter jejuni. Infect. Immun. 55:1806-1812.
CHAPTER 28
Prokhorova, T. A., P. N. Nielsen, J. Petersen, T. Kofoed, J. S. Crawford, C. Morsczeck, A. Boysen, and P. Schrotz-King. 2006. Novel surface polypeptides of Campylobacter jejuni as traveller’s diarrhoea vaccine candidates discovered by proteomics. Vaccine 24:6446-6455. Rahman, M. M., J. Guard-Petter, K. Asokan, C. Hughes, and R. W. Carlson. 1999. The structure of the colony migration factor from pathogenic Proteus mirabilis. A capsular polysaccharide that facilitates swarming. 1.Biol. Chem. 274:22993-22998. Reeser, R. J., R. T. Medler, S. J. Billington, B. H. Jost, and L. A. Joens. 2007. Characterization of Campylobacter jejuni biofilms under defined growth conditions. Appl. Environ. Microbiol. 73: 1908-1913. Richardson, P. T., and S. F. Park. 1997. Integration of heterologous plasmid DNA into multiple sites on the genome of Campylobacter coli following natural transformation. /. Bacteriol. 179:1809-1812. bRoberts, I. S. 1996. The biochemistry and genetics of capsular polysaccharide production in bacteria. Annu. Rev. Microbiol. 50:285-315. Rowe, S., N. Hodson, G. Griffiths, and I. S. Roberts. 2000. Regulation of the Escherichia coli K5 capsule gene cluster: evidence for the roles of H-NS, BipA, and integration host factor in regulation of group 2 capsule gene clusters in pathogenic E. coli. J. Bacteriol. 182:274 1-2745. Schmidt, M. A., and K. Jam. 1982. Phosholipid substitution of capsular (K) polysaccharide antigens from Escherichia coli causing extra-intestinal infections. FEMS Microbiol. Lett. 14:69-74. Schrotz-King, P., T. A. Prokhorova, P. N. Nielsen, J. S. Crawford, and C. Morsczeck. 2007. Campylobacter jejuni proteomics for new travellers’ diarrhoea vaccines. Travel. Med. Infect. Dis. 5: 106-1 09. Scott, D. A. 1997. Vaccines against Campylobacter jejuni.]. Infect. Dis. 176(Suppl. 2) :S 183-S 188. Shea, S. M. 1971. Lanthanum staining of the surface coat of cells. Its enhancement by the use of fixatives containing Alcian blue or cetylpyridinium chloride. J. Cell Biol. 51:611-620. Sizemore, D. R., B. Warner, J. Lawrence, A. Jones, and K. P. Killeen. 2006. Live, attenuated Salmonella typhimurium vectoring Campylobacter antigens. Vaccine 24:3793-3 803. Snelling, W. J., M. Matsuda, J. E. Moore, and J. S. Dooley. 2005a. Campylobacter jejuni. Lett. Appl. Microbiol. 41:297-302. Snelling, W. J., J. P. McKenna, D. M. Lecky, and J. S. Dooley. 2005b. Survival of Campylobacter jejuni in waterborne protozoa. Appl. Environ. Microbiol. 715560-5571. Spinosa, M. R., C. Progida, A. Tala, L. Cogli, P. Alifano, and C. Bucci. 2007. Neisseria meningitidis capsule is important for intracellular survival in human cells. Infect. Immun. 75:35943603. Stintzi, A. 2003. Gene expression profile of Campylobacter jejuni in response to growth temperature variation. 1. Bacteriol. 185: 2009-2016. Stintzi, A., D. Marlow, K. Palyada, H. Naikare, R Panciera, L. Whitworth, and C. Clarke. 2005. Use of genome-wide expression profiling and mutagenesis to study the intestinal lifestyle of Campylobacter jejuni. Infect. Immun. 73:1797-18 10.
C. TETUNI CAPSULAR POLYSACCHARIDE
521
Stintzi, A., and L. Whitworth. 2003. Investigation of the Campylobacter jejuni cold-shock response by global transcript profiling. Genome Lett. 2:18-27. Su, J., J. Yang, D. Zhao, T. H. Kawula, J. A. Banas, and J. R. Zhang. 2007. Genome-wide identification of Francisella tularensis virulence determinants. Infect. Immun. 75:3089-3 101. Sussman, M. 1997. Capsules of Esherichia coli, p. 114-143. In M. Sussman (ed.), Esherichia coli: Mechanisms of Virulence. Cambridge University Press, Cambridge, United Kingdom. Szymanski, C. M., F. St Michael, H. C. Jarrell, J. Li, M. Gilbert, S . Larocque, E. Vinogradov, and J. R. Brisson. 2003. Detection of conserved N-linked glycans and phase variable lipo-oligosaccharides and capsules from Campylobacter cells by mass spectrometry and high resolution magic angle spinning NMR spectroscopy.]. Biol. Chem. 278:24509-24520. Taylor, C.M, and I. S. Roberts. 2005. Capsular polysaccharides and their role in virulence. Contrib. Microbiol. 1255-66. Trachoo, N., J. F. Frank, and N. J. Stern. 2002. Survival of Campylobacter jejuni in biofilms isolated from chicken houses. ]. Food Prot. 65:lllO-1116. Tzianabos, A. O., and D. L. Kasper. 2002. Role of T cells in abscess formation. Cum. Opin. Microbiol. 5:92-96. Walker, R. I. 2005. Considerations for development of whole cell bacterial vaccines to prevent diarrheal diseases in children in developing countries. Vaccine 16:3369-3385. Wassenaar, T. M., B. N. Fry, A. J. Lastovica, J. A. Wagenaar, P. J. Coloe, and B. Duim. 2000. Genetic characterization of Campylobacter jejuni 0 : 4 1 isolates in relation with Guillain-Barri syndrome. J. Clin. Microbiol. 38:874-876. Whitfield, C. 2006. Biosynthesis and assembly of capsular polysaccharides in Escherichia coli. Annu. Rev. Biochem. 75:39-68. Whitfield, C., and I. S. Roberts. 1999. Structure, assembly and regulation of expression of capsules in Escherichia coli. Mol. Microbiol. 31: 1307-13 19. Woodall, C. A., M. A. Jones, P. A. Barrow, J. Hinds, G. L. Marsden, D. J. Kelly, N. Dorrell, B. W. Wren, and D. J. Maskell. 2005. Campylobacter jejuni gene expression in the chick cecum: evidence for adaptation to a low-oxygen environment. Infect. Immun. 735278-5285. Wyszynska, A., A. Raczko, M. Lis, and E. K. Jagusztyn-Krynicka. 2004. Oral immunization of chickens with avirulent Salmonella vaccine strain carrying C. jejuni 72Dz192 cjaA gene elicits specific humoral immune response associated with protection against challenge with wild-type Campylobacter. Vaccine 22: 1379-1389. Yi, K., A. W. Rasmussen, S. K. Gudlavalleti, D. S. Stephens, and I. Stojiljkovic. 2004. Biofilm formation by Neisseria meningitidis. Infect. Immun. 72: 6132-613 8. Yun, M., S. Yoon, Y. Shin, K. H. Chun, and J. E. Nam Shin. 2004. Synthesis of 2’-azidoethyl trisaccharide, a-~-Gal-(1-2)-6deoxy-cr-~-altro-Hepp-(l+3)-P-~-GlcNAc, an 0-antigenic repeating unit of C. jejuni 0:23 and 0:36. Arch. Pharm. Res. 27: 143-150. Ziprin, R L., M. E. Hume, C. R. Young, and R. B. Harvey. 2002. Inoculation of chicks with viable non-colonizing strains of Cumpylobacter jejuni: evaluation of protection against a colonizing strain. Curr. Microbiol. 44:221-223.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 29
Campylo bacter Metabolomics EVELYN C.
SOO,
DAVIDJ. MCNALLY, JEAN-ROBERT BRISSON, AND CHRISTOPHER W. REID
Oresic et al., 2006; Sabatine et al., 2005). Although recent impressive advances in mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy technologies have greatly enhanced the level of resolution and sensitivity that can be achieved for more comprehensive metabolite analyses to be performed, these technologies still do not provide the capacity to unequivocally characterize all the metabolites in a biological system. One approach to better understanding metabolism in a defined biological system, therefore, is to use a focused approach and extract from the metabolome information specific to a metabolic pathway under study. By targeting a specific metabolic pathway, it would be possible to identify the class or classes of metabolites that are known or suspected to be involved, thereby decreasing the complexity of the metabolome and substantially reducing the number of metabolites to be studied. This in turn allows a more rational choice of analytical methods to be used and selective analytical strategies to be developed according to the metabolites that are targeted. Targeted metabolite analysis is by no means a new area of research; it has been widely used in clinical applications-for example, for the determination of plasma glucose levels for the clinical diagnosis of diabetes (Kuzuya, 2000; Kuzuya et al., 2002). However, the novel aspect of targeted metabolomic studies is its use to identify unknown metabolites that have direct involvement in a poorly characterized pathway and the potential to investigate the influence of genetic and/ or environmental perturbations on the biosynthesis of these metabolites. As discussed in detail in chapter 26, there has been considerable interest in the characterization of the novel glycan modifications on Campylobacter flagellin and in the elucidation of unknown functions
Metabolomics is essentially the analysis of the total complement of low-molecular-weight molecules (metabolites) in a specific biological system under defined conditions (Fiehn, 2002; Oliver et al., 1998). Over the last few years, there has been a tremendous growth in the use of metabolomics as an additional postgenomic tool to gain more comprehensive understanding of cellular processes and biological systems. This surge in interest in metabolomics is mainly because of its promise to deliver a novel approach to better understand poorly characterized metabolic pathways and the potential to characterize the impact of selected genetic and/or environmental perturbations on metabolism in a defined biological system. The ultimate goal in metabolomics is to achieve unbiased identification and quantification of all the metabolites in a defined biological system. However, the metabolome is an inherently complex biological system in terms of the vast number of compounds that are present, the tremendous variation in the physicochemical properties of the metabolites, and the presence of many of the metabolites at trace-level concentrations (Goodacre et al., 2004). Presently, there is no single analytical technique that provides sufficient resolution and sensitivity for the qualitative and quantitative analysis of all the metabolites in a biological system. Nevertheless, comprehensive metabolomic analysis has found wide application in the study of plants (De Vos et al., 2007; Dixon et al., 2006; Oksman-Caldentey and Saito, 2005; Schauer and Fernie, 2006; Ward et al., 2007) and more recently expanded in the drug development arena to evaluate drug efficacy (Harrigan and Yates, 2006; Lindon et al., 2007; van der Greef et al., 2006) and drug toxicity (Griffin and Bollard, 2004), in nutrition studies (German et al., 2005; Gu et al., 2007), and for biomarker discovery (Jordan and Cheng, 2007;
Evelyn C. So0 NRC-Institute for Marine Biosciences, 1411 Oxford St., Halifax, NS, Canada B3H 321. Jean-Robert Brisson and David J. McChristopher W. Reid NRC-Institute for Biological Sciences, 100 Sussex Dr., Ottawa, ON, Canada K1A OR6. Nally The Lash Miller Chemical Laboratories, Department of Chemistry, University of Toronto, 80 St. George St., Toronto, Ontario MSS 3H6, Canada.
523
of genes involved in the flagellin glycosylation process. Much of the earlier work on the identification of novel glycan modifications in Campylobacter was performed by means of a bottom-up approach where the flagellin protein is tryptically digested and the glycopeptides analyzed by capillary liquid chromatography-electrospray mass spectrometry (ESMS) and tandem MS (Logan et al., 2002; Thibault et al., 2001). More recently, the use of top-down analysis of intact flagellin protein has offered a more complete analysis of posttranslational modifications of Campylobacter flagellin and identified several additional glycans (McNally et al,, 2007; Schirm et al., 2005). Although the use of bottom-up and top-down MS analyses has yielded much structural information on the novel flagellar glycans of Campylobacter, the precise determination of the anomeric configuration and structure of these carbohydrates requires NMR analysis of the purified glycopeptides. Given the limited sensitivity offered by current NMR technologies, such detailed structural analysis of the novel flagellar glycans can be difficult to achieve (Logan et al., 2002). In addition, despite the considerable amount of work that has been performed on the functional characterization of genes on the flagellin glycosylation locus of Campylobacter (chapter 26), there still remains a lack of information on the precise roles of the genes and the identity of substrates involved in the biosynthesis of the novel glycan modifications. Much of the work in bacterial metabolomics has involved the study of well-established metabolic pathways such as the tricarboxylic acid cycle, glycolysis, and specific metabolic pathways of microorganisms used in industrial applications (Kern et al., 2007; Mashego et al., 2007; Soga et al., 2002; Villas-BGas and Bruheim, 2007). In contrast, the field of Campylobacter metabolomics is very much in its infancy, and considering the lack of information on many of the novel glycoconjugate biosynthesis pathways in Campylobacter, there is much scope to use targeted metabolomics approaches to further define the substrates and genes involved in these metabolic pathways.
(CE) and electrospray ionization MS was developed to target substrates suspected to be involved in the
biosynthesis pathway (So0 et al., 2004). Pse5Ac7Ac is a nine-carbon sugar that shows much structural similarity to N-acetylneuraminic acid (NeuSAc), a sugar that is commonly found in prokaryotes and for which the biosynthesis pathway is well defined (Angata and Varki, 2002; Vimr et al., 2004). In addition to this structural similarity, there is significant homology between the genes involved in their biosynthesis, and therefore, it would not be surprising if the biosynthesis of PseSAc7Ac and its related derivatives would resemble that of NeuSAc (Guerry et al., 1996; Parkhill et al., 2000). Therefore, in designing a focused metabolomics approach to study PseSAc7Ac biosynthesis in C. jejuni, knowledge of the NeuSAc biosynthesis pathway in prokaryotes was used to propose a target class of metabolites that would most likely be involved. As shown in Fig. 1, the biosynthetic pathway of NeuSAc in prokaryotes involves the biosynthesis of its CMP-activated precursor before its transfer to the acceptor substrate. Similarly, it was expected that the biosynthesis of the CMPactivated precursor of PseSAc7Ac would be a key feature of the PseSAc7Ac biosynthesis pathway. As such, CMP precursors within cell lysates of C. jejuni were considered primary targets in the development of a focused metabolomics approach to study PseSAc7Ac biosynthesis. In addition to the CMP precursors, however, other nucleotide-linked precursors were considered highly relevant to the PseSAc7Ac biosynthesis pathway because the biosynthesis of NeuSAc also requires UDP-CPD-G~CNAC as one of the starting metabolites. Therefore, rather than targeting for CMP sugars alone, the analytical strategy was ex-
UDP-cL-D-GlcNAc NeuCl ManNAc N e u B k
FLAGELLIN GLYCOSYLATION PATHWAY IN C. JEJUnrI 81-176 Pseudaminic acid (5,7-diacetamido-3,5,7,9tetradeoxy-L-glycero-L-manno-nonulonsonic acid, PseSAc7Ac) and its derivatives were recently identified as the novel carbohydrate moieties glycosylating the flagellin of C. jejuni 81-176 (Thibault et al., 2001). In an effort to probe the biosynthesis of PseSAc7Ac in C. jejuni 81-176 by metabolomics, an analytical method that uses capillary electrophoresis
NeuSAc Neu
UDP-sugar Cj 1293
sialyltransferasei
+
Acceptor substrate
r
C6-sugar Cj1317
PseSNAc7NAc Cj1311
CMP-NeuSAc
1 1
CMP-PseSNAc7NAc sialyltransferase
i Flagellin
Figure 1. Prokaryotic biosynthetic pathway of N-acetylneuraminic acid (Neu5Ac) and the corresponding putative biosynthetic pathway of pseudaminic acid (Pse5Ac7Ac).
CHAPTER 29
panded to probe the metabolome for a variety of sugar nucleotides. CE was considered the most appropriate separation tool for the analysis of sugar nucleotides because it is an analytical technique that is particularly suited to the separation of polar ionic compounds. Good resolution of a mixture of CMP, UDP, ADP, and GDP sugars standards was achieved, and detection of the sugar nucleotides by electrospray ionization MS afforded excellent sensitivity (So0 et al., 2004). However, in order to enhance selectivity and expand the capability of the method to detect for unknown sugar nucleotides within a metabolome, a novel approach that uses MS with precursor ion scanning for fragment ions characteristic of the nucleotide carriers was used. In this way, it was possible to look for unknown sugar nucleotides on the basis of the nucleotide carrier rather than looking for a specific sugar nucleotide, for which its mass-to-charge ratio (mlz) must be known. The precursor ion scanning methodology has two main advantages. First, it does not require prior knowledge of the precise nature, and hence mlz, of the sugar nucleotide to be detected, which is particularly useful when studying poorly characterized biosynthetic pathways. Second, by selecting for the nucleotide carriers, the method is capable of picking up all sugars that are linked to each of the nucleotide carriers and not only the single metabolite in question. This expands the number of metabolites that can be detected simultaneously, and by performing tandem MS experiments on the novel metabolites, structural information could readily be achieved to ascertain their identities. By using this focused metabolomics approach to study PseSAc7Ac biosynthesis in C. jejuni 81-176, an intracellular pool of CMP sugars was observed by the CE-ESMS and precursor ion scanning method (So0 et al., 2004). On the basis of the m l z of the CMP sugars, it was suspected that these metabolites were CMP-linked NeuSAc, PseSAc7Ac and its acetamidino derivative, PseAm. Tandem MS experiments confirmed these to be the identities of the substrates, and the study provided for the first time in vivo evidence that the biosynthesis of PseSAc7Ac and the related acetamidino derivative involved their CMP-linked precursors. Surprisingly, however, the analysis of wild-type C. jejuni 81-176 for other nucleotidelinked sugars failed to reveal other substrates relating to the biosynthesis of PseSAc7Ac and related derivatives. UDP-CX-D-G~CNAC is known to be involved in the biosynthesis of NeuSAc, and therefore, it was expected to find a UDP-linked C6 sugar within the cell lysates of wild-type C. jejuni 81-176 of relevance to PseSAc7Ac biosynthesis. However, because there is no information on how the PseSAc7Ac biosynthesis
CAMPYLOBACTER METABOLOMICS
525
pathway is regulated, it is plausible that these UDPlinked intermediates may become rapidly used by other carbohydrate biosynthesis pathways within the cell. As mentioned earlier, one of the novel application areas of metabolomics is the study of the effect of genetic perturbations on metabolism. Because the precise role of many of the genes on the flagellin glycosylation locus in C. jejuni 81-176 is unknown, there was much opportunity to study selected isogenic mutants of individual genes from the flagellin glycosylation locus and examine the impact of the insertional inactivation of the individual genes on the biosynthesis of PseSAc7Ac. As already indicated in chapter 26, such recent metabolomics studies have yielded invaluable information on PseSAc7Ac biosynthesis. The first such metabolomic study was focused on Cj1293, Cj1311, and Cj1317 (Fig. 1, Table l), which were known to affect motility of C. jejuni 81176 and were implicated in the flagellin glycosylation process (So0 et al., 2004). By screening the three selected isogenic mutants by CE-ESMS and the precursor ion scanning method, it was shown that the CMP precursors of PseSAc7Ac and PseAm were no longer present in the metabolome of these defined isogenic mutants, providing significant proof that the three genes had a direct role in the biosynthesis of the CMP-linked precursors. In addition, two unknown UDP-linked sugars were identified in isogenic mutants Cj1311 and Cj1317, which through tandem MS experiments were identified as UDP-mono and diacetamido-trideoxyhexose. Because these novel UDP sugars had accumulated as a result of insertional inactivation of Cj1311 and Cj1317, it was speculated that these metabolites were potentially novel substrates of the PseSAc7Ac biosynthetic pathway. Given the initial success of the use of focused metabolomics approaches to identify novel biosynthetic substrates relating to PseSAc7Ac biosynthesis and the identification of a role for three genes on the flagellin glycosylation locus of C. jejuni 81-176 in the biosynthesis of the CMP precursors of PseSAc7Ac and its related acetamidino derivative, a more extensive study of the flagellin glycosylation locus in C. jejuni 81-176 was initiated (McNally et al., 2006a). It was also decided in this subsequent study to explore the potential of the use of the metabolome as a source for isolating novel biosynthetic substrates of novel glycan modifications for precise structural characterization by NMR spectroscopy. Up until this study, MS had been heavily relied on in the metabolomics study for obtaining structural information on novel biosynthetic intermediates detected in C. jejuni 81-176, and MS does offer the ultimate in terms of sensitivity. However, in order to determine the
526
SOOETAL.
Table 1. Metabolomic analysis of mutants of NeuSAc biosynthesis gene homologues in C. jejuni 81-176 StrainiCj n0.O
Gene annotation
Phenotype of mutantC
Intracellular sugar nucleotidesd
81-176 1293 1311 1317
pseB pseF (homolog of neuB) pseI (homolog of neuA)
Motile Nonmotile Nonmotile Nonmotile
CMP-PseSAc7Ac, CMP-PseAm, CMP-NeuSAc None UDP mono and di-acetamido-2,4,6-trideoxyhexose UDP mono and di-acetamido-2,4,6-trideoxyhexose
“Cj number refers to the gene showing highest homology in C. jejuni 11168 and in which the mutation was made in C. jejuni 81-176. bFrom Guerry et al. (2006). ‘Phenotype as determined by motility on agar gel (Guerry et al., 2006). dFrom So0 et al. (2004).
precise anomeric configuration and structure of the sugar nucleotides, it is necessary to use NMR spectroscopy, and in contrast to MS, NMR spectroscopy is a relatively insensitive analytical technique with sample requirements in the microgram concentration range, which is 1,000 times greater than that required for MS. Nonetheless, the precise anomeric configuration and structure of sugars can only be accomplished through NMR spectroscopy, and this further validates the need to use more than one analytical technique to comprehensively study the metabolome. In this extended study of the genes on the flagellin glycosylation locus of C. jejuni 81-176 for a role in the biosynthesis of CMP-PS~SAC~AC, it was found that in addition to the previously identified genes Cj1293 (pseB), Cj1311 (pseF), and Cj1317 (pse1) (So0 et al., 2004), three other genes annotated as pseC, G, and H also played a direct role in the biosynthesis of CMP-PseSAc7Ac and CMP-PseAm in C. jejuni 81-176 (McNally et al., 2006a). These data supported the notion that flagellin glycosylation is necessary for motility because earlier studies of the corresponding isogenic mutants revealed a nonmotile phenotype (Logan et al., 2002). It is noteworthy that two further genes, pseA and pseC, displayed unique metabolomic profiles (Fig. 2). In earlier studies, an exclusive role had been identified for pseA in the biosynthesis of the acetamidino derivative of pseudaminic acid (PseAm) in C. jejuni 81-176 (Logan et al., 2002). The CE-ESMS and precursor ion scanning study of the corresponding isogenic mutant provided corroborating evidence for an exclusive role of pseA in PseAm biosynthesis because the insertional inactivation of this gene resulted in the ability of C. jejuni 81-176 to synthesize CMP-PseAm but had no effect on CMP-PseSAc7Ac biosynthesis. This observation was highly encouraging because it further demonstrated the potential of metabolomics to link genotypic and phenotypic data. With regard to pseC, the isogenic mutant displayed a unique metabolomic profile in that a single UDP sugar had accumulated, and this was shown by tandem MS to be UDPdiacetamido-2,4,6-trideoxyhexose.It was suspected
that this UDP sugar was important to the biosynthesis of PseSAc7Ac because recent functional characterization of the PseC revealed its role as a 4-aminotransferase in the production of UDP-6-deoxy-PL - A ~ ~ N A c ~from N UDP-4-keto-4,6-dideoxy-P-~AltNAc (Schoenhofen et al., 2006a). Therefore, the accumulation of this novel UDP-diacetamido-trideoxyhexose sugar as a result of insertional inactivation of pseC could be highly relevant in terms of a novel precursor preceding this step. In order to determine whether this novel UDPdiacetamido-2,4,6-trideoxyhexosesugar could be a plausible metabolite in the PseSAc7Ac biosynthesis pathway, it would be necessary to gain unequivocal structural information on the metabolite. As mentioned earlier, in order to determine the anomeric configuration and structure of carbohydrates, it is necessary to use NMR spectroscopy, and therefore, sufficient amounts of the metabolite must be isolated from isogenic mutant pseC. In considering this need to isolate UDP-diacetamido-2,4,6-trideoxyhexose, it would be highly desirable if the analytical strategy was generic and amenable to the isolation of other novel sugar nucleotides. For example, an analytical strategy that could also be readily applied to the isolation of CMP-PseAm from C. jejuni 81-176 for precise structural determination of the acetamidino derivative by NMR would be invaluable. Although CE-ESMS and precursor ion scanning is a highly selective and sensitive technique to detect novel, intracellular sugar nucleotides from the Campylobucter metabolome, the technique is not suitable for largescale isolations of biosynthetic intermediates because nanoliter volumes are involved. Liquid chromatography, on the other hand, would be better suited for large-scale purifications of the metabolites, and the use of hydrophilic interaction liquid chromatography (HILIC) in conjunction with MS and precursor ion scanning provided an excellent generic strategy for isolating sugar nucleotide intermediates (CMPPseAm and UDP-diacetamido-2,4,6-trideoxyhexose) from Campylobucter for structural elucidation by NMR spectroscopy (McNally et al., 2006a).
a) C.jejuni 81-176
c ) pseA
b) pseC 631.4 UDP-diacetamido-2,4,6-trideoxyhexose
637.4 CMP-P e5Ac7Am
P
638.4 CMP-Pse5Ac7Ac I
659.4
Id
,
589.
300 400 500 600 700 mlz
1660.6
300 400 500 600 700 800 mlz
300 400 500 600 700 800 mlz
4.0e6
m Q
3.0e6
3.0e6
v
~
3.0e5
Q m
i
=x m
5 2.0e6
-
2
4
5 2
c 2.0e5 $
6.71
1.Oe6
1.Oe6
1.Oe5
9.44
1 ,.-. . ~
0.0
0.0
2
4
6 8 Time (min)
Figure 2. CE-ESMS and precursor ion scanning for fragments related to CMP (mlz 322) or UDP (mlz 3 8 5 ) activated sugars in cell lysates. (a) Wild-type C. jejuni 81-176. (b) Isogenic mutant pseC. (c) Isogenic mutant pseA.
1 0 1 2
528
SOOETAL.
One- and two-dimensional NMR techniques have found wide applications in metabolomic investigations of animal, plant, fungal, and most recently bacterial systems. NMR is particularly well suited for metabolomics research because it is a rapid, highly reproducible, quantitative, and nondestructive analytical method. Although NMR may never rival MS in terms of sensitivity, advances in magnet technology (900 MHz magnets are now commercially available) and probe technology such as cryogenically cooled probes (cold probes) and ultra-low-volume flow probes (5-pl coil volume) have increased the sensitivity of NMR into the range needed for metabolomics research. Furthermore, NMR spectroscopy can be highly automated through the use of robotics such as sample changers and automatic tuning systems, thus lowering the cost of analysis per sample. These recent innovations have made NMR highly amenable and cost-effective for the high-throughput applications that are commonly encountered in metabolomics research. A review of the use of NMR in metabolomics research is beyond the scope of this chapter, and interested readers are directed to the literature (Dunn et al., 2005; Ellis et al., 2007; Fiehn, 2002; Griffin, 2003, 2004; Holmes et al., 2006; Lenz and Wilson, 2007; Lindon et al., 2007; Reo, 2002; Robertson et al., 2007; Serkova and Niemann, 2006; Ward et al., 2007). This section will instead focus on the methods that have been developed for the metabolomics investigation of glycan biosynthetic pathways in C. jejuni. The main challenges associated with the study of sugar nucleotide metabolites by NMR have been the instability of the sugar nucleotides and their presence at low concentrations within the bacterial cells. In order to overcome these challenges, two approaches-the in vivo metabolomics approach and the in vitro metabolomics approach-have been developed. These methods are practical and complementary; they are based on a NMR spectroscopy platform and were developed largely through a process of trial and error during the last few years. The NMR methods were also designed to provide complementary data to that provided by MS-based approaches. The in vivo approach involves both MS and NMR and relies on the use of CE-ESMS and precursor ion scanning (So0 et al., 2004) for initial screening of bacterial cell lysates for sugar nucleotide metabolites, followed by the large-scale purification of the targeted metabolites by liquid chromatography methods such as HILIC, as described earlier. Generally speaking, approximately 10 to 30 p g of pure metabolites can be isolated via HILIC from 10 liters of C. jejuni cell culture (McNally et al., 2006a). A variety of one- and two-dimensional homo- and heter-
onuclear NMR experiments can then be used to elucidate the structure of the metabolites and provide insight into gene function and the glycan biosynthetic pathway of interest. As a general method, purified metabolites are lyophilized and resuspended in a minimal amount of D,O and analyzed by NMR (McNally et al., 2006a). Sensitivity is almost always an issue for NMR studies of bacterial metabolites because the quantity of available purified metabolites can be small as a result of their low concentration within the cell. Typically, a Varian 600 MHz ('H) spectrometer equipped with a cold probe for optimal sensitivity has been used for such metabolomic studies (McNally et al., 2006a, 2007). Use of a higher magnetic field spectrometer is advantageous, not only because of the heightened sensitivity, but also because of greater spectral dispersion, which is important in NMR studies of carbohydrates where spectra are often crowded with overlapping signals. Fortunately, recent developments in probe technology such as cold probes have increased NMR instrument sensitivity by a factor of 2 to 4. This gain in sensitivity is tantamount to having two to four times the amount of metabolite available for NMR analysis. A cold probe provides greater sensitivity by reducing the overall noise contribution from external factors such as coils and electronics, which are the principle sources of noise in NMR. In fact, cold probes reduce the amount of noise originating from these sources so that the main noise contribution is from the sample itself. This reduction in noise results in a gain of usable NMR signal originating from the metabolites within the sample and in a significant increase in the signal-to-noise ratio (S/N). The increased sensitivity of a cold probe makes acquisition of data possible in a reasonable amount of time on samples as low as 10 pg of sugar nucleotides. Such an in vivo approach recently revealed the structure of the unknown UDP-linked sugar in isogenic mutant pseC as UDP-2,4-diacetamido-2,4,6trideoxy-a-D-Glc (UDP-CI-D-QU~NAC~NAC). Only 10 p g or 16 nmol (0.08 M) of the purified metabolite was required for the precise structural analysis (McNally et al., 2006a). Figure 3a shows the proton spectrum of metabolite sample that was analyzed with a 600 MHz (lH) instrument with a cold probe, which gave a S/N value of 1 O O : l . In contrast, Fig. 3b shows the same sample that was analyzed at 500 MHz with a standard probe, which gave a S/N of 39:l. In this case, the higher magnetic field instrument with the cold probe provided more than twice the S/N, thereby facilitating interpretation of the spectrum and structural elucidation of the metabolite. Cold probes do have their limitations, however. For
CHAPTER 29
F
H3C
CAMPYLOBACTER METABOLOMICS
529
HO
0A C H ,
1
1
1
I
5.5
I
I
I
I
5.0
I
I
I
I
4.5
I
I
I
I
4.0
I
I
I
I
I
I
I
I
3.0
3.5
l
l
1
1
2.5
1
1
1
1
2.0
1
1
1
1
I.5
1
1
1
1.o
’H (PPW Figure 3. Comparison of the ratio of signal to noise (S/N) for a UDP-6-deoxy-a-~-GlcNAc4NAc metabolite sample isolated from C. jejuni 81-176 showing the advantages of the use of a higher-field NMR spectrometer and cold probe. (a) ‘H NMR spectrum acquired at 600 MHz using a cold probe (S/N = 1OO:l). (b) ‘H NMR spectrum acquired at 500 MHz using a conventional 3-mm probe (SIN = 39:l). For (a) and (b), the metabolite sample was analyzed in a 3-mm tube (200 ~ 1 9 9 % D,O) with 128 scans.
instance, the presence of salt, such as those found in most biological buffers, can reduce the S/N. Salt can be especially problematic in the NMR analysis of sugar nucleotide metabolites, which generally need to be buffered because of their instability at p H values slightly below neutrality. However, if the sample can be concentrated and placed into a 3-mm NMR tube (150 to 200 p1) instead of a 5-mm tube (500 pl), the negative effects associated with salt can be reduced. Further, use of a 3-mm NMR tube also improves shimming (shimming improves signal resolution by increasing magnetic field homogeneity so that every
part of the sample in the NMR tube experiences exactly the same magnetic field), reduces spurious signals from the solvent, and decreases the intensity of the water signal, which also improves S/N. Before the advent of cold probe technology, use of a nanoprobe (a high-resolution magic angle spinning probe) had allowed structural analysis of metabolites in the 200 to 500 nmol range (0.3 to 0.5 mg) (Gilbert et al., 2000; Szymanski et al., 2003b; Thibault et al., 2001). Hence, the use of a cold probe decreased by more than a factor of 10 the amount of sample required for NMR analysis.
530
SOOETAL.
way is highly conserved among Cumpylobacter sp., it UDP-~-D-QU~NAC~NAC is an important metabolite in the 2,4-diacetamido-bacillosamine biosynthewas not surprising to observe this novel sugar on C. sis pathway (Schoenhofen et al., 2 0 0 6 ~ and ) ~ its accoli VC167 flagellin. In addition to PseSAc7AcYan cumulation in pseC had not been expected because it acetamidino derivative suspected to be “PseAm” was has been thought that the inactivation of pseC would also observed in the study to be glycosylating the flalead to an accumulation of a novel precursor directly gellin of C. coli VC167, but both structural and related to PseSAc7Ac biosynthesis. This observation serological evidence clearly suggested that this led to the speculation of a cross-talk between the 0- “PseAm” in C. coli VC167 was novel and distinct linked flagellin glycosylation and N-linked bacillosafrom that found in C. jejuni 81-176 flagellin (Logan mine biosynthesis pathways (McNally et al., 2006a), et al., 2002). The flagellin glycosylation locus of C. and as already mentioned in chapter 26, the inaccoli VC167 contains a ptm locus that is conserved tivation of pseC is believed to cause the diversion of among many Campylobucter strains but absent from the biosynthetic product of PseB to the 2,4-diaC. jejuni 81-176, and earlier work looking at the cetamido-bacillosamine pathway, leading to the acfunction of this ptrn gene locus in C. coli VC167 cumulation of UDP-6-deoxy-a-~-GlcNAc4NAc in isclearly implicated the ptm gene products in the bioogenic mutant pseC. synthesis of this novel acetamidino variant (Logan et For CMP-PseAm, it was possible through foal., 2002). As had been the case with C. jejuni 81cused metabolomics approaches that use HILIC-MS 176, there is limited information on the precise nato isolate sufficient quantities of the metabolite for ture of this novel acetamidino modification in C. coli precise structural determination by NMR. Cold VC167 and the genes involved in its biosynthesis, but probe technology was used once again for enhanced by focused metabolomics approaches, there was sensitivity; the CMP-PseAm metabolite was identified much opportunity to gain further insight into the flaas CMP-5-acetamido-7-acetamidino-3,5,7,9-tetrade- gellin glycosylation process of this pathogen. oxy-L-glycero-a-L-munno-nonulosonicacid (CMPThe CE-ESMS and precursor ion scanning of PseSAc7Am) (Fig. 4). The results of this work were wild-type C. coli VC167 for intracellular sugar nuparticularly important because it provided for the cleotides immediately revealed a difference in the first time the unequivocal identification of a second sugar nucleotide profile compared with that obtained glycosyl modification on C. jejuni 81-176 flagellin. for C. jejuni 81-176. On the basis of the knowledge The ability to elucidate the precise structure of of the presence of PseSAc7Ac and a related acetamPseSAc7Am in C. jejuni 81-176 flagellin via its coridino derivative on C. coli VC167 flagellin, it was not responding CMP-activated sugar revealed an exciting surprising to find during the CE-ESMS and precursor opportunity to investigate whether subtle differences ion scanning experiments a corresponding pool of in the configuration of related sugars on CampyloCMP-linked precursors. However, in addition to bacter flagellin could contribute toward the antigenic these expected CMP sugars, there was present an apdifferences observed between different Campylobacpreciable amount of an unknown CMP sugar at m l z ter strains (Logan et al., 2002). 651 (McNally et al., 2007). It was not known whether this novel CMP sugar corresponded to a modification on the flagellin or if it was involved in other glycoconjugate biosynthesis pathways, because FLAGELLIN GLYCOSYLATION PATHWAY IN earlier work on the structural characterization of the CAMPYLOBACTER COLI VC167 flagellin glycans in C. coli VC167 had not revealed the presence of this unknown carbohydrate modifiThere has been extensive work on the characcation (Logan et al., 2002). Because the CE-ESMS terization of novel glycan modifications on Cumpyand precursor ion scanning method is not exclusive lobacter coli VC167 flagellin and genes involved in to the detection of activated sugars relating to flagelthe glycosylation process (Logan et al., 2002). In the lin glycosylation, there was a possibility that this earlier studies, PseSAc7Ac was identified on the flagellin of C. coli VC167, and given that the pse pathCMP sugar could be related to other glycosylation
Figure 4. NMR spectroscopy of CMP-5-acetamido-7-acetamidino-3,5,7,9-tetradeoxy-~-gZyc~o-~-manno-nonulosonic acid (CMP-Pse5NAc7Am, I) purified from cell lysates. (a) ’H NMR spectrum (256 transients). (b) One-dimensional TOCSY of I H3ax (80 ms). (c) Onedimensional TOCSY of I H7 (80 ms). (d) One-dimensional TOCSY of I H9 (80 ms). (e) One-dimensional TOCSY of I NAc N H (30 ms). (f) One-dimensional TOCSY of I Am NH (30 ms). (g) The 13C HSQC spectrum (165 transients, 128 increments, lJC,H-150 Hz, 24 h). From McNally et al. (2006a).
CHAPTER 29
I
\
2or
531
li 3ax
HN
CH3
CAMPYLOBACTER METABOLOMICS
1-7
-
--
Am-CH,
0
- 1-9
NAc-CH,
30
e
P
I-3eq
-
I-3a~
0
1-5
.TEAA 1-7 9
1-4mmR-5 . f R-3,eR o 70. e 1-8 1-6 e R-2 80- , 4.5 4.0 .
i
l
l
*D
1
1
1
1
1
3.5
1
.
1
1
.
1
.
3.0
'H ( P P W
1
2.5
.
1
.
1
2.0
1
1
.
,
1.5
1
1
1
1.o
532
SOOETAL.
D-gulucto-nonulosonic acid (CMP-LegSAm7Ac). pathways in C. coli VC167. Therefore, to investigate The metabolite identified at mlz 638 during this further, a top-down analysis of C. coli VC167 flagellin was performed to determine whether this HILIC-MS of wild-type C. coli VC167 and that had novel modification had been overlooked in earlier been suspected to be related to PseSAc7Ac is likely bottom-up experiments, and the top-down analysis to be the corresponding diacetamido sugar, CMP5,7 - diacetamido - 3 ,5,7,9 - tetradeoxy - D -glycero - D clearly demonstrated the presence of the corresponding sugar as a novel flagellar glycan modification that gulacto-nonulosonic acid (CMP-LegSAc7Ac) (Mcwas not observed previously (McNally et al., 2007). Nally et al., 2007). Sugar nucleotides can be unstable molecules that As had been shown with C. jejuni 81-176, medegrade rapidly at pH values near neutrality, and this tabolomics offers much opportunity to investigate can be problematic because degraded metabolites unknown gene functions and identify novel biosyntend to contaminate samples and complicate the thetic substrates of poorly characterized biosynthetic interpretation of NMR spectra. Selective onepathways, and to elucidate precise structures of novel dimensional NMR experiments have been found to flagellin glycans through the structural analysis of be particularly useful for analyzing samples that contheir corresponding biosynthetic precursors. CEtain complex mixtures of sugar nucleotide comESMS and precursor ion scanning offers a rapid pounds (Brisson et al., 2002; Kneidinger et al., screening method for intracellular sugar nucleotides but greater selectivity is afforded with the use of 2003). One-dimensional NMR experiments such as HILIC as a separation technique, and as discussed the one-dimensional total correlation spectroscopy earlier, it offers more scope in terms of large-scale (TOCSY) experiment allow the selective excitation of isolations of metabolites compared with CE. During targeted signals in the NMR spectrum and reveal the sugar nucleotide analysis of C. coli VC167 by only those signals that are spin-correlated with the excited resonance. By means of this approach, comHILIC-MS and precursor ion scanning, the novel plex spectra can be simplified because one can extract CMP sugar at m l z 651 was found to be present in resonances originating exclusively from the metabothe metabolome; this is consistent with the observations made by CE-ESMS (Fig. 5a). The HILIC-MS lites of interest and avoid signals from contaminants. data also clearly revealed the absence of CMPBy means of selective one-dimensional NMR experPseSAc7Am; the notable difference in retention beiments, the novel CMP sugar identified in C. coli havior between CMP-Pse5Ac7Am and this “CMPVC167 was shown to be a mixture of novel derivaPseAm” metabolite on the HILIC column suggested tives of legionaminic acid and was subsequently idenphysicochemical differences between the two CMPtified as CMP-Leg5-E-(N-methylacetimidoyl)7Ac and linked acetamidino intermediates (Fig. 5a & 5b). CMP-LegS-Z-(N-methylacetimidoyl)7Ac (McNally et NMR structural characterization of “ C M P - P S ~ A ~ al., ~ ~ 2007). Together with molecular modeling in C. coli VC167 would yield invaluable information studies, the metabolites were unequivocally identified as CMP-5-E/Z-N-(N-methylacetimidoyl)-7on the identity of this novel metabolite and also would provide unequivocal structural characterizaacetamidino - 3,5,7,9 - tetradeoxy-D-glycero-D-gulactotion of the corresponding sugar, which had been obnonulosonic acid (CMP-ElZ-LegSAmNMe7Ac). The served on the flagellin. The HILIC-MS analysis also identification of the legionaminic acid and its derivrevealed, in addition to CMP-PS~SAC~AC, another atives in C. coli VC167 was highly unexpected, and CMP sugar at r n l z 638, which was thought likely to it was evident from structural elucidation of the novel be related to PseSAc7Ac. metabolites that the ptm genes, which had been thought to be involved in the biosynthesis of pseuLarge-scale purifications of “CMP-PseAm” and the novel CMP sugar (mlz 651) by HILIC-MS prodaminic acid derivatives, were in fact involved in levided microgram-level quantities of the individual gionaminic acid biosynthesis. Therefore, to probe the metabolites for structural determination by NMR function of the flagellin glycosylation genes in C. coli spectroscopy. Contrary to what had been expected, VC167, an extensive metabolomic study of isogenic the NMR data revealed the presence of a new class mutants of the ptm gene locus was performed, along of legionaminic acid sugars in C. coli VC167. The with the isogenic mutants in four additional open CMP sugar at rnlz 637, which had been suspected to reading frames to probe their function in the flagellin be a variant of “CMP-PseAm,” was identified as the glycosylation process. acetamidino derivative of legionaminic acid, with the HILIC-MS and precursor ion scanning was used acetamidino group at the C5 position, as opposed to to determine changes in the complement of CMP the C7 position, as in pseudaminic acid. The metabsugars in C. coli VC167 after insertional inactivation olite was unequivocally identified as CMP-5-acetaof individual genes in the ptm gene locus and the midino-7-acetamido-3,5,7,9 -tetradeoxy-D-glycerogenes in addition to four open reading frames (hom-
m
Li
n
3
CHAPTER 29
d
CAMPYLOBACTER METABOLOMICS
_=
I
9 0
d hl
d
18
533
534
SO0 ETAL.
ologs of Cj1319, Cj1320, Cj1324, and Cj1325) as a means to determine their role in the biosynthesis of legionaminic acid and their derivatives (McNally et al., 2007). Mutations of the ptm gene locus (ptmAF) and study of the corresponding isogenic mutants confirmed a role for these genes in the biosynthesis of the CMP precursors of LegSAc7Ac, LegSAm7Ac, and EIZ-LegSAmNMe7Ac. The absence of CMPLegSAm7Ac in the ptm isogenic mutants was consistent with findings from earlier work where the flagellin of the mutants was no longer glycosylated with an acetamidino derivative, which at the time was thought to be “PseAm” (Logan et al., 2002) but which is now known to be LegSAm7Ac (McNally et al., 2007). Although the flagellin of these ptm mutants was not examined for the absence of LegSAc7Ac and EIZ-Leg5AmNMe7Acy there is little doubt that the biosynthesis of these legionaminic acid sugars also involves the ptm gene locus. Complementation of each of the ptm genes in trans restored the ability of C. coli VC167 to synthesize the CMPlinked legionaminic acid derivatives. The analysis of isogenic mutant Cj1324 revealed for the first time a functional role of this gene in the biosynthesis of CMP-LegSAm7Ac and CMP-EIZ-LegSAmNMe7Ac (Fig. 5c). Given that Cj1324 encodes a protein that displays significant homology with PseA from C. jejuni 81-176, which has been demonstrated to be involved in the biosynthesis of CMP-PseSAc7Am from CMP-Pse5Ac7AcYit is likely that Cj1324 plays a similar role in the biosynthesis of CMP-LegSAm7Ac and CMP-EIZ-LegSAmNMe7Ac from CMP-LegSAc7Ac. By probing the function of Cj1324 in the metabolomic study, it was possible to annotate Cj1324 as ptmG. With regard to the insertional inactivation of Cj 1325, the metabolomics data revealed that Cj1325 only had a direct impact on the biosynthesis of CMPEIZ-LegSAmNMe7Ac, and therefore Cj1325 was annotated as ptmH. In contrast to ptmG and H , insertional inactivation of Cj1319 and Cj1320 had no impact on the ability of C. coli VC167 to synthesize any of the CMP precursors, which demonstrated that these two genes are not involved in the biosynthesis of PseSAc7Ac or the legionaminic acid sugars (McNally et al., 2007). In addition to the above-mentioned genes, the effect of knocking out the pseB gene in C. coli VC167 was also investigated in the study (McNally et al., 2007). PseB is known to be directly involved in the pseudaminic acid biosynthesis pathway, and metabolomic analysis of the corresponding isogenic mutant confirmed its role in C. coli VC167. The biosynthesis of CMP-PseSAc7Ac was affected by the insertional inactivation of pseB, but C. coli VC167 retained the capacity to synthesize the CMP precur-
sors of the legionaminic acid, supporting the notion that the biosynthesis of pseudaminic acid and legionaminic acid sugars involve separate pathways. When both the pseB and ptmD genes were inactivated, the biosynthesis pathways of both CMP-PseSAc7Ac and the CMP precursors legionaminic acid sugars were absent from the metabolome. It was speculated that the enzyme products of ptmG and pseA were specific to the biosynthesis of CMP-LegSAm7Ac and CMP-Pse5Ac7AmY respectively, and this observation was also confirmed in this metabolomic analysis of C. coli VC167. Complementation of the ptmG with the pseA gene from C. jejuni 81-176 resulted in the exclusive conversion of CMPPseSAc7Ac to CMP-Pse5Ac7AmYand there was no evidence of the biosynthesis of CMP-LegSAm7Ac from CMP-LegSAc7Ac as a result of this complementation. These data once again provided evidence that the pseudaminic acid and legionaminic acid biosynthesis pathways were distinct and confirms a role for the ptm gene locus in legionaminic acid biosynthesis. The focused metabolomics studies of flagellin glycosylation in C. jejuni 81-176 and C. coli VC167 were extensive and examined unknown gene hnctions, characterized novel biosynthetic substrates and novel flagellar glycans, and elucidated poorly understood metabolic pathways. As shown, even with focused metabolomics studies, the development of highly selective analytical tools is crucial to critically assess the metabolome for subtle metabolite changes that allow unknown gene functions to be assigned, novel biosynthetic substrates to be identified, and unique metabolic pathways to be better understood. In addition, because of the complexity and uniqueness of flagellin glycosylation pathways in Campylobacter, a combination of both MS and NMR spectroscopy was a requirement to define the metabolic pathways. The metabolomic study of flagellin glycosylation has not only yielded important information on the biosynthetic process and highlighted the capacity of Campylobacter to synthesize the variety of novel sugars, but also demonstrated the value of metabolomics as a complementary postgenomic tool for interrogating complex biological systems.
IN VITRO METABOLOMICS ANALYSIS BY NMR SPECTROSCOPY During the recent characterization of the pseudaminic acid pathway, several novel UDP-hexos-4dose metabolites were identified and found to be extremely labile molecules that rapidly degraded into their gem-diol forms in solution (Schoenhofen et al., 2006c; McNally et al., 2006b). There is interest in
CHAPTER 29
o
%
~
HO
H,O
H&+JoH
L
7
'
2
HO HO H
H
*
H
-
NAc 0-UDP
HO
OH
i
PseC PseH ; PseG ; Psel I PseF I
H
HO
NAc 0-UDP
H
HO
535
et al., 2002). The sample can also be subjected to further analysis by one- and two-dimensional NMR techniques to complete the structural analysis. Product purification, which can lead to the degradation of labile compounds, is thus avoided. To characterize minor reaction products that are present in low abundance, a scaled-up reaction that has been performed concurrently on the bench can be examined by NMR with a cold probe. Figure 6 shows the PseB reaction, which is the first enzyme of the pseudaminic acid pathway. PseB is considered unique because it is a 4,6-dehydratase/ 5-epimerase that converts UDP-CX-D-G~C (step 1) to UDP-2-acetamido-2,6-dideoxy-~-~-urubino-hexos-4dose (step 2) (Schoenhofen et al., 2 0 0 6 ~ )By . using 'H NMR as a real-time analytical probe to continuously follow the PseB reaction, it was revealed that upon accumulation of step 2, PseB catalyzes an additional C5 epimerization forming UDP-2-acetamido2,6-dideoxy-a-~-xylo-hexos-4-ulose (step 3) (Schoenhofen et al., 2006c; McNally et al., 2006b). In C. jejuni, step 3 is used to make D-Qu~NAc~NAc,
determining the precise structure and ring conformation of nucleotide-activated hexos-4-ulose sugars because of their central role in the biosynthesis of lipopolysaccharide, capsular polysaccharide, and antibiotics in bacteria. In order to analyze these unstable sugar nucleotide metabolites with NMR, an in vitro metabolomics approach was developed. This approach involved performing an enzymatic reaction in the NMR tube and monitoring the progress of the reaction directly with NMR spectroscopy in real time (McNally et al., 2006b). Proton NMR (lH NMR) is particularly useful because it can be used as a realtime analytical probe to continuously follow enzymatic reactions directly in the NMR tube (Pellecchia, 2005; Pfoestl et al., 2003). By monitoring the enzymatic reaction directly with NMR, it can be stopped at any point in time by removal of the enzyme when the concentration of a desirable product is maximized. Selective one-dimensional NMR experiments can then be used to elucidate the structure of nucleotide-activated sugar metabolites in a mixture of compounds and to detect minor components (Brisson
UDP-a-D-QuiNAc4NAc ,-~g!ELPg!~--
CAMPYLOBACTER METABOLOMICS
NAc
4 0-CMP
CMP-Pse Figure 6 . First steps of the pseudaminic acid biosynthetic pathway in C. jejuni. From McNally (2006b). Pyranose rings are shown as their predominant chair conformation determined from nuclear Overhauser effects (NOES)and J(H,H) coupling constants. Step 1, UDP-~-D-GICNAC; step 2, UDP-2-acetamido-2,6-dideoxy-~-~-u~u6i~o-hexos-4-ulose; step 2', gem-diol form of 2; 3, UDP-2-acetamido-2,6-dideoxy-a-~-xylo-hexos-4-ulose; step 3', gem-diol form of 3 (from McNally et al., 2006b); step 4, CMP-PseSAc7Ac.
536
SOOETAL.
which is one of seven sugars that comprise the Nlinked glycan hepatasaccharide that modifies over 30 proteins in this bacterium (Szymanski and Wren, 2005; Young et al., 2002) and is further discussed in the next section. From Fig. 7, it can be seen that
during the first few minutes of the PseB reaction (Fig. 7a and b), a decrease in step 1 is observed, along with the simultaneous appearance of steps 2 and 2'. To characterize step 2 with NMR, a scaled-up PseB reaction was stopped when the concentration of step 2
1.c
0.8
- 0.6 !
8
CI
c
-
0.4
0.2 0.0 0 8
1
3 Time (h)
2
5
4
6
Time 0.08 h
2'
Time 6 h
h .
.... .... . . . . .... . . . .
5.90
5.80
5.70
I . . ' ' l . . . . I . . . . I . . . . I . . . . I . . .
5.60
5.50
5.40
'H ( P P m Figure 7. PseB reaction monitored directly with 'H NMR spectroscopy (500 MHz, 'H) in aqueous reaction buffer (25"C, 10% D,O, 25 mM UDP-CPD-G~CNAC, 225 pg of PseB, 25 mM NaPO,, 50 mM NaC1, pH 7.2). From McNally et al. (2006b). (a) Integrals for anorneric protons plotted over time. (b, c) 'H NMR spectra of the PseB reaction showing the anomeric region acquired at 0.08 and 6 h, respectively (from McNally et al., 2006b).
CHAPTER 29
was highest (Fig. 7c). The enzyme was removed and the sample was concentrated and exchanged into D,O (McNally et al., 2006b). Two-dimensional total correlation spectroscopy (2D-TOCSY), correlation spectroscopy (COSY), and nuclear Overhauser effect spectroscopy (NOESY) experiments as well as selective one-dimensional TOCSY and one-dimensional nuclear Overhauser effect spectroscopy (1D-NOESY) NMR experiments were then used to elucidate the structure of step 2. By using the in vitro metabolomics approach to analyze the PseB reaction directly with NMR, first structural description for the UDP-hexos-4-ulose sugars, steps 2 and 3 were achieved. The precise structural elucidation of these sugars led to a more complete understanding of the pseudaminic acid pathway in C. jejuni, complemented X-ray crystallographic studies, and provided invaluable information for the development of enzyme inhibitors as therapeutic agents (McNally et al., 2006b). Recently, the in vitro approach was used to analyze the entire pseudaminic acid pathway in a one-pot reaction within the NMR tube (Schoenhofen et al., 2006b). Currently the use of this NMR approach to screen the entire pseudaminic acid pathway for small molecule inhibitors as therapeutic agents is under evaluation.
CAMPYLOBACTER METABOLOMICS
537
N-LINKED PROTEIN GLYCOSYLATION PATHWAY IN C . JEJUNI NCTC 11168 The N-glycan of C. jejuni is a conserved heptasaccharide with the structure GalNAc-a1,4-GalNAc-
al,4-[Glc-~l,3]-GalNAc-al,4-GalNAc-al,4-GalNAcal,3-Bac2,4diNAc-pl, where Bac2,4diNAC is 2,4diacetamido-2,4,6-trideoxyglucopyranose (Young et al., 2002) and plays a major role in host adherence, invasion, and colonization (Karlyshev et al., 2004; Szymanski et al., 2002, 2003a). A more detailed description of the pathway is presented in chapter 25. The similarities between the N-linked protein glycosylation (Pgl) pathway in Campylobacter and the eukaryotic N-glycosylation machinery make it an ideal model system for investigating the intricacies of the lipid-mediated steps in the biosynthesis of the oligosaccharides (Fig. 8). Although the analysis and profiling of lipids is not new, the burgeoning interest in lipids, and by extension lipidomics, has been facilitated by the enhanced capabilities of modern MS instruments and interfaces, and an increase in the development of global lipid analytical methods (Bijlsma et al., 2006; Guan et al., 2006; Guerardel et al.,
'-
c
Figure 8. Comparison of N-linked protein glycosylation in (a) eukaryotes and (b) the bacterium, C. jejuni. In the eukaryotic system, the oligosaccharide is assembled on the cytoplasmic face of the endoplasmic reticulum (ER) on the polyprenyl carrier dolichol phosphate. It is then flipped to the luminal side of the ER, where it is transferred to acceptor proteins with the Ser/ Thr-X-Asn sequon by the oligosaccharytl transferase (OST) complex. In C. jejuni, the heptasaccharide is assembled on the cytoplasmic face of the cytoplasmic membrane on the polyprenyl carrier undecaprenyl phosphate. The completed heptasaccharide is then flipped to the periplasmic face of the membrane by PglK, where it is transferred to acceptor proteins with the extended sequon Ser/Thr-XpAsn-Glu/Asp by PglB.
538
SO0 ET AL.
2006; Hermansson et al., 2005; Levery, 2005; Houjou et al., 2005). Lipid-linked oligosaccharide (LLO) profiling of the C. jejuni N-glycan pathway has provided unique insight into the synthesis of this oligosaccharide (C. W. Reid et al., unpublished data). Levels of LLOs were analyzed with an affinity-capture method that used between 10l2 to 1013 bacterial cells. This method allows for rapid enrichment of LLOs that use an immobilized lectin (soybean agglutinin) with specificity toward the GalNAc residues in the C. jejuni LLO from the N-glycan pathway. In order to characterize the membraneassociated steps in the Pgl pathway, both wild-type C. jejuni and isogenic mutants of the pathway were used to investigate the assembly of LLO intermediates. LLOs of the N-glycan pathway in wild-type C. jejuni NCTC 11168 were undetectable even when a fivefold greater cell density was used as starting material. The lack of accumulating LLOs in wild-type C. jejuni suggests that in the presence of polypeptide acceptors, LLOs are used immediately as they are formed. Analyses of isogenic mutants in the oligosaccharyltransferase (pglB) and membrane flippase (pglK) revealed an accumulation of undecaprenylpyrophosphate (Und-PP)-heptasaccharide, with a significantly greater amount found in the flippase mutant (Fig. 9a). MS-MS experiments in the negative-ion mode also provided further evidence for the assembly of the C. jejuni N-glycan on undecaprenyl phosphate (Und-P) (Fig. 9b, Table 2). In addition, analysis of pglB and pglK mutants for nucleotide-linked sugar precursors showed no detectable accumulation of UDP-linked sugar intermediates. In contrast, analysis of the GalNAc transferase mutants pglH and pglJ, the acetyltransferase mutant pglD, and the glucosyltransferase mutant pglI, showed no measurable accumulation of LLO intermediates. This result was independently verified by ion-exchange enrichment of LLOs to ensure that the negative result was not the result of a lack of recognition by the lectin for truncated LLOs. When these glycosyltransferase mutants were screened for accumulation of nucleotide-sugar precursors of N-glycan synthesis, accumulation of UDP-diacetamido-trideoxy-Hex was observed in the pglJ and pglD mutants (Fig. 9c). The lack of accumulating LLOs was unexpected considering that analysis of glycopeptides from pglH and pglJ mutants in C. jejuni showed no detectable transfer of truncated N-glycan to glycoproteins (Kelly et al., 2006). In contrast, analysis of the reconstituted C. jejuni Pgl pathway in Escherichia coli demonstrated the addition of truncated oligosaccharides onto proteins in the pglH and pglJ mutants because the requirement of block transfer is not
1
6313
Figure 9. Metabolomic analysis of the N-linked protein glycosylation pathway in C. jejuni. (a) CE-MS precursor ion scans for undecaprenyl-phosphate (Und-P-H,O, m l z 907.6) of affinitycaptured lipid extracts of pglK indicating accumulation of Und-PPheptasaccharide 1. (b) CE-MS-MS analysis of m l z 1168.0 in the negative ion mode, confirming that the N-glycan heptasaccharide is assembled on undecaprenyl-phosphate. (c) CE-MS spectrum demonstrating accumulation of UDP-diacetamido-trideoxyhexose (mlz 631.5) in pglJ cell lysates.
maintained in this background (Linton et al., 2005). Additionally, in vitro experiments have demonstrated that PglJ will function in the absence of PglH (Glover et al., 2005). Therefore, the lack of detectable truncated LLOs in the C. jejuni pglH and pglJ mutants must be due to feedback inhibition at preceding steps, which is supported by the detection of UDPdiacetamido-trideoxyhexose in pglJ cell lysates. This conclusion is supported by the observed allosteric regulation in the yeast oligosaccharytl transferase complex (Karaoglu et al., 2001). Because undecaprenol is in limited abundance in the bacterial cell and is involved in a plethora of cellular activities, some of which are vital for cell survival, it is not surprising that a feedback regulation of the Pgl pathway
CHAPTER 29
CAMPYLOBACTER METABOLOMICS
539
Table 2. Results of N-glycan intermediate analyses in C. jejuni Strain NCTC 11168 PdB PdD PglH Pd1
Pdl PdK
Function
Glycoprotein phenotype"
LLOS
UDP-Bac2,rl.diNAc
Oligosaccharyl transferase Acetyltransferase Processive GalNAc transferase Glc transferase Second GalNAc transferase Flippase
Heptasaccharide Unmodified Minor heptasaccharide Unmodified Hexasaccharide Unmodified Unmodified
None Yes None None None None Yes
None None Yes None None Yesd None
"From Kelly et al. (2006). "ased on precursor ion scans for nucleotide-linked sugars (UDP) from cell lysates. Minor levels of N-glycosylation due to complementation by another acetyltransferase. dObserved higher levels of accumulation compared with the pglD mutant.
(chapter 25) would exist to prevent the accumulation of LLOs and sequestering of undecaprenol from other biological processes. The application of affinity capture MS to the analysis of membrane-associated processes in carbohydrate biosynthesis pathways provides a simple and rapid method to gather structural information of the oligosaccharides under study. Unlike conventional techniques, radioisotope labeling and chromatographic separation are not required for analysis (Lehrman, 2007). Additionally, by varying the lectin used for affinity capture, this method can be applied to a wide range of systems. Analysis of bacterial complex carbohydrate biosynthetic pathways such as pilin 0-glycosylation and peptidoglycan may provide new insights into the synthesis and control of these virulence factors. Once again, targeted metabolomic profiling has been demonstrated to be a powerful tool for bacterial glycobiology. It provides a highthroughput method for assessing the role of genes in a biosynthetic pathway and allows for identification of potential regulation mechanisms by intermediates in the pathway. With the recent advances in glycoengineering in bacteria, affinity-capture CE-MS also allows for a more comprehensive assessment of the efficiency and potential bottlenecks associated with producing potential protein glycoconjugates in bacteria.
SUMMARY AND FUTURE DIRECTIONS Focused metabolomics studies offer a rapid and highly selective approach to characterizing novel biosynthetic pathways. As shown in this chapter, it is often necessary to use more than one analytical technique to truly characterize the metabolites involved in the selected pathway. By targeting a set of metabolites, it is possible to significantly reduce the complexity of the metabolome, and this in turn facilitates the identification of unknown metabolites such that
specific information on the selected biosynthesis pathway can be achieved. In addition to identifying metabolite structures, focused metabolomics approaches can yield invaluable information on the unknown function of genes in selected metabolic pathways. In the case of flagellin glycosylation, the metabolome can also be used as a source for substrates relating to novel flagellin glycans. Both the in vitro and in vivo metabolomics approaches provide the opportunity to thoroughly investigate glycan biosynthetic pathways in C. jejuni as well as other bacterial systems. These approaches have their strengths and limitations, which should be considered when choosing a suitable method. The in vivo approach requires that isogenic mutants are available, involves laborious metabolite purification, and makes extensive use of NMR for structural elucidation, which can be challenging because of the limited sensitivity of NMR. The use of MS-based methods with precursor ion scanning, however, does offer a rapid approach to quickly identify metabolites of interest, and this in turn facilitates the exploration of biosynthetic pathways and is a powerful method of functionally characterizing genes. The in vitro approach requires that genes of interest can be cloned and that gene products can be expressed and purified. However, this approach avoids metabolite purification, which facilitates the identification of most metabolites (even labile and unstable ones). NMR sensitivity is generally less of an issue because metabolites are made enzymatically within the NMR tube and more can be produced simply by adding starting material to the NMR tube. An entire biosynthetic pathway can be analyzed in the NMR tube using a one-pot reaction, and this approach results in a practical system that can be used to quickly screen enzyme inhibitors, which is useful in the development of small molecule inhibitors as therapeutic agents. As would be expected, the success of such focused metabolomics studies and the ability to move to broader metabolomics studies will rely heavily on
540
SOOETAL.
the capability of continuously adding to the complement of analytical techniques such that further classes of metabolites relevant to glycoconjugate biosynthesis pathways can be characterized. In addition, further enhancing the sensitivity of analytical approaches to tackle trace-level metabolites and extraction procedures that would minimize the loss of metabolites from the biological sample would be important for both the qualitative and quantitative analysis of novel metabolites. With regard to Campylobacter, there is much scope for continuing with metabolomics-based strategies to further characterize the unknown function of genes and flagellar glycan structures unique to different Campylobacter strains and isolates. This in turn would contribute toward an understanding of the biological roles of these novel and diverse flagellin glycan structures. Last, there is much potential for applying focused metabolomics strategies to characterizing other unique glycoconjugate biosynthesis pathways in Campylobacter.
REFERENCES Angata, T., and A. Varki. 2002. Chemical diversity in the sialic acids and related alpha-keto acids: an evolutionary perspective. Chem. Rev. 102:439-469. Bijlsma, S., I. Bobeldijk, E. R. Verheij, R. Ramaker, S. Kochhar, I. A. Macdonald, B. van Ommen, and A. K. Smilde. 2006. Large-scale human metabolomics studies: a strategy for data (pre-) processing and validation. Anal. Chem. 78567-574. Brisson, J. R., S. C. Sue, W. G. Wu, G. McManus, P. T. Nghia, and D. Uhrin. 2002. NMR of carbohydrates: 1D homonuclear selective methods, p. 59-93. In J. Jimenez-Barber0 and T. Peters (ed.), NMR Spectroscopy of Glycoconjugates. Wiley-VCH, Weinheim, Germany. De Vos, R. C., S. Moco, A. Lommen, J. J. Keurentjes, R. J. Bino, and R. D. Hall. 2007. Untargeted large-scale plant metabolomics using liquid chromatography coupled to mass spectrometry. Nut. Protoc. 2:778-791. Dixon, R. A., D. R. Gang, A. J. Charlton, 0. Fiehn, H. A. Kuiper, T. L. Reynolds, R. S. Tjeerdema, E. H. Jeffery, J. B. German, W. P. Ridley, and J. N. Seiber. 2006. Applications of metabolomics in agriculture. J. Agric. Food Chem. 5453984-8994. Dunn, W. B., N. J. Bailey, and H. E. Johnson. 2005. Measuring the metabolome: current analytical technologies. Analyst 130: 606-625. Ellis, D. I., W. B. Dunn, J. L. Griffin, J. W. Allwood, and R. Goodacre. 2007. Metabolic fingerprinting as a diagnostic tool. Phamacogenomics 8: 1243-1266. Fiehn, 0. 2002. Metabolomics-the link between genotypes and phenotypes. Plant Mol. Biol. 48:155-171. German, J. B., S. M. Watkins, and L. B. Fay. 2005. Metabolomics in practice: emerging knowledge to guide future dietetic advice toward individualized health. J. Am. Diet. Assoc. 105:14251432. Gilbert, M., J. R. Brisson, M. F. Kanvaski, J. Michniewicz, A. M. Cunningham, Y. Wu, N. M. Young, and W. W. Wakarchuk. 2000. Biosynthesis of ganglioside mimics in Cumpylobacter jejuni OH4384. Identification of the glycosyltransferasegenes, enzymatic synthesis of model compounds, and characterization of
nanomole amounts by BOO-mHz (1)h and (13)c NMR analysis.
1.Biol. Chem. 275:3896-3906. Glover, K. J., E. Weerapana, and B. Imperiali. 2005. In vitro assembly of the undecaprenylpyrophosphate-linked heptasaccharide for prokaryotic N-linked glycosylation. Proc. Natl. Acad. Sci. USA 102~14255-14259. Goodacre, R., S. Vaidyanathan, W. B. Dunn, G. G. Harrigan, and D. B. Kell. 2004. Metabolomics by numbers: acquiring and understanding global metabolite data. Trends Biotechnol. 22:245252. Griffin, J. L. 2004. Metabolic profiles to define the genome: can we hear the phenotypes? Philos. Trans. R SOC. Lond. B Biol. Sci. 359: 857-8 71. Griffin, J. L. 2003. Metabonomics: NMR spectroscopy and pattern recognition analysis of body fluids and tissues for characterisation of xenobiotic toxicity and disease diagnosis. Cum Opin. Chem. Biol. 7:648-654. Griffin, J. L., and M. E. Bollard. 2004. Metabonomics: its potential as a tool in toxicology for safety assessment and data integration. Cum Drug Metab. 5:389-398. Gu, H., H. Chen, Z. Pan, A. U. Jackson, N. Talaty, B. Xi, C. Kissinger, C. Duda, D. Mann, D. Raftery, and R. G. Cooks. 2007. Monitoring diet effects via biofluids and their implications for metabolomics studies. Anal. Chem. 79:89-97. Guan, X. L., X. He, W. Y. Ong, W. K. Yeo, G. Shui, and M. R. Wenk. 2006. Non-targeted profiling of lipids during kainateinduced neuronal injury. FASEB J. 20:1152-1161. Guerardel, Y., L. Y. Chang, E. Maes, C. J. Huang, and K. H. Khoo. 2006. Glycomic survey mapping of zebrafish identifies unique sialylation pattern. Glycobiology 16:244-257. Guerry, P., P. Doig, R. A. Alm, D. H. Burr, N. Kinsella, and T. J. Trust. 1996. Identification and characterization of genes required for post-translational modification of Campylobacter coli VC167 flagellin. Mol. Microbiol. 19:369-378. Guerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Pattarini, G. Majam, P. Thibault, and s. Logan. 2006. Changes in flagellin glycosylation affect Campylobacter autoagglutination and virulence. Mol. Microbiol. 60:299-3 11. Harrigan, G. G., and L. A. Yates. 2006. High-throughput screening, metabolomics and drug discovery. IDrugs. 9:188-192. Hermansson, M., A. Uphoff, R. Kakela, and P. Somerharju. 2005. Automated quantitative analysis of complex lipidomes by liquid chromatography/mass spectrometry. Anal. Chem. 77:21662175. Holmes, E., H. Tang, Y. Wang, and C. Seger. 2006. The assessment of plant metabolite profiles by NMR-based methodologies. Planta Med. 72:771-785. Houjou, T., K. Yamatani, M. Imagawa, T. Shimizu, and R. Taguchi. 2005. A shotgun tandem mass spectrometric analysis of phospholipids with normal-phase and/or reverse-phase liquid chromatography/electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom. 19:654-666. Jordan, K. W., and L. L. Cheng. 2007. NMR-based metabolomics approach to target biomarkers for human prostate cancer. Expelt Rev. Proteomics. 4:389-400. Karaoglu, D., D. J. Kelleher, and R. Gilmore. 2001. Allosteric regulation provides a molecular mechanism for preferential utilization of the fully assembled dolichol-linked oligosaccharide by the yeast oligosaccharyltransferase. Biochemistry 40: 1219312206. Karlyshev, A. V., P. Everest, D. Linton, S. Cawthraw, D. G. Newell, and B. W. Wren. 2004. The Campylobacter jejuni general glycosylation system is important for attachment to human epithelial cells and in the colonization of chicks. Microbiology 150: 1957-1964.
CHAPTER 29
Kelly, J., H. Jarrell, L. Millar, L. Tessier, L. M. Fiori, P. C. Lau, B. Allan, and C. M. Szymanski. 2006. Biosynthesis of the Nlinked glycan in Campylobacter jejuni and addition onto protein through block transfer. J. Bacteriol. 188:2427-2434. Kern, A., E. Tilley, I. S. Hunter, M. Legisa, and A. Glieder. 2007. Engineering primary metabolic pathways of industrial microorganisms. ]. Biotechnol. 129:6-29. Kneidinger, B., K. O’Riordan, J. Li, J. R. Brisson, J. C. Lee, and J. S. Lam. 2003. Three highly conserved proteins catalyze the conversion of UDP-N-acetyl-D-glucosamine to precursors for the biosynthesis of 0 antigen in Pseudomonas aeruginosa 011 and capsule in Staphylococcus aureus type 5 . Implications for the UDP-N-acetyl-L-fucosamine biosynthetic pathway. J. Biol. Chem. 278:3615-3627. Kuzuya, T. 2000. Early diagnosis, early treatment and the new diagnostic criteria of diabetes mellitus. Br. J. Nutr. 84(Suppl. 2): S177-Sl8 1. Kuzuya, T., S. Nakagawa, J. Satoh, Y. Kanazawa, Y. Iwamoto, M. Kobayashi, K. Nanjo, A. Sasaki, Y. Seino, C. Ito, K. Shima, K. Nonaka, and T. Kadowaki. 2002. Report of the Committee on the Classification and Diagnostic Criteria of Diabetes Mellitus. Diabetes Res. Clin. Pract. 55:65-85. Lehrman, M. A. 2007. Teaching dolichol-linked oligosaccharides more tricks with alternatives to metabolic radiolabeling. Glycobiology 17:75R-85R. Lenz, E. M., and I. D. Wilson. 2007. Analytical strategies in metabonomics. 1. Proteome. Res. 6:443-458. Levery, S. B. 2005. Glycosphingolipid structural analysis and glycosphingolipidomics: methods in enzymology, p. 300-369. In A. L. Burlingame (ed), Mass Spectrometry: Modified Proteins and Glycoconjugates. Elsevier Academic Press, Amsterdam, The Netherlands. Lindon, J. C., E. Holmes, and J. K. Nicholson. 2007. Metabonomics in pharmaceutical R&D. FEBS J. 274:1140-1151. Linton, D., N. Dorrell, P. G. Hitchen, S. Amber, A. V. Karlyshev, H. R. Morris, A. Dell, M. A. Valvano, M. Aebi, and B. W. Wren. 2005. Functional analysis of the Campylobacter jejuni Nlinked protein glycosylation pathway. Mol. Microbiol. 5:16951703. Logan, S. M., J. F. Kelly, P. Thibault, C. P. Ewing, and P. Guerry. 2002. Structural heterogeneity of carbohydrate modifications affects serospecificity of Campylobacter flagellins. Mol. Microbiol. 46:587-597. Mashego, M. R., K. Rumbold, M. De Mey, E. Vandamme, W. Soetaert, and J. J. Heijnen. 2007. Microbial metabolomics: past, present and future methodologies. Biotechnol. Lett. 29:1-16. McNally, D. J., A. J. Aubry, J. P. Hui, N. H. Khieu, D. Whitfield, C. P. Ewing, P. Guerry, J. R. Brisson, S. M. Logan, and E. C. Soo. 2007. Targeted metabolomics analysis of Campylobacter coli VC167 reveals legionaminic acid derivatives as novel flagellar glycans. J. Biol. Chem. 282: 14463-14475. McNally, D. J., J. P. Hui, A. J. Aubry, K. K. Mui, P. Guerry, J. R. Brisson, S. M. Logan, and E. C. Soo. 2006a. Functional characterization of the flagellar glycosylation locus in Campylobacter jejuni 81-176 using a focused metabolomics approach. 1.Biol. Chem. 281:18489-18498. McNally, D. J., I. C. Schoenhofen, E. F. Mulrooney, D. M. Whitfield, E. Vinogradov, J. s. Lam, s. M. Logan, and J. R. Brisson. 2006b. Identification of labile UDP-ketosugars in Helicobacter pylori, Campylobacter jejuni and Pseudomonas aeruginosa: key metabolites used to make glycan virulence factors. Chembiochemistry 7:1865-1868. Oksman-Caldentey, K. M., and K. Saito. 2005. Integrating genomics and metabolomics for engineering plant metabolic pathways. Curr. Opin. Biotechnol. 16:174-179.
CAMPYLOBACTER METABOLOMICS
541
Oliver, S. G., M. K. Winson, D. B. Kell, and F. Baganz. 1998. Systematic functional analysis of the yeast genome. Trends Biotechnol. 16:373-378. Oresic, M., A. Vidal-Puig, and V. Hanninen. 2006. Metabolomic approaches to phenotype characterization and applications to complex diseases. Expert Rev. Mol. Diagn. 6575-585. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pellecchia, M. 2005. Solution nuclear magnetic resonance spectroscopy techniques for probing intermolecular interactions. Chem. Biol. 12:961-971. Pfoestl, A., A. Hofinger, P. Kosma, and P. Messner. 2003. Biosynin thesis of dTDP-3-acetamido-3,6-dideoxy-alpha-~-galactose Aneurinibacillus themzoaerophilus L420-91T. J. Biol. Chem. 278:26410-26417. Reid, C. W., J. Stupak, M. M. Chen, B. Imperiali, J. Li, and C. M. Szymanski. Affinity-captive tandem mass spectrometric characterization of polyprenyl-linked oligosaccharides: a novel tool to study protein N-glycosylation pathways. Anal. Chem., in press. Reo, N. V. 2002. NMR-based metabolomics. Drug Chem. Toxicol. 25~37.5-382. Robertson, D. G., M. D. Reily, and J. D. Baker. 2007. Metabonomics in pharmaceutical discovery and development. J. Proteome Res. 6526-539. Sabatine, M. S., E. Liu, D. A. Morrow, E. Heller, R. McCarroll, R. Wiegand, G. F. Berriz, F. P. Roth, and R. E. Gerszten. 2005. Metabolomic identification of novel biomarkers of myocardial ischemia. Circulation 112:3868-3 875. Schauer, N., and A. R. Fernie. 2006. Plant metabolomics: towards biological function and mechanism. Trends Plant Sci. 11508516. Schirm, M., I. C. Schoenhofen, S. M. Logan, K. C. Waldron, and P. Thibault. 2005. Identification of unusual bacterial glycosylation by tandem mass spectrometry analyses of intact proteins. Anal. Chem. 77:7774-7782. Schoenhofen, I. C., V. V. Lunin, J. P. Julien, Y. Li, E. Ajamian, A. Matte, M. Cygler, J. R. Brisson, A. Aubry, S. M. Logan, S. Bhatia, W. W. Wakarchuk, and N. M. Young. 2006a. Structural and functional characterization of PseC, an aminotransferase involved in the biosynthesis of pseudaminic acid, an essential flagellar modification in Helicobacter pylori. J. Biol. Chem. 281: 8907-8916. Schoenhofen, I. C., D. J. McNally, J. R. Brisson, and S. M. Logan. 2006b. Elucidation of the CMP-pseudaminic acid pathway in Helicobacter pylori: synthesis from UDP-N-acetylglucosamine by a single enzymatic reaction. Glycobiology 16:8C-14C. Schoenhofen, I. C., D. J. McNally, E. Vinogradov, D. Whitfield, N. M. Young, S. Dick, W. W. Wakarchuk, J. R. Brisson, and S. M. Logan. 2006c. Functional characterization of dehydratase/ aminotransferase pairs from Helicobacter and Campylobacter: enzymes distinguishing the pseudaminic acid and bacillosamine biosynthetic pathways. J. Biol. Chem. 281:723-732. Serkova, N. J., and C. U. Niemann. 2006. Pattern recognition and biomarker validation using quantitative 1H-NMR-based metabolomics. Expert Rev. Mol. Diagn. 6:717-731. Soga, T., Y. Ueno, H. Naraoka, Y. Ohashi, M. Tomita, and T. Nishioka. 2002. Simultaneous determination of anionic intermediates for Bacillus subtilis metabolic pathways by capillary electrophoresis electrospray ionization mass spectrometry. Anal. Chem. 74:2233-2239.
542
SO0 ETAL.
Soo, E. C., A. J. Aubry, S . M. Logan, P. Guerry, J. F. Kelly, N. M. Young, and P. Thibault. 2004. Selective detection and identification of sugar nucleotides by CE-electrospray-MS and its application to bacterial metabolomics. And. Cbem. 76:619-626. Szymanski, C. M., D. H. Burr, and P. Guerry. 2002. Cumpylobacter protein glycosylation affects host cell interactions. Infect. Immun. 70:2242-2244. Szymanski, C. M., S . M. Logan, D. Linton, and B. W. Wren. 2003a. Campylobacter-a tale of two protein glycosylation systems. Trends. Microbiol. 11:233-238. Szymanski, C. M., F. St Michael, H. C. Jarrell, J. Li, M. Gilbert, S. Larocque, E. Vinogradov, and J. R. Brisson. 2003b. Detection of conserved N-linked glycans and phase-variable lipooligosaccharides and capsules from Campylobacter cells by mass spectrometry and high resolution magic angle spinning NMR spectroscopy. ]. Biol. Chem. 278:24509-24520. Szymanski, C. M., and B. W. Wren. 2005. Protein glycosylation in bacterial mucosal pathogens. Nat. Rev. Microbiol. 3:225-237. Thibault, P., S. M. Logan, J. F. Kelly, J. R. Brisson, C. P. Ewing, T. J. Trust, and P. Guerry. 2001. Identification of the carbo-
hydrate moieties and glycosylation motifs in Campylobacterjejuni flagellin.]. Biol. Cbem. 276:34862-34870. van der Greef, J., T. Hankemeier, and R. N. McBurney. 2006. Metabolomics-based systems biology and personalized medicine: moving towards n = 1 clinical trials? ~hQmQCOgeflOmiCS 7: 1087-1094. Villas-BBas, S. G., and P. Bruheim. 2007. Cold glycerol-saline:the promising quenching solution for accurate intracellular metabolite analysis of microbial cells. Anal. Biocbem. 370:87-97. Vimr, E. R., K. A. Kalivoda, E. L. Deszo, and S. M. Steenbergen. 2004. Diversity of microbial sialic acid metabolism. Microbiol. Mol. Biol. Rev. 68:132-153. Ward, J. L., J. M. Baker, and M. H. Beale. 2007. Recent applications of NMR spectroscopy in plant metabolomics. FEBS J. 274:1126-1131. Young, N. M., J. R. Brisson, J. Kelly, D. C. Watson, L. Tessier, P. H. Lanthier, H. C. Jarrell, N. Cadotte, F. St Michael, E. Aberg, and C. M. Szymanski. 2002. Structure of the N-linked glycan present on multiple glycoproteins in the gram-negative bacterium, Campylobacter jejuni. J. Biol. Chem. 277:4253042539.
V. GENES AND GENE EXPRESSION
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 30
Regulation of Flagellar Gene Expression and Assembly DAVIDR. HENDRIXSON
system that posttranslationally modifies flagellin subunits with pseudaminic acid residues that are essential for filament formation (Goon et al., 2006; Guerry et al., 2006; Karlyshev et al., 2002; Linton et al., 2000; Thibault et al., 2001; chapter 26). Whereas synthesis of the flagellum in campylobacters is relatively understudied compared with other motile organisms, a picture is beginning to form that shows that these bacteria possess both deviations of common regulatory pathways found in other bacterial species and processes unique to themselves to result in their own characteristic mechanisms for expression of flagellar genes and building the flagellar organelle.
The earliest descriptions of campylobacters reported spiral-shaped or curved bacteria exhibiting a rapid, darting, corkscrew pattern of movement. Further examination of most Campylobacter species revealed that a single flagellum at one or both poles of the bacterium contributes to this characteristic mode of motility. Although rare, a few Campylobacter species produce multiple polar flagella (Phillips and Lee, 1983). Flagella and flagellar motility are not only important colonization determinants for promoting commensalism with animal hosts (Hendrixson, 2006; Hendrixson and DiRita, 2004; Nachamkin et al., 1993; Wassenaar et al., 1993; Wosten et al., 2004), but also are important virulence factors that enable the bacterium to invade human intestinal cells and promote other interactions to result in diarrheal disease in humans (Black et al., 1988; Grant et al., 1993; Wassenaar et al., 1991; Yao et al., 1994). The flagella give campylobacters an unusual motile ability in relatively viscous environments compared with the other bacteria (Ferrero and Lee, 1988). This property likely contributes to camplyobacters being able to penetrate, colonize, and persist in the thick mucous lining the intestinal surface and crypts of hosts (Lee et al., 1986). In addition, flagella appear to have a second function in secreting proteins that are not involved in motility but may instead contribute to invasion of human cells and other steps in pathogenesis of disease (Konkel et al., 2004; Poly et al., 2007; Song et al., 2004). The formation of flagella in bacteria is not a simple process. Rather, it requires the temporal coordination of expression of approximately 40 to 50 flagellar genes. Upon translation, the secretion of flagellar proteins must be in an ordered state so that they can interact properly to ensure the correct assembly of a functional rotating flagellar filament. Complicating the process of flagellar biosynthesis in some campylobacters is an 0-linked glycosylation
IDENTIFICATION OF COMPONENTS INVOLVED IN FLAGELLAR MOTILITY OF CAMPYLOBACTERS Bioinformatic analysis of genomic sequences of Campylobacter species combined with the development of transposon mutagenesis technologies and other mutational approaches contributed to identifying proteins of campylobacters required for flagellar gene regulation, biosynthesis, and motility (Bleumink-Pluym et al., 1999; Colegio et al., 2001; Fouts et al., 2005; Golden and Acheson, 2002; Hendrixson et al., 2001; Hendrixson and DiRita, 2003; Parkhill et al., 2000; Yao et al., 1994). Because motility has mostly been analyzed in Campylobacter jejuni and to a lesser extent in C. coli, these species are the model systems for studying motility in campylobacters. These bacteria appear to possess many homologues of proteins that are required for flagellar motility in other bacteria, but much research is still required to determine whether the Campylobacter proteins perform their proposed analogous functions. In addition, campylobacters produce proteins required for motility that are absent in other well-
David R. Hendrixson
Department of
Microbiology, University of Texas Southwestern Medical Center, Dallas, TX 75390. 545
546
HENDRIXSON
studied motile bacteria. Thus, campylobacters have portions of conserved pathways mixed with unique mechanisms of promoting flagellar gene regulation, biosynthesis, and motility. Identification of Flagellar Proteins by Analysis of Genomic Sequences Comparison of the genomic sequences of different Cumpylobucter species has revealed a particularly striking feature of the genome: the lack of organization of flagellar genes into operons (Fouts et al., 2005; Parkhill et al., 2000). In many bacteria, flagellar genes are grouped together into operons (commonly three to eight genes per operon) so that transcription of multiple flagellar genes are linked to the same promoter (Chilcott and Hughes, 2000; Dasgupta et al., 2003; Liu and Ochman, 2007; Prouty et al., 2001). This organization is thought to allow for the controlled coordinate expression of a large number of flagellar genes required for a specific stage in flagellar biosynthesis. In campylobacters, however, flagellar genes are scattered throughout the genome with just a small number of examples of groupings of more than three genes into operons. This disorganization would be expected to create a challenge for the bacterium in regulating the transcription of multiple genes in an ordered, temporal fashion. Obviously, the bacteria have overcome this problem to efficiently and appropriately control expression of various genetic loci in stages so that a flagellum is constructed properly. In terms of content, many flagellar proteins that are required for the basic functions of flagellar biosynthesis and motility in other bacteria are present in campylobacters (Table 1).The cytoplasmic flagellar motor and switch complex proteins that provide energy for motility and controlling the direction of flagellar rotation are present. Camplyobacters appear to produce an inner-membrane-localized export apparatus and a cytoplasmic ATPase, FliI, which together form a secretion system for flagellar proteins. This flagellar secretory system is similar to the type I11 secretion systems of pathogenic bacteria that secrete virulence proteins to influence the activities of eukaryotic host cells (Fan et al., 1997; Hueck, 1998; Kihara et al., 2001; Minamino et al., 1994; Minamino and Macnab, 1999; Ohnishi et al., 1997). The flagellar basal body proteins, including the ring structures in the peptidoglycan and outer membrane and the rod proteins, appear to be produced by campylobacters. The flagellar rod is a periplasmic structure that links the flagellar export apparatus in the inner membrane to the flagellar hook at the bacterial surface (Glenn-Calvo et al., 1994; Power et al., 1992).
The rod penetrates the peptidoglycan and outer membrane by passing through the central pore of the ring structures. Because the rod and flagellar hook are hollow structures, they form a conduit through which flagellin proteins are secreted from the cytoplasm, through the periplasm, and to the tip of nascent flagellum, where they are polymerized onto the growing structure. The genome of C. jejuni and C. coli contain genes for multiple flagellins, but only FlaA and FlaB have been shown to be a part of the flagellar filament (Guerry et al., 1991; Wassenaar et al., 1994). FlaA is the predominant flagellin, whereas FlaB appears to be sparse, constituting less than 20% of the flagellar filament (Guerry et al., 1991). At least one protein that was initially annotated in the first genomic sequence as a flagellin, FlaC, does not appear to play a role in motility but is secreted by the flagellar export apparatus and may be an effector to influence interactions with human intestinal epithelial cells (Parkhill et al., 2000; Song et al., 2004). Chemotaxis is an essential property of flagellar motility that influences the movement of bacteria toward appropriate environmental and host niches that support ideal bacterial growth and away from components that are less beneficial for growth or harmful to the organism (reviewed in Blair, 1995). Like most other motile bacteria, the genome of C. jejuni encodes a repertoire of 10 chemoreceptor proteins (also referred to as methyl-accepting chemotaxis [MCP] proteins) and six cytoplasmic signal transduction proteins forming the Che signaling cascade (Fouts et al., 2005; Marchant et al., 2002; Parkhill et al., 2000). Each MCP is hypothesized to specifically sense one or more substrates. Upon stimulation, a MCP transmits signals to the cytoplamic Che signaling cascade that then assimilates these signals to influence the flagellar switch proteins to appropriately control the direction of flagellar rotation. Detailed information regarding mechanisms of chemosensory signal transduction in C. jejuni that control chemotactic movement is provided in chapter 20. Requirements for Flagellar Motility in C . jejuni Revealed by Random Mutagenesis Screens Although much information regarding flagellar motility has been gleaned by analyzing predictions from genomic sequences, the field of flagellar motility in campylobacters was moved forward by the development of new genetic tools and strategies for studying these bacteria. Until the last decade, a major limitation for studying campylobacters has been the lack of suitable genetic systems for the bacteria, including a random transposon mutagenesis system. In 2000 and 2001, three groups independently developed in
CHAPTER 30
vitro or in vivo transposon mutagenesis systems that allowed for the creation of random transposon insertions in the C. jejuni genome (Colegio et al., 2001; Golden et al., 2000; Hendrixson et al., 2001). A derivative of the mariner family of transposons, Himarl, was used to create both an in vitro and in vivo transposon mutagenesis system (Golden et al., 2000; Hendrixson et al., 2001). A second in vitro transposon mutagenesis system used a derivative of Tn522 from Staphylococcus aureus (Colegio et al., 2001). In vitro transposon mutagenesis systems in C. jejuni proved particularly useful for generating random transposon insertions in the genome as a result of the bacterium being naturally competent for DNA uptake and transformation (Wang and Taylor, 1990). For this method, transposition reactions are performed in vitro by mixing genomic DNA from C. jejuni, the transposon DNA containing an antibiotic resistance gene, and the purified transposase. After repairing the transposon-chromosomal DNA junctions, the DNA is added to bacteria for uptake and recombination. Transposon mutants are selected for by growth on agar containing the appropriate antibiotic. The isolation of transposon mutants of C. jejuni with defects in flagellar motility confirmed that many of the proposed flagellar genes revealed by annotation of genomic sequences are involved in this process (Table 1; Colegio et al., 2001; Golden and Acheson, 2002; Hendrixson et al., 2001; Hendrixson and DiRita, 2003). In some cases, these genes have been studied in more detail by generating mutants in which the gene has been insertionally inactivated or removed from the chromosome; the resulting nonmotile mutants verified that the transposon mutagenesis approaches are valid in identifying genes required for motility. Through these transposon mutagenesis screens and other quasi-random mutagenesis approaches, genes involved in motility were identified that are not common to other motile bacteria and would not be predicted at first glance of secondary sequence analyses of predicted proteins from genomic sequences (Bleumink-Pluym et al., 1999; Colegio et al., 2001; Golden and Acheson, 2002; Hendrixson et al., 2001; Hendrixson and DiRita, 2003; Yao et al., 1994). One such screen identified the Cet energy taxis system of C. jejuni and a pair of genes, flgp and flgQ, that are involved in motility (Hendrixson et al., 2001; Sommerlad and Hendrixson, 2007). Production of FlgP and FlgQ and their involvement in motility appears to be unique to Campylobacter species. Disruption of either of these genes results in a mutant that produces flagella but is nonmotile. The stability of FlgP as an outer membrane protein appears to depend on FlgQ. However, the specific role of FlgP and FlgQ in mo-
FLAGELLAR GENE EXPRESSION AND ASSEMBLY
547
tility is unknown. Another gene required for flagellar motility, pflA, was identified in two separate studies that used a method to generate quasi-random mutants by inserting an antibiotic-resistance cassette into digested genomic libraries (Bleumink-Pluym et al., 1999; Yao et al., 1994). Whereas the exact function of PflA is unknown, lack of this protein causes a loss in motility despite producing flagella, similar to the flgP and fZgQ mutant phenotypes (Yao et al., 1994). These studies combined reinforce the idea that analysis of genomic sequences alone cannot predict all genes required for a specific function, even one as well studied in other bacteria as flagellar motility. The use of random mutants and unbiased genetic screens for phenotypes continues to be a powerful strategy to uncover proteins required for functions of interest.
FLAGELLAR GENE REGULATION IN OTHER BACTERIA
Salmonella Species Seminal works for understanding regulatory pathways for flagellar gene expression and assembly of proteins into a flagellum largely focused on those of Salmonella species. These early studies revealed that expression of flagellar genes, translation of flagellar proteins, and assembly of flagella are connected so that flagellar components are synthesized in a tightly controlled manner relative to when the components are needed in the construction of the flagellum (Hughes et al., 1993; Karlinsey et al., 2000; Kutsukake et al., 1990). Disruption of a flagellar gene that is required early in the process of flagellar biosynthesis has a negative effect on the expression of genes required at later steps of flagellar biogenesis (Kutsukake et al., 1990). Analysis of expression of flagellar genes in Salmonella revealed that the bacterium has organized these genes into three broad classes on the basis of their regulated, temporal expression (Karlinsey et al., 2000; Kutsukake et al., 1990). Class I genes consist of the flhDC operon, which encodes the master regulator for flagellar gene expression. Expression of class I1 genes, which include fZiA (encoding u2'),an alternative u factor for RNA polymerase, and flgM, which codes for the negative regulator of u2',is positively regulated by the FlhDC master regulator after detection of appropriate activating signals. Also included as class I1 genes are those encoding components of the flagellar export apparatus, and the basal body and hook components. The class I11 genes include those that encode the flagellins, which are the last proteins secreted after the flagellar
548
HENDRIXSON
Table 1. Known and proposed proteins of C. jejuni that function in flagellar motility Numerical Category
Protein"
annotation"
Proposed functionb
Reference(s)"
NCTC 11168 81-176 Regulation of expression
RpoN FliA FlgS FlgR
FlgM FliF FlhA FlhB FliO Flip FliQ FliR FliH FliI FliE Basal body components FlgC FlgB FlgF FlgG Fld FlgI FlgA FlgH FlgE Flagellar hook components FlgD FlgE2 FliK FlgK FlgL Flagellar filament components FlaA FlaBd FliD FliS FlaC Motor and switch components MotA MotB FliG FliM FliY PflA FlgP FkQ FliN Chemotaxis signal processing CheA Chew CheY CheV CheB CheR Chemoreceptors (MCPs) DocBd 0144cd DocCd 0448cd 0951cd lllOcd 1506cd 1564d CetB CetA 1191cd
Flagellar protein secretion
0670c 0061c 0793 1024c
0696 0099 0814 1043
1464 0318 0882c 0335 0352 0820c 1675 1179c 0320 0195 0526c 0527c 0528c 0697 0698 1463 1462 0769c 0687c 1729c 0042 0043 0041 1466 0887c 1339c 1338c 0548 0549 0720c 0337c 0336c 03 19 0060c 0059c 1565c 1026c 1025c 0351 0284c 0283c 1118c 0285c 0924c 0923c 0020c 0144c 0262c 0448c 0951c 111oc 1506c 1564 1189c 1190c 1191c
1457 0340 0890 0357 0765 0837 1671 1194 0342
Two-component system histidine kinase Two-component system d4-dependent response regulator Anti-a factor 10, 28 Inner membrane MS ring Inner membrane export apparatus protein 2, 5, 9, 10, 15, 21 Inner membrane export apparatus protein 5, 9, 10, 18, 20 Inner membrane export apparatus protein Inner membrane export apparatus protein 4, 5, 10 Inner membrane export apparatus protein Inner membrane export apparatus protein 9, 10 Negative regulator of FliI -
0226 0551 0552 0553 0720 0721 1456 1455 0790 0710 0025 0080 008 1 0079 1459 0894 1339 1338 0573 0574 0743 0359 0358 0341 0098 0097 1550 1045 1044 0357 03 10 0309 1136 03 11 093 1 0930 0046 0180 0289 0473 0975 1128 1498 1548 1204 1205 1206
Cytoplasmic ATPase MS ring-rod junction protein Proximal rod protein Proximal rod protein Proximal rod protein Distal rod protein Rod assembly protein Peptidoglycan-localizedP-ring Chaperone for FlgI Outer membrane localized L ring Flagellar hook protein Flagellar hook-associated protein Flagellar hook-associated protein Flagellar hook-length control protein Proximal hook-filament junction protein Distal hook-filament junction protein Major flagellin protein Minor flagellin protein Filament capping protein Flagellin chaperone Flagellin protein Flagellar motor stator protein Flagellar motor stator protein Flagellar rotor/switch protein Flagellar motor switch C-ring protein Flagellar motor switch protein Unknown; mutant makes paralyzed flagella Unknown; mutant makes paralyzed flagella Unknown; mutant makes paralyzed flagella Flagellar motor switch C-ring protein Chemotaxis histidine kinase CheA-MCP coupling protein Chemotaxis response regulator CheA/ Chew hybrid protein MCP methylesterase MCP methyltransferasae MCP MCP MCP MCP MCP MCP MCP MCP Sensor protein involved in energy taxis MCP involved in energy taxis CetB-like protein
4, 5, 9, 10, 13, 28 2, 9, 10, 13, 28 10, 28 10, 12, 13, 28
as4 uZ8
4, 9 18 5, 18 25 5 9 4, 10, 18 9 10 5 1, 5 4, 5 5, 7, 9, 14, 15, 22, 23, 27 7, 15, 22, 23, 27 5, 18 5 25 9 4, 9 5 4, 5 1, 29 8, 24 24 3, 4, 5, 9 5, 9, 30 9 11 11 11 11 11 11 11 11 5, 9 9 9
Continued on following page
CHAPTER 30
-
FLAGELLAR GENE EXPRESSION AND ASSEMBLY
549
Table 1. Continued Numerical Category
Protein"
annotationa
Proposed function
Reference+)"
NCTC 11168 81-176 Flagellin glycosylation
Unknown functions"
PseB
1293
PseC
1294
PseE
1337
PseF
1311
PseG PseH PseI
1312 1313 1317
PseAd
1316c
PseDd
1333
CjOO62c FlhG FlhF Cj0248 Cj0390 FlaG FlhXf CjO883c FliWg Mafl FliL Cj1497c
0062c 0063c 0064c 0248 0390 0547 0848c 0883c 1075 1318 1408 1497c
1310 Pseudaminic acid biosynthesis; sugar nucleotide dehydratase 1311 Pseudaminic acid biosynthesis; sugar nucleotide aminotransferase 1337 Flagellin modification; putative pseudaminic acid transferase 1328 Pseudaminic acid biosynthesis; CMPpseudaminic acid transferase 1329 Pseudaminic acid biosynthesis 1330 Pseudaminic acid biosynthesis 1334 Pseudaminic acid biosynthesis; pseudaminic acid synthase 1333 Pseudaminic acid biosynthesis; PseAm synthase 1336 Flagellin modification; putative PseAM transferase 0100 Motility integral membrane protein 0101 Flagellar biosynthesis regulator 0102 Flagellar biosynthesis protein 0275 Signal transduction protein 0413 Unknown; possible transmembrane protein 0572 Possible flagellar protein 0864 FlhB cytoplasmic domain-like protein 0891 Rrf2 family transcriptional regulator 1093 Possible flagellin chaperone -h Motility accessory factor 1407 Putative flagellar protein 1489 Unknown
4, 6 8
8, 14, 16, 17 8
8 8 1, 19, 26 26
8 9 5, 10 9 4 15 9, 10 5 16 5
"Numerical annotation of the protein is based on the annotation of the genomes of C. jejuni NCTC 11168 (Parkhill et al., 2000) and C. jejuni 81-176 (Fouts et al., 2005). bThe predicted function of the protein is based on homology to corresponding proteins of other motile bacteria. In most cases, the specific proposed function of the protein has not been verified in Campylobucter species. 'References listed include those that have examined the respective protein for a role in motility. References include: 1, Bleumink-Plyum et al. (1999); 2, Carrillo et al. (2004); 3, Chang and Miller (2006); 4, Colegio et al. (2001); 5, Golden and Acheson (2002); 6, Goon et al. (2003); 7, Grant et al. (1993); 8, Guerry et al. (2006); 9, Hendrixson et al. (2001); 10, Hendrixson and DiRita (2003); 11, Hendrixson and DiRita (2004); 12, Hendrixson (2006); 13, Jagannathan et al. (2001); 14, Jones et al. (2004); 15, Kalmokoff et al. (2006); 16, Karlyshev et al. (2002); 17, Karlyshev and Wren (2005); 18, Konkel et al. (2004); 19, Linton et al. (2000); 20, Matz et al. (2002); 21, Miller et al. (1993); 22, Misawa and Blaser (2000); 23, Nachamkin et al. (1993); 24, Sommerlad and Hendrixson (2007); 25, Song et al. (2004); 26, Thibault et al. (2001); 27, Wassenaar et al. (1991); 28, Wosten et al. (2004); 29, Yao et al. (1994); 30, Yao et al. (1997); A dash indicates a protein that has yet to be analyzed for a role in flagellar motility of C. jejuni. "These proteins have been shown not to be involved in flagellar motility by in vitro analysis. These proteins may constitute part of the flagellum but are not required for motile behavior or have a role in motility in alternative environments or conditions other than the ones studied in the initial analyses. 'The functions of these proteins are largely unknown. Proposed functions are loosely based on the homology that these proteins or their domains have with proteins that are involved in motility in other bacteria. 'FlhX may be involved in motility as predicted by Pallen et al. (2005). A homologue in H. pylori is required for motility in the absence of the cytoplasmic domain of FlhB (Wand et al., 2006). gThis protein displays homology to the FliW proteins of Treponemu pullidurn and Bacillus subtilis (Titz et al., 2006). 'Mafl is not encoded in the C. jejuni 81-176 genome.
hook that polymerize into the flagellar filament. Expression of class I11 genes is dependent on a2'. Assembly of the flagellum commences at the bacterial inner membrane and extends outward to the tip of the organelle (Macnab, 2003). Because secretion of flagellins is through the hollow conduit formed by the flagellar export apparatus, periplasmic rod, and surface hook structure, it is not beneficial for the bacterium to make flagellins until this complete secretion system is constructed. Thus, the bacterium needs a regulatory mechanism to activate tran-
scription of the cr28-dependent flagellin genes only after this entire secretion apparatus is in place. The inclusion of FlgM as part of the flagellar regulatory transcriptional cascade accomplishes this task (Hughes et al., 1993; Karlinsey et al., 2000). Before formation of the flagellar export apparatus, rod, and hook, FlgM remains cytoplasmic and complexed to a2',which inhibits both the association of a2' with RNA polymerase holoenzyme and the activity of the holoenzyme so that transcription of class I11 genes is inoperative (Chadsey et al., 1998; Gillen and
550
HENDRIXSON
Hughes, 1991). Once the complete flagellar secretory system is constructed, FlgM is secreted out of the cytoplasm through the system and into the extracellular milieu (Hughes et al., 1993). u28is then relieved from inhibition and can function in transcription of class I11 genes. Synthesis of the flagellins and their export through the flagellar secretion system proceed with the flagellins polymerizing at the tip of the flagellar hook and continuing outward to complete the filament. Vibrio and Pseudomonas Species Once these regulatory pathways for flagellar gene expression and assembly had begun to be elucidated in Salmonella species, research was expanded into other bacteria to determine whether these pathways were conserved. Initial analyses in species of Vibrio and Pseudomonas revealed that the same global theme of organization of the temporal expression of flagellar genes on the basis of when the encoded proteins are assembled into the flagellar organelle exists. However, the genetic regulatory mechanisms for controlling expression of flagellar genes are altered in these bacteria to include another alternative u factor, d4,and additional transcriptional factors that add another level of control to create a four-tiered flagellar regulatory cascade. At the top of this regulatory cascade are the class I genes flrA (for Vibrio species; Klose and Mekalanos, 1998) and fleQ (for Pseudomonas species; Arora et al., 1997), which encode the master regulator in each bacterium. These regulators receive a specific unknown signal, and in conjunction with oS4,they activate transcription of class I1 flagellar genes. These genes include those encoding u2',components of the flagellar export apparatus, and a two-component regulatory system consisting of a sensor histidine kinase and a d4dependent response regulator (Arora et al., 1997; Dasgupta et al., 2000, 2003; Jyot et al., 2002; Klose and Mekalanos, 1998; Prouty et al., 2001). Once the sensor kinase component (FlrB for Vibrio species and FleS for Pseudomonas species) is activated via phosphorylation after sensing an as yet undescribed signal, a phosphorelay event occurs to terminate in phosphorylation of the cognate response regulator (FlrC for Vibrio species and FleR for Pseudomonas species). Activation of these response regulators allows d4 to then stimulate expression of class 111 flagellar genes, which include those for the flagellar hook and basal body genes (Dasgupta et al., 2003; Prouty et al., 2001). In Vibrio cholerae, one class I11 gene encodes the major flagellin. Upon successful assembly of the flagellar rod and hook and secretion of FlgM, u28is then capable of assisting in transcription of class IV
genes, which include those for the motor proteins, and the flagellins (Dasgupta et al., 2003; Prouty et al., 2001).
FLAGELLAR GENE REGULATION IN CAMPYLOBACTER Initial Findings Early studies focusing on antigens of C. jejuni that are recognized by convalescent human antisera after infection revealed that the major flagellin FlaA is the foremost immunodominant antigen (Wenman et al., 1985). Thus, much early work regarding flagellar motility in campylobacters largely centered on the genetic organization and expression of the flagellin genes of C. jejuni and C. coli. During the cloning and analysis of the flaA locus of C. coli, it was noticed that a second highly homologous flagellin gene appeared to be at an adjacent location (Logan et al., 1989). Further examination of the region downstream of flaA revealed the presence of flaB, which shares 91.5% identity with flaA (Guerry et al., 1990). Simultaneous research in C. jejuni revealed the same genomic organization of and similarity between flaA and flaB (Nuijten et al., 1990). Despite this tight organization, transcriptional analysis revealed that both genes are expressed from separate promoter elements. The promoter for flaA is characteristic of a a2*-dependent promoter, whereas the one for flaB contains typical binding sites for oS4(Guerry et al., 1990, 1991; Nuijten et al., 1990). Elucidation of the C. jejuni Transcriptional Regulatory Cascade for aS4-Dependent Flagellar Genes Insights into flagellar gene regulation were revealed by both site-directed and transposon mutagenesis approaches in C. jejuni (Carrillo et al., 2004; Hendrixson et al., 2001; Hendrixson and DiRita, 2003; Jagannathan et al., 2001; Wosten et al., 2004). Disruption or deletion of rpoN (encoding d4) and fliA (encoding a2')were found to result in nonmotile phenotypes. The development of transcriptional reporter systems and microarray analysis for C. jejuni and the application of real-time PCR confirmed that u28is involved in transcription of flaA. In addition, expression of genes for minor flagellar components and a nonflagellar protein, Cj0977, that is involved in virulence are also dependent on uZ8(Carrillo et al., 2004; Goon et al., 2006; Wosten et al., 2004). d4is required not only for transcription of fiaB but also for a second operon, flgDE2, encoding two flagellar hook-associated proteins (Carrillo et al., 2004;
CHAPTER 30
FLAGELLAR GENE EXPRESSION A N D ASSEMBLY
551
predicted to encode components that are potentially Hendrixson and DiRita, 2003; Wosten et al., 2004). secreted by the flagellar export apparatus. Thus, their Genomic analysis of promoter elements in C. jejuni expression is likely not necessary until the flagellar strains predict that other flagellar genes such as flgB, export apparatus is synthesized. Therefore, identififlgc, flgE, flgE flgG flgM flgL flgK f l g L PiE, and fliK contain as4-binding sites and may also be d4 cation of the flagellar export apparatus to be involved in transcription of as4-dependent flagellar genes is a dependent (Carrillo et al., 2004; Fouts et al., 2005; logical finding. The last class of mutants had transpoParkhill et al., 2000; Wosten et al., 2004). These latson insertions in flhF, encoding a putative GTPase ter flagellar genes are all predicted to encode comthat is found in many bacteria with polarly localized ponents of the flagellar basal body and hook (Table 1). flagella such as Helicobacter pylori, Vibrio, and PseuDevelopment of a transcriptional reporter sysdomonas species (Correa et al., 2005; Murray and Kazmierczak, 2006; Niehus et al., 2004; Pandza et tem on the basis of the C. jejuni astA gene and in al., 2000). The function of FlhF appears to be varied vitro transposon mutagenesis contributed much to the beginnings of elucidating the flagellar transcripin the different bacteria. Whereas it is required for the polar placement of flagella in Pseudomonas spetional regulatory cascade and mechanisms of gene regulation in C. jejuni (Hendrixson, 2006; Hendrixcies and may aid in regulating flagellar number in Pseudomonas putida (Murray and Kazmierczak, son and DiRita, 2003). astA encodes the enzyme arylsulfatase in the jejuni genome, which functions 2006; Pandza et al., 2000), BhF mutants in H. pylori and V. cholerae do not make flagella because of the to cleave sulfate groups from aryl compounds (Fouts et al., 2005; Parkhill et al., 2000; Yao and Guerry, involvement of FlhF in the expression of select d41996). When C. jejuni expressing astA is grown on dependent flagellar genes (Correa et al., 2005; Nieagar with the chromogenic substrate 5-bromo-4hus et al., 2004). Verification that the genes of C. jejuni identified chloro-3-indoyl sulfate (X-S), the substrate is cleaved to produce a blue color; strains that lack or do not in these screens are required for expression of d4express astA remain colorless or white. In addition, dependent flagellar genes was accomplished by insera simple quantitative assay is available to measure the tionally inactivating or deleting them from the bacterial genome. By means of the astA transcriptional levels of arylsulfatase in a bacterial lysate by monitoring the release of nitrophenol from nitrophenyl reporter fusion, microarray, or quantitative real-time sulfate via spectrophotometeric analysis (Hendrixson PCR analyses, these mutants have large defects in exand DiRita, 2003; Henderson and Milazzo, 1979; pression of flgDE2 or flaB (Carrillo et al., 2004; HenYao and Guerry, 1996). drixson and DiRita, 2003; Wosten et al., 2004). Other proposed as4-dependent flagellar genes remain By performing transposon mutagenesis with a to be analyzed to determine whether they are simistrain that contains a transcriptional fusion to fZgDE2, larly regulated. By assimilating the information rea blue-white screening procedure to identify genes of C. jejuni required for expression of a aS4-dependent garding expression of d4-dependent flagellar genes, a transcriptional regulatory cascade can be proposed operon was created. Three classes of transposon mufor C. jejuni (Fig. 1).The formation and the possible tants were defective for expression of the flgDE2:: secretory function of the flagellar export apparatus astA transcriptional fusion. One class of mutants contained transposon insertions in rpoN (encoding d4), are likely the initial key elements atop this pathway. flgS or fgR. The latter two genes respectively encode It is hypothesized that an uncharacterized signal ema sensor histidine kinase and a response regulator anating from the flagellar export apparatus, possibly similar to the FlrBC and FleSR systems of Vibriocholwith FlhF, is sensed by the FlgS sensor kinase to result in autophosphoryaltion of the protein. A phoserae and Pseudomonas aeruginosa (Arora et al., 1997; Dasgupta et al., 2000, 2003; Jyot et al., 2002; Klose phorelay event likely occurs to result in the phosphorylation and activation of FlgR as a transcriptional and Mekalanos, 1998; Prouty et al., 2001). FlgR is regulator. This hypothesis has been supported by homologous to members of the NtrC family of prodemonstrating that FlgS can phosphorlyate FlgR in teins that interact with as4in the RNA polymerase vitro (Wosten et al., 2004). Upon activation, FlgR holoenzyme at specific promoters and promote open may then interact with d4in the RNA polymerase DNA complex formation to initiate expression of holoenzyme complex to promote open DNA comd4-dependent genes in bacteria. The second class of plex formation at target promoters for initiation of mutants was found to have transposon insertions in transcription. Several questions remain to be anflhA, flhB, flip, and fliR. These genes encode proteins swered regarding this putative cascade, with one of that make up the inner-membrane-localized flagellar export apparatus. In addition to flaB and flgDE2, the the most important questions centering on what the other proposed as4-dependent flagellar genes are all specific signal may be that originates from the flagel-
c.
552
HENDRIXSON
formation of the flagellar export apparatus and expression of 054dependent flagellar genes
OM
secretion of FlgM and relief of inhibition of expression of cPdependent flagellar genes
completion of the flagellar filament
1I rod
. I
IM
Flagellar Export Apparatus
I
Flagellar Export
Apparatus
f
t
FlgM
? FlgM
\@
l? a28
1
Flagellar
secretion of FlaA and other filament proteins
autophosphorylation
I
phosphorelay
I
&'-dependent expression of flagellar genes (major flagellin and minor flagellar components)
u"-dependent expression of flagellar genes (rod, hook genes)
Figure 1. Proposed regulatory cascade for expression of flagellar genes in campylobacters. The regulatory cascade is based on data acquired through analysis of C. jejuni strains. (Left) C. jejuni may lack a master regulator for expression of early flagellar genes. These early genes, which include those encoding d4,a28,FlgM, FlgS, FlgR, FlhF, and components of the flagellar export apparatus (FlhA, FlhB, Flip, FliR, FliO, FIiQ, and FliF), may be constitutively expressed. After formation of the flagellar export apparatus, FlgS may sense an undetermined signal to autophosphorylate and begin a signal transduction cascade, terminating in activation of FlgR and expression of d4-dependent flagellar genes. Expression of #-dependent flagellar genes results in production of the flagellar basal body and hook proteins, which complete formation of the flagellar secretory system. (Middle) Until the flagellar secretory system has formed, FlgM may inhibit the activity of u2*for expression of target genes. This inhibitory effect may be strain dependent in campylobacters. After formation of the secretory system, FlgM is likely transported out of the cytoplasm through this system. v28 is then relieved from inhibition and can function in expression of target genes that includes fluA, encoding the major flagellin. (Right) Secretion of FlaA and other filament proteins occurs through the conduit formed by the flagellar export apparatus, rod, and hook to result in polymerization of the flagellar filament. OM, outer membrane; IM, inner membrane.
lar export apparatus to activate FlgS. In addition, it is unclear what constitutes the dependency of FlhF on this regulatory cascade. Considering that many of these proteins and similar regulatory pathways are present in various species of Pseudomonas, Vibrio, and Helicobacter, analysis of these elements in C. jejuni may provide an understanding of mechanisms governing regulation of flagellar genes in a broad set of bacteria. Regulation of Expression of flaA by u28 As described above, interruption of any gene encoding a protein required in the early steps of flagel-
lar biosynthesis such as the flagellar export apparatus, hook, or rod proteins in Salmonella species results in lack of transcription of a28-dependent flagellar genes such as those encoding the flagellins (Hughes et al., 1993; Kutsukake et al., 1990). Prevention of a2* activity in these mutants is mediated by FlgM, an anti-a factor that binds to a28 and inhibits both the formation and activity of the U ~ ~ - R Npolymerase A holoenzyme (Chadsey et al., 1998; Hughes et al., 1993; Ohnishi et al., 1992). C. jejuni contains genes for both a28 and FlgM, and it would be expected that FlgM inhibits a28in any mutants that fail to produce a functional flagellar export apparatus, basal body, or hook structures to
CHAPTER 30
secrete FlgM. In initial analysis with C. jejuni 81-176, expression of flaA::astA was found to not be significantly reduced in the ArpoN, AflgR, AflgS, or export apparatus mutants (Hendrixson and DiRita, 2003). Additional experiments suggested that FlgM is produced in this strain, but that it only weakly inhibits the transcriptional activity of a28(Hendrixson and DiRita, 2003). In addition, fluA expression is only decreased twofold in this strain in the absence of uZ8, suggesting the possibility that a uzs-independent pathway for expression of flaA may exist. Analysis of other C. jejuni strains, however, revealed that flaA expression is significantly hindered in mutants defective for secretion of flagellar components. A mutant of C. jejuni NCTC 11168 lacking FlhA has lower detectable levels of flaA transcripts (Carrillo et al., 2004). In C. jejuni 81116, QoN, flgS, or flgR mutants that would be predicted to be unable to form a flagellar secretion system because of lack of expression of flagellar rod and hook genes result in a large reduction of expression of fldA (Wosten et al., 2004). Thus, the requirement for linking expression of fluA to the construction of the flagellar export apparatus and hook may have strain-to-strain variances even within Campylobacter species. In strains where the inhibitory activity of FlgM is present, cytoplasmic FlgM may inhibit u28until the expression of d4-dependent flagellar genes and formation of the flagellar export apparatus, basal body, and hook are complete. Secretion of FlgM out of the cytoplasm through this nascent flagellar structure would follow, relieving uZ8from inhibition to allow for expression of flaA and production of the complete flagellar filament (Fig. 1). Expression of Early Flagellar Components May Not Be Dependent on a Master Transcriptional Regulator A curious aspect regarding the regulation of early flagellar genes in campylobacters to come out of these global transposon mutagenesis approaches and analyses of genomic sequences is the apparent lack of a master transcriptional regulator. These master regulators are usually atop the flagellar regulatory cascade and are responsible for initiating the expression of the first set of flagellar genes to begin flagellar biosynthesis. Genomic sequences of campylobacters suggest the absence of this type of regulator and the transposon mutagenesis screens to identify genes required for motility or transcription of us4-dependent flagellar genes has yet to identify a regulator (Carrillo et al., 2004; Fouts et al., 2005; Hendrixson and DiRita, 2003; Parkhill et al., 2000; Wosten et al., 2004). However, it is possible that a master transcrip-
FLAGELLAR GENE EXPRESSION AND ASSEMBLY
553
tional regulator different from FlhDC of Salmonella entericu serovar Typhimurium, FlrA of V. cholerue, or FleQ of P. aeruginosa may exist in campylobacters or be an essential protein for viability. If a master regulator for flagellar gene transcription is indeed absent, it creates an interesting question regarding how expression of early components of flagellar biosynthesis (i.e., the flagellar export apparatus), genetic regulatory factors (such as uzs, d4, and the FlgSR two-component system), and other components such as the motor complex are coordinately and appropriately controlled. The question remains unanswered, but the organization of some of these genes may provide insight. Many of these genes are grouped into operons with genes involved in other processes that may be essential for growth and metabolism. Thus, expression of these early flagellar genes may be constitutive by being coexpressed with these other genes. Therefore, a master regulator would not be required for transcription of these flagellar genes if their expression is constitutive.
PHASE VARIATION OF FLAGELLAR MOTILITY IN CAMPYLOBACTERS The regulatory mechanisms to control expression of flagellar genes ensure that the bacterium produces a flagellar protein at the correct stage in flagellar biosynthesis. Flagellar motility in campylobacters is also affected by phase variation (Caldwell et al., 1985; Nuijten et al., 1989). In bacteria, phase variation alters the expression of genes or production of proteins through a random, reversible change in the length of usually short DNA sequence repeats. Some bacterial DNA sequence repeats consist of homopolymeric tracts, and others are composed of a discrete DNA sequence that repeats multiple times. These repeats may be in a promoter so that changes in the length of the repeats affect transcription of a gene. In other cases, the repeats are in the coding sequence of a gene, and alteration of these repeats disrupts the reading frame to interrupt translation of the protein. Thus, phase variation can exert transcriptional or translational controls. Phase variation of flagellar motility has been noticed for both C. jejuni and C. coli, but the genes that are subjected to phase variation are apparently different between the two species. Analysis of one nonmotile phase "off" variant of C. coli identified a phase-variable tract of thymine residues in the 5' end of the coding sequence of fiThA, encoding a component of the flagellar export apparatus (Park et al., 2000). The wild-type sequence of flhA contains a poly T tract of eight thymine residues, whereas the
554
HENDFUXSON
phase “ ~ f f )variant ’ contains seven residues, which results in a frameshift of the coding sequence to generate a premature stop codon. The use of a transcriptional fusion of the promoterless astA reporter gene to the as4-dependent flgDE2 operon in wild-type C. jejuni enabled the identification of spontaneous white colonies derived from wild-type bacteria when grown on agar with X-S (Hendrixson, 2006). Because expression of d4dependent flagellar genes is affected in these variants, it was hypothesized that at least one component of the regulatory pathway to activate this set of genes could be subjected to phase variation. Unlike C. coli, this factor is not flhA; C. jejuni does not possess the phase-variable poly T tract found in C. coli flhA. Instead, immunoblot analysis of phase “off” variants revealed the absence of the FlgR as4-dependent response regulator. Within the figR sequences of phase “off” variants, one of two homopolymeric tracts, a poly A tract of six residues and a poly T tract of eight residues, located approximately one-third into the coding sequence, was found to have lost a single nucleotide, causing a premature stop codon to terminate translation of FlgR. Revertants of nonmotile flgR phase “off” variants in which the poly T tract had been shortened revealed correct restoration of the reading frame by addition of a thymine in this tract. However, different strategies to repair phase “off” variants that shortened the poly A tract were found. Some phase revertants had correctly repaired the poly A tract by addition of an adenine to the site. In contrast, other phase “on” revertants had inserted a nucleotide 2 to 7 nucleotides upstream of the poly A tract to restore the reading frame of fgR with only minor amino acid changes that did not affect activity of FlgR as a transcriptional regulator. Phase variation of flgR from the phase “off” to phase “on” state was also shown to be biologically relevant because nonmotile flgR phase “off” variants, which have defects for commensal colonization of the chick intestinal tract, are able to revert to the motile flgR phase ‘‘on” state in vivo to result in high levels of colonization. Other phase “off” variants identified in this study were found to have wild-type flgR sequences and FlgR protein levels, suggesting that other factors of C. jejuni are subjected to phase variation to affect expression of as4-dependent flagellar genes, flagellar biosynthesis, and motility. Some C. jejuni strains contain a paralogous set of up to seven genes (originally annotated as maf genes) that were proposed to encode motility accessory factors (Karlyshev et al., 2002; Parkhill et al., 2000). Upon additional analysis, one Maf protein, Maf5, was found to be involved in O-linked glycosylation of flagellin subunits and was renamed PseE
(Guerry et al., 2006; McNally et al., 2006). Two maf genes, mafl and maf4, have a poly G tract of 11 nucleotides within their coding sequence to promote the phase state. However, in the wild-type motile NCTC 11168 strain, these genes were found to be in the phase “off” state by lacking a guanidine residue in the poly G tract. Further analysis revealed that interruption of maf5lpseE resulted in a nonmotile, aflagellated phenotype (Karlyshev et al., 2002). Partial restoration of the maf5lpseE mutant to the motile state occurred by spontaneous switching of mafl to the phase state by the addition of a guanidine to the poly G tract. Thus, Mafl may at least partially functionally substitute for Maf5 /PseE in glycosylation of flagellin in strain NCTC 11168. However, the role of maf genes in motility, phase variation of motility, and flagellin glycosylation is not universal to all C. jejuni strains because many strains do not contain the complete set of mafgenes that are present in NCTC 11168.
FLAGELLAR ASSEMBLY IN CAMPYLOBACTERS Assembly Comparisons to Other Bacteria The actual process by which the flagellar components are secreted and interact with each other to form the flagellar filament has not been studied in much depth in campylobacters. In comparison to other flagellated bacteria, campylobacters appear to produce the set of proteins that form a typical flagellar export apparatus and a complete basal body structure consisting of the periplasmic rod and ring proteins that together are thought to be necessary for protein secretion. In a few studies with flgE mutants that lack the flagellar hook protein, the bacteria are aflagellate, demonstrating that the hook is necessary for flagellar filament assembly (Hendrixson and DiRita, 2003; Kinsella et al., 1997; Konkel et al., 2004). Thus, many of the basic requirements for flagellar protein secretion and assembly required in other bacteria are likely to hold true for campylobacters. Polar Assembly of Flagella and Regulation of Flagellar Number As in Vibrio, Helicobacter, and Pseudomonas species, a major question in the biology of campylobacters is how flagellar biosynthesis is restricted to the poles of the bacteria. The best insight into the process of polar localization has been revealed by analysis of the FlhF protein in P. aeruginosa (Murray and Kazmierczak, 2006). Mutants of P. aeruginosa
CHAPTER 30
lacking flhF are nonmotile as a result of the mislocalization of the flagellum to lateral sites on the bacterial surface, which causes the bacteria to rotate or twirl around a fixed point. These results suggest that FlhF is a major factor in influencing the placement of the flagellum at polar sites in this bacterium, but the mechanism by which the protein is involved in this polar localization is not known. FlhF proteins of different bacteria contain a domain that is most homologous to the GTP-binding domain of the GTPases of the signal-recognition particle system, Ffh and FtsY. These proteins function together to target nascently translated proteins to the Sec translocase for insertion into the inner membrane (Froderberg et al., 2003; Luirink et al., 1994; Powers and Walter, 1997; Ribes et al., 1990; Seluanov and Bibi, 1997; Valent et al., 1998). Considering this homology between FlhF and the signal-recognition particle GTPases, it is conceivable that FlhF could participate in ensuring that secretion of some flagellar proteins occurs specifically at the poles of P. aeruginosa. In contrast to the phenotype of the flhF mutant in P. aeruginosa, a flhF mutant in C. jejuni does not assemble flagella and is defective for expression of as4-dependent flagellar genes (Hendrixson and DiRita, 2003), similar to the phenotypes of flhF mutants in H. pylori and V. cholerae (Correa et al., 2005; Niehus et al., 2004). Thus, FlhF appears to perform an alternative role in flagellar biosynthesis in these organisms by being involved in the expression of flagellar genes. It is possible that FlhF of C. jejuni, H. pylori, and V. cholerae may have a second function by being involved in the correct secretion and polar placement of flagellar proteins. These organisms may have incorporated FlhF into their flagellar regulatory cascades to ensure flagellar genes are transcribed only if FlhF is present to influence the placement of the flagellum at the poles. More research is required to understand the role of FlhF in flagellar gene regulation and biosynthesis. A second important question regarding flagellar biosynthesis in campylobacters is how they ensure that only one flagellum is constructed at each pole. Flagellar number appears to be controlled in P. aeruginosa and V. cholerae by the FleN and FlhG proteins, respectively (Correa et al., 2005; Dasgupta and Ramphal, 2001; Dasgupta et al., 2000). Despite these proteins being homologous to each other, they appear to regulate flagellar number by different mechanisms. A fleN mutant of P. aeruginosa produces three to six flagella compared with the single flagellum produced by the wild-type strain (Dasgupta et al., 2000). FleN limits flagellar numbers by binding to the P. aeruginosa master regulator FleQ, resulting in a reduction of transcription of FleQ-dependent op-
FLAGELLAR GENE EXPRESSION AND ASSEMBLY
555
erons (Dasgupta and Ramphal, 2001). These operons contain genes encoding the flagellar export apparatus and basal body components. FleN does not inhibit FleQ from binding target promoters but appears to downregulate the activity of the protein by another mechanism (Dasgupta and Ramphal, 2001). FlhG of V. cholerae, however, appears to inhibit the expression of flrA, encoding the master regulator, by an unknown mechanism (Correa et al., 2005). Thus, FlhG limits the production of FlrA, which lowers the expression of target genes to limit V. cholerae to one flagellum. cjOO63c in the C. jejuni NCTC 11168 genome (Parkhill et al., 2000) and cjj81276j-0102 in the C. jejuni (Fouts et al., 2005) encode homologues of FleN of P. aeruginosa and FlhG of V. cholerae. However, evidence is lacking for a cognate master regulator that the FleN/FlhG homologue may influence by an analogous mechanism in P. aeruginosa or V. cholerae. Thus, if the FleN/FlhG homologue controls flagellar number in campylobacters, it may do so by an alternative mechanism. Influence of the Flagellin Glycosylation System in Flagellar Assembly A peculiar characteristic of the flagella of Campylobacter species such as C. coli, C. jejuni, and C. fetus is that the flagellin proteins are modified by a 0-linked protein glycosylation system (Doig et al., 1996; Goon et al., 2003; Guerry et al., 1996, 2006; Linton et al., 2000; Logan et al., 2002; McNally et al., 2006, 2007; Thibault et al., 2001). At least two different carbohydrates and their associated derivatives decorate the flagellins, depending on the species (Logan et al., 2002; McNally et al., 2007; Thibault et al., 2001). One modification consists of pseudaminic acid (PseSAc7Ac) or its acetamidino-substituted derivative, PseAm (Thibault et al., 2001). The pathway for pseudaminic acid biosynthesis is encoded by a set of genes (annotated as the pse genes) that have been extensively analyzed in C. jejuni (Goon et al., 2003; Guerry et al., 2006; McNally et al., 2006). A second set of genes (annotated as the ptm genes) in certain species such as C. coli allows for the generation of legionaminic acid and associated derivatives (Guerry et al., 1996; McNally et al., 2007). Detailed analyses have been performed on the mechanisms of 0-linked flagellar glycosylation (chapter 26). An interesting aspect of this system is that flagellar biosynthesis is dependent on glycosylation, and there may be some dependency on the PseSAc7Ac epitope rather than other eptiopes such as PseAm or legionaminic acid (Goon et al., 2003; Guerry et al., 2006;
556
HENDRIXSON
Logan et al., 2002; McNally et al., 2006, 2007; Thibault et al., 2001).
CONCLUSIONS Much progress has been made in the last decade in identifying proteins of campylobacters required for flagellar motility and understanding the roles of these proteins in flagellar gene regulation, biosynthesis of the organelle, and chemotaxis. Campylobacters appear to contain similarities with other bacterial flagellar systems to regulate expression of flagellar genes and construct the flagellum. Yet these bacteria continue to demonstrate that they utilize proteins and have developed processes unique to themselves to result in a distinct pathway for construction of flagella and the process of motility. Future studies will be sure to provide insights into flagellar organelle development in campylobacters. REFERENCES Arora, S . K., B. W. Ritchings, E. C. Almira, S . Lory, and R Ramphal. 1997. A transcriptional activiator, FleQ, regulates mucin adhesion and flagellar gene expression in Pseudomonas aeruginosa in a cascade manner. J. Bacteriol. 1795574-5581. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157472-479. Blair, D. F. 1995. How bacteria sense and swim. Annu. Rev. Microbiol. 49:489-522. Bleumink-Pluym, N. M., F. Verschoor, W. Gaastra, B. A. van der Zeijst, and B. N. Fry. 1999. A novel approach for the construction of a Campylobacter mutant library. Microbiology. 145: 2145-215 1. Caldwell, M. B., P. Guerry, E. C. Lee, J. P. Burns, and R. I. Walker. 1985. Reversible expression of flagella in Campylobacter jejuni. Infect. Immun. 50:941-943. Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R. Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K. Amoako, S . Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Gemone-wide expression analyses of Campylobacter jejuni NCTClll68 reveals coordinate regulation of motility and virulence by PhA. J. Biol. Chem. 279:20327-20338. Chadsey, M. S., J. E. Karlinsey, and K. T. Hughes. 1998. The flagellar anti-a factor FlgM actively dissociates Salmonella tyA holoenzyme. Genes. Dev. 12: phimurium O ~ ~ - R Npolymerase 3 123-3 136. Chang, C., and J. F. Miller. 2006. Campylobacter jejuni colonization of mice with limited enteric flora. Infect. Immun. 74: 5261-5271. Chilcott, G. S., and K. T. Hughes. 2000. Coupling of flagellar gene expression to flagellar assembly in Salmonella enterica serovar Typhimurium and Escherichia coli. Microbiol. Mol. Biol. Rev. 64: 694-708. Colegio, 0. R., T. J. Griffin, N. D. Grindley, and J. E. Galan. 2001. In vitro transposition system for efficient generation of random mutants of Campylobacter jejuni. J. Bacteriol. 183: 2384-238 8.
Correa, N. E., F. Peng, and K. E. Klose. 2005. Roles of the regulatory proteins FlhF and FlhG in the Vibrio cholerae flagellar transcription hierarchy. J. Bacteriol. 187:6324-6332. Dasgupta, N., S. K. Arora, and R. Ramphal. 2000. fleN, a gene that regulates flagellar number in Pseudomonas aeruginosa. J. Bacteriol. 182:3 57-3 64. Dasgupta, N., and R Ramphal. 2001. Interaction of the antiactivator FleN with the transcriptional activator FleQ regulates flagellar number in Pseudomonas aeruginosa. J. Bacteriol. 183: 6636-6644, Dasgupta, N., M. C. Wolfgang, A. L. Goodman, S . K. Arora, J. Jyot, S . Lory, and R. Ramphal. 2003. A four-tiered transcriptional regulatory circuit controls flagellar biogenesis in Pseudomonas aeruginosa. Mol. Microbiol. 50:809-824. Doig, P., N. Kinsella, P. Guerry, and T. J. Trust. 1996. Characterization of a post-translational modification of Campylobacter flagellin: identification of a sero-specific glycosyl moiety. Mol. Microbiol. 19:379-3 87. Fan, F., K. Ohnishi, N. R. Francis, and R. M. Macnab. 1997. The Flip and FliR proteins of Salmonella typhimurium, putative components of the type I11 flagellar export apparatus are located in the flagellar basal body. Mol. Microbiol. 26:1035-1046. Ferrero, R. L., and A. Lee. 1988. Motility of Campylobacterjejuni in a viscous environment: comparison with conventional rodshaped bacteria. ]. Gen. Microbiol. 13463-59. Froderberg, L., E. Houben, J. C. Samuelson, M. Chen, S.-K. Park, G. J. Phillips, R Dalbey, J. Lurink, and J.-W. de Gier. 2003. Versatility of inner membrane protein biogenesis in Escherichia coli. Mol. Microbiol. 47:1015-1027. Fouts, D. E., E. F. Mongodin, R E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel., L. M. Brinkac, R. T. DeBoy, C. T. Parker, S . C. Daugherty, R. J. Dodson, A. S . Durkin, R. Madupu, S . A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Gillen, K. L., and K. T. Hughes. 1991. Molecular characterization of flgM, a gene encoding a negative regulator of flagellin synthesis in Salmonella typhimurium. J. Bacteriol. 173:6453-6459. Glenn-Calvo, E., W. Bar, and M. Frosch. 1994. Isolation and characterization of the flagellar hook of Campylobacter jejuni. FEMS Microbiol. Lett. 123:299-304. Golden, N. J., and D. W. Acheson. 2002. Identification of motility and autoagglutination Campylobacter jejuni mutants by random transposon mutagenesis. Infect. lmmun. 70:1761-1771. Golden, N. J., A. Camilli, and D. W. Acheson. 2000. Random transposon mutagenesis of Campylobacter jejuni. Infect. lmmun. 685450-5453. Goon, S., C. P. Ewing, M. Lorenzo, D. Pattrini, G. Majam, and P. Guerry. 2006. A ds-regulated nonflagella gene contributes to virulence of Campylobacter jejuni 8 1-176. Infect. lmmun. 74: 769-772. Goon, S., J. F. Kelly, S . M. Logan, C. P. Ewing, and P. Guerry. 2003. Pseudaminic acid, the major modification on Campylobacter flagellin, is synthesized via the Cj1293 gene. Mol. Microbiol. 50:659-671. Grant, C. C. R., M. E. Konkel, W. Cieplak, Jr., and L. S. Tomkins. 1993. Role of flagella in adherence, internalization, and translocation of Carnpylobacterjejuni in nonpolarized and polarized epithelial cell cultures. Infect. Immun. 61:1764-1771. Guerry, P., R A. Alm, M. E. Power, S . M. Logan, and T. J. Trust. 1991. Role of two flagellin genes in Campylobacter motility. J. Bacteriol. 173:4757-4764. Guerry, P., P. Doig, R. A. Alm, D. H. Burr, N. Kinsella, and T. J. Trust. 1996. Identification and characterization of genes re-
CHAPTER 30
quired for post-translational modification of Campylobacter coli VC167 flagellin. Mol. Microbiol. 19:369-378. Guerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Pattarini, G. Majam, P. Thibault, and S. Logan. 2006. Changes in flagellin glycosylation affect Campylobacter autoagglutination and virulence. Mol. Microbiol. 60:299-311. Guerry, P., S. M. Logan, S. Thornton, and T. J. Trust. 1990. Genomic organization and expression of Campylobacter flagellin genes. J. Bacteriol. 172:1853-1860. Henderson, M. J., and F. H. Milazzo. 1979. Arylsulfatase in Salmonella typhimurium: detection and influence of carbon source and tyramine on its synthesis.J. Bacteriol. 139:80-87. Hendrixson, D. R. 2006. A phase-variable mechanism controlling the Campylobacter jejuni FlgR response regulator influences commensalism. Mol. Microbiol. 61:1646-1659. Hendrixson, D. R., B. J. Akerley, and V. J. DiRita. 2001. Transposon mutagenesis of Campylobacter jejuni identifies a bipartite energy taxis system required for motility. Mol. Microbiol. 40: 214-224. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Hendrixson, D. R., and V. J. DiRita. 2003. Transcription of d4dependent but not uz8-dependent flagellar genes in Campylobacter jejuni is associated with formation of the flagellar secretory apparatus. Mol. Microbiol. 50:687-702. Hueck, C. J. 1998. Type 111 protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Bid. Rev. 62: 379-433. Hughes, K. T., K. L., Gillen, M. J. Semon, and J. E. Karlinsey. 1993. Sensing structural intermediates in bacterial flagellar assembly by export of a negative regulator. Science 262: 12771280. Jagannathan, A., C. Constantinidou, and C. W. Penn. 2001. Roles of rpoN, fliA, and flgR in expression of flagella in Campylobacter jejuni. J. Bacteriol. 183:2937-2942. Jones, M. A., K. L. Marston, C. A. Woodall, D. J. Maskell, D. Linton, A. V. Karlyshev, N. Dorrell, B. W. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni NCTC 11168 to high-level colonization of the avian gastrointestinal tract. Infect. lmmun. 72:3769-3776. Jyot, J., N. Dasgupta, and R. Ramphal. 2002. FleQ, the major flagellar gene regulator in Pseudomonas aeruginosa, binds to enhancer sites located either upstream or atypically downstream of the RpoN binding site. J. Bacteriol. 1845251-5260. Kalmokoff, M., P. Lanthier, T.-L. Tremblay, M. FOSS,P. C. Lau, G. Sanders, J. Autin, J. Kelly, and C. M. Szymanski. 2006. Proteomic analysis of Campylobacter jejuni 11168 biofilms reveals a role for the motility complex in biofilm formation. J. Bacteriol. 188~4312-4320. Karlinsey, J. E., S. Tanaka, V. Bettenworth, S. Yamaguchi, W. Boos, S . 4 . Aizawa, and K. T. Hughes. 2000. Completion of the hook-basal body complex of the Salmonella typhimurium flagellum is coupled to FlgM secretion and fliC transcription. Mol. Microbiol. 37: 1220-123 1. Karlyshev, A. V., D. Linton, N. A. Gregson, and B. W. Wren. 2002. A novel paralogous gene family involved in phase-variable flagella-mediated motility in Campylobacter jejuni. MicrobiolO ~ Y . 148:473-480. Karlyshev, A. V., and B. W. Wren. 2005. Development and application of an insertional system for gene delivery and expression in Campylobacter jejuni. Appl. Environ. Microbiol. 71: 4004-40 13. Kihara, M., T. Minamino, S. Yamaguchi, and R. M. Macnab. 2001. Intergenic suppression between the flagellar MS ring pro-
FLAGELLAR GENE EXPRESSION AND ASSEMBLY
557
tein FliF of Salmonella and FlhA, a membrane component of its export apparatus. J. Bacteriol. 183:1655-1662. Kinsella, N., P. Guerry, J. Cooney, and T. J. Trust. 1997. The flgE gene of Campylobacter coli is under the control of the alternaJ. Bacteriol. 179:4647-4653. tive u factor uS4. Klose, K. E., and J. J. Mekalanos. 1998. Distict roles of an alternative sigma factor during both free-swimming and colonizing phases of the Vibrio cholerae pathogenic cycle. Mol. Microbiol. 28:501-520. Konkel, M. E., J. D. Klena, V. Rivera-Amill, M. R. Monteville, D. Biswas, B. Raphael, and J. Mickelson. 2004. Secretion of virulence proteins from Campylobacter jejuni is dependent on a functional flagellar export apparatus. J. Bacteriol. 186:32963303. Kutsukake, K., Y. Ohya, and T. Iino. 1990. Transcriptional analysis of the flagellar regulon of Salmonella typhimurium. J. Bacteriol. 172:741-747. Lee, A., J. L. O’Rourke, P. J. Barrington, and T. J. Trust. 1986. Mucus colonization as a determinant of pathogenecity in intestinal infection by Campylobacter jejuni: a mouse cecal model. Infect. lmmun. 51536-546. Linton, D., A. V. Karlyshev, P. G. Hitchen, H. R. Morris, A. Dell, N. A. Gregson, and B. W. Wren. 2000. Multiple N-acetyl neuraminic acid synthetase (neuB) genes in Campylobacter jejuni: identification and characterization of the gene involved in sialyation of lipo-oligosaccharide. Mol. Microbiol. 35: 1120-1 134. Liu, R, and H. Ochman. 2007. The origins of flagellar gene operons and secondary flagellar systems. J. Bacteriol. 189:70987104. Logan, S. M., J. F. Kelly, P. Thibault, C. P. Ewing, and P. Guerry. 2002. Structural heterogeneity of carbohydrate modifications affects serospecificity of Campylobacter flagellins. Mol. Microbiol. 46587-597. Logan, S. M., T. J. Trust, and P. Guerry. 1989. Evidence for posttranslational modification and gene duplication of Campylobacter flagellin. J. Bacteriol. 171:3031-3038. Luirink, J., C. M. ten Hagen-Jongman, C. C. van der Weijden, B. Oudega, S. High, B. Dobberstein, and R Kusters. 1994. An alternative protein targeting pathway in Escherichia coli: studies on the role of FtsY. EMBO J. 13:2289-2296. Macnab, R. M. 2003. How bacteria assemble flagella. Annu. Rev. Microbiol. 5 7:77- 100. Marchant, J., B. Wren, and J. Ketley. 2002. Exploiting genome sequence: predictions for mechanisms of Campylobacter chernotaxis. Trends Microbiol. 10:155-159. Matz, C., A. H. M. van Vliet, J. M. Ketley, and C. W. Penn. 2002. Mutational and transcriptional analysis of the Campylobacer jejuni flagellar biosynthesis gene flhB. Microbiology. 148:16791685. McNally, D. J., A. J. Aubry, J. P. Hui, N. H. Khieu, D. Whitfield, C. P. Ewing, P. Guerry, J. R. Brisson, S. M. Logan, and E. C. Soo. 2007. Targeted metabolomics analysis of Campylobacter coli VC167 reveales legionaminic acid derivatives as novel flagellar glycans. J. Biol. Chem. 282:14463-14475. McNally, D. J., J. P. Hui, A. J. Aubry, K. K. Mui, P. Guerry, J. R. Brisson, S. M. Logan, and E. C. Soo. 2006. Functional characterization of the flagellar glycosylation locus in Campylobacter jejuni 81-176 using a focused metabolomics approach. J. Biol. Chem. 28 1:18489-1 8498. Miller, S., E. C. Pesci, and C. L. Pickett. 1993. A Campylobacter jejuni homolog of the LcrD/FlbF family of proteins is necessary for flagellar biosynthesis. Infect. Immun. 61:2930-2936. Minamino, T., T. Iino, and K. Kutuskake. 1994. Molecular characterization of the Salmonella typhimurium flhB operon and its protein products. J. Bacteriol. 176:7630-7637.
558
HENDRIXSON
Minamino, T., and R. M. Macnab. 1999. Components of the Salmonella flagellar export apparatus and classification of export substrates. J. Bacteriol. 181: 1388-1394. Misawa, N., and M. J. Blaser. 2000. Detection and characterization of autoagglutination activity by Campylobacter jejuni. Infect. Immun. 68:6168-6175. Murray, T. S., and B. I. Kazmierczak. 2006. FlhF is required for swimming and swarming in Pseudomonas aeruginosa. J. Bacteriol. 188:6995-7004. Nachamkin, I., X.-H. Yang, and N. J. Stern. 1993. Role of Campylobacter jejuni flagella as colonization factors for three-dayold chicks: analysis with flagellar murants. Appl. Environ. Microbiol. 59:1269-1273. Niehus, E., H. Gressmann, F. Ye, R. Schlapbach, M. Dehio, C. Dehio, A. Stack, T. F. Meyer, S. Suerbaum, and C. Josenhans. 2004. Genome-wide analysis of transcriptional hierarchy and feedback regulation in the flagellar system of Helicobacter pylori. Mol. Microbiol. 52:947-961. Nuijten, P. J. M., N. M. C. Bleumink-Pluym, W. Gaastra, and B. A. M. van der Zeijst. 1989. Flagellin expression in Campylobacter jejuni is regulated at the transcriptional level. Infect. Immun. 5 7: 1084- 1088. Nuijten, P. J., F. J. van Asten, W. Gaastra, and B. A. van der Zeijst. 1990. Structural and functional analysis of two Campylobacter jejuni flagellin genes. ]. Biol. Chem. 265:17798-17804. Ohnishi, K., F. Fan, G. J. Schoenhals, M. Kihara, and R. M. Macnab. 1997. The FliO, Flip, FliQ, and FliR proteins of Salmonella typhimurium: putative components for flagellar assembly.]. Bacteriol. 179:6092-6099. Ohnishi, K., K. Kutsukake, H. Suzuki, and T. Iino. 1992. A novel transcriptional regulation mechanism in the flagellar regulon of Salmonella typhimurium: an anti-sigma factor inhibits the activity of the flagellum-specific sigma factor, vF.Mol. Microbiol. 6 : 3149-3157. Pallen, M. J., C. W. Penn, and R. R. Chaudhuri. 2005. Bacterial flagellar diversity in the post-genomic era. Trends Microbiol. 13: 143-149. Pandza, S. M., M. Baetens, C. H. Park, T. Au, M. Keyhan, and A. Matin. 2000. The G-protein FlhF has a role in polar flagellar placement and general stress response induction in Pseudomonas putida. Mol. Microbiol. 36:414-423. Park, S.-F., D. Purdy, and S. Leach. 2000. Localized reversible frameshift mutation in the flhA gene confers phase variability to flagellin gene expression in Campylobacter coli.J. Bacteriol. 182: 207-210. Parkhill, J., B. W. Wren., K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K, Gagels, A. V. Karlyshev, S. Mode, M. J. Pallen, C. W. Penn, M. A. Quail, M.-A. Rajandream, K. M. Rutherford, A. H. M van Vliet, S. Whitehead, and B. G. Barrel]. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Phillips, M. W., and A. Lee. 1983. Isolation and characterization of a spiral bacterium from the crypts of rodent gastrointestinal tracts. Appl. Environ. Microbiol. 45:675-683. Poly, F., C. Ewing, S. Goon, T. E. Hickey, D. Rockabrand, G. Majam, L. Lee, J. Phan, N. J. Savarino, and P. Guerry. 2007. Heterogeneity of a Campylobacter jejuni protein that is secreted through the flagellar filament. Infect. Immun. 75:3859-3867. Power, M. E., R. A. Alm, and T. J. Trust. 1992. Biochemical and antigenic propertites of the Campylobacter flagellar hook protein. ]. Bacteriol. 174:3874-3883. Powers, T., and P. Walter. 1997. Co-translational protein targeting catalyzed by the Escherichia coli signal recognition particle and its receptor. EMBO J. 16:4880-4886.
Prouty, M. G., N. E. Correa, and K. E. Klose. 2001. The novel d4and v28-dependent flagellar gene transcription1 hierarchy of Vibrio cholerae. Mol. Microbiol. 39:1595-1609. Ribes, V., K. Romisch, A. Giner, B. Dobberstein, and D. Tollervey. 1990. E. coli 4.5s RNA is part of a ribonucleoprotein particle that has properties related to signal recognition particle. Cell 63: 591-600. Seluanov, A., and E. Bibi. 1997. FtsY, the prokaryotic signal recognition particle receptor homologue, is essential for biogenesis of membrane proteins. J. Biol. Chem. 272:2053-2055. Sommerlad, S. M., and D. R. Hendrixson. 2007. Analysis of the roles of FlgP and FlgQ in flagellar motility of Campylobacter jejuni. J. Bacteriol. 189:179-186. Song, Y. C., S. Jin, H. Louie, D. Ng, R. Lau, Y. Zhang, R. Weerasekera, S. A1 Rashid, L. A. Ward, S. D. Der, and V. L. Cban. 2004. FlaC, a protein of Campylobacter jejuni TGH9011 (ATCC43431) secreted through the flagellar apparatus, binds epithelial cells and influences cell invasion. Mol Microbiol. 53: 54 1-553. Thibault, P., S. M. Logan, J. F. Kelly, J. R. Brisson, C. P. Ewing, T. J. Trust, and P. Guerry. 2001. Identification of the carbohydrate moieties and glycosylation motifs in Campylobacter jejuni flagellin. J. Biol. Chem. 276:34862-34870. Titz, B., S. V. Rajagopala, C. Ester, R. Hauser, and P. Uetz. 2006. Novel conserved assembly factor of the bacterial flagellum. J. Bacteriol. 188:7700-7706. Valent, Q. A., P. A. Scotti, S. High, J.-W. de Gier, G. von Heigne, G. Lentzen, G. Wintermeyer, B. Oudega, and J. Luirink. 1998. The Escherichia coli SRP and SecB targeting pathways converge at the translocon. EMBO]. 17:2504-2512. Wand, M. E., R E. Sockett, K. J. Evans, N. Doherty, P. M. Sharp, K. R. Hardie, and K. Winzer. 2006. Helicobacter pylori FlhB functions: the FlhB C-terminal homologue HP1575 acts as a “spare part” to permit flagellar export when the HP0770 FlhBCC domain is deleted. J. Bacteriol. 188:7531-7541. Wang, Y., and D. E. Taylor. 1990. Natural transformation in Campylobacter species. 1.Bacteriol. 172:949-955. Wassenaar, T. M., N. M. C. Bleumink-Pluym, D. G. Newell, P. J. Nuijten, and B. A. M. van der Zeijst. 1994. Differential flagellin expression in a flaA flaB+ mutant of Campylobacter jejuni. Infect. Immun. 62:3901-3906. Wassenaar, T. M., N. M. C. Bleumink-Pluym, and B. A. M. van der Zeijst. 1991. Inactivation of Campylobacter jejuni flagellin genes by homologous recombination demonstrates that flaA but not flaB is required for invasion. EMBO J. 10:2055-2061. Wassenaar, T. M., B. A. M. van der Zeijst, R. Ayling, and D. G. Newell. 1993. Colonization of chicks by motility mutants of Cumpylobacter jejuni demonstrates the importance of flagellin A expression. J. Gen. Microbiol. 139:1171-1175. Wenman, W. M., J. Chai, T. J. Louie, C. Goudreau, H. Lior, D. G. Newell, A. D. Pearson, and D. E. Taylor. 1985. Antigenic analysis of Campylobacter flagellar protein and other proteins. ]. Clin.Microbiol. 2 1:108-1 12. Wosten, M. M. S. M., J. A. Wagenaar, and J. P. M. van Putten. 2004. The FlgS/FlgR two-component signal transduction system regulates the flu regulon in Campylobacter jejuni. J. Biol. Chem. 279~16214-16222. Yao, R, D. H. Burr, P. Doig, T. J. Trust, H. Niu, and P. Guerry. 1994. Isolation of motile and non-motile insertional mutants of Cumpylobacter jejuni: the role of motility in adherence and invasion of eukaryotic cells. Mol. Microbiol. 14:883-893. Yao, R, D. H. Burr, and P. Guerry. 1997. CheY-mediated modulation of Campylobacter jejuni virulence. Mol. Microbiol. 23: 1021-1031. Yao, R., and P. Guerry. 1996. Molecular cloning and site-specific mutagenesis of a gene involved in arylsulfatase production in Campylobacter jejuni. 1.Bacterol. 178:33 35-3 3 3 8.
Cumpyiobucter, 3rd ed.
Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 31
Natural Competence and Transformation in Campylobacter REBECCA s. WIESNER AND VICTOR J. DIRITA
nutrient source, (ii) acquisition of new sequences to aid in repair of damaged chromosomes, and (iii) acquisition of new genes to increase genetic diversity and fitness. These roles are not mutually exclusive, and it is likely that a combination of factors has led to the evolution and maintenance of transformation systems in competent bacteria. For detailed discussions of these potential roles, several reviews are available (Dubnau, 1991b, 1999; Lorenz and Wackernagel, 1994; Redfield, 1988, 1993; Redfield et al., 1997; Stewart and Carlson, 1986). Although the evolutionary advantage of competence is not definitively understood and may even differ among microbes, a case for it as a means of generating strain diversity in Campylobacter can be made on the basis of the large amount of genome variation found within this species. This is described briefly below and in much more detail in chapter 2.
Bacteria have three distinct mechanisms for transferring genetic material between individuals: conjugation, transduction, and transformation. In conjugation, DNA is transferred directly from cell to cell; during transduction, DNA is transferred by a bacteriophage; in transformation, naked DNA is directly taken up by the cell. The ability of bacteria to take up DNA from the environment and incorporate it heritably into their genomes is called competence. In this chapter, transformation will refer to the mechanism of DNA recognition or uptake, and competent or competence will refer to the physiological state in which transformation occurs. Over 40 naturally competent bacterial species are known, and that number continues to grow (Lorenz and Wackernagel, 1994). Almost all competent bacteria studied to date use components of the type I1 secretiodtype IV pilus biogenesis family of proteins for transformation, with slight differences between the transformation machinery in each system (Dubnau, 1999) (Table 1).The two known exceptions are Helicobacter pylori, which requires components of a type IV secretion/conjugation system for transformation (not included in Table l),and Campylobacter jejuni, in which some strains use both type 11-like and type IV-like secretion systems, as detailed in this chapter (Bacon et al., 2000; Hofreuter et al., 1998). In addition to subtle differences in the transformation machinery from organism to organism, differences exist in the regulation of competence and whether there is specificity for the type of DNA that can be taken up. These topics are also addressed in this chapter.
POPULATION STRUCTURE OF C . JEJUNI Numerous epidemiological studies that use a wide variety of typing methods have revealed extensive genetic variation among strains of C. jejuni. There are over 60 Penner heat-stable antigen serotypes and more than 100 Lior heat-labile antigen serotypes (Lior et al., 1982; Penner and Hennessy, 1980). Phenotypic variation among strains has been observed for many other traits, including invasion, sialylation of lipooligosaccharide (LOS), colonization of chickens, natural transformation, toxin production, and serum resistance (Blaser et al., 1986; Florin and Antillon, 1992; Korolik et al., 1998; Lindblom et al., 1990; Linton et al., 2000; McFarland and Neill, 1992; Wang and Taylor, 1990; Wilson et al., 2003; Wooldridge and Ketley, 1997). One study analyzed housekeeping fragments of seven genes (asd,
WHY COMPETENCE? Three explanations for the evolution of competence have been proposed: (i) the use of DNA as a
Rebecca S. Wiesner and Victor J. DiRita Department of Microbiology and Immunology & Unit for Laboratory Animal Medicine, University of Michigan, 5641 Medical Sciences 11, Ann Arbor, MI 48103.
559
560
WIESNER AND DIRITA
Table 1. Conserved proteins associated with transformation in naturally competent bacteria
-
Protein associated with transformation in the followine. bacterial suecies PSTC class or other conserved protein PSTC class I
Protein function
Traffic NTPase
I1 I11
Membrane protein Pseudopilins
IV
Pre-pilin peptidase/ methylase Secretin
V Other conserved proteins Lipoprotein DNA receptor Membrane channel DEAD family protein ssDNA binding protein
C. jejuni
CtsE CtsP(?) CtsF CtsG CtsT Cj1078
N. gonorrhoeae (31)
B. subtilis (17)
S. pneumoniae (52)
H. influenzae (68)
Cj0825
PilF PilT PilG PilE ComP PilV PiIH-PilL PilD
CtsD
PilQ
Not applicable Not applicable ComE (gram positive) (gram positive)
Not identified CjOOllc? Not identified Not identified DprA
PilP ComE ComA Not identified Not identified
Not applicable ComEA ComEC ComFA Smf
atpA, ddlA, efts, fumC, nuoH, and yphC) from 33 geographically distinct isolates of C. jejuni by multilocus sequence typing (Suerbaum et al., 2001). The number of unique alleles identified among these genes ranged from 9 to 15, and only two pairs of strains had identical sequences for all seven fragments (Suerbaum et al., 2001). For C. jejuni, the homoplasy ratio-an indication of the degree of clonality ranging from 0 (high) to 1 (low)-was determined to be between 0.36 and 0.48, indicating frequent intraspecies recombination (Suerbaum et al., 2001). These studies suggest that C. jejuni has a weakly clonal population structure and that much of the observed genetic heterogeneity is due to frequent intraspecies recombination. The amount of genetic diversity observed among strains of C. jejuni is similar to that observed in Neisseria meningitidis and is greater than that observed in Streptococcus pneumoniae, both of which are naturally competent for transformation (Enright and Spratt, 1998; Holmes et al., 1999). Comparative genomic hybridization studies, in which arrays from the first sequenced strain, NCTC 11168, were probed with DNA from several other strains, corroborated the conclusion that extensive genetic diversity occurs between strains. C. jejuni strain 43431 lacks 88 open reading frames (ORFs) found in strain NCTC 11168 and contains 130 apparently unique ORFs, while the 81-176 genome contains 87 unique ORFs relative to NCTC 11168 (Poly et al., 2004, 2005). Only one of the unique genes was present in both 81-176 and 43431 (Poly et al., 2004, 2005). By means of subtractive hybrid-
ComGA
CglA/CilDl
PilB
ComGB ComGC ComGD ComGE ComGG ComC
CglB/CilD2 CglC CglD
PilC PilA PilG PilJ
CilC
PilD
Not applicable CilEl / CelA CelB ComFA (CflA) CilB/DprA
Not identified ComE Rec2 Not identified DprA
ization, 23 genes present in strain 81116 but not in strain NCTC 11168 were identified (Ahmed et al., 2002). Another comparative genomic hybridization study comparing C. jejuni human isolates to strain NCTC 11168 revealed extensive genetic diversity between strains; 78 to 84% of the annotated genes examined from strain NCTC 11168 were common to all strains analyzed (Dorrell et al., 2001; Pearson et al., 2003). Of the genes in the variable gene pool, 50% were located in seven hypervariable plasticity regions (Pearson et al., 2003). A large-scale study that used comparative genomic hybridization of 5 1strains integrated with the data from three other microarray studies expanded the number of such variable C. jejuni genes to 559 (Dorrell et al., 2001; Leonard et al., 2003; Pearson et al., 2003; Taboada et al., 2004). Of these, 249 genes are uniquely variable in a single strain, and 350 are divergent across multiple strains (Taboada et al., 2004). Approximately half of the variable genes mapped to the variable regions mentioned above (Taboada et al., 2004). Overall, it appears that gene divergence in C. jejuni is primarily observed in a small number of variable genomic regions, suggesting that these multigene insertions and deletions may be the result of homologous recombination (Taboada et al., 2004). The variant nature of C. jejuni genomes has been confirmed with the release of nucleotide sequences of several other strains (Fouts et al., 2005a; Hofreuter et al., 2006) The stability of the C. jejuni genome has been examined in vivo after passage through chick intestines for 6 days. Of 12 strains examined, 2 were
CHAPTER 31
found to have different pulsed-field gel electrophoresis patterns than the inoculum, suggesting that intragenomic recombination occurs during intestinal colonization (Hanninen et al., 1999). A separate study of colonizing and noncolonizing strains of C. jejuni and Campylobacter coli by restriction endonuclease profiling detected small changes in the genomes of one C. jejuvzi and one C. coli strain after chicken colonization, providing further evidence of genomic rearrangement during colonization of the host (Korolik et al., 1998). Three different mechanisms may contribute to generating this type of genetic diversity. The first is local sequence change such as nucleotide substitutions or small insertions or deletions of one or a few nucleotides. Another mechanism is DNA rearrangement? where related sequences in the genome undergo recombination to create novel fusion genes, and duplicate or delete DNA segments. Finally, diversity can be generated by horizontal acquisition of DNA through mechanisms such as natural transformation, conjugation, and transduction (Arber, 2000). Although intrastrain mechanisms such as recombination play an important role in generating genome diversity in C. jejuni (Scott et al., 2007), its ability to carry out natural transformation likely accounts for some of this as well. Horizontal gene transfer by transformation between C. jejuni strains has been demonstrated. One study examined the flagellin gene locus, which consists of two genes, flaA and flaB. The former is required for motility, while the latter is thought to act as a donor for recombination into flaA (Alm et al., 1993; Wassenaar et al., 1994). With two C. jejuni mutants carrying different antibiotic resistance cassettes in flaA or flaB, both intragenomic recombination and intergenomic recombination were observed at this locus (Wassenaar et al., 1995). Intergenomic recombination occurred by coculturing the two strains and subsequently selecting for isolates with resistance to both antibiotics; when DNase was added during the period of coculture, the number of double-resistant mutants was significantly reduced, indicating that horizontal gene transfer occurred through transformation (Wassenaar et al., 1995). A similar study in C. coli also suggested that both intraand intergenomic events lead to recombination of the flagellin genes (Alm et al., 1993). Evidence of intergenomic rearrangements at the flagellin locus also comes from sequence analysis of the flaA gene in 18 C. jejuni strains (Harrington et al., 1997). DNA rearrangements in the flagellin locus were also observed during experimental colonization of chick ceca (Nuijten et al., 2000). Chicks were colonized with a flaA mutant and inoculated with wildtype strain 8 1116 several weeks after initial infection.
COMPETENCE AND TRANSFORMATION
561
Motile bacteria were analyzed and the flaA flaB locus was sequenced. Two different rearrangements were found, one that appeared to have resulted from an intragenomic recombination event and another that may have arisen through natural transformation or recombination during cellular division (Nuijten et al., 2000). The second type of rearrangement of the flagellin locus was observed more often in bacteria isolated from chicks after infection with wild-type C. jejuni (analyzed at day 35) than in bacteria isolated from chicks before infection with the wild-type strain (day 29) (Nuijten et al., 2000). This supports the hypothesis that the donor DNA for this rearrangement is from the wild type, although other explanations are that revertants to motility were simply selected and ultimately grew out or that a replication-dependent recombination event fortuitously occurred around the same time that the wild-type bacteria were added (Nuijten et al., 2000). Further support for natural transformation of C. jejuni strains during experimental infection of chickens came from an experiment monitoring exchange of antibiotic resistance cassettes inserted in two nonessential genes, hip0 and htrA (de Boer et al., 2002). Strains resistant to both markers were identified among the intestinal flora as early as 2 days after infection of 7-day-old chickens, and these remained in the intestinal tract for the duration of the experiment (29 days). Interspecies recombination also occurs in Campylobacter. Molecular typing of C. jejuni strains uncovered several strains with alleles identical or very similar (97% nucleotide identity) to those of C. coli strains, indicating that horizontal gene transfer can occur between species (Dingle et al., 2001; Schouls et al., 2003). Recombination within a population increases genetic diversity by generating new allele combinations, which can confer selective advantage to individuals within the population (Zhang et al., 2002). Although the observations described above are from experimental settings, an example of interstrain gene transfer that is more relevant to public health is the spread of antibiotic resistance. Ciprofloxacin resistance in C. jejuni has risen as the antibiotic has increasingly been used in the food supply of animals (Engberg et al., 2001; McDermott et al., 2002; Smith et al., 1999). Given the natural competence of most Campylobacter species, antibiotic resistance is probably spread in part through transformation.
TRANSFORMATION IN CAMPYLOBACTER Several species of Cumpylobacter are naturally competent for transformation, including C. jejuni
562
WIESNER A N D DIRITA
and C. coli (Wang and Taylor, 1990). In one study, five of five C. coli strains were found to be competent for transformation, with frequencies of approximately (per recipient cell), while three out of six C. jejuni strains were competent with a transformation frequency of approximately However, only donor DNA from one of the C. jejuni strains was used as a source of transforming DNA in those experiments (Wang and Taylor, 1990). In another study, strains of C. jejuni were tested for their abilities to be transformed by using isogenic DNA; three of these were competent for transformation (Wilson et al., 2003).
PHYSIOLOGY OF TRANSFORMATION Regulation of transformation differs among bacterial genera. Some bacteria, such as Neisseria gonorrhoeae, are competent throughout the growth curve, while competence development is a highly regulated process in others, such as Bacillus subtilis. Competence of C. jejuni is observed throughout the growth curve, but transformation frequencies are highest (lop4)during early exponential growth and decrease to approximately lop6 at entry to stationary phase (Wilson et al., 2003). C. coli is also competent for transformation throughout the growth curve, with slightly higher transformation frequencies observed in bacteria collected at early-log phase than those collected at late-log phase (Wang and Taylor, 1990). The microaerophilic lifestyle of C. jejuni prompted investigation into the effects of CO, levels on transformation during growth in a liquid shake culture. This revealed an inverse relationship between CO, level and transformation efficiency. Transformation frequencies of cells grown in 10% CO, decreased from to as C. jejuni proceeded through its growth curve, whereas those of cells grown at 0.7% CO, were transformed at frequencies as high as lo-, to (Wilson et al., 2003). Growth is more favorable for C. jejuni in the higher CO, concentrations (5 and lo%), so the increased transformation efficiency observed at 0.7% CO, raises the question of whether competence may be upregulated in response to general growth stress.
DNA DISCRIMINATION A key distinction among competent bacterial species is whether or not they discriminate between sources of donor DNA. Gram-positive bacteria such as B. subtilis and Streptococcus pneumoniae readily
take up foreign DNA (ie., from a donor source other than their own species). Several gram-negative species, including N.gonorrhoeae, Haemophilus influenzae, and Actinobacillus actinomycetemcomitans, discriminate against foreign DNA by recognition of a DNA element known as a DNA uptake sequence (DUS) present in multiple copies in their genomes. Other gram-negative species, such as Acinetobacter cakoaceticus, can be transformed by foreign DNA (Danner et al., 1980; Goodman and Scocca, 1988; Palmen et al., 1993; Smith et al., 1995; Wang et al., 2002). In general, C. jejuni is poorly electrotransformed with plasmid DNA from Escherichia coli when compared with plasmid DNA derived from C. jejuni (Miller et al., 1988; Wassenaar et al., 1993). One reason for this could be differences in restriction/modification systems between E. coli and C. jejuni. An EcoRIlike modification system was identified, but DNA methylated at the EcoRI site in vitro was unable to be efficiently electroporated into C. jejuni, suggesting the presence of multiple restriction and modification systems (Labigne-Roussel et al., 1987). Examination of the methylation state of DNA from 12 strains of C. jejuni, C. coli, C. hyointestinalis, C. fetus, and Campylobacter upsaliensis revealed GATC methylation in only three strains (Edmonds et al., 1992). Genome sequence analysis of five Carnpylobacter strains-C. jejuni RM1221, C. jejuni NCTC 11168, Campylobacter lari RM2100, C. upsaliensis RM3195, and C. coli RM2228-demonstrated differences in the complement of restriction-modification systems not only between the species, but also between strains of the same species (Fouts et al., 2005b). Strain-to-strain diversity in the number and composition of restriction-modification systems was also observed (Dorrell et al., 2001; Poly et al., 2004, 2005). The route of DNA entry during electroporation differs from the route of entry of DNA during the process of natural transformation, and although the presence of restriction systems may pose a barrier to electrotransformation, the role of these restriction systems in natural transformation of C. jejuni is not known. In other competent bacteria, incoming DNA may be protected from the action of restriction endonucleases and exonucleases (Lorenz and Wackernagel, 1994; Smith et al., 1981). It is not known whether this is the case during natural transformation of C. jejuni. Early studies of C. coli suggested that Carnpylobacter can discriminate self DNA from foreign DNA; this may occur by recognizing a DUS, although modification has not been ruled out (Wang and Taylor, 1990). C. coli efficiently takes up DNA derived
CHAPTER 31
from C. coli but not from E. coli (Wang and Taylor, 1990). In addition, uptake of radiolabeled C. coli DNA is inhibited by addition of excess unlabeled C. coli or C. jejuni DNA but not by unlabeled E. coli DNA (Wang and Taylor, 1990). Despite the apparent lack of uptake of E. coli DNA by C. coli, plasmids derived from both E. coli and C. coli transform C. coli with similar efficiency (Wang and Taylor, 1990). This discrepancy could be accounted for by a difference in sensitivity between the transformation assay and the radiolabeled uptake assays. Competition studies that used C. jejuni strain 81-176 showed that unmarked DNA derived from C. jejuni added to a transformation reaction reduces the number of transformants observed, while DNA from E. coli or C. coli does not. This again suggests that Campylobacter can discriminate self DNA from heterologous DNA (Wilson et al., 2003). Bioinformatic analysis of the C. jejuni strain NCTC 11168 genome identified no repeated sequences of the length and frequency that would be expected for a DUS (S. K. Highlander et al., unpublished data), so the mechanism of DNA discrimination in this species remains an open question. The lack of competition from C. coli DNA is surprising because several C. jejuni strains, including 8 1176, can be transformed with DNA derived from C. coli (Guerry et al., 1994; Wang and Taylor, 1990). The C. coli strain used by Wilson et al. was not among those tested, and perhaps the differences can be explained by the diversity in restrictionmodification systems present in different Campylobacter strains.
MOLECULAR BIOLOGY AND GENETICS OF COMPETENCE IN CAMPYLOBACTER JEJUNI Transformation mechanisms in bacteria typically require components of a large family of structurally and functionally related proteins involved in diverse functions such as type IV pilus biogenesis, type I1 secretion, and twitching motility. These proteins have been termed PSTC proteins (for pilus/secretion/ twitching motility/competence), and- the; fall into several classes (Dugnau, 1999). Table 1 shows the various known and suspected orthologues from several competent species, including C. jejuni. Proteins comprising the transformation machinery have been most extensively studied in the gram-positive bacterium B. subtilis and the gram-negative bacterium N. gonorrhoeae, although homologues are found in other competent species (Table 1); the nomenclature in these systems does not allow an easy correlation of proteins with homologous functions. One bacterium, H. pylori, does not use members of the PSTC
COMPETENCE AND TRANSFORMATION
563
family for transformation and instead uses components of a type IV secretion system, similar to at least one strain of C. jejuni; type IV secretion will be treated separately from the PSTC systems. Reviews on basic transformation mechanisms, as well as type I1 and type IV secretion systems in different bacteria, are available (Cascales and Christie, 2003; Chen and Dubnau, 2003; Chen et al., 2005; Dubnau, 1991a, 1999; Filloux, 2004; Johnson et al., 2006; Peabody et al., 2003; Sandkvist, 2001a, 2001b; Tortosa and Dubnau, 1999). Here, we restrict our discussion as much as possible to how the knowledge from other systems contributes to understanding competence and transformation in C. jejuni. C. jejuni strains sequenced to date have chromosomal genes encoding PSTC homologues often found in type I1 secretion systems. These are termed Cts, for Campylobacter transformation system (Fig. 1). The association of thkse genes with transformation was established in a genetic screen in strain 81176 for transposon mutants unable to take up a selectable marker (Wiesner et al., 2003). Six cts genes-ctsD, ctsE, ctsF, ctsP, ctsR, and ctsX-are encoded at a single locus (Wiesner et al., 2003), and the gene order is conserved in sequenced genomes of C. jejuni and other Campylobacter species (Fig. 1).CtsP and CtsE are similar to ATPases that presumably provide the energy required to build the apparatus or move molecules through it (Planet et al., 2001; Sandkvist, 2001b). CtsD is weakly homologous to a family of proteins in gram-negative bacteria, called secretins, that serve as outer membrane pores for extracellular delivery of macromolecules (Sandkvist, 2001b). The role of secretins in competence is not clear. PilQ, a secretin protein of Neisseria meningitidis, binds DNA with a preference for singlestranded DNA and is proposed to serve as a general DNA binding protein in the transformation process (Assalkhou et al., 2007). Transformation in Neisseria species requires a DUS, and the nonspecific nature of DNA binding by PilQ therefore suggests that it acts at a step in the pathway before sequence specificity is imposed (Assalkhou et al., 2007). CtsF is a conserved inner membrane protein found in all described type I1 secretion systems; a role for its activity has yet to be proposed (Sandkvist, 2001b). This protein may provide a physical link between cytoplasmic and periplasmic components of the type I1 apparatus. The structure of a member of this family of proteins, solved by three-dimensional electron microscopy, resulted in a model of the protein with two substantial domains, one cytoplasmic and the other periplasmic, separated by a domain that spans the inner membrane (Collins et al., 2007). The two other genes in this cts locus, ctsR and ctsX, are both required for transformation, yet they
564
WIESNER AND DIRITA
RM1221
81-176
84-25
- 7
T
-*
T
NCTC 11168
Figure 1. Conserved locus encoding cts type I1 secretion system genes from several strains of C. jejuxi. See text for details.
share no homology with PSTC genes from other microbes (Wiesner et al., 2003). CtsX is a membrane protein with its amino terminus inside the cell and carboxy terminus in the periplasm. Bacterial twohybrid studies on protein-protein interactions among the Cts PSTC proteins suggest that CtsX interacts with the putative ATPase CtsP, but the consequences of this interaction have not been determined. The interaction of one conserved PSTC protein (CtsP) with another that appears to be unique to Campylobacter (CtsX) suggests at least one novel feature of the transformation mechanism in C. jejuni. PSTC proteins are often part of assembly systems for surface pili that function in adherence; such pili are called type IV pili (not to be confused with type IV secretion, discussed below). The major subunit of type IV pili is derived from a prepilin molecule, which is processed by cleavage of its unique signal sequence and subsequent modification of the amino-terminal residue on the mature protein (Peabody et al., 2003; Sandkvist, 2001b). PSTC proteins called pseudopilins are generally thought to be evolutionarily related to pili. Pseudopilins have cleavable signal sequences processed by the same peptidase that
processes pilins. Further, when overexpressed, pseudopilins can form an oligomeric structure similar to a pilus (Vignon et al., 2003). A major hypothesis for the role of pseudopilins is that they act like pistons that may physically move substrates out of the secretion complex. The movement of pseudopilins is hypothesized to be driven by dynamic assembly and disassembly of the higher-order pseudopilin structure (Chen et al., 200.5), similar to how retraction of bona fide type IV pili may occur (Wolfgang et al., 2000). Extension and retraction of these pilus and pseudopilus structures through monomer assembly and disassembly is proposed to be similar to actin-based movement in eukaryotic cells (Wolfgang et al., 2000). Three ORFs in the C. jejuni NCTC 11168 share a signal sequence homologous with those of pilins and pseudopilins. Because C. jejuni evidently does not express a surface pilus (Wiesner et al.), these are presumed to encode pseudopilins associated with the type I1 system. In fact, two of them, ctsG and ctsT, were identified in the screen for transformation mutants (Wiesner et al., 2003), and the third, cj1078, is encoded directly downstream of ctsT (Wiesner et al., 2003). A prepilin peptidase of the type found in both
CHAPTER 31
competence and pilus biogenesis systems is also encoded in the C. jejuni genome (cj082.5 in the NCTC 11168 sequence database). A molecular basis for understanding how Cts proteins, or similar systems in other microbes, direct various steps in transformation is not yet established. Competence proteins associated with DNA binding, uptake, and recombination in B. subtilis have been localized to the poles of the cell, and a site of DNA binding can also be observed at the poles (Hahn et al., 2005). These findings suggest that-at least in B. subtilis-the transformation machinery is localized to a specific site of action. Whether this localization is a general feature of naturally transformable bacteria is of obvious interest, as is the mechanism by which it might occur. Other genes not typically encoded as part of type I1 secretion systems and unlinked from the major cts locus in C. jejuni were also identified by genetic screening for transformation-defective mutants (Wiesner et al.). One of these is CtsW, annotated in the C. jejuni genome database as a purine/pyrimidine phosphoribosyltransferase, and sharing a similar domain with the ComFC family of proteins. ComFC is a conserved but poorly understood constituent of the competence pathway in many bacteria (LondonoVallejo and Dubnau, 1993). Unlike other tested cts mutants, including those of the type I1 system in C. jejuni, a mutant strain lacking CtsW still takes up DNA into a DNase-resistant form, yet is unable to be transformed (Wiesner et al., 2003). This suggests that CtsW operates at a later step in the process than the other Cts gene products. One protein involved in C. jejuni transformation, CjOOllc, was unexpectedly identified by a biochemical screening for DNA binding proteins as a protein that might control expression of the cmeDEF operon (whose products confer resistance to some antimicrobial compounds; Jeon and Zhang, 2007). It is homologous to the Bacillus ComEA protein, a double-stranded DNA binding protein proposed to serve as a DNA receptor by spanning the membrane and binding DNA with its C-terminal domain (Inamine and Dubnau, 1995; Provvedi and Dubnau, 1999). CjOOllc is a periplasmic protein that binds both single-stranded and double-stranded DNA, with a higher affinity for the latter. Mutants lacking CjOOI Ic exhibited transformation frequencies an order of magnitude below that of wild type, although they were much less defective than many cts mutants of C. jejuni or comEA mutants of B. subtilis (Inamine and Dubnau, 1995; Wiesner et al., 2003). This prompted the hypothesis that, unlike B. subtilis, there is at least one other DNA receptor in C. jejuni (Jeon
COMPETENCE AND TRANSFORMATION
565
and Zhang, 2007). It should be noted, however, that study of CjOOllc has thus far been carried out with strain NCTC 11168, whereas the cts studies were carried out with strain 81-176.
pVir-ENCODED TYPE IV SECRETION SYSTEM Some strains of C. jejuni, including 81-176, carry a plasmid called pVir that encodes proteins homologous to type IV secretion systems in other bacteria (Bacon et al., 2002). These are more commonly used for conjugation or, in a few cases, delivery of proteins or protein-DNA complexes into eukaryotes. Other genes on pVir contribute to pathogenicity traits of C. jejuni 81-176, which presumably provides selection for maintaining the plasmid (Bacon et al., 2000). Two pseudogenes with homology to type IV secretion genes were also identified on the apparent relic of an integrated plasmid in the genome of strain 81-176 (Hofreuter et al., 2006). The best characterized bacterial type IV secretion system is the Vir system from the plant pathogen Agrobacterium tumefaciens, which mediates translocation of DNA and proteins from the bacterium into plant cells (reviewed in Cascales and Christie, 2003). The prevailing view of this system is that DNA transfer occurs as a by-product of transferring proteins to which DNA is covalently linked (Chen et al., 2005). A type IV secretion system used in transformation was first observed in Helicobacter pylori, which has chromosomally encoded Com proteins similar to the A. tumefaciens Vir proteins (Hofreuter et al., 1998). These are essential for transformation, although how they mediate uptake of DNA, rather than conjugation (which seems the opposite process), is not well understood. A second type IV secretion system in H. pylori, encoded at the cag pathogenicity island, is an important virulence determinant (Odenbreit et al., 2000) but does not contribute to transformation. As with the Cts PSTC proteins encoded on the C. jejuni chromosome, the Vir/ Com proteins encoded on pVir clearly contribute to natural competence in C. jejuni but do not suggest a definitive mechanism, at least on the basis of results of work on other microbes. For example, a pilus structure in A. tumefaciens-like the pili and pseudopili of the type I1 secretion systems-is hypothesized to push substrates through a pore. The gene encoding this, virB2, has a homologue in the H. pylori com system (comBZ), but it is found neither among the virlcom homologues on the C. jejuni pVir plasmid nor on the C. jejuni chromosome. Further, the putative VirB11
566
WIESNER AND DIRITA
ATPase necessary for transformation in H. pylori is dispensable for transformation in C. jejuni (Bacon et al., 2000; Wiesner et al., 2003). Thus, although pVir encodes transformation genes similar to the type IV secretion system transformation system of H. pylori, there are clearly significant differences in how these gene products contribute to the process in the two genera. Although mutation of virBl1 did not affect the transformation efficiency, an insertion in another type IV secretion homologue on pVir, comB3 (subsequently renamed virB2 0), reduced transformation efficiency approximately fourfold relative to the wildtype levels (Bacon et al., 2000). Mutation of a comB3 homologue (later renamed comB10) in H. pylori reduced the transformation efficiency to 1% of the wild-type levels (Hofreuter et al., 1998, 2001). The less severe phenotype observed by mutation of cornB3 in C. jejuni suggests either that there is another protein in the cell that can substitute for ComB3, or that that the type IV secretion system is not the primary transformation system in C. jejuni. The pVir competence system is unique for being plasmid based, as other microbes do not have competence genes on extrachromosomal elements. That some strains of C. jejuni have both pVir and the chromosomal cts genes-providing two mechanisms for transformation-is unusual and raises questions of how (or, perhaps, whether) the function of the virl corn type IV and cts type I1 secretion systems are coordinated by the cell. However, the relative magnitude of transformation defects in strains lacking either components of the pVir-encoded transformation system or the chromosomal type 11-like Cts system (Wiesner et al., 2003), as well as the broad conservation of the cts genes across C. jejuni strains (Fig. l), suggests that the Cts system provides the major mechanism for DNA uptake and transformation in C. jejuni.
CANDIDATE GENE ANALYSES TO IDENTIFY TRANSFORMATION PROTEINS
C. jejuni ORFs homologous to those required for various steps in transformation or development of competence in other species have been tested for their roles in C. jejuni transformation on an individual basis by analysis of mutants. One of these is DprA, whose importance in transformation has been well documented in other species, and whose basic mechanism of action is being uncovered (Ando et al., 1999; Campbell et al., 1998; Karudapuram et al., 1995; Smeets et al., 2000). In H. pylori, DprA is in-
volved in transformation with both chromosomal and plasmid DNA, while in H. influenzae, DprA is only required for transformation with chromosomal DNA (Ando et al., 1999; Karudapuram et al., 1995; Smeets et al., 2000). DprA is one of the proteins demonstrated by fluorescent imaging to be localized to the pole in B. subtilis, although its primary sequence suggests that it would be a cytosolic protein (Kramer et al., 2007). It is proposed to be part of a large complex of interacting proteins that make up the putative polar-localized transformation system in B. subtilis. C. jejuni NCTC 11168 with an insertion mutation in dprA displays transformation frequencies approximately 100 to 1,000 times lower than the parental strain (Takata et al., 2005). dprA mutants of C. jejuni strains NCTC 11168 and strain 480-which is unable to be transformed through natural competence but can be transformed by electroporation-are both poorly electroporated with chromosomal and plasmid DNA (Takata et al., 2005). DprA is postulated to serve as a recombination mediator protein that enables the recombination protein RecA (also required for transformation; see below) to bind to single-stranded DNA before formation of recombination joint molecules (MortierBarriere et al., 2007). The ubiquitous presence of dprA homologues within bacterial genomes, even of those species for which transformation has not been observed, may suggest that the protein serves a basic function in the cell. On the other hand, if DprA mediates recombination specifically for transformation, as suggested (Mortier-Barriere et al., 2007), its widespread presence in bacteria may mean that transformation mechanisms exist in species that have not yet been shown to exhibit natural competence. The candidate gene approach identified recombinase RecA as having a function in transformation. recA mutants were unable to be transformed to streptomycin resistance in three C. jejuni strains tested, 81-176, VC83, and 81-116 (Guerry et al., 1994). The galE gene product is also required for transformation because mutations in galE reduce the transformation efficiency to less than 5% of the wild-type level (Fry et al., 2000). Because GalE is involved in LOS biosynthesis, a C. jejuni galE mutant produces LOS molecules with smaller lipid A cores than those of wild type (Fry et al., 2000). The transformation defect could result from changes in LOS itself, if LOS per se participates in the transformation process. Alternatively, and probably more likely, the galE mutant may have reduced transformation efficiency as a result of an altered outer membrane composition that may interfere with the assembly or stability of the transformation complex.
CHAPTER 31
N-LINKED PROTEIN GLYCOSYLATION AND TRANSFORMATION Modification of C. jejuni proteins by addition of an N-linked glycan is due to the activity of a number of genes found in the pgl locus (Szymanski et al., 2003). N-linked protein glycosylation is an atypical posttranslational modification in prokaryotes, and although its mechanism is well understood in C. jejuni, the biological advantage it may confer is not. Many complex phenotypes are affected by mutation in the pgl system, including competence (Larsen et al., 2004). The pVir-encoded protein ComB3IVirB10 is glycosylated at two asparagine residues. Alteration of one of these sites (N97) led to a fourfold decrease in transformation efficiency relative to wild type, a defect similar to that observed with a virBl0 null mutant (Larsen et al., 2004). Glycosylation of the other site (N32) does not appear to be important for VirBlO function because a mutant with the N32A substitution transformed with wild-type efficiency (Larsen et al., 2004). It appears that VirB10 needs to be glycosylated at N97 in order for it to function in transformation, although the reason for this is not clear. Because ComB3IVirB10 requires glycosylation at one of two sites to function in transformation, the role in transformation of two key pgl genes was studied (Larsen et al., 2004). Mutants carrying kanamycin cassette insertions in pglB and pglE exhibited transformation efficiencies approximately 100 times lower than that of the wild type (Larsen et al., 2004). These mutants also expressed undetectable levels of ComB3 /VirB10, suggesting that glycosylation is required for protein expression or stability. The effect on ComB31VirB10 may be indirect, however, because unlike the result in a pgl mutant, removing the glycosylation sites from VirBlO did not dramatically reduce its steady-state levels. The magnitude of the transformation defect of the pgl mutants cannot be explained by loss of ComB3 /VirB10 glycosylation alone, suggesting that other proteins involved in transformation may need to be glycosylated to be fully functional.
CONCLUSIONS Transformation is a complex process that requires an ordered pathway of DNA recognition, uptake, and, usually, recombination. Even in the beststudied microbes, we have only a general picture of how the process occurs. DNA is taken into the cell and ultimately incorporated into the chromosome, and like other competent microbes, C. jejuni uses ma-
COMPETENCE AND TRANSFORMATION
567
chinery consisting of proteins that span multiple cell compartments. The similarity of the components of this process to components of pilus assembly and protein secretion systems clearly suggests that moving molecules either in or out of the cell poses a mechanistically similar challenge. Acknowledgments. Supported in part by a grant from the USDA Food Safety Program. R.S.W. was a trainee of the Genetics Training Program at University of Michigan. We thank K. T. Young and Lindsay Davis for helpful comments. Analysis of C. jejuni genomes was carried out with tools available from the National Microbial Pathogen Data Resource (http://www.nmpdr.org/index.php).
REFERENCES Ahmed, H. I., G. Manning, T. M. Wassenaar, S. Cawthraw, and D. G. Newell. 2002. Identification of genetic differences hetween two Campylobacter jejuni strains with different colonization potentials. Microbiology 148: 1203-1212. Alrn, R A., P. Guerry, and T. J. Trust. 1993. Significance of duplicated flagellin genes in Campylobacter.]. Mol. Biol. 230:359363. Ando, T., D. A. Israel, K. Kusugami, and M. J. Blaser. 1999. HP0333, a member of the dprA family, is involved in natural transformation in Helicobacter pylori. J. Bacteriol. 181:55725580. Arber, W. 2000 Genetic variation: molecular mechanisms and impact on microbial evolution. FEMS Microbiol. Rev. 24:l-7. Assalkhou, R., S. Balasingham, R. F. Collins, S. A. Frye, T. Davidsen, A. V. Benam, M. BjsrHs, J. P. Derrick, and T. Tsnjum. The outer membrane secretin PilQ from Neisseria meningitidis binds DNA. Microbiology 153:1593-1603. Bacon, D. J., R. A. Alrn, D. H. Burr, L. Hu, D. J. Kopecko, C. P. Ewing, T. J. Trust, and P. Guerry. 2000. Involvement of a plasmid in virulence of Campylobacter jejuni 8 1-176. Infect. Immun. 68:4384-4390. Bacon, D. J., R. A. Alm, L. Hu, T. E. Hickey, C. P. Ewing, R. A. Batchelor, T. J. Trust, and P. Guerry. 2002. DNA sequence and mutational analyses of the pVir plasmid of Campylobacter jejuni 8 1-176. Infect. Immun. 70:6242-6250. Blaser, M. J., G. P. Perez, P. F. Smith, C. Patton, F. C. Tenover, A. J. Lastovica, and W. I. Wang. 1986. Extraintestinal Campylobacter jejuni and Campylobacter coli infections: host factors and strain characteristics.J. Infect. Dis. 153552-559. Campbell, E. A., S. Y. Choi, and H. R. Masure. 1998. A competence regulon in Streptococcus pneumoniae revealed by genomic analysis. Mol. Microbiol. 27:929-939. Cascales, E., and P. J. Christie. 2003. The versatile bacterial type IV secretion systems. Nut. Rev. Microbiol. 1:137-149. Chen, I., P. J. Christie, and D. Dubnau. 2005. The ins and outs of DNA transfer in bacteria. Science 310:1456-1460. Chen, I., and D. Dubnau. 2003. DNA transport during transformation. Front. Biosci. 8:s544-s556. Collins, R. F., M. Saleem, and J. P. Derrick. 2007. Purification and three-dimensional electron microscopy structure of the Neisseria meningitidis type IV pilus biogenesis protein PUG. 1. Bacteriol. 189~6389-6396. Danner, D. B., R. A. Deich, K. L. Sisco, and H. 0. Smith. 1980. An eleven-base-pair sequence determines the specificity of DNA uptake in Haemophilus transformation. Gene 11:311-318. de Boer, P., J. A. Wagenaar, R. P. Achterberg, J. P. Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:351-359.
568
WIESNER AND DIRITA
Dingle, K. E., F. M. Colles, D. R. Wareing, R. Ure, A. J. Fox, F. E. Bolton, H. J. Bootsma, R. J. Willems, R. Urwin, and M. C. Maiden. 2001. Multilocus sequence typing system for Campylobacter jejuni. J. Clin. Microbiol. 39: 14-23. Dorrell, N., J. A. Mangan, K. G. Laing, J. Hinds, D. Linton, H. Al-Ghusein, B. G. Barrell, J. Parkhill, N. G. Stoker, A. V. Karlyshev, P. D. Butcher, and B. W. Wren. 2001. Whole genome comparison of Campylobacter jejuni human isolates using a lowcost microarray reveals extensive genetic diversity. Genome Res. 11~1706-1715. Dubnau, D. 1999. DNA uptake in bacteria. Annu. Rev. Microbiol. 53 :217-244. Dubnau, D. 1991a. Genetic competence in Bacillus subtilis. Microbiol. Rev. 55 :395-424. Dubnau, D. 1991b. The regulation of genetic competence in Bacillus subtilis. Mol. Microbiol. 5:ll-18. Edmonds, P., B. M. Hall, W. R. Edwards, and K. M. Hartline. 1992. Presence of methylated adenine in GATC sequences in chromosomal DNAs from Campylobacter species. J. Bacteriol. 174~8156-8157. Engberg, J., F. M. Aarestrup, D. E. Taylor, P. Gerner-Smidt, and I. Nachamkin. 2001. Quinolone and macrolide resistance in Campylobacter jejuni and C. coli: resistance mechanisms and trends in human isolates. Emerg. Infect. Dis. 7:24-34. Enright, M., and B. G. Spratt. 1998. A multilocus sequence typing scheme for Streptococcus pneumoniae: identification of clones associated with seriou invasive disease. Microbiology 144:30493060. Filloux, A. 2004. The underlying mechanisms of type I1 protein secretion. Biocbirn. Biopbys. Acta 1694:163-179. Florin, I., and F. Antillon. 1992. Production of enterotoxin and cytotoxin in Campylobacter jejuni strains isolated in Costa Rica. J. Med. Microbiol. 37:22-29. Fouts, D. E., E. F. Mongodin, R. E. Mandrell, W. G . Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005a. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Fouts, D. E., E. F. Mongodin, R E. Mandrell, W. G . Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005b. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:72-85. Fry, B. N., S. Feng, Y. Y. Chen, D. G. Newell, P. J. Coloe, and V. Korolik. 2000. The galE gene of Campylobacter jejuni is involved in lipopolysaccharide synthesis and virulence. Infect. Immun. 68:2594-2601. Goodman, S. D., and J. J. Scocca. 1988. Identification and arrangement of the DNA sequence recognized in specific transformation of Neisseria gonorrboeae. Proc. Natl. Acad. Sci. USA 85:69826986. Guerry, P., P. M. Pope, D. H. Burr, J. Leifer, S. W. Joseph, and A. L. Bourgeois. 1994. Development and characterization of recA mutants of Campylobacter jejuni for inclusion in attenuated vaccines. Infect. Immun. 62:426-432. Hahn, J., B. Maier, B. J. Haijema, M. Sheetz, and D. Dubnau. 2005. Transformation proteins and DNA uptake localize to the cell poles in Bacillus subtilis. Cell 12259-71. Hanninen, M. L., M. Hakkinen, and H. Rautelin. 1999. Stability of related human and chicken Campylobacter jejuni genotypes
after passage through chick intestine studied by pulsed-field gel electrophoresis. Appl. Enuiron. Microbiol. 65:2272-2275. Harrington, C. S., F. M. Thomson-Carter, and P. E. Carter. 1997. Evidence for recombination in the flagellin locus of Campylobacter jejuni: implications for the flagellin gene typing scheme. J. Clin. Microbiol. 35:2386-2392. Hofreuter, D., S. Odenbreit, and R. Haas. 2001. Natural transformation competence in Helicobacter pylori is mediated by the basic components of a type IV secretion system. Mol. Microbiol. 41 :379-39 1. Hofreuter, D., S. Odenbreit, G. Henke, and R. Haas. 1998. Natural competence for DNA transformation in Helicobacterpylori: identification and genetic characterization of the comB locus. Mol. Microbiol. 28:1027-1038. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Holmes, E. C., R. Urwin, and M. C. Maiden. 1999. The influence of recombination on the population structure and evolution of the human pathogen Neisseria meningitidis. Mol. Biol. Euol. 16: 741-749. Inamine, G. S., and D. Dubnau. 1995. ComEA, a Bacillus subtilis integral membrane protein required for genetic transformation, is needed for both DNA binding and transport.]. Bacteriol. 177: 3045-3051. Jeon, B., and Q. Zhang. 2007. CjOOllc, a periplasmic single- and double-stranded DNA binding protein, contributes to natural transformation in Campylobacter jejuni. 1.Bacteriol. 189:73997407. Johnson, T. L., J. Abendroth, W. G. Hol, and M. Sandkvist. 2006. Type I1 secretion: from structure to function. FEMS Microbiol. Lett. 255:175-186. Karudapuram, S., X. Zhao, and G. J. Barcak. 1995. DNA sequence and characterization of Haemophilus influenzae dprA+, a gene required for chromosomal but not plasmid DNA transformation. /. Bacteriol. 177:3235-3240. Korolik, V., M. R. Alderton, S. C. Smith, J. Chang, and P. J. Coloe. 1998. Isolation and molecular analysis of colonising and noncolonising strains of Campylobacter jejuni and Campylobacter coli following experimental infection of young chickens. Vet. Microbiol. 60:239-249. Kramer, N., J. Hahn, and D. Dubnau. 2007. Multiple interactions among the competence proteins of Bacillus subtilis. Mol. Microbiol. 65:454-464. Labigne-Roussel, A,, J. Harel, and L. Tompkins. 1987. Gene transfer from Escbericbia coli to Campylobacter species: development of shuttle vectors for genetic analysis of Campylobacter jejuni. I. Bacteriol. 1695320-53 23. Larsen, J. C., C. Szymanski, and P. Guerry. 2004. N-linked protein glycosylation is required for full competence in Campylobacter jejuni 8 1- 176.1. Bacteriol. 186:6508-65 14. Leonard, E. E., T. Takata, M. J. Blaser, S. Falkow, L. Tompkins, and E. C. Gaynor. 2003. Use of an open-reading frame-specific Campylobacter jejuni DNA microarray as a new genotyping tool for studying epidemiologically related isolates. J. Infect. Dis. 187:691-694. Lindblom, G. B., L. E. Cervantes, E. Sjogren, B. Kaijser, and G. Ruiz-Palacios. 1990. Adherence enterotoxigenicity invasiveness and serogroups in Campylobacter jejuni, Campylobacter coli, and Campylobacter laridis. APMIS 98:179-184. Linton, D., A. V. Karlyshev, P. G. Hitchen, H. R. Morris, A. Dell, N. A. Gregson, and B. W. Wren. 2000. Multiple N-acetyl neuraminic acid synthetase (neuB) genes in Campylobacter jejuni: identification and characterization of the gene involved in sialylation of lipo-oligosaccharide. Mol. Micro5iol. 35:1120-1134.
CHAPTER 3 1
Lior, H., D. L. Woodward, J. A. Edgar, L. J. Laroche, and P. Gill. 1982. Serotyping of Campylobacter jejuni by slide agglutination based on heat-labile antigenic factors. 1. Clin. Microbiol. 15: 76 1-768.
Londono-Vallejo, J. A., and D. Dubnau. 1993. comF, a Bacillus subtilis late competence locus, encodes a protein similar to ATPdependent RNAiDNA helicases. Mol. Microbiol. 9:119-131. Lorenz, M. G., and W. Wackernagel. 1994. Bacterial gene transfer by natural genetic transformation in the environment. Microbiol. Rev. 58563-602. McDermott, P. F., S. M. Bodeis, L. L. English, D. G. White, R. D. Walker, S. Zhao, S. Simjee, and D. D. Wagner. 2002. Ciprofloxacin resistance in Campylobacter jejuni evolves rapidly in chickens treated with fluoroquinolones. J. Infect. Dis. 185:837840.
McFarland, B. A., and S. D. Neill. 1992. Profiles of toxin production by thermophilic Campylobacter of animal origin. Vet. Microbiol. 30~257-266. Miller, J. F., W. J. Dower, and L. S. Tompkins. 1988. High-voltage electroporation of bacteria: genetic transformation of Campylobacter jejuni with plasmid DNA. Proc. Natl. Acad. Sci. USA 85~856-860.
Mortier-Barriere, I., M. Velten, P. Dupaigne, N. Mirouze, 0. Pietrement, S. McGovern, G. Fichant, B. Martin, P. Noirot, E. Le Cam, P. Polard, and J. P. Claverys. 2007. A key presynaptic role in transformation for a widespread bacterial protein: DprA conveys incoming ssDNA to RecA. Cell 130:824-836. Nuijten, P. J. M., A. J. G. van den Berg, I. Formentini, B. A. M. van der Zeijst, and A. A. C. Jacobs. 2000. DNA rearrangements in the flagellin locus of an flaA mutant of Campylobacter jejuni during colonization of chicken ceca. Infect. Immun. 68:71377140.
Odenbreit, S., J. Puls, B. Sedlmaier, E. Gerland, W. Fischer, and R. Haas. 2000. Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IV secretion. Science 287:14971500.
Palmen, R., B. Vosman, P. Buijsman, C. K. Breek, and K. J. Hellingwerf. 1993. Physiological characterization of natural transformation in Acinetobacter calcoaceticus.]. Gen. Microbiol. 139: 295-305.
Peabody, C. R., Y. J. Chung, M. R. Yen, D. Vidal-Ingigliardi, A. P. Pugsley, and M. H. Saier, Jr. 2003. Type I1 protein secretion and its relationship to bacterial type IV pili and archaeal flagella. Microbiology 149:3051-3072. Pearson, B. M., C. Pin, J. Wright, K. I'Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS Lett. 554:224-23 0.
Penner, J. L., and J. N. Hennessy. 1980. Passive hemagglutination technique for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heat-stable antigens. 1. Clin. Microbiol. 12: 732-737.
Planet, P. J., S. C. Kachlany, R. DeSalle, and D. H. Figurski. 2001. Phylogeny of genes for secretion NTPases: identification of the widespread tadA subfamily and development of a diagnostic key for gene classification. Proc. Natl. Acad. Sci. USA98:2503-2508. Poly, F., D. Threadgill, and A. Stintzi. 2005. Genomic diversity in Campylobacter jejuni: identification of C. jejuni 8 1-176-specific genes. J. Clin. Microbiol. 43:2330-2338. Poly, F., D. Threadgill, and A. Stintzi. 2004. Identification of Campylobacter jejuni ATCC 4343 1-specificgenes by whole microbial genome comparisons. 1. Bacteriol. 186:4781-4795. Provvedi, R., and D. Dubnau. 1999. ComEA is a DNA receptor for transformation of competent Bacillus subtilis. Mol. Microbiol. 31:271-280.
COMPETENCE AND TRANSFORMATION
569
Redfield, R. 1993. Genes for breakfast: the have-your-cake-andeat-it-too of bacterial transformation. 1.Hered. 84:400-404. Redfield, R J. 1988. Evolution of bacterial transformation: is sex with dead cells ever better than no sex at all? Genetics 119:213221.
Redfield, R. J., M. R. Schrag, and A. M. Dean. 1997. The evolution of bacterial transformation: sex with poor relations. Genetics 146:27-38.
Sandkvist, M. 2001a. Type I1 secretion and pathogenesis. Infect. lmmun. 69:3523-3535. Sandkvist, M. 2001b. Biology of type I1 secretion. Mol. Microbiol. 40:271-283.
Schouls, L. M., S. Reulen, B. Duim, J. A. Wagenaar, R. J. Willems, K. E. Dingle, F. M. Colles, and J. D. Van Embden. 2003. Comparative genotyping of Campylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host range, and recombination. 1. Clin. Microbiol. 41:15-26. Scott, A. E., A. R. Timms, P. L. Connerton, C. LOCCarrillo, K. Adzfa Radzum, and I. F. Connerton. 2007. Genome dynamics of Campylobacter jejuni in response to bacteriophage predation. PLoS Pathog. 3x119. Smeets, L. C., J. J. Bijlsma, E. J. Kuipers, C. M. VandenbrouckeGrauls, and J. G. Kusters. 2000. The dprA gene is required for natural transformation of Helicobacter pylori. FEMS lmmunol. Med. Microbiol. 27:99-102. Smith, H. O., D. B. Danner, and R. A. Deich. 1981. Genetic transformation. Ann. Rev. Biochem. 50:41-68. Smith, H. O., J. F. Tomb, B. A. Dougherty, R. D. Fleischmann, and J. C. Venter. 1995. Frequency and distribution of DNAuptake signal sequences in the Huemopbibs influenzae Rd genome. Science 269538-540. Smith, K. E., J. M. Besser, C. W. Hedburg, F. T. Leano, J. B. Bender, J. H. Wicklund, B. P. Johnson, K. A. Moore, and M. T. Osterholm. 1999. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992-1998. N. Engl. ]. Med. 340:15251532.
Stewart, G. J., and C. A. Carlson. 1986. The biology of natural transformation. Ann. Rev. Microbiol. 40:211-235. Suerbaum, S., M. Lohrengel, A. Sonnevend, F. Ruberg, and M. Kist. 2001. Allelic diversity and recombination in Campylobacter jejuni. 1.Bacteriol. 183:2553-2559. Szymanski, C. M., S. M. Logan, D. Linton, and B. W. Wren. 2003. Campylobacter-a tale of two protein glycosylation systems. Trends Microbiol. 11:233-23 8. Taboada, E. N., R. R. Acedillo, C. D. Carrillo, W. A. Findlay, D. T. Medeiros, 0. L. Mykytczuk, M. J. Roberts, C. A. Valencia, J. M. Farber, and J. H. E. Nash. 2004. Large-scale comparative genomics meta-analysis of Campylobacter jejuni reveals low level of genome plasticity. 1. Clin. Microbiol. 42:45664576.
Takata, T., T. Ando, D. A. Israel, T. M. Wassenaar, and M. J. Blaser. 2005. Role of dprA in transformation of Campylobacter jejuni. FEMS Microbiol. Lett. 252:161-168. Tonjum, T., N. E. Freitag, E. Namork, and M. Koomey. 1995. Identification and characterization of pilG, a highly conserved pilus-assembly gene in pathogenic Neisseria. Mol. Microbiol. 16: 45 1-464.
Tortosa, P., and D. Dubnau. 1999. Competence for transformation: a matter of taste. Curr. Opin. Microbiol. 2588-592. Vignon, G., R. Kohler, E. Larquet, S. Giroux, M. C. Prevost, P. Roux, and A. P. Pugsley 2003. Type IV-like pili formed by the type I1 secreton: specificity, composition, bundling, polar localization, and surface presentation of peptides. ]. Bacteriol. 185: 3416-3428.
570
WIESNER AND DIRITA
Wang, Y., S. D. Goodman, R. J. Redfield, and C. Chen. 2002. Natural transformation and DNA uptake signal sequences in Actinobacillus actinomycetemcomitans. J. Bacteriol. 184:34423449. Wang, Y., and D. E. Taylor. 1990. Natural transformation in Campylobacter species. J. Bacteriol. 172:949-955. Wassenaar, T. M., N. M. C. Bleumink-Pluym, D. G. Newell, P. J. M. Nuijten, and B. A. M. van der Zeijst. 1994. Differential flagellin expression in a flaA- flaB+ mutant of Campylobacter jejuni. Infect. Immun. 62:3901-3906. Wassenaar, T. M., B. N. Fry, and B. A. van der Zeijst. 1993. Genetic manipulation of Campylobacter: evaluation of natural transformation and electro-transformation. Gene 132:13 1-135. Wassenaar, T. M., B. N. Fry, and B. A. M. van der Zeijst. 1995. Variation of the flagellin gene locus of Campylobacter jejuni by recombination and horizontal gene transfer. Microbiology 141: 95-101.
Wiesner, R. S., D. R. Hendrixson, and V. J. DiRita. 2003. Natural transformation of Campylobacter jejuni requires components of a type I1 secretion system. J. Bacteriol. 185:5408-5418. Wilson, D. L., J. A. Bell, V. B. Young, S. R. Wilder, L. S. Mansfield, and J. E. Linz. 2003. Variation of the natural transformation frequency of Campylobacter jejuni in liquid shake culture. Microbiology 149:3 603-3 6 15. Wolfgang, M., J. P. van Putten, S. F. Hayes, D. Donvard, and M. Koomey. 2000. Components and dynamics of fiber formation define a ubiquitous biogenesis pathway for bacterial pili. EMBO J. 19:6408-6418. Wooldridge, K. G., and J. M. Ketley. 1997. Campylobacter-host cell interactions. Trends Microbiol. 5:96-102. Zhang, Y. X., K. Perry, V. A. Vinci, K. Powell, W. P. C. Stemmer, and S. B. del Cardayre. 2002. Genome shuffling leads to rapid phenotypic improvements in bacteria. Nature 415:644-646.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser Q 2008 ASM Press, Washington, DC
Chapter 32
Survival Strategies of Campylobacter jejuni: Stress Responses, the Viable but Nonculturable State, and Biofilms SARAH
L.
SVENSSON,
EMILISAFRIRDICH,AND ERIN C. GAYNOR
Campylobacter jejuni is an important and extremely prevalent worldwide enteric pathogen. However, C. jejuni also has significantly more fastidious growth and survival requirements than other, less fragile enteric pathogens such as Escherichia coli, Vibrio, Salmonella, and Shigella spp. Numerous hallmark and well-established stress response genes are noticeably absent from the sequenced C. jejuni genomes, including (but not limited to) those encoding the stationary phase and stress response sigma factor RpoS, the oxidative stress response regulatory elements SoxRS and OxyR, and osmotic shock protectants like BetAB. Nonetheless, C. jejuni must survive and even thrive in many diverse environments, including those encountered during commensal relationships with poultry and other zoonotic hosts, within disease-susceptible human hosts, and conditions experienced ex vivo in the natural environment and throughout the food preparation chain. Each environment poses significant stressful challenges that C. jejuni must overcome in order to be a successful pathogen. Insight into responses that C. jejuni mounts to counter challenges posed by stressful environments has accelerated remarkably over the past several years, thanks in part to experimental advances such as improved genetic manipulation techniques, wholegenome sequencing efforts, and global expression and proteomics analysis tools. Responses identified range from those that are conserved among bacterial species to some that are more C. jejuni specific, and include both specific and global responses as well as whole-population differentiation strategies. In this chapter, we will focus primarily on work performed over the past several years, although previous work will be included where appropriate. The reader is also
directed to other reviews (Alter and Scherer, 2006; Murphy et al., 2006; Park, 2002, 2005), which may cover some of the earlier-identified and other responses more extensively than space permits here. This chapter is divided into three sections. The first covers specific stresses and responses, is organized by the type of stress encountered, and discusses specific gene products or global response pathways that participate in countering the stress condition. Table 1 lists characterized and putative stress response genes. The second and third sections cover two different types of pleiotropic whole-population differentiation strategies or outcomes: the viable but nonculturable (VBNC) and coccoid forms of Campylobacter, and C. jejuni biofilms. Also incorporated into the latter two sections are the types of stresses that induce the differentiation, the genes involved in their formation, and the effect of differentiation on survivability or growth of C. jejuni.
SPECIFIC STRESSES AND RESPONSES Heat Stress
C. jejuni is considered a modest thermophile, with ideal growth temperatures in the human (37°C) to bird (42°C) range. Considering its lack of obvious cold shock genes such as cspA, C. jejuni can both survive and maintain metabolic activity for a remarkably long time at refrigeration temperatures (Chan et al., 2001). However, apart from a role for sodB and katA genes (see below) in freeze-thaw survival, little is understood regarding cold stress responses. In contrast, although heat stress accelerates the spiral-tococcoid transition and temperatures above 55°C rap-
~
Sarah L. Svensson, Emilisa Frirdich, and Erin C. Gaynor Department of Microbiology and Immunology, The University of British Columbia, 2558-2350 Health Sciences Mall, Vancouver, BC V6T 1 2 3 Canada.
571
572
SVENSSON ET AL.
Table 1. Characterized and putative stress response genes in C. jejuni Environment/stress Heat shock
Oxidative/Aerobic
Starvation Osmotic
Gene
Reference(s)
groESlgroEL clpP dnaK grpE dnaJ hrcA htrA lon clpB hslU
Thies et al. (1999b), Stintzi (2003) Parkhill et al. (2000) Thies et al. (1999a), Stintzi (2003) Stintzi (2003) Konkel et al. (1998), Stintzi (2003) Parkhill et al. (2000) Brondsted et al. (2005) Thies et al. (1998), Stintzi (2003) Stintzi (2003) Stintzi (2003) Bras et al. (1999) Andersen et al. (2005) Phongsisay et al. (2007) Pesci et al. (1994), Purdy et al. (1999), Stead and Park (2000) Grant and Park (1995), Day et al. (2000), Stead and Park (2000) Baillon et al. (1999) Ishikawa et al. (2003) van Vliet et al. (1999) Brondsted et al. (2005) Andersen et al. (2005) Phongsisay et al. (2007) Gaynor et al. (2005) Gaynor et al. (2004) Gaynor et al. (2004) van Vliet et al. (2001) Gaynor et al. (ZOOS) Candon et al. (2007) S. L. Svensson and E. C. Gaynor, unpublished Candon et al. (2007) Phongsisay et al. (2007) Raphael et al. (2005) Lin et al. (2002) Lin et al. (2005) Elvers et al. (2005),Pittman et al. (2007) Pittman et al. (2007) Elvers et al. (2004) Phongsisay et al. (2007)
racRS bspR htrB sodB katA ahpC dPS
perR htrA hspR htrB spoT sdb dcuA fdd spoT PPkl cjl226c
PPkl Bile
Nitrosative
Low pH
htrB cbrR cmeABC cmeR nssR nrfA cgb htrB
idly result in cell death (Klancnik et al., 2006; Nguyen et al., 2006; Tangwatcharin et al., 2006), several key C. jejuni heat shock responses have been identified and characterized that contribute to maintenance of viability. Heat shock proteins As with most bacteria, C. jejuni harbors homologs of numerous molecular chaperones and proteases collectively referred to as heat shock proteins (HSPs). These include GroESL, ClpP, DnaK, GrpE, DnaJ, HrcA, and HtrA/DegP (Parkhill et al., 2000). A 1998 study described 24 proteins that were synthesized in response to 46°C heat shock. One of these was identified as DnaJ, which was shown via mutant analyses to be required for heat shock survival and chick colonization (Konkel et al., 1998). Subsequent
studies also identified the lon protease gene, groESL, clpB, and dnaK, as upregulated in response to thermal stress (Thies et al., 1998, 1999a, 1999b, 1 9 9 9 ~ ) . More recently, a whole-genome array study (Stintzi, 2003) identified groESL, grpE, dnaK, dna], clpB, lon, and the hslU putative proteasome gene as upregulated at 42 versus 37°C. Several membrane proteins were also induced by raising the temperature, suggesting that membrane remodeling may play a role in heat tolerance. Finally, a study found that the htrA gene encoding a likely periplasmic protease is required for growth above 44°C and in the presence of puromycin, implying a role for HtrA in the misfolded protein response (Brondsted et al., 2005). The AhtrA mutant also exhibited higher expression levels of dnaK and clpB and was defective for aerobic survival, host cell adherence, and invasion, implicating HtrA in virulence and transmission properties in addition to heat stress response.
CHAPTER 32
Regulatory genes involved in heat shock response To date, only two regulatory factors directly affecting heat stress have been characterized. The first is the RacR (reduced ability to colonize) response regulator (Braset al., 1999). Mutant analyses demonstrated that RacR is required for optimal growth at 42°C and for colonization of chicks. Although several proteins with altered abundance in the AracR mutant versus wild type were noted, the RacR regulon is currently uncharacterized. More recently, microarray and proteomics analyses revealed that HspR is a negative regulator of several C. jejuni heat shock genes, including dnaK, clpB, and groESL (Andersen et al., 2005). A AhspR mutant also exhibited heat (44°C) and aerobic sensitivity, adherence and invasion defects, motility defects, and downregulation of several flagellar genes. The C. jejuni genome also harbors a homolog of the putative but uncharacterized HrcA repressor. Novel and alternative heat stress responses Heat shock in general is severely detrimental to C. jejuni; in fact, a recent study demonstrated that it can also impair C. jejuni’s ability to adhere to and invade host cells (Mihaljevic et al., 2007). However, improved response to heat shock may occur via potentially novel mechanisms. Starvation, for instance, was shown to increase the ability of C. jejuni to withstand heat shock (Cappelier et al., 2000; Klancnik et al., 2006); the Cappelier et al. study also demonstrated that this tolerance depends on de novo protein synthesis during the starvation period. Spent media from midlog or stationary phase cells also conferred heat shock protection, as described in more detail below in “Acid Stress” (Murphy et al., 2003a). Interestingly, C. jejuni also exhibited heat stress resistance differences in brucella broth from two different suppliers, with Difco media being more protective than that from BBL (Murphy et al., 2005). Finally, it has been shown that the htrB gene, which is located in the lipooligosaccharide (LOS) locus and likely encodes a lipid A biosynthesis lauroyl acyltransferase (referred to as LpxL in E. coli [Raetz and Whitfield, 2002]), was upregulated during several stress conditions, including 44°C heat shock (Phongsisay et al., 2007). HtrB is essential in C. jejuni, and a Salmonella enterica serovar Typhimurium AhtrB mutant, although viable, is temperature sensitive (ts). The C. jejuni htrB gene rescued the ts defect of that mutant, further implicating HtrB and potentially lipid A acylation in resistance to heat as well as other stresses (see below).
SURVIVAL STRATEGIES OF c. rmiw
573
Oxidative (Reactive Oxygen Species) Stress
C. jejuni is both microaerophilic and capnophilic, with a reported optimal growth environment of 12% CO, and 6% 0,. However, C. jejuni also regularly encounters suboptimal atmospheres ranging from fully anaerobic to fully aerobic. Anaerobiosis is most frequently encountered in the intestinal tracts and ceca (when present) of its animal hosts; anaerobic adaptation primarily involves respiration and metabolism genes (chapter 3 ) . Atmospheric oxygen levels and reactive oxygen species, however, pose a challenge C. jejuni must overcome in a number of different milieus. Reactive oxygen species, including superoxide, hydrogen peroxide (H20,), and halogenated 0, molecules, form as a result of aerobic respiration both inside and outside an animal host and also comprise a major host defense mechanism against bacterial intruders. Because several unique aerobic and 0, stress responses have been identified, they are reviewed in the next section; this section briefly covers reactive oxygen species stress response genes. Superoxide dismutase (SOD) catalyzes the breakdown of superoxide to H,02 and dioxygen. Unlike E. coli, which contains several SOD genes (Cabiscol et al., 2000), C. jejuni harbors only one, sodB, which was the first reactive oxygen species response gene described for C. jejuni (Pesci et al., 1994; Purdy and Park, 1994). Studies of AsodB mutants in C. jejuni and C. coli revealed roles for SOD in human epithelial cell invasion, survival in the presence of the superoxide generator methyl viologen, survival on model foods, and survival during the initial stages of H202shock (Pesci et al., 1994; Purdy et al., 1999). Also in contrast to E. coli, C. jejuni and C. coli harbor only one gene encoding catalase (katA), which converts H 2 0 2to H,O and 0,. Mutant studies identified roles for catalase in C. jejuni H,O, resistance and intramacrophage but not intraepithelial cell survival (Day et al., 2000; Grant and Park, 1995). SOD and catalase were also jointly required for resistance to freeze-thaw stress under aerobic conditions (Stead and Park, 2000), indicating that detoxification of superoxide formed during freeze-thaw is important for C. jejuni survival. C. jejuni also contains an alkyl hydroperoxide reductase, encoded by ahpC, which confers increased resistance to cumene hydroperoxide and aerobic stress but not H,O, and, along with katA, was upregulated during iron limitation (Baillon et al., 1999). As noted, C. jejuni lacks OxyR, which regulates ahpC and katA in response to oxidative stress in E. coli (Cabiscol et al., 2000). Interestingly, the C. jejuni Fur homolog PerR was found to repress ahpC and katA in an iron-dependent manner, making it a
574
SVENSSON ET AL.
functional but not homologous substitute for OxyR (van Vliet et al., 1999). Finally, a recent study identified the constitutively expressed dps gene, encoding a putative DNA-binding protein from starved cells, as being important for H26, resistance, but only in the presence of iron (Ishikawa et al., 2003). Because the C. jejuni Dps binds iron but not DNA, the authors hypothesized that in C. jejuni, this protein may provide protection by sequestering iron to prevent hydroxyl radical formation occurring via the Fenton reaction. Additional coverage of C. jejuni and iron is provided in chapter 33. Aerobic (0,)Stress
al., 2005). In the stringent response, the alarmone ppGpp, produced by RelA or SpoT, binds RNA polymerase, altering gene expression to counter the stress condition. Although the aerobic stress defect of the AspoT mutant was unexpected, it may be explained by downregulation in the mutant of several key respiration and metabolism genes (i.e., an oxidoreductase complex, the nap operon, and the nuo operon encoding NADH dehydrogenase, which is involved in the aerobic respiratory chain). A more complete description of these genes is provided in chapter 3. Thus, in the absence of other stress elements mentioned above, the stringent response may play a more global role in regulating stress response in C. jejuni.
Outcomes and effects of short-term aerobic stress
Long-term aerobic adaptation
Although 0, is stressful for C. jejuni, few recent reports have described specific phenotypic consequences of aerobiosis, and those that have often differ in their conclusions as to whether 0, is detrimental, neutral, or possibly beneficial to the bacterium. For instance, Klancnik et al. (2006) found that 24-h exposure to 0, accelerated the transition to the VBNC and/or coccoid form, and as noted above, Stead and Park (2000) described a requirement for SOD and catalase to combat the deleterious cumulative effects of aerobiosis and freeze-thaw stress. In contrast, Mihaljevic et al. (2007) found that compared with both heat shock and starvation, 15-h 0, exposure had a minimal deleterious effect on C. jejuni culturability and viability and in fact led to an increase in the bacterium’s ability to adhere to and invade host epithelial cells. Likewise, Murphy et al. (2003b) found that aerobiosis enhanced the acid adaptive tolerance response described more fully below in “Acid Stress.”
Interestingly, C. jejuni can grow in the ambient atmosphere. This can occur via cultivation on supportive plates containing horse blood and 0, scavengers such as pyruvate and ferrous sulfate (Hodge and Krieg, 1994; Tangwatcharin et al., 2007), or via adaptation to an aerobic environment, as was demonstrated by Jones et al. (1993), who showed that C. jejuni exposed to an aerobic environment for 2 days subsequently grew better aerobically versus microaerobically. However, aerobic adaptation can result in a diminished capacity of C. jejuni to colonize the microaerobic or anaerobic in vivo environment. For instance, two studies found that the laboratory-passaged, genome-sequenced NCTC 11168 strain variant exhibited chick colonization, motility, and morphology defects compared with its clonal parental isolate (Carrillo et al., 2004; Gaynor et al., 2004). This may have occurred via vertical evolution or phase variation during exposure of the original clinical isolate to the aerobic environment, similar to an in vivo phase variation adaptation described for the flagellar regulator flgR (Hendrixson, 2006; chapter 30). Indeed, aerobic adaptation of the original NCTC 11168 isolate yielded defects similar to those observed for the sequenced variant, and although clonal, numerous gene expression differences between the sequenced and original variants were identified in respiration and metabolism genes (i.e., downregulation in the sequenced variant of the sdh operon and dcuA, described in chapter 3), as well as an impaired ability of the sequenced variant to respond to anaerobic conditions (Gaynor et al., 2004).
Genes involved in short-term aerobic tolerance In addition to sodB, katA, and ahpC described above, several other genes have been identified as important for short-term 0, survival. Two of these are the heat shock protease htrA and regulator hspR described above in “Heat Stress,” suggesting a connection between the heat shock pathway and aerobic tolerance (Andersen et al., 2005; Brondsted et al., 2005); htrB was likewise upregulated during aerobic stress (Phongsisay et al., 2007). The fdxA ferredoxin gene upstream of ahpC was also found to be important for aerotolerance; however, in contrast to ahpC, fdxA was induced by high iron levels (van Vliet et al., 2001). SPOT,which regulates the C. jejuni stringent response, was also found to be important for low (5%) CO, growth and aerobic survival (Gaynor et
Starvation and Stationary Phase Stress Outside its natural zoonotic hosts, C. jejuni frequently encounters nutrient-poor environments such as water, which has served as a vehicle for several
CHAPTER 32
larger-scale campylobacteriosis outbreaks (Auld et al., 2004; Clark et al., 2003; O’Reilly et al., 2007). Nutrient deprivation is also one of the stresses incurred during stationary phase growth in a laboratory or natural environment, and would also be expected to occur in an intravacuolar environment as described below and in chapter 16. Detriments and possible benefits of starvation In contrast to E. coli, which gains certain stress resistance properties during stationary phase through the action of the stationary phase sigma factor RpoS, C. jejuni has been described as lacking this “traditional” response as a result of its lack of RpoS. Indeed, stationary phase C. jejuni exhibited increased sensitivities to heat, oxidative, and acid stress (Kelly et al., 2001; Murphy et al., 2003b; Park, 2005), AahpC mutants exhibited increased oxidative sensitivity during stationary phase (Baillon et al., 1999), and starvation was shown to be more detrimental to culturability and viability than either 0, or heat shock stress (Mihaljevic et al., 2007). Interestingly, the latter study also showed that host cell adherence and invasion decreased dramatically for starved bacteria; however, in contrast to nonstarved bacteria, where only 10% of adhered bacteria actually invaded, the number of starved bacteria that invaded was approximately equal to the number of adhered bacteria. In contrast to the above studies, certain benefits of brief starvation have also been noted, such as a de novo protein-synthesis-dependent increase in nutrient uptake and heat resistance (Cappelier et al., 2000; Klancnik et al., 2006), and stationary phase cells, although more acid susceptible, were also optimally suited for mounting a low-pH-induced adaptive tolerance response (see below) (Murphy et al., 2003b). Finally, an interesting GASP (growth advantage in stationary phase)-like adaptation phenomenon was observed, where bacteria that had undergone a severe decline in viability during late stationary phase and then began growing again exhibited a heritable advantage in surviving early stationary phase as well as certain other stresses (Martinez-Rodriguez et al., 2004). Genes important for stationary phaselstarvation stress resistance Two global factors were identified as important for these responses. First, the stringent response, which was rapidly induced during nutrient limitation, was required for both stationary phase and lownutrient media survival (Gaynor et al., 2005). In
-
SURVIVAL STRATEGIES OF C. TEIUNI
575
other bacteria, the stringent response to starvation is tightly linked with RpoS; thus, C. jejuni provides an example of an RpoS-independent requirement for this response. Another factor, polyphosphate (poly P), produced by the ppk2 polyphosphate kinase 1 gene product (Candon et al., 2007), is a polymer of multiple phosphate residues that has been hypothesized to play numerous intracellular roles (Brown and Kornberg, 2004). In C. jejuni, poly P was shown to increase in abundance during stationary phase and to be important for low-nutrient survival (Candon et al., 2007). The AspoT and Appkl mutants are also defective for intraepithelial cell survival, consistent with the hypothesis that C. jejuni exists intracellularly within a nutrient-poor vacuole. The mechanisms by which these factors affect starvation stress are under investigation, although gene expression changes in AspoT described in above in “Aerobic (0,)Stress” may also affect the starvation response. It is also interesting to hypothesize that perhaps stress resistance induced by brief starvation may involve induction of these global responses. Finally, a Acj1226c sensor kinase mutant also exhibits a decrease in culturability during stationary phase; however, as described below in “C. jejuni Biofilms,” this may reflect an accelerated planktonic-to-biofilm conversion rather than a decrease in bacterial viability (S. L. Svensson and E. C. Gaynor, unpublished data). Osmotic Stress
C. jejuni must counter hypo-osmotic and hyperosmotic stress during both environmental transmission and in vivo existence. Several responses to lownutrient stress, as would be encountered in water, were described above in “Starvation and Stationary Phase Stress”; however, little else is known about hypo-osmotic stress responses, even though C. jejuni can survive for long periods of time in a hypoosmotic environment. In contrast, several food study-based reports have documented the sensitivity of C. jejuni to hyperosmotic stress as would be encountered during food processing, desiccation, saltwater survival, and several in vivo environments (Alter and Scherer, 2006; Birk et al., 2004; Doyle and Roman, 1982; Park, 2002). Each report confirmed that C. jejuni is more sensitive to salt than most other food-borne bacterial pathogens and is typically unable to grow at above 2% NaC1. Three genes have been identified as important for hyperosmotic stress resistance. The first, htrB, was upregulated during 1.5% NaCl stress and rescued the salt sensitivity of the S. enterica serovar Typhimurium AhtrB mutant (Phongsisay et al., 2007). The poly P synthesis gene ppkl was likewise critical for both survival at late
576
SVENSSON ET AL.
growth stages in broth containing 0.25M NaC1, as well as growth from single colonies on plates containing 0.17 M NaCl (Candon et al., 2007). Finally, the Acj1226c sensor kinase mutant also displayed growth defects on plates containing 0.17 M NaCl (S. L. Svensson and E. C. Gaynor, unpublished data). Specific mechanisms by which these genes participate in hyperosmotic resistance are unknown, but it is interesting to speculate on potential roles for lipid A acylation on the basis of the htrB data and the fact that lpxL/htrB mutants exhibit defects in membrane fluidity, stabilization, and permeability in other bacteria (Raetz and Whitfield, 2002), poly P as possibly playing a membrane stabilization or compatible solute role, and the Cj1226c sensor kinase as potentially modulating solute transport or cell envelope permeability.
Acid Stress As a food- and waterborne pathogen, C. jejuni must pass through several hostile environments in the digestive system before colonizing the intestinal tract. Unlike the gastric pathogen Helicobacter pylori, which harbors urease genes allowing it to survive prolonged low pH exposure, C. jejuni is quite acid sensitive, and the low stomach pH is considered to be a major host defense mechanism preventing C. jejuni colonization (Doyle and Roman, 1981). However, several reports have provided insight into how C. jejuni might transiently survive acid shock. One study demonstrated that acid survival is strain dependent and can be significantly affected by the assay medium. For example, Mueller-Hinton broth conferred a 5-log advantage over brain heart infusion or Campylobacter enrichment broth (Murphy et al., 2005). One interpretation of these data is that the type of infected food ingested may significantly influence whether a given amount of C. jejuni will subsequently establish intestinal colonization. The same group also identified two potential means of C. jejuni protein induction conferring acid protection. In one study, spent media from midlog and stationary phase cultures was found to contain a heat-stable, strainspecific, non-density-dependent protein factor that provided protection against pH 4.5 shock to naive C. jejuni cultures (Murphy et al., 2003a). Another study found that pretreatment at pH 5.5, particularly when conducted aerobically, induced a protein-synthesisdependent adaptive tolerance response allowing prolonged survival at pH 4.5 (Murphy et al., 2003b). Finally, HtrB was also shown to be acid inducible and important for acid shock survival (Phongsisay et al., 2007).
Nitrosative Stress Nitric oxide (NO) is a potent antimicrobial that C. jejuni encounters in numerous locations inside an animal host. N O can be produced by both macrophages and epithelial cells, frequently arises as a result of inflammation, and can cause death of both extracellular and intracellular bacteria. Little was known about C. jejuni’s defense mechanisms against nitrosative stress until a recent study by Elvers et al. (2004) identified a single-domain hemoglobin termed Cgb (Campylobacter globin) that was found to be specifiFally upregulated during nitrosative stress and is important for detoxifying nitrosating agents (i.e., S-nitrosoglutathione) but not other oxidants such as H 2 0 2 and methyl viologen. The same group subsequently found that the Crp-Fnr homolog NssR specifically regulates the nitrosative stress response in c. jejuni, controlling the expression of both cgb and at least four other genes, including another truncated hemoglobin (Elvers et al., 2005). Gene regulation was further shown to occur via binding of NssR to an Fnr-like binding sequence in the promoters of the regulated genes, demonstrating a specific control mechanism for this new stress response regulator. Even more recently, the constitutively expressed NrfA nitrate reductase was shown to provide protection from N O stress under both aerobic and anaerobic conditions (Pittman et al., 2007). Chapter 3 provides further discussion of nitrate and nitrite respiration pathways. Bile Stress Bile, composed of bile salts and phospholipids, is secreted by the liver into the intestinal tract, where it functions as a detergent to break down fats and other substances, including bacterial membranes. As such, it is also provides an effective antimicrobial defense that invading pathogens such as C. jejuni must overcome during intestinal colonization. Several specific c. jejuni gene products have been identified as participating in bile resistance, one of which is the CmeABC multidrug efflux system composed of a periplasmic fusion protein, an efflux transporter, and an outer membrane protein (Lin et al., 2002). CmeABC was identified via homology to other efflux transporters and is important for C. jejuni resistance to several structurally unrelated but specific compounds, including fluoroquinolones, p-lactams, and most notably bile salts. Chick colonization required a functional CmeABC system, which was also found to be expressed and immunogenic in infected birds (Lin et al., 2003). Subsequent studies by the same group identified CmeR as a specific, promoter-binding tran-
CHAPTER 32
scriptional repressor of cmeABC (Lin et al., 2005a), and found that bile salts induced expression of the efflux pump by inhibiting binding of CmeR to the cmeABC promoter (Lin et al., 2005b). X-ray crystallography analyses of CmeR should yield additional insight into DNA and ligand-binding domains (Gu et al., 2007; Su et al., 2007). A second regulatory protein, CbrR (Campylobacter bile resistance regulator), was identifia via its invoGemGnt in deoxycholate (bile salt) resistance and was also shown to be important for chick colonization (Raphael et al., 2005). CbrR has two response regulator domains and a modestly conserved GGDEF motif known to be important for cyclic di-GMP production in other organisms, but no obvious cognate sensor kinase. Other groups have also begun investigating whole-genome expression and proteomics responses to bile stress (Fox et al., 2007; Okoli et al., 2007), revealing the upregulation of several predicted factors such as flagella and transcription factors as well as novel genes that may encode additional bile stress response proteins and/or components of the CmeR and CbrR regulons. Finally, more recent studies have shown that efflux pump inhibitors significantly decrease the minimal inhibitory concentration of bile salts, suggesting their potential utility as novel antimicrobial therapeutics (Lin and Martinez, 2006; Quinn et al., 2007). Additional details regarding antibiotic resistance in C. jejuni are provided in chapter 14.
VBNC AND COCCOID FORMS OF CAMPYLOBACTER Under environmental stress and unfavorable growth conditions that are potentially lethal, C. jejuni has been proposed to enter a VBNC state (Moore, 2001; Murphy et al., 2006; Pearson et al., 1993; Rollins and Colwell, 1986). This is defined as a state of dormancy during which growth on bacteriological media normally used for culture of the organism ceases, yet the bacteria retain viability with minimal metabolic activity (reviewed in Barer and Harwood, 1999; Moore, 2001; Oliver, 2005). Bacteria entering a VBNC state often exhibit certain characteristics, including a reduction in size (such as transformation from a rod to a coccoid form), a decrease in metabolic activity and macromolecular biosynthesis with respiration occurring at a lower rate, maintenance of ATP levels, continued gene expression, and a modification of cell wall structures (Oliver, 2005). Culturability is restored after a period of resuscitation, as is infectivity in the case of pathogenic organisms, giving rise to an important reservoir of infection and public health risk.
SURVIVAL STRATEGIES OF C. rEETUNI
577
Several environmental stresses such as a lack of nutrients, suboptimal growth temperatures, oxygen tension, pH, osmolarity changes, and high pressure have been shown to lead to changes in Campylobacter cell morphology and composition that result in dormancy (Boucher et al., 1994; Chaveerach et al., 2003; Hazeleger et al., 1995; Lazaro et al., 1999; Medema et al., 1992; Reezal et al., 1998; Rollins and Colwell, 1986). The ability to enter a VBNC state was also found to vary depending on the strain (Cools et al., 2003; Lazaro et al., 1999; Medema et al., 1992; Tholozan et al., 1999). The key to determining whether an organism is in a VBNC state and not dead is the viability count. Viability assays measure an aspect of metabolic activity or cellular integrity (Oliver, 2005, and references therein). The most common method of measuring viability and cellular respiration is by examining a cell's capacity to reduce tetrazolium salts to formazans (Kogure et al., 1979), which was first applied to C. jejuni by Boucher et al. (1994). The viable count is then measured against the total number of bacteria, which can be assayed by DAPI (4',6-diamidino-2phenylindole) or acridine orange staining. Characterization of the Campylobacter VBNC State Morphology and viability changes associated with VBNC formation
Carnpylobacter has several morphological forms (Vandamme, 2000). The characteristic spiral form predominates in young, actively growing cultures (Butzler and Skirrow, 1979). Under stress, the organism transitions through several intermediate forms to a coccoid morphology (Griffiths, 1993; Ng et al., 1985). The mechanism by which this transition occurs has not been elucidated but is thought to be the formation of donut- or ring-shaped cells (Alonso et al., 2002; Ng et al., 1985; Thomas et al., 1999). Several studies support the conclusion that the transformation process from spiral to coccoid in C. jejuni and C. coli does not require gene transcription and de novo protein synthesis, pointing to a passive and not an active process (Boucher et al., 1994; Hazeleger et al., 1995; Hudock et al., 2005; Jacob et al., 1993; Thomas et al., 1999). The conversion of spiral to coccoid parallels a decrease in culturability (Butzler and Skirrow, 1979; Chou et al., 1983; Ng et al., 1985; Rollins et al., 1983; Smibert, 1978), and when viability is maintained, this change is associated with a transition to VBNC (Boucher et al., 1994; Medema et al., 1992). Spiral forms predominate in cultures of C. jejuni and C. coli grown on solid media and in liquid culture at 37°C for 24 h, but in as little as 48 h, the
578
SVENSSON ET AL.
culture primarily contains bacteria with coccoid morphology (Alonso et al., 2002; Buck et al., 1983; Griffiths, 1993). At 4"C, the appearance of coccoid cells is greatly slowed (Holler et al., 1998; Jacob et al., 1993; Lazaro et al., 1999). Tangwatcharin et al. (2006) compared VBNC formation at 4 and 60°C and found that under their experimental conditions, the culture changed completely to a VBNC form in 25 days at 4°C and 15 min at 60°C. Considerable debate exists in the literature whether Campylobacter truly enters a VBNC state or whether the coccoid form is a degenerate state of the organism. A good explanation for these variable results, as suggested by Hazeleger et al. (1995), is that there exist different types of coccoid cells with different characteristics that depend on the conditions of formation. Coccoid cells formed at higher temperatures and in nutrientrich environments show more degradation (Moran and Upton, 1986, 1987) and a more rapid decline in culturability, likely as a result of the higher metabolic rate of these organisms (Boucher et al., 1994; Hazeleger et al., 1995; Rollins and Colwell, 1986). In contrast, coccoid cells formed at lower temperatures and at low nutrient concentrations (conditions that mimic the environment to which C. jejuni would be exposed after being shed from a host) appear to retain similar characteristics to the spiral bacteria. Aeration increases the rate of coccoid formation but is not required for transformation to occur (Boucher et al., 1994). A thorough study by Lazaro et al. (1999) looked at the survival of C. jejuni at 4 and 20°C in phosphate-buffered saline and examined potential viability in the absence of culturability. Viability was assessed by examining cellular structures by microscopy, metabolic activity, the maintenance of DNA, and the total cell protein profile. Differences in protein profile were noted, but no unifying conclusions were drawn. Intact genomic DNA was recovered after 4 months at 4°C and 2 months at 20°C, which is more than 1 month after becoming nonculturable. Bacterial cells with intact DNA and cellular integrity were defined as VBNC, which could be detected up to 7 months at 4°C. They also noted that spiral cells can lose culturability while retaining the capacity for respiration. As such, nonculturable cells can be classified as both viable and nonviable, and not necessarily specifically correlated with shape. VBNC formation and coccoid formation can therefore be considered as related but separate phenomena (Lazaro et al., 1999). Hudock et al. (2005) examined the stability of DNA as a measure of viability by pulsedfield gel electrophoresis of genomic DNA from spiraland coccoid-shaped C. jejuni and found that the degree of DNA degradation increased with the age of
the cells at 37°C. In contrast, at 4"C, the bacteria showed no genomic DNA degradation even in the absence of culturability and were mainly spiral shaped, further supporting the argument that the true VBNC state occurs at 4°C and is not limited to coccoid cells. This is important to keep in mind because coccoid and VBNC are frequently yet incorrectly used synonymously. Changes in cell wall structure associated with VBNC formation Changes in Gram stain have been observed between the spiral and coccoid forms, indicating a change in cell wall composition (Moran and Upton, 1986). In that study of C. jejuni grown at 42"C, the spiral cells yielded a typical gram-negative reaction, while the coccoid cells only stained lightly with the safranin counterstain. Similar observations were seen when C. jejuni grown at 37°C was stained with crystal violet: spiral bacteria stained darkly while coccoid cells stained only faintly (E. Frirdich and E. C. Gaynor, unpublished data). Examination of the coccoid form by electron microscopy has revealed a loss of cell wall integrity in some studies (Buck et al., 1983; Ng et al., 1985), whereas others have shown that most of the coccoid cells retained their structural integrity (Jones et al., 1991), although some cells with an enlarged periplasmic space and blebbing membranes could be detected (Jacob et al., 1993). It has been proposed that this is a result of cellular disintegration (Kusters et al., 1997). Lazaro et al. (1999) discuss an alternate explanation for these gaps and bleb formation in viable cells on the basis of a similar phenomenon in Vibrio parahuemolyticus (Jiang and Chai, 1996). They argue that blebs may represent a method of cell volume adjustment, resulting in a minimal cell maintenance volume with an increased surface area for nutrient absorption Uiang and Chai, 1996; Lazaro et al., 1999). The membrane fatty acid composition of coccoid bacteria at 4°C and to a lesser extent at 12°C was found to resemble that of spiral bacteria, while that of coccoid bacteria formed at 25°C differed significantly and was thought to result in a leaky membrane (Hazeleger et al., 1995). This may account for the decrease in DNA and ATP content in coccoid cells formed at this temperature (Hazeleger et al., 1994, 1995). No changes in membrane protein content in C. jejulzi have been observed as a result of the spiral to coccoid transition (Hazeleger et al., 1995; Moran and Upton, 1986; Thomas et al., 1999). An electrophoretic study of the whole-cell protein profile and LOS banding pattern of C. coli likewise showed no
CHAPTER 32
change between the spiral and coccoid form over an incubation period of 19 and 25 days in water at 37 and 4"C, respectively (Jacob et al., 1993). Coccoid C. jejuni were also found to be more resistant than the spiral forms to cell lysis by the membrane perturbing agent EDTA and the detergents sodium dodecyl sulfate and Triton X-100 (Moran and Upton, 1986). This may be the result of a change in surface charge that would preclude binding to the membrane or in surface accessibility. The peptidoglycan (PG) of the coccoid forms was also less sensitive to lysozyme treatment, indicating that there was either a change in PG structure or that changes to the cell surface may be affecting lysozyme accessibility to the PG. Changes in cell morphology would require remodeling of the PG layer, which is the major determinant of bacterial cell shape (Cabeen and Jacobs-Wagner, 2005). Amano and Shibata (1992) examined the amount of PG in spiral and coccoid C. jejuni, C. coli, and C. fetus and found that the amount of PG isolated correlated with the amount of spiral bacteria, with only very little to none extracted from coccoid forms. C. fetus did not change to a coccoid form, so PG could always be isolated from that species. The PG in coccoid forms could possibly be degraded, which would explain the reports that coccoid cells produced at 37°C lack cell integrity and lead to leakage of cytoplasmic contents (Buck et al., 1983; Moran and Upton, 1987). It would be interesting to isolate PG from coccoid cells produced at 4"C, which are thought to be more representative of the VBNC state of the organism. H. pylori, a close relative of Campylobacter spp. (Vandamme, 2000), also undergoes a shape change from spiral to coccoid that has been associated with a VBNC state (Benaissa et al., 1996; Bode et al., 1993; Nilius et al., 1993). Analysis of the PG from H. pylori cells undergoing a transition from spiral to coccoid showed an alteration in PG structure (Costa et al., 1999), which has been attributed to the activity of the amiA gene product, a PG hydrolase (Chaput et al., 2006). The amiA gene of H. pylori represents the first gene identified to date shown to be involved in the spiral-coccoid transition. The C. jejuni genome also codes for a copy of the amiA gene that is interestingly located two genes upstream of SPOT; as noted above, spoT modulates the stringent response and is important for late stage culturability. Resuscitation and Virulence Formation of a VBNC state only becomes a true means of surviving stress if the organism can return to a metabolically active state. Many studies have been designed to examine this issue for Campylobac-
-
SURVIVAL STRATEGIES OF c. r E r w
579
ter and have led to varying conclusions. This is likely due to the vast differences in experimental design (strain variability, parameters used for VBNC formation, animal model used), as mentioned previously. Several groups have shown that C. jejuni VBNC are unable to revert to a culturable form capable of colonization (Beumer et al., 1992; Hald et al., 2001; Hazeleger et al., 1995; Medema et al., 1992; van de Giessen et al., 1996; Ziprin and Harvey, 2004; Ziprin et al., 2003). Other studies showed reversion after passage through the intestines of chicks (Cappelier et al., 1999a; Stern, 1994; Talibart et al., 2000), rats (Saha and Sanyal, 1991), and mice (Baffone et al., 2006; Federighi et al., 1998; Jesudason et al., 1989; Jones et al., 1991). Recovery by inoculation into embryonated eggs was also demonstrated (Cappelier et al., 1999b; Chaveerach et al., 2003). Cappelier et al. (1999b) showed that the VBNC cells had lost their adhesion properties (a measurement of their pathogenic potential) but that this was regained after recovery. This transient loss of infectivity was also demonstrated by Saha and Sanyal (1991). Tholozan et al. (1999) carried out an in-depth physiological characterization of VBNC cells formed in microcosm water (low nutrients and hypo-osmotic conditions; Rollins and Colwell, 1986) at 4°C. By measuring cell volume, adenylate cyclase charge, internal pH, intracellular potassium concentration, and membrane potential values, they demonstrated a progressive decrease in the ability of the cells to maintain internal homeostasis. However, the cells were able to regain their pathogenic potential in a mouse model (Federighi et al., 1998). A recent study by Baffone et al. (2006) demonstrated that C. jejuni VBNC generated at 4°C in water was capable of surviving for several months in this state and regained culturability after passage through the mouse intestine, as long as more than lo4 respiring VBNC cells per ml were present. This indicates that a high amount of respiring bacteria in a VBNC state is required to trigger recovery of the cells and for subsequent colonization. In summary, much remains to be learned about the VBNC state of Campylobacter, including elucidating the signal or signals that lead to VBNC and coccoid formation, identifying the genes involved in remodeling the cell wall, and characterizing the conditions required for bacterial recovery. Campylobacter dormancy as a mechanism of surviving stress poses an important avenue of research, considering that this may be the key in understanding the environmental transmission of this organism. In addition, conditions that increase the survival of Campylobacter at low temperatures are of importance to the food industry when considering that food is a major ve-
580
SVENSSON ET AL.
hicle for transmission and low temperatures are used for preservation (Buswell et al., 1998).
C. JEJUNI BIOFILMS
The paradox of why C. jejuni is such a successful zoonotic pathogen despite its fastidious growth requirements in the laboratory may also be explained by its tendency to persist in distinct lifestyles in the natural environment. More than 99% of bacterial species exist outside of the laboratory environment not as free-swimming, solitary cells but as part of communities called biofilms (Ehrlich et al., 2005). Biofilms are surface-associated consortia of microorganisms encased in a protective polymeric matrix, whose residents possess distinct phenotypic differences from their planktonic counterparts. In contrast to cells grown in broth culture in the laboratory, residents of a biofilm exhibit marked differences in metabolism, changes in cell physiology, and, importantly, increased stress tolerance (O’Toole et al., 2000). As discussed in detail below, C. jejuni has been shown to form biofilms under conditions similar to environments both inside and outside animal hosts that are encountered during pathogenesis, such as low nutrient availability and high concentrations of 0,. Furthermore, the changes C. jejuni undergoes during biofilm formation do in fact appear to confer increased stress tolerance in the absence of an extensive repertoire of classical stress responses. Although our understanding of the specific mechanisms underlying biofilm formation in C. jejuni is still relatively limited, application of molecular and genetic tools is demonstrating how a global realignment of physiology and metabolism during a sessile biofilm existence can provide stress tolerance. Biofilms Contribute to Stress Tolerance in C. jejuni Survival in aquatic environments The majority of research to date on the contribution of biofilms to stress tolerance in C. jejuni has been in the context of survival in aquatic environments. Dissemination through its two major reservoirs-contaminated poultry products and untreated drinking water-both require C. jejuni to endure the challenges of low nutrients, poor osmotic support, and ambient 0, concentrations posed by such environments. Nonetheless, despite the relative fragility of C. jejuni, it is able to tolerate these challenges suf-
ficiently during transmission to be a significant public health concern (Terzieva and McFeters, 199 1). The most commonly cited source for C. jejuni infection is poultry products and uncooked food contaminated by raw poultry juices, with up to 98% of commercial samples containing live campylobacters (Kazwala et al., 1990). However, on hatching, chicks do not yet normally contain C. jejuni as part of their resident microflora (Kazwala et al., 1990). Drinking water within poultry raising facilities has been shown to be a source for infection of newly hatched birds, and the presence of C. jejuni within biofilms on drinking nipples correlates strongly with colonization of a particular flock (Zimmer et al., 2003). Thus, biofilms are of relevance to the persistence of C. jejuni within commercial poultry flocks. In the majority of studies, increased survival of C. jejuni within biofilms in environments mirroring the aquatic microcosms encountered during transmission has been reported. Greater stress tolerance has been demonstrated on abiotic materials such as stainless steel and polyvinyl chloride (PVC) (Sanders et al., 2007; Trachoo et al., 2002), and although an increase in culturability from approximately one day to up to 1 week within sessile cultures of C. jejuni has been demonstrated (Buswell et al., 1998; Rollins and Colwell, 1986; Trachoo et al., 2002), a single study indicated that biofilm-grown C. jejuni was less stress tolerant than planktonic bacteria (Dykes et al., 2003). However, in studies that use non-culturebased methods such as fluorescence in situ hybridization, viable cells were detected up to 42 days at 4°C after biofilm inoculation, implying that the presence of VBNC organisms within biofilms may cause a dramatic underestimation of the number of surviving bacteria (Keevil, 2003; Lehtola et al., 2006). This result may account for the decrease in survival of biofilm C. jejuni recorded by Dykes et al., who used culture-based methods to measure stress tolerance. In addition, numerous studies have demonstrated a twoto fourfold increase in the ability of C. jejuni to tolerate conditions of low nutrients and increased oxygenation when integrated into biofilms isolated from aquatic environments (Buswell et al., 1998; Trachoo et al., 2OO2), and that incubation in vitro with protozoan grazers such as Acanthamoeba spp. and Tetrahymena pyriformiscan, which are expected to be present within natural microcosm biofilms, increase the survival and/or culturability of C. jejuni in water (Axelsson-Olsson et al., 2005; Snelling et al., 2005). Taken together, these results confirm that presence within a biofilm does allow C. jejuni to persist in expectedly unfavorable conditions such as those encountered during transmission through aquatic environments.
CHAPTER 32
Resistance to antimicrobial agents Bacteria growing within biofilms, such as those in environmental reservoirs or within the human host during infection, are often more resistant to chemical disinfectants and antibiotics than those grown in broth culture in the laboratory (Berry et al., 2006; Fux et al., 2005; Patel, 2005). Reasons for this are thought to be multifactorial but include decreased efficacy due to metabolic downshift (Fux et al., 2005) and decreased penetration due to encasement in matrix material such as exopolysaccharide (Patel, 2005). Resistance to sanitizers is of crucial relevance to preventing C. jejuni persistence within poultry houses and abattoirs, as well as in the treatment of potable water contaminated by agricultural runoff. It has been reported that C. jejuni biofilms are in fact less sensitive than their planktonic counterparts to commonly used sanitizers such as tetrasodium phosphate (Somers et al., 1994), as well as quaternary ammonia compounds, peracetic acid, and chlorine-based disinfectants (Trachoo and Frank, 2002). It has also been shown that uptake of pathogens by protozoa present within biofilms in the natural environment contributes to resistance to chlorine disinfection (King et al., 1988; Snelling et al., 2005). To date, to our knowledge, there are no reports in the literature pertaining to increased antibiotic resistance occurring in a C. jejuni biofilm; however, a C. jejuni twocomponent sensor kinase (Acjl226c) mutant that overproduces biofilms is approximately twofold less susceptible to the antimicrobial peptides protamine and polymyxin B (S. L. Svensson and E. C. Gaynor, unpublished data). Certainly, more research in this area will aid in the selection and correct application of sanitizing procedures and antibiotic regimens for control of C. jejuni dissemination and infection. Adaptation to environments encountered in vivo As mentioned above, C. jejuni encounters numerous stresses in vivo within commensal and susceptible hosts in addition to those encountered outside the host during transmission and food processing. However, biofilm formation by C. jejuni during colonization or infection has not yet been observed. Nonetheless, the closely related pathogen Helicobacter pylori forms biofilm-like structures in gastric mucosa of humans (Carron et al., 2006; Coticchia et al., 2006), and nitrate-reducing species of Campylobacter have been identified within biofilms in the upper gastrointestinal tract of patients with Barrett’s esophagus (Macfarlane et al., 2007). Furthermore, recent work indirectly suggests that biofilms may contribute to survival in vivo. Targeted
*
SURVIVAL STRATEGIES OF C. TETUNI
581
knockout mutants in the stringent response modulator SPOT,encoding a ppGpp synthetase, and cj1226c, encoding a two-component sensor kinase, both of which display specific in vitro defects (see above) but form accelerated or enhanced biofilms (Fig. l),retain their capacities to colonize animal hosts (S. L. Svensson and E. C. Gaynor, unpublished data). In addition, the chick colonization defect of a Appkl poly P mutant, which demonstrates a trend for biofilm formation in vitro that positively correlates with cell density, can be suppressed by increasing the density of the chick inoculating dose (Candon et al., 2007). Taken together, these findings provide circumstantial evidence that biofilms may also contribute to the fitness of C. jejuni in commensal and human hosts, thus contributing to multiple aspects of the colonizationtransmission-virulence cycle. However, additional experiments are necessary to confirm and elaborate on these observations.
C . jejuni Strains Display Wide Variation in Their Ability to Form Biofilms Although specific host-related phenotypes may correlate well with virulence during gastrointestinal infection of humans, the ability of a strain to exist within a biofilm may aid stress survival and thus also contribute to the capacity of a particular isolate to cause disease. Strain-to-strain variation in the capability of different C. jejuni strains to form biofilms has been noted. For example, it appears that although widely studied laboratory strains such as NCTC 11168H and 81-176 readily form biofilms, certain clinical isolates seem to be impaired for biofilm formation (Joshua et al., 2006). In addition, Buswell et al. (1998) observed a significant twofold variation in biofilm formation between isolates from sources ranging from poultry, water, or human clinical sources. Interestingly, a recent assessment that used microarray genomotyping of C. jejuni strains from various sources suggested that the majority of human isolates appeared to be from nonlivestock sources (Champion et al., 2005). It is therefore tempting to suggest that strains that have robust mechanisms, such as biofilm formation, to survive in the environment also have a greater capacity to infect humans. However, a more complete survey of clinical and environmental isolates is necessary before further conclusions can be drawn. Toward Defining the Structure and Function of C. jejuni Biofilms Biofilm formation is a highly regulated and sequential process that follows defined steps in devel-
lox
1oox
B
Figure 1. Biofilms of Cumpylobucter jejuni 81-176 wild-type and isogenic mutants. (A) Static biofilms grown on borosilicate glass were stained with crystal violet and visualized macroscopically (insets) or were processed for scanning electron microscopy (large photographs; bars = 10 pm). The AspoT mutant forms more mature biofilms than wild type; conversely, dim mutants are defective for biofilm formation. In these strains, biofilm levels correlate with levels of a calcofluor white-reactive exopolysaccharide (EPS), which is overproduced in AspoT and underproduced in the dim mutants (McLennan et al., 2007). (B) Crystal violet-stained biofilms were visualized macroscopically (insets) or by light microscopy (large photographs; magnification, X10 and X100). The AcjZ226c mutant exhibits an extreme hyperbiofilm phenotype but displays wild-type levels of the CFW-reactive EPS (S. L. Svensson and E. C. Gaynor, unpublished data), suggesting multiple means by which biofilm formation can be upregulated in C. jejuni.
582
CHAPTER 32
opment that have been elucidated in detail in organisms such as Pseudomonas aeruginosa (O’Toole et al., 2000). These steps include surface attachment of free-swimming bacteria via cell adhesins, proliferation and spreading into microcolonies, and maturation into matrix-encapsulated, three-dimensional mushroom-like structures with extensive water channels and heterogeneity in the characteristics of resident bacteria. Dispersal of bacteria into the surrounding milieu completes the process, allowing planktonic bacteria to exit and colonize new niches. Our understanding of this process in C. jejuni, compared with other pathogens, is at present in its infancy. However, the molecular factors mediating adhesion, maturation, and matrix production, in addition to the specific phenotypic characteristics that define cells within the biofilm, are beginning to be elucidated. Autoagglutination Biofilms are characteristic of microorganisms attached to a surface, whether this surface is abiotic or organic. C. jejuni is able to autoagglutinate, indicating the presence of surface structures mediating cellcell interactions. However, although there appears to be a strong relationship between autoagglutination and host-related phenotypes in C. jejuni (Golden and Acheson, 2002; Guerry et al., 2006; Misawa and Blaser, 2000), it is not yet understood how this behavior relates to biofilm formation, if at all (Ulett et al., 2007). Nonetheless, the virulent strain 8 1-176 has strong autoagglutination activity, which is dependent on a factor that was heat labile, protease sensitive, and subject to acid extraction, but resistant to treatments affecting lipids, DNA, and sugars, suggesting a protein underlying the phenotype (Misawa and Blaser, 2000). Furthermore, although a nonmotile isogenic strain resembled the parental strain for autoagglutination activity, mutants lacking flagellar structures had drastically reduced autoagglutination activity. Other groups have also identified requirements for flagellar expression and chemotaxis (Golden and Acheson, 2002), flagellar glycosylation (Guerry et al., 2006), and luxS-mediated quorum sensing in biofilm formation (Guerry et al., 2006; Jeon et al., 2003). Molecular themes underlying C. jejuni biofilm physiology Studies that use both global and targeted molecular genetics approaches have identified genes highly expressed in cells growing in biofilms and those required for biofilm formation. Numerous groups have demonstrated genome-wide changes in gene expres-
SURVIVAL STRATEGIES OF c. rEruNI
583
sion dependent on present lifestyle (ie., sessile or biofilm versus planktonic), and this has been followed up with phenotypic analyses of targeted knockout mutants. Microarray- and proteomics-based transcriptional profiling has identified expression patterns associated with sessile (plate grown) versus planktonic bacteria (Sampathkumar et al., 2006). Immobilized bacteria undergo a shift away from general metabolism and motility toward a focus on stress defense and uptake functions. Oxidative stress response genes such as katA, trxB, and ahpC were induced in sessile bacteria, and although protein synthesis genes were downregulated, a large number of ABC transporter genes putatively involved in amino acid transport were also more highly expressed in plate-grown bacteria. However, these transporters may also be functioning as efflux pumps, further contributing to the stress tolerance that has been demonstrated in sessile C. jejuni. Finally, genes such as WUUD, involved in LOS modifications, were also induced transcriptionally, indicating changes in surface characteristics. Interestingly, the authors also noted that some of the expression patterns found in sessile bacteria overlapped with those seen by other investigators during colonization, suggesting that similar themes may arise within an animal host. In marked contrast to other bacteria, where flagellar genes are typically required for biofilm initiation but are then turned off once the biofilm is established, further proteomic analysis of biofilmgrown bacteria by Kalmokoff et al. (2006) highlighted the continued importance of the motility complex in C. jejuni biofilms. Specifically, it was found that proteins composing the motility organelle, such as the FlaA and FlaB flagellins, cap and basal body proteins, and the chemotaxis regulator CheA, were all more highly expressed in biofilm cells, as were two known adhesins, Pebl and FlaC. Pebl has also been shown to function as an ABC transporter glutamate or aspartate binding protein (Leon-Kempis Mdel et al., 2006), consistent with the abovementioned upregulation of ABC transporters in biofilm bacteria. Furthermore, as with plate-grown bacteria, proteins known to be involved in stress tolerance, such as the GroEL and GroES heat shock chaperones, a ClpP homolog, and the oxidative stress proteins Tpx and Ahp, were also induced. The authors noted that expression of stress response proteins in actively growing biofilms is unusual compared with other bacteria; however, this may reflect the lifestyle of C. jejuni as a zoonotic pathogen. Broad changes in the metabolic focus between growth phases that mirrored those that have been reported in other pathogens were seen, and expression
584
SVENSSON ET AL.
profiles suggested that C. jejuni remains metabolically active within the biofilm. Finally, targeted mutations in some of the upregulated genes, such as those affecting the flagellar filament, as well as an aflagellate flhA mutant and flaC adhesin mutant, resulted in an inability to form pellicles in vitro. In-depth analyses of monospecies biofilms demonstrated that C. jejuni NCTC 11168 can form three distinct forms of biofilms: cell-cell aggregates, pellicles at the air-liquid interface, and glass-attached flocs (Joshua et al., 2006). Not surprisingly, each of these structures exhibited increased survival compared with their planktonic counterparts. Mutations in pis, as well as cjO688, encoding a putative phosphate acetyltransferase, resulted in the inability of C. jejuni to form flocs and pellicles. However, mutations in kpsM, pglH, or neuBl were not impaired in forming flocs, suggesting that biofilm formation is independent of capsular polysaccharide, protein glycosylation, and LOS, respectively. The protective matrix of microbial biofilms is often composed of polysaccharides, proteins, and DNA (Branda et al., 2005). However, the nature of the polymeric matrix encapsulating and protecting the C. jejuni biofilm remains poorly defined. The related pathogen H. pylori secretes a hydrophilic exopolysaccharide during biofilm growth (Stark et al., 1999), and other gram-negative enteric pathogens such as Salmonella spp. and E. coli also extensively utilize carbohydrates (Romling, 2005). Serendipitously, a AspoT stringent response mutant has been found to overproduce a novel calcofluor white (CFW)reactive exopolysaccharide, independent of both capsule and LOS (McLennan et al., 2007). Interestingly, the AspoT mutant was also found to overproduce biofilms, and CFW-dim mutants isolated from a transposon mutagenesis library were correspondingly poor biofilm formers (Fig. 1).Furthermore, although the stringent response mutant exhibited defects in some stress-related phenotypes, many were unaffected, supporting the notion that a biofilm lifestyle, and specifically the exopolysaccharide matrix, may confer resilience to C. jejuni in the face of environmental challenges. Control of Biofilm Formation with a Limited Repertoire of Regulatory Factors Phenomena regulating the sequential steps of biofilm formation have been extensively characterized in other pathogens such as P. aeruginosa, Vibrio cholerae, and E. coli and is under genetic control by a wide variety of regulatory schemes, including twocomponent regulatory systems (Fredericks et al., 2006), global regulators such as CsrA Uackson et al.,
2002), levels of the second messenger cyclic di-GMP (Kuchma et al., 2007), autoinducer-2-mediated quorum sensing (Rickard et al., 2006), and the stringent response (Balzer and McLean, 2002), all of which are present in the limited regulatory repertoire of C. jejuni. Nevertheless, while de novo protein synthesis is required for biofilm formation (Reeser et al., 2007) and the global expression profile experiments mentioned above demonstrate that C. jejuni cells growing as a biofilm are physiologically distinct from their planktonic counterparts, understanding of the regulatory circuits underlying biofilm formation in C. jejuni remains limited. Conditions promoting biofilm formation in C. jejuni Biofilm formation in C. jejuni is promoted by growth in conditions of low nutrient availability, such as those encountered in aquatic environments (Reeser et al., 2007). Recent data also suggest that bile salts such as deoxycholate also stimulate biofilm formation in C. jejuni (S. L. Svensson and E. C. Gaynor, unpublished data), again suggesting that biofilms may be a survival mechanism important within the gastrointestinal tract of commensal or susceptible hosts. In contrast, presence of osmolytes such as sucrose, glucose, or sodium chloride significantly decreased biofilm formation (Reeser et al., 2007). Thus, specific conditions mirroring stresses that C. jejuni may encounter during pathogenesis appear to influence presently unidentified regulatory circuits that control biofilm formation. Stringent control The stringent response in C. jejuni, mediated by SPOT,has been shown to be important for survival of low nutrient stress (Gaynor et al., 2005). Furthermore, recent work has suggested that like in other bacteria, the stringent response intersects with metabolism of Poly P (Candon et al., 2007), a ubiquitous molecule involved in stress tolerance. However, inactivation of either SPOTor p p k l in C. jejuni resulted in enhanced biofilm formation, which is in contrast to observations in other bacteria. At present, it is unknown whether altered levels of ppGpp and/or Poly P in these strains acts as a specific signal stimulating initiation of biofilm formation, or whether the bacteria may be continually stressed and initiating compensatory responses that promote differentiation into the more resilient biofilm state. Furthermore, the handful of regulatory networks in C. jejuni may play broader and/ or distinct roles that encompass aspects of biofilm formation instead of other absent regula-
CHAPTER 32
tory schemes, such as those mediated by the RpoS stationary phase sigma factor or the secondary ppGpp synthetase RelA. AI-2-mediated quorum sensing Early studies of C. jejuni biofilms demonstrated that their formation and stress tolerance are greatly promoted in the presence of other microflora (Buswell et al., 1998). To this end, it is unknown whether this is because of the physical environment presented by coinhabitants or aspects of quorum sensing, which is known to mediate biofilm attachment and dispersal in other bacteria. C. jejuni encodes a homolog of the ZuxS autoinducer-2 synthase and has been shown to produce AI-2 in a ZuxS-dependent fashion (Elvers and Park, 2002). Furthermore, inactivation of LuxS activity has been shown to reduce both autoagglutination (Guerry et al., 2006) and biofilm formation (Reeser et al., 2007), and this the defect can be complemented by addition of cell-free extracts from multispecies biofilms (Reeser et al., 2007). However, although ZuxS is required for efficient transcription of flagellar genes (Jeon et al., 2003), which are known to be important for biofilm formation in C. jejuni, the factors that detect AI-2 appear to be absent from the genome. It is therefore unclear whether quorum sensing plays a regulatory role in biofilm formation in C. jejuni. Two-component regulatory systems and cyclic di-GMP Within its small complement of two-component regulatory systems, C. jejuni lacks obvious homologs of systems that control aspects of biofilm formation in other pathogens, such as the GacSA system of P. aeruginosa (Parkins et al., 2001). However, initiation of biofilm formation may be under environmental control of the two-component system encoded by cj1226c and cj1227c because mutation of the 1226 sensor kinase results in marked acceleration and enhancement of biofilm formation (Fig. 1; S. L. Svensson and E. C. Gaynor, unpublished data). Control of biofilm-related phenotypes by two-component systems is also supported by the involvement of the phosphate acetyltransferase encoded by cj0688 in biofilm formation (Joshua et al., 2006). The small molecule acetyl phosphate plays a role in the regulation of biofilm formation in E. coZi (Wolfe et al., 2003), possibly through direct phosphorylation of response regulators (McCleary and Stock, 1994; McCleary et al., 1993). In addition, C. jejuni encodes an orphan response regulator (cbrR) containing a putative GGDEF domain, which may contribute to lev-
SURVIVAL STRATEGIES OF c. r E r m I
58s
els of cyclic di-GMP (Raphael et al., 2005), a second messenger that is a key mediator of biofilm formation in other pathogens (Romling and Amikam, 2006). A role for the CbrR GGDEF domain in control of C. jejuni biofilm phenotypes remains to be identified. Certainly, identification of specific environmental conditions stimulating biofilm formation and the regulatory networks controlling maturation of biofilms will provide insight into how this alternative lifestyle complements C. jejuni’s relatively limited stress response networks. CONCLUSIONS AND FUTURE DIRECTIONS C. jejuni clearly harbors both specific and global means to counter stress, and these means can be effected by either specific gene products or wholepopulation differentiation strategies. However, much remains to be discovered regarding this important facet of C. jejuni pathogenesis. For instance, genes and pathways involved in responding to several stresses C. jejuni must certainly encounter in numerous settings have not yet begun to be elucidated. An example of this is envelope stress (Raivio, 2005), which is known to involve membrane remodeling, phospholipid dynamics, and periplasmic protein quality control in other bacteria but represents a nascent area of investigation for C. jejuni. It will also be interesting to explore the roles of genes or factors predicted to participate in stress response pathways either because of a previously established role in pathogenesis or because of high homology to stress response factors in other bacteria. An example of the former is the dccRS two-component signal transduction system, which affects colonization by an as yet unidentified mechanism (MacKichan et al., 2004); an example of the latter is SsrA (transfer-messenger RNA), which has been shown in other bacteria to help counter stresses by rescuing stalled ribosomes (Withey and Friedman, 2003), but thus far remains unstudied in C. jejuni. Finally, it is important to point out that although this chapter specifically reviewed survival strategies and stress responses, many of the genes and pathways discussed affect multiple or all aspects of the pathogenesis cycle, encompassing colonization, transmission, and virulence. As such, future insight into additional stress response pathways, as well as better in-depth understanding of those already identified, should provide significant new information on how this important pathogen both causes disease and persists so successfully in a variety of environmental milieus. Acknowledgments. S.L.S. and E.F. contributed equally to this chapter. We thank past and current members of the Gaynor laboratory for their direct or indirect contributions to work presented
586
SVENSSON ET AL.
here, including Meghan McLennan, Danielle Ringoir, Heather Candon, Olivia Champion, Ann Lin, Mizue Naito, Jenny Vermeulen, Mark McCabe, Ryan Walton, and Craig Hawkshaw. We also thank our collaborators, Christine Szymanski, Derek Wells, Brenda Allan, and Cres Fraley. Work in the Gaynor laboratory is supported by the following grants to E.C.G.: Canadian Institutes of Health Research (CIHR) operating grant MOP-68981, a Burroughs Wellcome Fund Career Award in the Biomedical Sciences, and a Michael Smith Foundation for Health Research Scholar Award. Salary support is provided by a Canada Research Chair award to E.C.G., a CIHR postdoctoral fellowship to E.F., and a Natural Sciences and Engineering Research Council (NSERC) award to S.L.S.
REFERENCES Alonso, J. L., S. Mascellaro, Y. Moreno, M. A. Ferrus, and J. Hernandez. 2002. Double-staining method for differentiation of morphological changes and membrane integrity of Campylobacter coli cells. Appl. Environ. Microbiol. 685151-5154. Alter, T., and K. Scherer. 2006. Stress response of Campylobacter spp. and its role in food processing. J. Vet. Med. B Infect. Dis. Vet. Public Health 53:351-357. Amano, K., and Y. Shibata. 1992. Structural studies of peptidoglycans in Campylobacter species. Microbiol. Immunol. 36:961967. Andersen, M. T., L. Brondsted, B. M. Pearson, F. Mulholland, M. Parker, C. Pin, J. M. Wells, and H. Ingmer. 2005. Diverse roles for HspR in Campylobacter jejuni revealed by the proteome, transcriptome and phenotypic characterization of an hspR mutant. Microbiology 151:905-915. Auld, H., D. MacIver, and J. Klaassen. 2004. Heavy rainfall and waterborne disease outbreaks: the Walkerton example. J. Toxicol. Environ. Health. A 67:1879-1887. Axelsson-Olsson, D., J. Waldenstrom, T. Broman, B. Olsen, and M. Holmberg. 2005. Protozoan Acanthamoeba polyphaga as a potential reservoir for Campylobacter jejuni. Appl. Environ. Microbiol. 71:987-992. Baffone, W., A. Casaroli, B. Citterio, L. Pierfelici, R. Campana, E. Vittoria, E. Guaglianone, and G. Donelli. 2006. Campylobacter jejuni loss of culturability in aqueous microcosms and ability to resuscitate in a mouse model. Int. J. Food Microbiol. 107:8391. Baillon, M. L., A. H. van Vliet, J. M. Ketley, C. Constantinidou, and C. W. Penn. 1999. An iron-regulated alkyl hydroperoxide reductase (AhpC) confers aerotolerance and oxidative stress resistance to the microaerophilic pathogen Campylobacter jejuni. J. Bacteriol. 181:4798-4804. Balzer, G. J., and R. J. McLean. 2002. The stringent response genes relA and spoT are important for Escherichia coil biofilms under slow-growth conditions. Can. J. Microbiol. 48:675-680. Barer, M. R., and C. R. Harwood. 1999. Bacterial viability and culturability. Adv. Microb. Physiol. 41:93-137. Benaissa, M., P. Babin, N. Quellard, L. Pezennec, Y. Cenatiempo, and J. L. Fauchere. 1996. Changes in Helicobacter pylon' ultrastructure and antigens during conversion from the bacillary to the coccoid form. Infect. lmmun. 64:2331-2335. Berry, D., C. Xi, and L. Raskin. 2006. Microbial ecology of drinking water distribution systems. Curr. Opin. Biotechnol. 17:297302. Beumer, R. R., J. de Vries, and F. M. Rombouts. 1992. Campylobacter jejuni non-culturable coccoid cells. Int. ]. Food Microbiol. 15:153-163. Birk, T., H. Ingmer, M. T. Andersen, K. Jorgensen, and L. Brondsted. 2004. Chicken juice, a food-based model system suit-
able to study survival of Campylobacter jejuni. Lett. Appl. Microbiol. 38:66-71. Bode, G., P. Malfertheiner, G. Lehnhardt, M. Nilius, and H. Ditschuneit. 1993. Ultrastructural localization of urease of Helicobacter pylori. Med. Microbiol. Immunol. 182:233-242. Boucher, S. N., E. R. Slater, A. H. Chamberlain, and M. R. Adams. 1994. Production and viability of coccoid forms of Campylobacter jejuni. J. Appl. Bacteriol, 77:303-307. Branda, S. S., S. Vik, L. Friedman, and R. Kolter. 2005. Biofilms: the matrix revisited. Trends Microbiol. 13:20-26. Bras, A. M., S. Chatterjee, B. W. Wren, D. G. Newell, and J. M. Ketley. 1999. A novel Campylobacter jejuni two-component regulatory system important for temperature-dependent growth and colonization. J. Bacteriol. 181:3298-3302. Brondsted, L., M. T. Andersen, M. Parker, K. Jorgensen, and H. Ingmer. 2005. The HtrA protease of Campylobacter jejuni is required for heat and oxygen tolerance and for optimal interaction with human epithelial cells. Appl. Environ. Microbiol. 71: 3205-3212. Brown, M. R., and A. Kornberg. 2004. Inorganic polyphosphate in the origin and survival of species. Proc. Natl. Acad. Sci. USA 101:16085-16087. Buck, G. E., K. A. Parshall, and C. P. Davis. 1983. Electron microscopy of the coccoid form of Campylobacter jejuni. J. Clin. Microbiol. 18:420-42 1. Buswell, C. M., Y. M. Herlihy, L. M. Lawrence, J. T. McGuiggan, P. D. Marsh, C. W. Keevil, and S. A. Leach. 1998. Extended survival and persistence of Campylobacter spp. in water and aquatic biofilms and their detection by immunofluorescentantibody and -rRNA staining. Appl. Environ. Microbiol. 64:733741. Butzler, J. P., and M. B. Skirrow. 1979. Campylobacter enteritis. Clin. Gastroenterol. 8:737-765. Cabeen, M. T., and C. Jacobs-Wagner. 2005. Bacterial cell shape. Nut. Rev. Microbiol. 3:601-610. Cabiscol, E., J. Tamarit, and J. Ros. 2000. Oxidative stress in bacteria and protein damage by reactive oxygen species. lnt. Microbiol. 3:3-8. Candon, H. L., B. J. Allan, C. D. Fraley, and E. C. Gaynor. 2007. Polyphosphate kinase 1 is a pathogenesis determinant in Campylobacter jejuni. J. Bacteriol. 18953099-8108. Cappelier, J. M., C. Magras, J. L. Jouve, and M. Federighi. 1999a. Recovery of viable but non-culturable Campylobacter jejuni cells in two animal models. Food Microbiol. 16:375-383. Cappelier, J. M., J. Minet, C. Magras, R. R. Colwell, and M. Federighi. 1999b. Recovery in embryonated eggs of viable but nonculturable Campylobacter jejuni cells and maintenance of ability to adhere to HeLa cells after resuscitation. Appl. Environ. Microbiol. 6 5 5 154-5 157. Cappelier, J. M., A. Rossero, and M. Federighi. 2000. Demonstration of a protein synthesis in starved Campylobacter jejuni cells. lnt. 1.Food Microbiol. 55:63-67. Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R. Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K. Amoako, S. Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Genome-wide expression analyses of Campylobacter jejuni NCTCl 1168 reveals coordinate regulation of motility and virulence by flhA. J. Biol. Chem. 279:20327-20338. Carron, M. A., V. R. Tran, C. Sugawa, and J. M. Coticchia. 2006. Identification of Helicobacter pylori biofilms in human gastric mucosa. J. Gastrointest. Surg. 10:712-8 17. Champion, 0. L., M. W. Gaunt, 0. Gundogdu, A. Elmi, A. A. Witney, J. Hinds, N. Dorrell, and B. W. Wren. 2005. Comparative phylogenomics of the food-borne pathogen Campylobacter
CHAPTER 32
jejuni reveals genetic markers predictive of infection source. Proc. Natl. Acad. Sci. USA 102:16043-10648. Chan, K. F., H. Le Tran, R. Y. Kanenaka, and S. Kathariou. 2001. Survival of clinical and poultry-derived isolates of Campylobacter jejuni at a low temperature (4 degrees C). Appl. Environ. Microbiol. 67:4186-4191. Chaput, C., C. Ecobichon, N. Cayet, S. E. Girardin, C. Werts, S. Guadagnini, M. C. Prevost, D. Mengin-Lecreulx, A. Labigne, and I. G. Boneca. 2006. Role of AmiA in the morphological transition of Helicobacter pylori and in immune escape. PLoS Patbog. 2:e97. Chaveerach, P., A. A. ter Huurne, L. J. Lipman, and F. van Knapen. 2003. Survival and resuscitation of ten strains of Campylobacter jejuni and Campylobacter coli under acid conditions. Appl. Enuiron. Microbiol. 69:711-714. Chou, S. P., R. Dular, and S. Kasatiya. 1983. Effect of ferrous sulfate, sodium metabisulfite, and sodium pyruvate on survival of Campylobacter jejuni. 1. Clin. Microbiol. 18:986-987. Clark, C. G., L. Price, R. Ahmed, D. L. Woodward, P. L. Melito, F. G. Rodgers, F. Jamieson, B. Ciebin, A. Li, and A. Ellis. 2003. Characterization of waterborne outbreak-associated Campylobacter jejuni, Walkerton, Ontario. Emerg. Infect. Dis. 9: 12321241. Cools, I., M. Uyttendaele, C. Caro, E. D’Haese, H. J. Nelis, and J. Debevere. 2003. Survival of Campylobacter jejuni strains of different origin in drinking water. 1.Appl. Microbiol. 94:886892. Costa, K., G. Bacher, G. Allmaier, M. G. Dominguez-Bello, L. Engstrand, P. Falk, M. A. de Pedro, and F. Garcia-del Portillo. 1999. The morphological transition of Helicobacter pylori cells from spiral to coccoid is preceded by a substantial modification of the cell wall. J. Bacteriol. 181:3710-3715. Coticchia, J. M., C. Sugawa, V. R. Tran, J. Gurrola, E. Kowalski, and M. A. Carron. 2006. Presence and density of Helicobacter pylori biofilms in human gastric mucosa in patients with peptic ulcer disease. ] Gastrointest. Surg. 10:883-889. Day, W. A., Jr., J. L. Sajecki, T. M. Pitts, and L. A. Joens. 2000. Role of catalase in Campylobacter jejuni intracellular survival. Infect. Immun. 68 :6337-6345. Doyle, M. P., and D. J. Roman. 1981. Growth and survival of Campylobacter fetus subsp. jejuni as a function of temperature and pH. J. Food Prot. 44596-601. Doyle, M. P., and D. J. Roman. 1982. Response of Campylobacter jejuni to sodium chloride. Appl. Environ. Microbiol. 43561565. Dykes, G. A., B. Sampathkumar, and D. R. Korber. 2003. Planktonic or biofilm growth affects survival, hydrophobicity and protein expression patterns of a pathogenic Campylobacter jejuni strain. lnt. J. Food Microbiol. 89:l-10. Ehrlich, G . D., F. Z. Hu, K. Shen, P. Stoodley, and J. C. Post. 2005. Bacterial plurality as a general mechanism driving persistence in chronic infections. Clin. Orthop. Relat. Res. (473):2024. Elvers, K. T., and S. F. Park. 2002. Quorum sensing in Campylobacter jejuni: detection of a luxS encoded signalling molecule. Microbiology 148:1475-1481. Elvers, K. T., S. M. Turner, L. M. Wainwright, G. Marsden, J. Hinds, J. A. Cole, R. K. Poole, C. W. Penn, and S. F. Park. 2005. NssR, a member of the Crp-Fnr superfamily from Campylobacter jejuni, regulates a nitrosative stress-responsive reguIon that includes both a single-domain and a truncated haemoglobin. Mol. Microbiol. 57:735-750. Elvers, K. T., G. Wu, N. J. Gilberthorpe, R. K. Poole, and S. F. Park. 2004. Role of an inducible single-domain hemoglobin in mediating resistance to nitric oxide and nitrosative stress in
SURVIVAL STRATEGIES OF c. rEruNI
587
Campylobacter jejuni and Campylobacter coli. 1.Bacteriol. 186: 5332-5341. Federighi, M., J. L. Tholozan, J. M. Cappelier, J. P. Tissier, and J. L. Jouve. 1998. Evidence of non-coccoid viable but nonculturable Campylobacter jejuni cells in microcosm water by direct viable count, CTC-DAPI double staining, and scanning electron microscopy. Food Microbiol. 15539-550. Fox, E. M., M. Raftery, A. Goodchild, and G. L. Mendz. 2007. Campylobacter jejuni response to ox-bile stress. FEMS lmmunol. Med. Microbiol. 49: 165-172. Fredericks, C. E., S. Shibata, S. Aizawa, S. A. Reimann, and A. J. Wolfe. 2006. Acetyl phosphate-sensitive regulation of flagellar biogenesis and capsular biosynthesis depends on the Rcs phosphorelay. Mol. Microbiol. 61:734-747. Fux, C. A., J. W. Costerton, P. S. Stewart, and P. Stoodley. 2005. Survival strategies of infectious biofilms. Trends Microbiol. 13: 34-40. Gaynor, E. C., S. Cawthraw, G. Manning, J. K. MacKichan, S. Falkow, and D. G. Newell. 2004. The genome-sequenced variant of Campylobacter jejuni NCTC 11168 and the original clonal clinical isolate differ markedly in colonization, gene expression, and virulence-associated phenotypes. ]. Bacteriol. 186: 503-5 17. Gaynor, E. C., D. H. Wells, J. K. MacKichan, and S. Falkow. 2005. The Campylobacter jejuni stringent response controls specific stress survival and virulence-associated phenotypes. Mol. Microbiol. 56:8-27. Golden, N. J., and D. W. Acheson. 2002. Identification of motility and autoagglutination Cumpylobacter jejuni mutants by random transposon mutagenesis. Infect. Immun. 70:1761-1771. Grant, K. A., and S. F. Park. 1995. Molecular characterization of katA from Campylobacter jejuni and generation of a catalasedeficient mutant of Campylobacter coli by interspecific allelic exchange. Microbiology 141(Pt. 6):1369-13 76. Griffiths, P. L. 1993. Morphological changes of Campylobacter jejuni growing in liquid culture. Lett. Appl. Microbiol. 17:152155. Gu, R., C. C. Su, F. Shi, M. Li, G. McDermott, Q. Zhang, and E. W. Yu. 2007. Crystal structure of the transcriptional regulator CmeR from Campylobacter jejuni. J. Mol. Biol. 372583593. Guerry, P., C. P. Ewing, M. Schirm, M. Lorenzo, J. Kelly, D. Pattarini, G. Majam, P. Thibault, and s. Logan. 2006. Changes in flagellin glycosylation affect Campylobacter autoagglutination and virulence. Mol. Microbiol. 60:299-3 11. Hald, B., K. Knudsen, P. Lind, and M. Madsen. 2001. Study of the infectivity of saline-stored Campylobacter jejuni for day-old chicks. Appl. Enuiron. Microbiol. 67:2388-2392. Hazeleger, W., C. Arkesteijn, A. Toorop-Bouma, and R. Beumer. 1994. Detection of the coccoid form of Campylobacter jejuni in chicken products with the use of the polymerase chain reaction. Int. 1.Food Microbiol. 24:273-281. Hazeleger, W. C., J. D. Janse, P. M. Koenraad, R. R. Beumer, F. M. Rombouts, and T. Abee. 1995. Temperature-dependent membrane fatty acid and cell physiology changes in coccoid forms of Campylobacter jejuni. Appl. Enuiron. Microbiol. 61: 2713-2719. Hendrixson, D. R. 2006. A phase-variable mechanism controlling the Campylobacter jejuni FlgR response regulator influences commensalism. Mol. Microbiol. 61: 1646-1659. Hodge, J. P., and N. R Krieg. 1994. Oxygen tolerance estimates in Campylobacter species depend on the testing medium.]. Appl. Bacteriol. 77:666-673. Holler, C., D. Witthuhn, and B. Janzen-Blunck. 1998. Effect of low temperatures on growth, structure, and metabolism of Campylobacter coli SP10. Appl. Enuiron. Microbiol. 6 4 5 8 1 4 8 7 .
588
SVENSSON ET AL.
Hudock, J. F., A. C. Borger, and C. W. Kaspar. 2005. Temperature-dependent genome degradation in the coccoid form of Campylobacter jejuni. Curr. Microbiol. 5O:llO-113. Ishikawa, T., Y. Mizunoe, S. Kawabata, A. Takade, M. Harada, S. N. Wai, and S. Yoshida. 2003. The iron-binding protein Dps confers hydrogen peroxide stress resistance to Campylobacter jejuni. J. Bacteriol. 185:1010-1017. Jackson, D. W., K. Suzuki, L. Oakford, J. W. Simecka, M. E. Hart, and T. Romeo. 2002. Biofilm formation and dispersal under the influence of the global regulator CsrA of Escherichia coli. J. Bacteriol. 184:290-3 0 1. Jacob, J., W. Martin, and C. Holler. 1993. Characterization of viable but nonculturable stage of C. coli, characterized with respect to electron microscopic findings, whole cell protein and lipooligosaccharide (LOS) patterns. Zentralbl. Mikrobiol. 148:310. Jeon, B., K. Itoh, N. Misawa, and S. Ryu. 2003. Effects of quorum sensing on flaA transcription and autoagglutination in Campylobacter jejuni. Microbiol. Immunol. 47:833-839. Jesudason, M. V., D. J. Hentges, and P. Pongpech. 1989. Colonization of mice by Campylobacter jejuni. Infect. lmmun. 57: 2279-2282. Jiang, X., and T. J. Chai. 1996. Survival of Vibrio parahaemolyticus at low temperatures under starvation conditions and subsequent resuscitation of viable, nonculturable cells. Appl. Environ. Microbiol. 62: 1300-1 3 05. Jones, D. M., E. M. Sutcliffe, and A. Curry. 1991. Recovery of viable but non-culturable Campylobacter jejuni. J. Gen. Microbiol. 137:2477-2482. Jones, D. M., E. M. Sutcliffe, R. Rios, A. J. Fox, and A. Curry. 1993. Campylobacter jejuni adapts to aerobic metabolism in the environment. J. Med. Microbiol. 38:145-150. Joshua, G. W., C. Guthrie-Irons, A. V. Karlyshev, and B. W. Wren. 2006. Biofilm formation in Campylobacter jejuni. Microbiology 1 5 2 ~87-396. 3 Kalmokoff, M., P. Lanthier, T. L. Tremblay, M. Foss, P. C. Lau, G. Sanders, J. Austin, J. Kelly, and C. M. Szymanski. 2006. Proteomic analysis of Campylobacter jejuni 11168 biofilms reveals a role for the motility complex in biofilm formation. J. Bacteriol. 188:4312-4320. Kazwala, R. R., J. D. Collins, J. Hannan, R. A. Crinion, and H. O’Mahony. 1990. Factors responsible for the introduction and spread of Campylobacter jejuni infection in commercial poultry production. Vet. Rec. 126:305-306. Keevil, C. W. 2003. Rapid detection of biofilms and adherent pathogens using scanning confocal laser microscopy and episcopic differential interference contrast microscopy. Water Sci. Technol. 47: 105-1 16. Kelly, A. F., S. F. Park, R. Bovill, and B. M. Mackey. 2001. Survival of Campylobacter jejuni during stationary phase: evidence for the absence of a phenotypic stationary-phase response. Appl. Environ. Microbiol. 67:2248-2254. King, C. H., E. B. Shotts, Jr., R. E. Wooley, and K. G . Porter. 1988. Survival of coliforms and bacterial pathogens within protozoa during chlorination. Appl. Environ. Microbiol. 54:30233033. Klancnik, A., N. Botteldoorn, L. Herman, and S. S. Mozina. 2006. Survival and stress induced expression of groEL and rpoD of Campylobacter jejuni from different growth phases. Int. J. Food Microbiol. 112:200-2007. Kogure, K., U. Simidu, and N. Taga. 1979. A tentative direct microscopic method for counting living marine bacteria. Can. J. Microbiol. 25:415-420. Konkel, M. E., B. J. Kim, J. D. Klena, C. R. Young, and R. Ziprin. 1998. Characterization of the thermal stress response of Campylobacter jejuni. Infect. lmmun. 66:3666-3672.
Kuchma, S. L., K. M. Brothers, J. H. Merritt, N. T. Liberati, F. M. Ausubel, and A. O’Toole G. 2007. BifA, a c-di-GMP phosphodiesterase, inversely regulates biofilm formation and swarming motility by Pseudomonas aeruginosa PA14. J. Bacteriol. 189: 8165-8 178. Kusters, J. G., M. M. Gerrits, J. A. Van Strijp, and C. M. Vandenbroucke-Grauls. 1997. Coccoid forms of Helicobacter pylori are the morphologic manifestation of cell death. Infect. Immun. 65 :3672-3 679. Lazaro, B., J. Carcamo, A. Audicana, I. Perales, and A. FernandezAstorga. 1999. Viability and DNA maintenance in nonculturable spiral Campylobacter jejuni cells after long-term exposure to low temperatures. Appl. Environ. Microbiol. 65:4677-468 1. Lehtola, M. J., T. Pitkanen, L. Miebach, and I. T. Miettinen. 2006. Survival of Campylobacter jejuni in potable water biofilms: a comparative study with different detection methods. Water Sci. Technol. 5457-61. Leon-Kempis Mdel, R., E. Guccione, F. Mulholland, M. P. Williamson, and D. J. Kelly. 2006. The Campylobacter jejuni PEBla adhesin is an aspartateightamate-binding protein of an ABC transporter essential for microaerobic growth on dicarboxylic amino acids. Mol. Microbiol. 60:1262-1275. Lin, J., M. Akiba, 0. Sahin, and Q. Zhang. 2005a. CmeR functions as a transcriptional repressor for the multidrug efflux pump CmeABC in Campylobacter jejuni. Antimicrob. Agents Chemother. 49:1067-1075. Lin, J., C. Cagliero, B. Guo, Y. W. Barton, M. C. Maurel, S. Payot, and Q. Zhang. 2005b. Bile salts modulate expression of the CmeABC multidrug efflux pump in Campylobacter jejuni. J. Bacteriol. 187:74 17-7424. Lin, J., and A. Martinez. 2006. Effect of efflux pump inhibitors on bile resistance and in vivo colonization of Campylobacter jejuni. J. Antimicrob. Chemother. 58:966-972. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Chemother. 46:2124-213 1. Lin, J., 0. Sahin, L. 0. Michel, and Q. Zhang. 2003. Critical role of multidrug efflux pump CmeABC in bile resistance and in vivo colonization of Campylobacter jejuni. Infect. Immun. 71:42504259. Macfarlane, S., E. Furrie, G. T. Macfarlane, and J. F. Dillon. 2007. Microbial colonization of the upper gastrointestinal tract in patients with Barrett’s esophagus. Clin. Infect. Dis.45:29-38. MacKichan, J. K., E. C. Gaynor, C. Chang, S. Cawthraw, D. G. Newell, J. F. Miller, and S. Falkow. 2004. The Campylobacter jejuni dccRS two-component system is required for optimal in vivo colonization but is dispensable for in vitro growth. Mol. Microbiol. 54:1269-1286. Martinez-Rodriguez, A., A. F. Kelly, S. F. Park, and B. M. Mackey. 2004. Emergence of variants with altered survival properties in stationary phase cultures of Campylobacter jejuni. Int. J. Food Microbiol. 90:321-329. McCleary, W. R., and J. B. Stock. 1994. Acetyl phosphate and the activation of two-component response regulators. J. Biol. Chem. 269:3 1567-3 1572. McCleary, W. R., J. B. Stock, and A. J. Ninfa. 1993. Is acetyl phosphate a global signal in Escherichia coli? J. Bacteriol. 175: 2793-2798. McLennan, M. K., D. D. Ringoir, E. Frirdich, S. L. Svensson, D. H. Wells, H. C. Jarrell, C. M. Szymanski, and E. C. Gaynor. 2007. Campylobacter jejuni biofilms up-regulated in the absence of the stringent response utilize a calcofluor white-reactive surface polysaccharide. J. Bacteriol. 190:1097-1107. Medema, G. J., F. M. Schets, A. W. van de Giessen, and A. H. Havelaar. 1992. Lack of colonization of 1 day old chicks by
CHAPTER 32
viable, non-culturable Campylobacter jejuni. J. Appl. Bacteriol. 72:512-516. Mihaljevic, R. R., M. Sikic, A. Klancnik, G. Brumini, S . S . Mozina, and M. Abram. 2007. Environmental stress factors affecting survival and virulence of Campylobacter jejuni. Microb. Pathog. 43: 120-125. Misawa, N., and M. J. Blaser. 2000. Detection and characterization of autoagglutination activity by Campylobacter jejuni. Infect. Immun. 68:6168-6175. Moore, J. E. 2001. Bacterial dormancy in Campylobacter: abstract theory or cause for concern? Int. J. Food Sci. Technol. 36593600. Moran, A. P., and M. E. Upton. 1986. A comparative study of the rod and coccoid forms of Campylobacter jejuni ATCC 29428. J. Appl. Bacteriol. 60:103-110. Moran, A. P., and M. E. Upton. 1987. Factors affecting production of coccoid forms by Campylobacter jejuni on solid media during incubation. J. Appl. Bacteriol. 625277537. Murphy, C., C. Carroll, and K. N. Jordan. 2005. The effect of different media on the survival and induction of stress responses by Campylobacter jejuni. J. Microbiol. Methods 62:161-166. Murphy, C., C. Carroll, and K. N. Jordan. 2006. Environmental survival mechanisms of the foodborne pathogen Campylobacter jejuni. J. Appl. Microbiol. 100:623-632. Murphy, C., C. Carroll, and K. N. Jordan. 2003a. Identification of a novel stress resistance mechanism in Campylobacter jejuni. J. Appl. Microbiol. 95:704-708. Murphy, C., C. Carroll, and K. N. Jordan. 2003b. Induction of an adaptive tolerance response in the foodborne pathogen, Campylobacter jejuni. FEMS Microbiol. Lett. 223:89-93. Ng, L. K., R. Sherburne, D. E. Taylor, and M. E. Stiles. 1985. Morphological forms and viability of Campylobacter species studied by electron microscopy. J. Bacteriol. 164:338-343. Nguyen, H. T., J. E. Corry, and C. A. Miles. 2006. Heat resistance and mechanism of heat inactivation in thermophilic campylobacters. Appl. Enuiron. Microbiol. 72:908-913. Nilius, M., A. Strohle, G. Bode, and P. Malfertheiner. 1993. Coccoid like forms (CLF) of Helicobacter pylori. Enzyme activity and antigenicity. Zentralbl. Bakteriol. 280:259-272. Okoli, A. S., T. Wadstrom, and G. L. Mendz. 2007. MiniReview: bioinformatic study of bile responses in Campylobacterales. FEMS Immunol. Med. Microbiol. 49:lOl-123. Oliver, J. D. 2005. The viable but nonculturable state in bacteria. J Microbiol. 43:93-100. O’Reilly, C. E., A. B. Bowen, N. E. Perez, J. P. Sarisky, C. A. Shepherd, M. D. Miller, B. C. Hubbard, M. Herring, S . D. Buchanan, C. C. Fitzgerald, V. Hill, M. J. Arrowood, L. X. Xiao, R. M. Hoekstra, E. D. Mintz, and M. F. Lynch. 2007. A waterborne outbreak of gastroenteritis with multiple etiologies among resort island visitors and residents: Ohio, 2004. Clin. Infect. Dis. 44506-612. O’Toole, G., H. B. Kaplan, and R. Kolter. 2000. Biofilm formation as microbial development. Annu. Rev. Microbiol. 54:49-79. Park, S. F. 2005. Campylobacter jejuni stress responses during survival in the food chain and colonization, p. 311-330. In J. M. Ketley and M. E. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Bioscience, Norfolk, United Kingdom. Park, S . F. 2002. The physiology of Campylobacter species and its relevance to their role as foodborne pathogens. Int. J. Food Microbiol. 74:177-188. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S . Holroyd, K. Jagels, A. V. Karlyshev, S . Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S . Whitehead, and B. G. Barrell. 2000. The genome
SURVIVAL STRATEGIES OF C. TETUNI
589
sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature. 403:665-668. Parkins, M. D., H. Ceri, and D. G. Storey. 2001. Pseudomonas aeruginosa GacA, a factor in multihost virulence, is also essential for biofilm formation. Mol. Microbiol. 40:1215-1226. Patel, R. 2005. Biofilms and antimicrobial resistance. Clin.Orthop. Relat Res.41-47. Pearson, A. D., M. Greenwood, T. D. Healing, D. Rollins, M. Shahamat, J. Donaldson, and R. R. Colwell. 1993. Colonization of broiler chickens by waterborne Campylobacter jejuni. Appl. Environ. Microbiol. 59:98 7-9 96. Pesci, E. C., D. L. Cottle, and C . L. Pickett. 1994. Genetic, enzymatic, and pathogenic studies of the iron superoxide dismutase of Campylobacter jejuni. Infect. Immun. 62:2687-2694. Phongsisay, V., V. N. Perera, and B. N. Fry. 2007. Expression of the htrB gene is essential for responsiveness of Salmonella typhimurium and Campylobacter jejuni to harsh environments. Microbiology 153:254-262. Pittman, M. S., K. T. Elvers, L. Lee, M. A. Jones, R. K. Poole, S . F. Park, and D. J. Kelly. 2007. Growth of Campylobacter jejuni on nitrate and nitrite: electron transport to NapA and NrfA via NrfH and distinct roles for NrfA and the globin Cgb in protection against nitrosative stress. Mol. Microbiol. 63575590. Purdy, D., S . Cawthraw, J. H. Dickinson, D. G. Newell, and S . F. Park. 1999. Generation of a superoxide dismutase (SOD)deficient mutant of Campylobacter coli: evidence for the significance of SOD in Campylobacter survival and colonization. Appl. Environ. Microbiol. 65:2540-2546. Purdy, D., and S . F. Park. 1994. Cloning, nucleotide sequence and characterization of a gene encoding superoxide dismutase from Campylobacter jejuni and Campylobacter coli. Microbiology 140(Pt. 5):1203-1208. Quinn, T., J. M. Bolla, J. M. Pages, and S . Fanning. 2007. Antibiotic-resistant Campylobacter: could efflux pump inhibitors control infection? J. Antimicrob. Chemother. 59:1230-1236. Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71:635-700. Raivio, T. L. 2005. Envelope stress responses and gram-negative bacterial pathogenesis. Mol. Microbiol. 56:1119-1128. Raphael, B. H., S. Pereira, G. A. Flom, Q. Zhang, J. M. Ketley, and M. E. Konkel. 2005. The Campylobacter jejuni response regulator, CbrR, modulates sodium deoxycholate resistance and chicken colonization. J. Bacteriol. 187:3662-3670. Reeser, R J., R. T. Medler, S. J. Billington, B. H. Jost, and L. A. Joens. 2007. Characterization of Campylobacter jejuni biofilms under defined growth conditions. Appl. Environ. Microbiol. 73: 1908-1913. Reezal, A., B. McNeil, and J. G. Anderson. 1998. Effect of lowosmolality nutrient media on growth and culturability of Campylobacter species. Appl. Environ. Microbiol. 64:4643-4649. Rickard, A. H., R. J. Palmer, Jr., D. S . Blehert, S. R. Campagna, M. F. Semmelhack, P. G. Egland, B. L. Bassler, and P. E. Kolenbrander. 2006. Autoinducer 2: a concentration-dependent signal for mutualistic bacterial biofilm growth. Mol. Microbiol. 60:1446-1456. Rollins, D. M., and R R Colwell. 1986. Viable but nonculturable stage of Campylobacter jejuni and its role in survival in the natural aquatic environment. Appl. Environ. Microbiol. 5 2 5 3 1538. Rollins, D. M., J. C. Coolbaugh, R. I. Walker, and E. Weiss. 1983. Biphasic culture system for rapid Campylobacter cultivation. Appl. Environ. Microbiol. 45 :284-28 9. Romling, U. 2005. Characterization of the rdar morphotype, a multicellular behaviour in Enterobacteriaceae. Cell. Mol. Life. Sci. 62:1234-1246.
590
SVENSSON ET AL.
Romling, U., and D. Amikam. 2006. Cyclic di-GMP as a second messenger. Curr. Opin. Microbiol. 9:218-228. Saha, S. K., and S. C. Sanyal. 1991. Better preservation of Campylobacter jejunilC. coli in a defined medium. lndian J. Med. Res. 93:26-28. Sampathkumar, B., S. Napper, C. D. Carrillo, P. Willson, E. Taboada, J. H. Nash, A. A. Potter, L. A. Babiuk, and B. J. Allan. 2006. Transcriptional and translational expression patterns associated with immobilized growth of Campylobacter jejuni. Microbiology 152567-577. Sanders, S . Q., D. H. Boothe, J. F. Frank, and J. W. Arnold. 2007. Culture and detection of Campylobacter jejuni within mixed microbial populations of biofilms on stainless steel. J. Food Prot. 70~1379-1385. Smibert, R. M. 1978. The genus Campylobacter. Annu. Rev. Microbiol. 32:673-709. Snelling, W. J., J. P. McKenna, D. M. Lecky, and J. S. Dooley. 2005. Survival of Campylobacter jejuni in waterborne protozoa. Appl. Environ. Microbiol. 715560-5571. Somers, E. B., J. L. Schoeni, and A. C. Wong. 1994. Effect of trisodium phosphate on biofilm and planktonic cells of Campylobacter jejuni, Escherichia coli 0157: H7, Listeria monocytogenes and Salmonella typhimurium. lnt. J. Food Microbiol. 22: 269-2676. Stark, R. M., G. J. Gerwig, R. S. Pitman, L. F. Potts, N. A. Williams, J. Greenman, I. P. Weinzweig, T. R. Hirst, and M. R. Millar. 1999. Biofilm formation by Helicobacter pylora. Lett. Appl. Microbiol. 28: 121-1 26. Stead, D., and S. F. Park. 2000. Roles of Fe superoxide dismutase and catalase in resistance of Campylobacter coli to freeze-thaw stress. Appl. Environ. Microbiol. 66:3 110-3 112. Stern, N. J. 1994. Mucosal competitive exclusion to diminish colonization of chickens by Campylobacter jejuni. Poult. Sci. 73: 402-407. Stintzi, A. 2003. Gene expression profile of Campylobacter jejuni in response to growth temperature variation. J. Bacteriol. 185: 2009-2016. Su, C. C., F. Shi, R. Gu, M. Li, G. McDermott, E. W. Yu, and Q. Zhang. 2007. Preliminary structural studies of the transcriptional regulator CmeR from Campylobacter jejuni. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 63:34-36. Talibart, R., M. Denis, A. Castillo, J. M. Cappelier, and G. Ermel. 2000. Survival and recovery of viable but noncultivable forms of Campylobacter in aqueous microcosm. lnt. J. Food Microbiol. 55:263-267. Tangwatcharin, P., S. Chanthachum, P. Khopaibool, J. R. Chambers, and M. W. Griffiths. 2007. Media for the aerobic resuscitation of Campylobacter jejuni. J. Food Prot. 70:1099-1109. Tangwatcharin, P., S. Chanthachum, P. Khopaibool, and M. W. Griffiths. 2006. Morphological and physiological responses of Campylobacter jejuni to stress. J. Food Prot. 69:2747-2753. Terzieva, S. I., and G. A. McFeters. 1991. Survival and injury of Escherichia coli, Campylobacter jejuni, and Yersinia enterocolitica in stream water. Can. J. Microbiol. 37:785-790. Thies, F. L., H. P. Hartung, and G. Giegerich. 1998. Cloning and expression of the Campylobacter jejuni Ion gene detected by RNA arbitrarily primed PCR. FEMS Microbiol. Lett. 165:329334. Thies, F. L., H. Karch, H. P. Hartung, and G. Giegerich. 1999a. Cloning and expression of the dnaK gene of Campylobacter je-
juni and antigenicity of heat shock protein 70. Infect. Immun. 67:1194-1200. Thies, F. L., H. Karch, H. P. Hartung, and G. Giegerich. 1999b. The ClpB protein from Campylobacter jejuni: molecular characterization of the encoding gene and antigenicity of the recombinant protein. Gene 230:61-67. Thies, F. L., A. Weishaupt, H. Karch, H. P. Hartung, and G. Giegerich. 1999c. Cloning, sequencing and molecular analysis of the Campylobacter jejuni groESL bicistronic operon. Microbiology 145(Pt. 1):89-98. Tholozan, J. L., J. M. Cappelier, J. P. Tissier, G. Delattre, and M. Federighi. 1999. Physiological characterization of viable-butnonculturable Campylobacter jejuni cells. Appl. Environ. Microbiol. 65: 1110-1 116. Thomas, C., D. J. Hill, and M. Mabey. 1999. Morphological changes of synchronized Campylobacter jejuni populations during growth in single phase liquid culture. Lett. Appl. Microbiol. 28~194-198. Trachoo, N., and J. F. Frank. 2002. Effectiveness of chemical sanitizers against Campylobacter jejuni-containing biofilms.J. Food Prot. 65 :1117-1 121. Trachoo, N., J. F. Frank, and N. J. Stern. 2002. Survival of Campylobacter jejuni in biofilms isolated from chicken houses. J. Food Prot. 65:lllO-1116. Ulett, G. C., J. Valle, C. Beloin, 0. Sherlock, J. M. Ghigo, and M. A. Schembri. 2007. Functional analysis of antigen 43 in uropathogenic Escherichia coli reveals a role in long-term persistence in the urinary tract. Infect. lmmun. 75:3233-3244. van de Giessen, A. W., C. J. Heuvelman, T. Abee, and W. C. Hazeleger. 1996. Experimental studies on the infectivity of nonculturable forms of Campylobacter spp. in chicks and mice. Epidemiol. Infect. 117:463+1670. van Vliet, A. H., M. L. Baillon, C. W. Penn, and J. M. Ketley. 1999. Campylobacter jejuni contains two fur homologs: characterization of iron-responsive regulation of peroxide stress defense genes by the PerR repressor. J. Bacteriol. 181:6371-6376. van Vliet, A. H., M. A. Baillon, C. W. Penn, and J. M. Ketley. 2001. The iron-induced ferredoxin FdxA of Campylobacter jejuni is involved in aerotolerance. FEMS Microbiol. Lett. 196: 189-193. Vandamme, P. 2000. Taxonomy of the family Campylobacteraceae, p. 3-26. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Withey, J. H., and D. I. Friedman. 2003. A salvage pathway for protein structures: tmRNA and trans-translation. Annu. Rev. Microbiol. 57:lOl-123. Wolfe, A. J., D. E. Chang, J. D. Walker, J. E. Seitz-Partridge,M. D. Vidaurri, C. F. Lange, B. M. Pruss, M. C. Henk, J. C. Larkin, and T. Conway. 2003. Evidence that acetyl phosphate functions as a global signal during biofilm development. Mol. Microbiol. 48~977-988. Zimmer, M., H. Barnhart, U. Idris, and M. D. Lee. 2003. Detection of Campylobacter jejuni strains in the water lines of a commercial broiler house and their relationship to the strains that colonized the chickens. Avian Dis. 47:lOl-107. Ziprin, R. L., R. E. Droleskey, M. E. Hume, and R. B. Harvey. 2003. Failure of viable nonculturable Campylobacter jejuni to colonize the cecum of newly hatched leghorn chicks. Avian Dis. 47:753-758. Ziprin, R. L., and R. B. Harvey. 2004. Inability of cecal microflora to promote reversion of viable nonculturable Campylobacter jejuni. Avian Dis. 48:647-650.
Campylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 33
Iron Metabolism, Transport, and Regulation ALAINSTINTZI, ARNOUD H. M.
VAN
VLIET,AND JULIAN M. KETLEY
organisms, and this is achieved by balancing metabolism, transport, and storage of iron (Schaible and Kaufmann, 2004; Wooldridge and Williams, 1993). Next to controlling its iron status, cells often couple expression of their oxidative stress defense systems to the iron status of the cell, thus ensuring optimal expression of its defense systems (Touati, 2000). Campylobacter species are microaerophilic bacteria that colonize the avian intestine and cecum (Ketley, 1997; van Vliet and Ketley, 2001) and are considered to be either commensal or opportunistic colonizers in poultry (Cogan et al., 2007). Fecal contamination of meat often occurs during slaughter, and human infection is usually acquired through the consumption of undercooked contaminated meat or cross-contaminated food products. The infectious dose can be as few as 500 to 800 organisms (Black et al., 1988), and the continuing increase in incidence of Campylobacter infection in industrialized countries is thought to be related to changes in eating habits rather than reflecting increased awareness or better diagnostic tools. Human infection with enteric campylobacters usually leads to diarrhea, which can range from inflammatory diarrhea to the milder clinical symptoms of watery, noninflammatory diarrhea. Complications after Campylobacter infection are uncommon, but such sequelae include Guillain-Barre syndrome, a neurological disorder (Hughes, 2004; Yuki and Koga, 2006). The avian and mammalian intestinal tracts are characterized by continuous changes in environmental conditions, with the mucosal surface comprising many microniches. The environmental pH and oxygen concentration influence the bioavailability of iron (Andrews et al., 2003), and iron availability may also fluctuate as a result of release of nutrients from the
The transition metal iron plays an essential role in the metabolism of almost all organisms. Iron is a versatile divalent cation that exists in both the ferrous (Fe2+)and ferric (Fe3+)states. The redox potential of the Fe2+-Fe3+pair spans from -500 mV to +300 mV, and this makes iron well suited for participating in electron transfer reactions. Iron can be a cofactor of enzymes, catalyzing basic functions like electron transport, redox reactions, energy metabolism, and DNA synthesis, and thus iron-containing proteins are mostly involved in basic cell metabolism (Andrews et al., 2003). In addition, the iron-porphyrin compound heme is an essential part of many enzymes involved in bacterial respiration, electron transport, and peroxide reduction, and iron-sulfur proteins participate in electron transport reactions, anaerobic respiration, amino acid metabolism, and energy metabolism (Andrews et al., 2003; Woolridge and Williams, 1993). The use of iron in combination with oxygen for electron transfer, however, poses the danger of development of reactive oxygen species (ROS) like superoxide anions (O,-), peroxides (RO,), and hydroxyl radicals ( O H ) through the Haber-Weiss reactions (Fe3++ O,-+ Fe2+ + 0, and Fe2+ + H,O, Fe3+ + OH- + O H ) (Storz and Imlay, 1999; Touati, 2000). These ROS are highly reactive and will damage lipids, proteins, and DNA by oxidation. Cells need to detoxify these compounds when formed, but also prevent the formation of ROS by limiting the amount of reactive iron in the cytoplasm. Thus, the cell needs to carefully control the intracellular concentration of iron because both iron limitation and iron overload delay growth and can cause cell death. The maintenance of iron homeostasis is considered to be of critical importance to all living
-
Alain Stintzi Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, 451 Smyth Rd., Ottawa, Ontario K1H 8M5, Canada. Arnoud H. M. van Vliet Institute of Food Research, Office E410, Colney Ln., Norwich NR4 7UA, United Kingdom. . Julian M. Ketley Department of Genetics, University of Leicester, University Rd., Leicester LE1 7RH, United Kingdom.
591
592
STINTZI ET AL.
damaged epithelium and diffusion of iron from food. In addition, the intestinal tract contains many bacterial species, and thus Campylobacter will have to compete for iron and other nutrients. This implies that Campylobacter in the avian and mammalian enteric environment may have to cope with both conditions of iron restriction and iron overload, and thus it may need systems to cope with the associated stresses. The rapid development of technologies like transcriptomics, proteomics, and genome sequencing has enabled the detailed analysis of the iron metabolism of Campylobacter at the genomic level (Holmes et al., 2005; Palyada et al., 2004). Similar advances have been made in biochemical and molecular analyses (Naikare et al., 2006; Ridley et al., 2006; van Vliet et al., 1998, 2002). In this chapter, we will discuss the wealth of knowledge about iron metabolism of Campylobacter by discussing mechanisms of iron transport, iron storage, and iron-responsive regulation of genes involved in iron metabolism. Furthermore, we will discuss the role of iron in the function and regulation of oxidative stress defense, and the recent advances in transcriptomic analyses of iron metabolism in Campylobacter. Most of the data discussed in this chapter have been obtained by using c. jejuni, but we will also discuss data obtained for C. coli; it is thought that the mechanisms involved in iron metabolism are essentially similar in both species.
AVAILABILITY OF IRON IN THE INTESTINE The availability of free iron inside mammalian and avian hosts is extremely limited as a result of the toxicity of iron in combination with oxygen. Most of the iron is sequestered within cells, where it may be complexed as a cofactor of various proteins or stored in the iron-storage protein ferritin. Alternatively, extracellular iron is complexed with the iron-binding glycoproteins transferrin (in serum) and the related protein lactoferrin (in mucosal secretions) (Ward and Conneely, 2004). Both transferrin and lactoferrin are usually only partially iron saturated, and thus they compete for free iron with bacteria (Abdallah and El Hage Chahine, 2000; Schaible and Kaufmann, 2004). The resulting concentration of free iron in the mammal is thought to be around M, which is far below the levels required for growth of bacteria M). This iron restriction of tissues is thought to represent a nonspecific host defense system against pathogenic bacteria (Schaible and Kaufmann, 2004). The iron status of the intestinal mucosa is incompletely understood, both in chickens and in humans.
Iron can be present in the ferrous (2+) or ferric (3 +) state; this depends on the pH and redox conditions of the environment whereby soluble ferrous iron can be rapidly oxidized to the insoluble ferric form. Potential sources of iron in the intestine are mucosal secretions and cells, food components, and siderophores from the intestinal microflora (Wooldridge and van Vliet, 2005). The mucous layer contains iron-bound lactoferrin, which will sequestrate the majority of the available ferric iron, thus making it unavailable to most bacteria (Ward and Conneely, 2004). Furthermore, iron may be released from intestinal cells and possibly from blood cells during inflammatory responses, and these may represent sources of heme and ferritin iron. Heme is an iron porphyrin-based prosthetic group found in many metalloproteins, and food contains iron in this heme form (including the hemoproteins myoglobin and hemoglobin) and in nonheme iron, which may well have been reduced to the ferrous form during gastric passage. Stress hormones like the catecholamine norepinephrine may act in a siderophore-like manner, facilitating uptake of iron by Campylobacter (Cogan et al., 2007; Freestone et al., 2003). Furthermore, bile acids and ascorbic acid are also capable of complexing ferrous iron and may stabilize ferrous iron (Conrad and Schade, 1968). Finally, the intestinal microflora produce siderophores that compete with lactoferrin and bind ferric iron with high affinity (Baig et al., 1986; Braun and Braun, 2002). In summary, within the context of iron, successful colonization of the intestine requires that Campylobacter cope with the iron-restricted conditions, compete with the intestinal flora for available iron, and overcome iron restriction mediated by the host’s innate defense mechanisms. IRON ACQUISITION In the iron-restricted environment that is the intestine and perhaps also in biofilms (see below), campylobacters have several potential sources for the iron; these include ferric iron bound to glycoproteins and siderophores, ferric iron in heme, and ferrous iron. Campylobacters possess a range of uptake systems that enable the acquisition of iron from these various sources. Utilization of Ferrous Iron Ferrous iron is utilized by many bacteria, and in Escherichia coli, the high-affinity ferrous transport system expressed under anaerobic conditions involves two proteins, FeoA and FeoB, and a probable transcriptional regulator, FeoC (Cartron et al., 2006;
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
Kammler et al., 1993). A general scheme of uptake of ferrous iron across the cell envelope is diagramed in Fig. 1. Although the role of FeoA is not yet clear, FeoB has been shown to be an inner membrane protein containing a cytoplasmic G protein domain and membrane-associated “gate” domain. Mutation of the FeoB homolog in Salmonella enterica serovar Typhimurium did not attenuate infection in a mouse model, but did affect intestinal colonization when competition with wild-type S. enterica serovar Typhimurium was involved (Tsolis et al., 1996). In contrast, intestinal colonization in a mouse model was disrupted with an E. coli K-12 feoB mutant (Stojiljkovic et al., 1993). In Helicobacter pylori, FeoB is required for colonization of the gastric mucosa in a mouse model (Velayudhan et al., 2000). This restricted set of examples would suggest that Feomediated ferrous iron uptake has differing significance to gastrointestinal colonization by different bacteria. Clearly, the context of other, primarily ferric, iron uptake systems is likely to be important. The genome of C. jejuni contains a homolog of feoB (Cj1398), and upstream there is a gene (Cj1397) with weak homology to feoA, which is in the same operon (Naikare et al., 2006); no feoC gene appears to be present. Interestingly, in several of the sequenced strains (Chaudhuri and Pallen, 2006), in-
593
eluding 81-176 and RM1221, feoB contains frameshift mutations or premature translation stops, whereas in all but one strain, Cj1397 is unaffected by such mutations. Despite prior evidence to the contrary (Raphael and Joens, 2003), C. jejuni is capable of utilizing ferrous iron via the Feo system (Naikare et al., 2006), with 81-176 producing a functional FeoB (Fig. 2). In contrast to the other iron uptake systems, feoB does not appear to be Fur dependent or even iron regulated (Holmes et al., 2005; Palyada et al., 2004); however, feoB does appear to be PerR regulated (see below). Importantly, the Feo system in C. jejuni plays a role in promoting intracellular survival and intestinal colonization of rabbits, chicks, and pigs (Naikare et al., 2006). In H. pylori, the CorA magnesium transporter provides an additional ferrous iron transport system (Hantke, 1997; Pfeiffer et al., 2002; Wainwright et al., 2OO1), although the biological relevance of this is not clear because in extracellular compartments, magnesium concentrations are far in excess of iron, abolishing CorA-mediated iron transport (Smith and Maguire, 1998). In contrast, within host cells, magnesium concentrations may be low. It is possible that C. jejuni requires the CorA homolog (Cj0726) for intracellular survival during infection. However, in one limited study, no evidence for CorA-mediated
@ Outer membrane Periplasm Cytoplasmic membrane
Figure 1. Iron transport across the gram-negative cell envelope. (Left) TonB- and ABC-dependent transport pathway. Transport through the high-affinity outer membrane receptor is energized by TonB-mediated transduction of energy from the proton motive force with participation from ExbB and ExbD. Passage across the cytoplasmic membrane involves proteins of an ABC transporter system whose ATPase activity energizes the process. (Right) Ferrous iron uptake pathway. Ferrous iron crosses the outer membrane via a nonspecific porin or after reduction of ferric iron. The inner membrane FeoB containing a G protein domain and cytoplasmic FeoA transport the ferrous iron into the cytoplasm.
594
STINTZI ET AL.
I
I
I
I
Figure 2. Iron transport systems of C. jejuni as determined by comparative genome analysis and experimental data (see text for details). Substrates are shown where known, and systems present in all strains are highlighted.
ferrous iron uptake was found (Raphael and Joens, 2003). In addition, corA is not iron regulated in campylobacters (Holmes et al., 2005; Palyada et al., 2004). Siderophore-Mediated Iron Scavenging High-affinity, low-molecular-weight iron chelators called siderophores are produced by many bacteria in order to appropriate ferric iron from the surrounding environment (Braun et al., 1998). Secreted siderophores, most being either catechols or hydroxamates, bind ferric iron and are transported into the cell by energy-dependent high-affinity systems. In gram-negative bacteria, the outer membrane receptor protein forms a ligand-gated pore (Ma et al., 2007) with the energy for transport across the outer membrane provided by the cytoplasmic membrane proton motive force transduced via the TonB-ExbBExbD protein complex, which spans the periplasmic space from cytoplasmic to outer membrane (Braun and Braun, 2002; Hannavy et al., 1990; Postle and Larsen, 2007). Once across the outer membrane, the ferric siderophore complex binds to a periplasmicbinding protein of a cognate ABC transport system, which then mediates the transfer of the complex across the inner membrane. Iron-uptake-associated systems constitute a distinct subset of the ubiquitous family of ABC transport systems in gram-negative bacteria (Andrews et al., 2003; Linton and Higgins, 1998). A general scheme of uptake of ferric iron across the cell envelope is shown in Fig. 1.
Analysis of genome sequences of campylobacters has not revealed the presence of genes associated with siderophore biosynthesis (Chaudhuri and Pallen, 2006; Parkhill et al., 2000). Although there was one early report of the production of uncharacterized siderophore activity by some C. jejuni strains (Field et al., 1986), the lack of an ability to synthesize siderophores is supported by other studies (Baig et al., 1986; Pickett et al., 1992). Siderophore piracy, or the ability to utilize exogenous siderophores produced by other microorganisms, was evident in campylobacters. The catechol siderophore enterobactin and the hydroxamate siderophore ferrichrome are utilized by C. jejuni (Field et al., 1986), although the hydroxamate siderophores aerobactin, desferrioxamine B, and rhodotorulic acid were not found to act as iron sources in cross-feeding assays (Field et al., 1986). Utilization of ferric enterobactin Enterobactin, which is produced by members of the mammalian and avian intestinal microbial flora, has the potential of being a significant source of iron to C. jejuni. An ABC transport system involved in an enterobactin uptake system has been described in both C. jejuni and C. coli (Park and Richardson, 1995; Richardson and Park, 1995) (Fig. 2). The ceuB and ceuC genes both encode integral membrane proteins, which presumably form the permease complex in the cytoplasmic membrane; ceuD encodes the ATP-binding protein component; and ceuE encodes a lipoprotein that functions as an atypical
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
periplasmic-binding protein. Mutation of individual ceuBCDE genes indicated that they are all required for efficient uptake of enterobactin in C. coli strain used (Richardson and Park, 1995). The outer membrane receptor specific for ferric enterobactin uptake has been shown to be CfrA (Palyada et al., 2004) (Fig. 2), encoded by a gene previously shown to be iron regulated (Guerry et al., 1997), and present only in a subset of C. jejuni strains. Inspection of available genome sequences for C. jejuni and C. coli reveals that cfrA is present in all strains but 81116, 81-176, and HB93-13 (Chaudhuri and Pallen, 2006; Hofreuter et al., 2006; Pearson et al., 2007). In contrast, the cognate ceu ABC transport system is present in all genomes, but in C. jejuni subspp. doylei, only a remnant of the system is left. Adjacent to the c f ~ A locus in NCTC 11168 is one of the three tonB genes (tonB.3) found in this strain (Parkhill et al., 2000). Mutation of tonB3 in NCTC 11168 abrogates ferric enterobactin utilization (A. Stintzi, unpublished data). Similarly, mutation of the equivalent gene in the C. coli strain VC167 T 1 resulted in the inability to utilize ferric enterobactin, as well as hemin and ferrichrome (Guerry et al., 1997). As stated above, it is likely that ferric enterobactin is an important source of iron for campylobacters in the intestine. Indeed, mutation of cfrA or ceuE in C. jejuni strain NCTC 11168 dramatically reduced the ability to colonize the avian intestine (Palyada et al., 2004), suggesting that ferric enterobactin is required for colonization by campylobacters. Earlier studies have indicated that enterobactin was not an important source of iron in the avian gut by C. coli because a ceuE mutant was not impaired in chick colonization experiments (Cawthraw et al., 1996). Moreover, C. jejuni strain 81-176, an efficient colonizer of the chicken intestine, does not possess CfrA (Hofreuter et al., 2006) and therefore would not be expected to utilize ferric enterobactin. The importance of enterobactin, therefore, is likely to be more complex and depend on genomic context. Utilization of ferrichrome Ferrichrome is a hydroxamate predominantly produced by certain soil fungi; enteric microorganisms are not known to produce this siderophore. Campylobacters are able to acquire iron from ferrichrome (Field et al., 1986), but given the ecology of this siderophore (or related siderophores), the importance to campylobacters is likely either to be for survival outside of the intestine or due to the existence of members of this hydroxamate group in food. Genes with homology to fiuABD, which encode the outer membrane receptor and part of the ABC trans-
595
port system of the E. coli ferrichrome uptake system, have been identified in a set of C. jejuni strains (Galindo et al., 2001). The role of these sequences in campylobacters, however, remains unclear because the described sequences are remarkably GC rich, a gene encoding an ATP-binding component of the system was not present, and a functional role for these genes in ferrichrome transport was not demonstrated (Galindo et al., 2001); none of the Campylobacter full genome sequences contains homologs of these genes (Chaudhuri and Pallen, 2006). The p19 system The p19 system consists of an ABC transport system (cjlG59-cjl663) with an associated gene (cjl658) that encodes a putative membrane protein. The gene (cjl659) that encodes the 19-kDa periplasmic protein, p19, previously determined to be nonimmunoreactive (Janvier et al., 1998), was found to be iron regulated (van Vliet et al., 1998). Analysis of available C. jejuni genome sequences suggests that the system is conserved (Chaudhuri and Pallen, 2006). Both Cj1658 and Cj1659 have likely homologs in an iron uptake pathogenicity island of Yersinia pestis (Carniel, 2001) and a homolog of Cj1659 named ChpA has been described in the magnetotactic bacterium MV-1 (Dubbels et al., 2004). ChpA is essential for the formation of Fe,O,-containing magnetosomes and is suggested to function as part of a copper-dependent high-affinity iron uptake system (Dubbels et al., 2004). The p19 system in C. jejuni, however, has been found to be involved in the utilization of iron associated with rhodotorulic acid (A. Stintzi and J. M. Ketley, unpublished data), a fungal hydroxamate siderophore; it is not yet clear which outer membrane receptor is involved (Fig. 2). Utilization of Heme Compounds Given the insolubility and toxicity of free Fe3+, intracellular iron in the host is often associated with proteins in the form of hemin (Genco and Dixon, 2001). A number of bacteria are known to utilize free heme, hemoglobin, or heme-binding protein complexes such as hemopexin or albumin (Cope et al., 1995; Henderson and Payne, 1994; Hornung et al., 1996; Mills and Payne, 1997; Ochsner et al., 2000; Stojiljkovic et al., 1996; Torres and Payne, 1997). For bacteria to acquire iron from hemoproteins, heme must first be removed from the protein complex. This process is not siderophore mediated, but it may involve specific degradative enzymes, and in some systems, a heme-sequestering protein, termed a hemophore, delivers heme to the cell surface receptor
596
STINTZI ET AL.
(Ghigo et al., 1997; Lewis et al., 1998). Examples of receptors for heme or major circulating hemoproteins include the hemoglobin/ hemoglobin-haptoglobin receptor complex HpuAB from Neisseria meningitidis (Lewis et al., 1998), the HasR heme receptor and cognate hemophore HasA of Serratia marcescens (Letoffe et al., 2004), and the hemopexin receptor HxuA of Haemophilus influenzae (Cope et al., 1998). In gram-negative bacteria, heme transport across the cell envelope is TonB dependent and requires an ABC transport system (Genco and Dixon, 2001). In eukaryotes, iron is liberated from heme by an heme oxygenase (HO), and the subsequent release of iron from ferric-biliverdin requires a biliverdin reductase (Liu and Ortiz de Montellano, 2000). In bacteria, different HOs have been described, including H e m 0 from Neisseria spp. (Zhu et al., 2000), ChuS from E. coli 0157:H7 (Suits et al., 2005), and HutZ from Vibrio cholerae (Wyckoff et al., 2004). C. jejuni has been shown to utilize heme, heme bound to the acute-phase protein hemopexin, hemoglobin, and hemoglobin complexed with haptoglobin (Pickett et al., 1992). A 70-kDa outer membrane protein with homology to TonB-dependent receptors was shown to be associated with C. jejuni utilization of heme compounds (Pickett et al., 1992; van Vliet et al., 1998). The gene chuA (Cj1614), which encodes this 70-kDa protein, is adjacent to the chuBCD genes (Cj1615-Cj1617) that are predicted to encode components of an ABC transport system (Fig. 2). Mutation of chuA resulted in the inability to utilize hemin or hemoglobin when present as a sole iron source, whereas mutation of genes in the adjacent ABC transport system only partially attenuated growth under these conditions (kdley et al., 2006), suggesting some redundancy in ABC transport system specificity in NCTC 11168. Transcribed divergently to chuA is Cj1613. Mutation of this gene also leads to an inability to grow in the presence of hemin or hemoglobin as a sole source of iron (Ridley et al., 2006). Further analysis of Cj1613 showed it to be capable of oxidatively degrading heme in the presence of the electron donor ascorbic acid, indicating that Cj1613c (renamed ChuZ) functions as a heme oxygenase (hdley et al., 2006). Genotyping experiments (Ridley et al., 2006), supported by analysis of available genome sequences (Chaudhuri and Pallen, 2006) indicate that the Chu system is highly conserved among campylobacters. One other system has been associated with the utilization of heme. Cj0178 has significant homology to both the HasA and HasR proteins of Pseudomonas aeruginosa, which constitutes a heme uptake system (Ochsner et al., 2000), and Cj0177 has homology to a P. aeruginosa gene of unknown function, phuW,
which encodes a lipoprotein and is part of the TonBdependent heme-uptake locus (Ochsner et al., 2000). Structural analysis of purified Cj0177 has indicated that Cj0177 binds heme and that this binding is important in dimerization of the protein (Chan et al., 2006). In contrast to Chu, no direct functional link has been made for the CjO177/CjO178 system with the utilization of heme or hemoglobin, and strains like 81-176 can utilize heme in the absence of the CjO177/CjO178 system (C. E. Miller and J. M. Ketley, unpublished data). Therefore, the binding of heme in Cj0177 may be functionally important in the utilization of iron from another source. Hemolysin-mediated erythrocyte lysis is a strategy used by some bacteria to liberate heme for subsequent use as an iron source. Hemolytic activity has been observed in C. jejuni, and two genes have been associated with this activity. The ABC transport system associated with enterobactin transport across the inner membrane includes the periplasmic binding protein CeuE (see above). When placed in an E. coli background, ceuE conferred a hemolytic phenotype (Park and Richardson, 1995), but it is not known whether CeuE can function as a hemolysin in C. jejuni. Although the periplasmic location of CeuE argues against a role as a hemolysin, it is possible that as a lipoprotein (see above) (Richardson and Park, 1995), it has a second role as a membrane-bound hemolysin. An outer membrane phospholipase A (PldA) has been shown to be a hemolytic determinant in at least one strain of C. coli (Grant et al., 1997). A pldA homologue is present in the genome of C. jejuni NCTC 11168 (Cjl351) (Parkhill et al., 2000), where it is upstream of the ceu operon and probably transcribed separately, and can also be found in the genome of other campylobacters. The H. pylori PldA has been shown to be involved in the variation of membrane lipid composition, the phenotypic consequences of which include changes in resistance to acid stress (Tannas et al., 2001) and an effect on gastric colonization (Dorrell et al., 1999). The inflammatory response that follows tissue and cellular invasion by campylobacters will also increase the availability of heme compounds (chapter 17).The resulting release of erythrocytes and other cells into both the intestinal lumen and damaged intestinal tissues provides a source of heme compounds, especially if hemolysin production is evident. Utilization of Host Iron-Binding Glycoproteins Members of the transferrin iron-binding protein family are glycosylated, bilobed, monomeric proteins of approximately 80 kDa (Gray-Owen and Schryvers, 1996). The N-terminal and C-terminal lobes are
CHAPTER 33
-
IRON METABOLISM. TRANSPORT. AND REGULATION
structurally similar, and the interdomain cleft contains the iron-binding site. Each protein binds reversibly two atoms of ferric iron in conjunction with two bicarbonate anions (Abdallah and El Hage Chahine, 2000; Anderson et al., 1987). Transferrin is primarily found in serum (Gray-Owen and Schryvers, 1996) and lactoferrin in exocrine secretions, where it has bacteriostatic and bacteriocidal properties (Anderson et al., 1987; Singh et al., 2002). Avian ovotransferrin, which is present in blood serum and egg white, demonstrates 5 1% homology with human transferrin and 49% with human lactoferrin (Giansanti et al., 2002). The use of lactoferrin- or transferrin-bound iron has been described for pathogenic Neisseria (GrayOwen and Schryvers, 1996) and a range of other pathogens, including Pasteurella spp. (Gray-Owen and Schryvers, 1996), H. influenzae (Herrington and Sparling, 1985; Pidcock et al., 1988), P. aeruginosa (Sriyosachati and Cox, 1986), Bordetella spp. (Gorringe et al., 1990; Menozzi et al., 1991; Redhead et al., 1987), H. pylori (Dhaenens et al., 1997), Staphylococcus aureus (Park et al., 2005), and Candida albicans (Knight et al., 2005). Utilization of lactoferrin and transferrin by H. pylori remains controversial (Dhaenens et al., 1999; Husson et al., 1993; Velayudhan et al., 2000) and may be a strain-dependent phenomenon. Specific cell surface receptors that bind transferrin or lactoferrin in Neisseria spp. consist of two proteins: transferrin or lactoferrin-binding protein A (TbpA/LbpA) and B (TbpB/LbpB). Protein A demonstrates homology to siderophore receptor outer membrane proteins (Biswas and Sparling, 1995; Cornelissen et al., 1992), and protein B is a bilobed lipoprotein (Anderson et al., 1994; Pettersson et al., 1998). An interaction between transferrin and TbpAB induces conformational change, allowing the uptake of iron across the outer membrane via the TbpA pore (Gomez et al., 1998). Iron is then bound in the periplasm by FbpA (Chen et al., 1993) and delivered to the inner membrane ABC transport system FbpBC, which transports it into the cytoplasm (Adhikari, 1996). C. jejuni was previously considered incapable of utilizing iron bound to members of the transferrin protein family when using a plate-based assay (Pickett et al., 1992). Recent studies that utilize liquid medium have shown that holo-lactoferrin, holotransferrin, or holo-ovotransferrin can be used as a source of iron by C. jejuni (Miller et al., 2008). In the genome of C. jejuni NCTC 11168, there are no close homologs of genes that would encode a transferrin-binding lipoprotein, but one gene (Cj0178) shares some identity with TonB-dependent outer membrane components of transferrin and lactoferrin-binding proteins. Direct evidence for a
597
role is evidenced by the observation that mutation of Cj0178 markedly affects the ability to utilize lactoferrin or transferrin as a sole iron source (Miller et al., 2008). The putative ferric-iron uptake outer membrane receptor protein Cj0178 is expressed from a gene downstream of Cj0177, which has been linked with heme uptake (Chan et al., 2006) (see below). Mutation of Cj0178 results in reduced colonization ability in a chick-cecum model and rabbit ileal-loop model, demonstrating the necessity for, and therefore the importance of, Cj0178 during colonization and infection in vivo (Palyada et al., 2004; Stintzi et al., 2005). Upstream of Cj0177 and Cj0178, and transcribed in the opposite direction, there is another set of genes (CjO173-CjO176) that encodes an ABC transport system tentatively named cfbpA-C (Tom-Yew et al., 2005; van Vliet et al., 2002), the function of which has been linked with Cj0178 (Miller et al., 2008) (Fig. 2). Interestingly, like the ferric enterobactin uptake system (CfrA-Ceu), the CjO173-CjO178 system is not encoded by the genome of 81116 and 81-176 (Chaudhuri and Pallen, 2006; Pearson et al., 2007), yet has a role in intestinal colonization (Palyada et al., 2004; Stintzi et al., 2005). TonB and Associated Genes in C . jejuni A C. coli tonB homolog has been demonstrated to be essential for utilization of enterobactin, ferrichrome, and heme-complexed iron (Guerry et al., 1997), suggesting that one TonB has specificity for iron acquisition outer membrane receptors. Interestingly, inspection of the NCTC 11168 genome sequence revealed the presence of three tonB homologs, CjOl81 (tonBI), Cj1630 (tonB2), and the presumed ortholog of the gene identified in C. coli, tonB3 (Cj0753). The tonBl and tonB2 genes are both apparently transcriptionally coupled to a pair of exbBlexbD genes (Cj0179/CjO180 and Cj1628/ Cj1629, respectively). The tonB3 gene is not adjacent to any ex6 genes, but a third pair of exbBlexbD genes (CjOlO9/CjOllO) is present in the genome adjacent to an ATP synthase operon. All three tonB genes are iron regulated (Holmes et al., 2005; Palyada et al., 2004). The presence of multiple tonB genes with differing specificity for iron uptake systems is not unusual; for example, in V. cholerae, different tonB gene products mediate transport of different iron siderophores (Occhino et al., 1998; Seliger et al., 2001). It is not known whether multiple tonBlexbBlexbD gene sets are present in the C. coli strain in which the tonB3 ortholog was shown to be required for uptake of hemin and siderophores. However, inspection
598
STINTZI ET AL.
of available genome sequences shows that not all campylobacters contain three TonB systems as, for example, strains 81-176 and 81116 do not contain the tonBZ and tonB.3 systems (Chaudhuri and Pallen, 2006; Hofreuter et al., 2006; Pearson et al., 2007). Studies with individual tonB mutants have shown that TonB3 is essential for the utilization of ferric enterobactin and that the ChuA outer membrane receptor can be energized by either TonB3 or TonBl (A. Stintzi, unpublished data). Thus, there is some specificity in TonB function (Fig. 2), but at present it remains unclear how heme transport is catalyzed in strains such as 81116 and 81-176.
Other C. jejuni Genes with a Potential Role in Iron Uptake In strain NCTC 11168 (Parkhill et al., 2000), the gene Cj0444 appears to be a pseudogene that encodes a TonB-dependent outer membrane ligand gated pore likely to be associated with iron acquisition. Although this seems to be the case for strains RM1221, 84-25, and CF93-6, the gene is likely to encode a full-length protein in 81116 and 81-176 (Chaudhuri and Pallen, 2006; Pearson et al., 2007). The role of this potential receptor is now under investigation. Iron Storage Iron storage forms an important part of iron homeostasis (Andrews, 1998). In iron-replete conditions, maximizing iron stores provides an advantage when iron availability is limited. In addition, storage of excess iron in a nontoxic form reduces the potential impact of iron stress (see below). Bacteria produce two proteins, ferritin and bacterioferritin, that are capable of iron storage. These distantly related proteins form hollow spherical multimers that can contain more than 2,000 iron atoms in the form of a ferric-hydroxyphosphate core (Andrews, 1998). Ferritin is a nonheme protein with an ortholog found in eukaryotic cells, whereas the bacterioferritins contain heme and are restricted to prokaryotes (Andrews, 1998). In H. pylori, the ferritin protein Pfr was shown to be important for iron storage (Bereswill et al., 1998; Waidner et al., 2002), and a pfr mutant is more susceptible to iron toxicity in a pH-dependent manner and was found to be unable to colonize the gastric mucosa in a Mongolian gerbil model (Waidner et al., 2002). The genome of C. jejuni contains both a ferritin gene, cft (Cj0612), and a putative bacterioferritin gene, dps (Cj1534). The Cft protein has been shown to contain iron, and the growth of a cft mu-
tant was inhibited under low iron conditions, indicating a role as a functional iron storage protein (Wai et al., 1996). The cft mutant was also more sensitive to oxidative stress than its parent (Wai et al., 1996). In a recent assessment of the response of campylobacters to bile, ferritin (but not bacterioferritin; see below) was found to be produced at higher levels and therefore implicated in tolerance to bile (Fox et al., 2007). C. jejuni bacterioferritin is similar to the neutroPhil-activating protein of H. pylori. Neutrophilactivating protein has been shown to bind iron (Tonello et al., 1999), but it has also been reported to be a DNA-binding protein with similarity to the Dps family of DNA-binding proteins (Almiron et al., 1992). Neutrophil-activating protein expression is repressed under conditions of iron starvation, at least partly in a Fur-dependent manner (see below) (Cooksley et al., 2003). Mutation of the H. pylori dps gene results in increased sensitivity to oxygen (Tonello et al., 1999). Dps protein from C. jejuni strain 81-176 was seen to be able to bind up to 40 atoms of iron per monomer but did not appear to bind DNA, and mutation of the dps gene increases sensitivity to H202 in iron-replete, but not iron-limited, conditions (Ishikawa et al., 2003). Therefore, in C. jejuni, bacterioferritin/Dps appears to play a role in protection against oxidative stress when intracellular iron levels are high.
IRON AND OXIDATIVE STRESS As indicated above, iron and oxidative stress are intimately linked (Fig. 3). The intracellular environment is reducing and favors the formation and stability of Fe2+, with all the associated risks, and thus the intracellular concentration of iron is controlled at the level of uptake, utilization, and storage (see “Iron
OH.
ROH’
Figure 3. Overview of pathways involved in generation and inactivation of different forms of reactive oxygen species in C. jejuni, including superoxides (O,-), hydrogen peroxide (H,O,), and alkyl peroxides (RHO,).
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
Regulatory Proteins” below) (van Vliet et al., 2002). Furthermore, reactive oxygen species like superoxide and peroxides are produced by normal metabolic processes and need to be removed before they can react with iron (Storz and Imlay, 1999). Cumpylobucter is exposed to atmospheric levels of oxygen during transmission and to ROS (reactive oxygen species) when the pathogen is in contact with the human immune system-for example, professional phagocytes (Day et al., 2000). Three types of oxidative stress are connected with iron metabolism, and these stresses, and the Cumpylobucter defense against them, will be discussed next. Superoxide Stress Defense The main systems defending cells against superoxides operate via enzymatic inactivation of those superoxides. This is mediated by superoxide dismutases (SODS), which catalyze the dismutation of superoxides into hydrogen peroxide and oxygen (20,- + 2H+ H20, + 0,) (Fig. 3), with the hydrogen peroxide produced by SOD being subsequently inactivated by catalase (Touati, 2000; van Vliet et al., 2002). There are several classes of SOD enzymes in nature, and these are generally classified according to their metal cofactors, which can be copper-zinc, nickel, manganese, or iron (Lynch and Kuramitsu, 2000). Cumpylobucter expresses a single SOD, which is cofactored by iron (SodB) (Pesci et al., 1994; Purdy and Park, 1994). The SodB enzyme is a cytoplasmic protein and has been characterized in both C. jejuni and C. coli. Mutants defective in the sodB gene did not show reduced growth under laboratory conditions, but they displayed reduced intracellular survival in epithelial cells and were attenuated in chicken colonization experiments (Pesci et al., 1994; Purdy et al., 1999). A C. coli sodB mutant also displayed severely reduced survival in milk and on chicken skin, and reduced aerobic survival during stationary phase and freeze-thaw stress (Pesci et al., 1994; Purdy et al., 1999; Stead and Park, 2000). The regulation of SodB expression has not been studied in detail, although one transcriptomic study reported iron-responsive repression of sodB transcription (Palyada et al., 2004). However, in a separate study combining transcriptomics and proteomics, SodB expression was not seen as significantly altered by iron (Holmes et al., 2005).
-
Peroxide Stress Defense Peroxides can be formed during normal metabolism or by superoxide dismutation by SOD (see above) (Fig. 3 ) ; they must be inactivated to prevent
599
formation of hydroxyl radicals (Chelikani et al., 2004). C. jejuni expresses several peroxidases, of which two have been more than superficially characterized. Catalase (KatA) is a heme-cofactored enzyme that converts hydrogen peroxide to oxygen and water (2H20, 2 H 2 0 + OJ. Cumpylobucter expresses a single catalase enzyme, and C. jejuni cells that lack this catalase were only isolated from macrophages when nitric oxide synthetase and NADPH oxidase were inhibited (Day et al., 2000). In contrast, catalase is not required for survival in epithelial cells or resistance to freeze-thaw stress (Day et al., 2000; Grant and Park, 1995; Stead and Park, 2000). The role of the catalase in chicken colonization differs between C. coli and C. jejuni. Although it is not required for C. coli chicken colonization and survival on chicken skin (Purdy et al., 1999), a kutA mutant of C. jejuni is significantly affected in its ability to colonize the chick cecum (A. Stintzi, unpublished data). Expression of catalase in C. jejuni is controlled by the PerR protein (van Vliet et al., 1999), is repressed by iron, and is induced by exposure to H20, (Park, 1999; van Vliet et al., 1999). The second C. jejuni peroxidase that has been characterized is the alkyl hydroperoxide reductase (AhpC), also known as thiol-specific antioxidant (Tsa or TsaA). The AhpC enzyme converts alkyl hydroperoxides to the corresponding alcohols (ROOH + NADH + H+ ROH + NAD’ + H,O) (Fig. 3). In E. coli and many other bacteria, Ahp consists of the AhpC catalytic subunit and the AhpF flavoprotein, which recycles the AhpC protein (Poole et al., 2000). C. jejuni only expresses an AhpC homolog, but does not contain an uhpF gene. It is predicted that the function of AhpF is probably mediated by the thioredoxin reductase (TrxB) with thioredoxin, similar to H. pylori, and this is supported by the ironresponsive regulation of the trxB gene (Holmes et al., 2005; Palyada et al., 2004). Inactivation of the uhpC gene in C. jejuni resulted in increased sensitivity to cumene hydroperoxide, but not to H202, suggesting that the two peroxidases have complementary functions (Baillon et al., 1999). Conversely, overexpression of AhpC in a C. jejuni perR mutant made the cells hyperresistant to cumene hydroperoxide (van Vliet et al., 1999). Inactivation of uhpC resulted in significantly reduced aerotolerance in the mutant strain when compared with the wild-type strain (Baillon et al., 1999), indicating that AhpC may contribute to environmental survival of C. jejuni. Finally, the C. jejuni genome contains several other peroxidases that may contribute to oxidative stress survival of Cumpylobucter. These include the thioredoxin-linked thiol peroxidase Tpx, which belongs to a family of recently isolated bacterial antioxidant enzymes
-
-
600
STINT21 ET AL.
(Wang et al., 2006). Tpx expression is similarly regulated to AhpC and KatA (Holmes et al., 2005; Palyada et al., 2004), and in the related pathogen H. pylori, this enzyme is involved in oxidative stress resistance (Wang et al., 2006). Its role in C. jejuni awaits further analysis. Other Oxidative Stress Defense Mechanisms Several other Campylobacter systems contribute directly or indirectly to oxidative stress defense in Campylobacter. One example of this is iron storage by the C. jejuni ferritin Cft (Wai et al., 1995, 1996), which stores iron in an inactive, reusable form inside the ferritin cavity, thus making it unavailable for generation of ROS. Absence of Cft leads to increased sensitivity against both superoxide and peroxide stress inducers (Wai et al., 1996), a role that is similar to that described for the Pfr ferritin in H. pylori (Waidner et al., 2002). C. jejuni also expresses a bacterioferritin/Dps ortholog, which can bind iron and contributes to resistance to hydrogen peroxide, probably via a similar mechanism as ferritin (Ishikawa et al., 2003). The ferredoxin FdxA does not contribute to resistance to specific peroxide stress, but its absence significantly reduces aerotolerance of C. jejuni (van Vliet et al., 2001). Finally, Campylobacter also expresses mechanisms defending the cell against nitrosative stress caused by nitric oxide (NO). NO plays an important role in biological systems as a signaling molecule and a defense molecule (Poole and Hughes, 2000). NO is generated by the immune system and can react with superoxide to form the potentially toxic compound peroxynitrite and toxic radicals. C. jejuni senses NO stress via the NssR regulatory system (Elvers et al., 2005), and the NssR regulon includes two genes encoding members of the bacterial hemoglobin family, the truncated globin Ctb and the single domain globin Cgb. These two globins have different functions because Cgb contributes to detoxification of NO (Pittman et al., 2007), whereas Ctb is thought to be involved in 0, metabolism and flux in C. jejuni (Wainwright et al., 2005).
IRON REGULATORY PROTEINS Iron regulation in almost all gram-negative bacteria and many gram-positive bacteria is mediated by proteins of the Fur family (Lee and Helmann, 2007). The Fur family contains several metal-responsive repressor proteins, which downregulate transcription of its target genes by metal-dependent binding to conserved DNA sequences located in the promoters of metal-regulated genes, thus blocking access of
RNA polymerase to the promoter. Binding only occurs when the intracellular concentration of metal exceeds a certain threshold, allowing the formation of a dimer or multimer (Escolar et al., 1999; Hantke, 2001). The Fur family contains proteins regulating gene expression in response to iron, zinc, manganese, and other metals (Lee and Helmann, 2007). The Fur (ferric uptake regulator) protein is an iron-responsive regulator whose targets usually include genes encoding iron uptake systems, directly repressing iron acquisition once the intracellular iron concentration becomes too high and thus preventing accumulation of iron to toxic concentrations. Other targets of Fur often include genes encoding proteins involved in iron storage, oxidative stress defense, or acid resistance (Bijlsma et al., 2002; Hantke, 2001; van Vliet et al., 2002). The C. jejuni genome contains two genes encoding Fur orthologs: the Cj0400 (Fur) protein, which regulates iron uptake genes (Chan et al., 1995; Holmes et al., 2005; Palyada et al., 2004; van Vliet et al., 1998; Wooldridge et al., 1994), and the Cj0322 (PerR) protein, which regulates oxidative stress defense (van Vliet et al., 1999). fur
The C. jejuni fur gene was originally identified functionally via reporter gene analysis and complementation of an E. coli fur mutant (Wooldridge et al., 1994), and independently by sequencing upstream of the C. jejuni lysS gene (Chan et al., 1995). Since then, the role of Fur in Campylobacter gene regulation has been investigated by several approaches (Holmes et al., 2005; Palyada et al., 2004; van Vliet et al., 1998, 2000). The C. jejuni fur gene is located inside an operon that also contains the downstream lyss and glyA housekeeping genes, and that has an upstream gene encoding a hypothetical protein (Chan et al., 1995; van Vliet et al., 2000). This genetic organization is also observed in Campylobacter upsaliensis (Bourke et al., 1996). The region directly upstream of the fur gene contains several putative Fur boxes and other putative operator sequences, and although recombinant C. jejuni Fur was reported to bind to oligonucleotides containing these putative Fur boxes (Chan et al., 1995), the 370bp region directly upstream of fur failed to show any promoter activity when fused to a lac2 reporter gene in C. jejuni (van Vliet et al., 2000). Thus, it is likely that the C. jejuni fur gene does not have its own promoter; it may be transcribed from promoters located upstream of the two genes directly upstream of fur (van Vliet et al., 2000), thus potentially leading to modulation of expression of Fur as in E. coli (Zheng et al., 1999). Inactivation of the fur gene in C. jejuni
CHAPTER 33
IRON METABOLISM. TRANSPORT, AND REGULATION
resulted in a decrease in growth rate when compared with the wild-type strain, both under iron-restricted and iron-sufficient conditions (van Vliet et al., 1998), but this may be due to polar effects and awaits further investigation. Inactivation of the fur gene resulted in significant changes in transcription and protein expression, including genes and proteins involved in iron transport, metabolism, and oxidative stress defense (Fig. 4), but also affected pathways not involved in iron metabolism (Holmes et al., 2005; Palyada et al., 2004; van Vliet et al., 1998). These will be further discussed in other sections of this chapter.
PerR Initial characterization of the C. jejuni fur mutant indicated that there were proteins that displayed iron-responsive regulation independent of Fur (van Vliet et al., 1998). These proteins were identified as the AhpC and KatA proteins (Baillon et al., 1999), and analysis of the C. jejuni genome sequence (Parkhill et al., 2000) allowed the identification of a second Fur homolog, named PerR. The PerR system was first described in Bacillus subtilis as the regulator of genes encoding AhpC and catalase (Bsat et al., 1998), and analysis of a C. jejuni perR mutant showed similar forms of regulation (Fig. 4). The C. jejuni perR mutant constitutively expressed very high amounts of the iron-repressed KatA and AhpC proteins, which made the C. jejuni perR mutant hyperresistant to peroxide stress induced by cumene hydroperoxide and
I
n
1
I beW% I storage
metabolism
Figure 4. Roles of iron and Fur and PerR regulatory proteins in controlling iron metabolism and oxidative stress resistance in C. jejuni. Arrows indicate connecting pathways.
601
hydrogen peroxide (van Vliet et al., 1999). Interestingly, expression and activity of KatA were still partially iron regulated in the perR mutant, but not in the fur perR double mutant (van Vliet et al., 1999), suggesting that Fur and PerR may coregulate genes (Fig. 4) and may be involved in responding to stimuli other than iron or peroxides, as in B. subtilis (Fuangthong et al., 2002). Transcription of perR is Fur and iron regulated (Holmes et al., 2005; Palyada et al., 2004), but conversely, PerR does not seem to regulate transcription of the fur gene (van Vliet et al., 2000) (Fig. 4; M. Reuter et al., unpublished data). The mechanism of peroxide stress sensing by PerR has not been elucidated in C. jejuni, but the mechanism may be similar to that of B. subtilis PerR, which senses peroxide stress by oxidation of the metal cofactor and subsequent oxidation of histidine residues in the protein (Herbig and Helmann, 2001; Lee and Helmann, 2006). PerR is thought to both use iron and manganese as possible cofactors, and the C. jejuni genome contains the CjOl4lc, CjO142c, and CjO143c genes encoding putative orthologs of a manganese ABC transporter (Jakubovics and Jenkinson, 2001). Other putative targets of PerR include the FeoB ferrous iron transporter and the TrxB thioredoxin reductase, thought to be involved in recycling the AhpC protein (M. Reuter et al., unpublished data).
TRANSCRIPTIONAL RESPONSES TO IRON AVAILABILITY Iron scarcity in the host represents a major stress for incoming pathogens and is considered to be a key signal that leads to the expression of virulence and colonization factors. With the advances of the genomic era, microarray technology has been instrumental in the identification of iron-regulated genes in various microbes, including C. jejuni (Deng et al., 2006; Ducey et al., 2005; Grifantini et al., 2003; Holmes et al., 2005; Madsen et al., 2006; Merrell et al., 2003; Miethke et al., 2006; Palyada et al., 2004; Roehrig et al., 2007; Tuanyok et al., 2005; Whitby et al., 2006; Zhou et al., 2006). Indeed, in the past 3 years, two independent studies have catalogued the genes belonging to the iron stimulon by wholegenome transcriptional profiling (Holmes et al., 2005; Palyada et al., 2004). Both of them compared the transcriptome profiles of C. jejuni cultured in iron-rich and iron-limited conditions during the exponential growth phase. In addition, Palyada and collaborators (2004) studied the changes in transcript levels as a function of time after the addition of ferrous iron to iron-restricted Cumpylobucter cells.
602
STINTZI ET AL.
These studies confirmed the tight regulation of the iron acquisition systems and also provided novel observations on the transcriptional response of other genes to iron levels. The time-course study showed that a drastic reprogramming of gene expression takes place during adaptation to a change in iron availability. Although not necessarily complete, the total number of iron-responding genes reaches approximately 460 genes, representing 27% of the genome. As to the timing of these transcriptional changes, there seems to be a strict chronological order, with some genes being expressed early and others later once adaptation has occurred. Furthermore, the transcript level of more than half of the ironresponsive genes was found to be transiently affected. Indeed, only 208 genes were differentially expressed at midlog phase by comparing bacteria grown in ironrich and iron-limited minimal medium. Interestingly, 65 of these 208 genes were identified as being iron regulated by Holmes and collaborators (2005). The discrepancies between the two studies likely reflect the different experimental designs and growth media used. Undeniably, the use of a defined minimal growth media (MEMa in Palyada et al., 2004) instead of a rich broth (Mueller-Hinton in Holmes et al., 2005) certainly affected Cumpylobacter physiology and consequently the transcriptome changes in response to iron availability. This difference in transcriptome profiles as a function of the growth conditions is not surprising. Indeed, it confirms the intrinsic link between iron metabolism and bacterial physiology. In agreement with this statement, most of the genes identified as iron responsive by Palyada et al. (2004), but not by Holmes et al. (2005), encode proteins involved in general bacterial physiology such as energy metabolism, cofactor biosynthesis, and ribosomal protein synthesis and modification. On the other hand, the genes commonly identified by both studies (Table 1) are likely to play a key role in iron metabolism because these genes are controlled by the iron level independently of the growth conditions. Looking at global gene expression profiles of other bacteria in response to iron starvation, most of the genes identified as differentially expressed in C. jejuni are commonly found to be iron regulated. As expected, the genes encoding proteins involved in iron acquisition were upregulated in iron-limited compared with iron-rich conditions. Many of the differentially expressed genes have confirmed or putative roles in iron acquisition. These include the hemin uptake system (ChuABCD) (hdley et al., 2006), one putative ferric binding transporter system (CjO173cCjOl75c), the ferric enterobactin transporter (CfrA and CeuE) (Palyada et al., 2004), one putative ABC
transporter system (Cj1661 to Cj1663), two proteins potentially involved in iron acquisition from rhodotorulic acid (Cj1658 and p19), and all the components of the energy-transducing TonB-ExbB-ExbD complex, which are known to be required for iron transport across the bacterial membrane (Postle and Larsen, 2007). Similar to other bacteria, the transcript level of those C. jejuni genes encoding proteins involved in oxidative stress defense was found to increase in response to iron restriction. This set of proteins includes the catalase KatA (which converts hydrogen peroxide to water and oxygen), the alkyl hydroperoxide reductase AhpC (which reduces alkyl hydroperoxide to alcohols), the peroxide stress regulator PerR, the thioredoxin TrxB, the thioredoxin reductase TrxA, the ferritin Cft, and oxidoreductases. Several genes encoding other transporters and membrane proteins were found to be upregulated in response to iron restriction. However, the function of these transporters in C. jejuni iron metabolism is obscure and would require further investigation. Beside genes encoding membrane proteins and oxidative stress defense systems, few other genes were found to be upregulated under iron-limited conditions. One gene of interest is fldA, which encodes a flavodoxin. In other bacteria, such as E. coli, Histophilus ovis, and Haemophilus parasuis, fldA forms an operon with fur, and the transcript level of fldA-fur is iron regulated (Bigas et al., 2006; Ekins and Niven, 2002; Zheng et al., 1999). In C. jejuni, fur is not iron regulated, and fldA is located apart from fur (Parkhill et al., 2000; van Vliet et al., 1998). Moreover, fldA forms an operon with Cj1383c and Cj1384c, which were both found to be similarly upregulated in response to iron restriction (Palyada et al., 2004). Although the function of these three genes in iron metabolism remains speculative, it is interesting to note that the expression of E. coli fldA has also been shown to be induced by the oxidant paraquat, suggesting a role for FldA in oxidative stress defenses (Zheng et al., 1999). In fact, it has been suggested that the function of FldA might be to keep the Fe-S clusters of metalloproteins reduced, thus protecting them against superoxide attack (Zheng et al., 1999). In cyanobacteria, flavodoxins are able to replace Fe-containing ferredoxin in various processes, such as photosynthesis, allowing bacterial growth under iron deficiency (Meimberg et al., 1999). Undeniably, flavodoxin plays an important role in bacterial physiology, yet its exact function in C. jejuni remains to be characterized. Interestingly, H. pylori FldA has been recently shown to be reduced by the flavodoxin:quinone reductase FqrB (St Maurice et al., 2007). Moreover, FldA and FrqB were
CHAPTER 33
*
IRON METABOLISM. TRANSPORT. AND REGULATION
603
Table 1. C. jejuni genes differentially expressed between iron-rich and iron-restricted growth conditions" Category and gene name Genes with increased expression level in iron-limited medium Iron transport cfrA ceuE chuABCD CjO173c-CjO174c-CjO175c Cj0177 exbB1-exbD1-tonBl ex6B2-ex6D2-tonB2 tonB3 Cj1658-pl9 Cj 1661-Cj 1662-Cj 1663 Oxidative stress defenses cft
ahpC katA perR tPX trxA-trxB Cj1386 CjO559 Cj1287c Other transport, membrane, or periplasmic proteins Cj0203 Cj0236c Cj0343c CjO722c CjO723c Cj0909 Cj1163c Cj0420 CjO982c (cjaA) Cj1725 Others Cj1383c-Cj1384c Cj1613c
fld panB-panC rpmA Genes with decreased expression level in iron-limited medium CjOOllC cj0012c (YK) Cj0073c Cj0358 Cj0735 Cj1356c Cj1357c (nrfA)-Cjl358c (nrfH) napABGHL
sdhAB
Proposed function
Ferric-enterobactin receptor Ferric-enterobactin uptake periplasmic-binding protein Heme transporter complex Putative ABC ferric-binding transporter system Putative iron transport protein Outer-membrane energy transducer system Outer-membrane energy transducer system TonB transport protein Membrane proteins potentially involved in iron transport Putative ABC transporter system Ferritin Alkyl hydroperoxide reductase Catalase Peroxide stress regulator Thiol peroxidase Thioredoxin (TrxB) and thioredoxin reductase (TrxA) Ankyrin-repeat containing protein Probable oxidoreductase Probable malate oxidoreductase Probable citrate transporter Probable integral membrane protein Probable integral membrane protein Probable DNA methylase Probable integral membrane zinc-metalloprotease Putative membrane or periplasmic protein Possible cation transport protein Probable periplasmic protein Periplasmic-binding protein component of an ABC-type cysteine transporter system Probable periplasmic protein Hypothetical proteins Hypothetical protein Flavodoxin Probable 3-methyl-2-oxobutanoate hydroxymethyltransferase (PanB) and pantoate-beta-alanine ligase (PanC) 50s ribosomal protein L27 Possible nonspecific DNA-binding protein Rubredoxin (Rbo)/ rubrerythrin (Rbr)-like protein Hypothetical protein Probable cytochrome C551 peroxidase Probable periplasmic protein Probable integral membrane protein Periplasmic nitrite reductase system Nitrate reductase system Succinate dehydrogenase proteins
"The table lists those genes that are controlled by the iron level independently of the growth conditions as identified by both Palyada, et al. (2004) and Holmes et al. (2005) and therefore are likely to play a key role in iron metabolism.
shown to form an enzymatic complex with pyruvate: ferredoxine oxidoreductase, enabling the generation of pyruvate via CO, fixation in the presence of acetylCoA and NADPH (St Maurice et al., 2007). Interestingly, the C. jejuni ortholog of FrqB, Cj0559, and
two enzymes involved in acetyl-CoA biosynthesis, PanB and PanC, were also found to be upregulated under iron-restricted conditions. On the basis of these observations, it is tempting to propose that iron regulates C. jejuni capnophilic ability, allowing the
604
STINTZI ET AL.
bacterium to fully capitalize the abundance of CO, in the host gastrointestinal tract. Sixteen genes were found to have their transcript level decreased in iron-limited medium by both studies. Interestingly, at least 11 of them encode ironcontaining proteins or enzymatic complexes, including the Rbo/Rbr-like protein (Yamasaki et al., 2004), a probable cytochrome (25.51 peroxidase, the periplasmic nitrite and nitrate reductase systems (Pittman et al., 2007), and the succinate dehydrogenase. The downregulation of these iron-containing proteins might allow C. jejuni to survive iron scarcity by shifting its metabolism and/or shuttling the limited amount of available iron to essential functions.
TRANSCRIPTIONAL PROFILE OF THE IRON METABOLISM GENES The transcriptional profiles of C. jejuni during colonization of the chick and rabbit gastrointestinal tracts were recently reported (Stintzi et al., 2005; Woodall et al., 2005). These two studies provided a blueprint of the C. jejuni transcriptome while residing in the host gastrointestinal tract, allowing the identification of potential colonization and/or virulence determinants. The first report used a mammalian model of gastroenteritis, the rabbit ileal loop model, to reveal the C. jejuni in vivo transcriptome (Stintzi et al., 2005). In this work, C. jejuni was directly inoculated into ligated 20-cm sections of the ileum, and the bacterial total RNA was recovered 48 h after inoculation. The second report used the chick colonization model, which consisted of orally inoculating 10-h-old chicks with C. jejuni and recovering the bacterial total RNA 12 h after inoculation (Woodall et al., 2005). Both studies compared the transcript levels of C. jejuni grown in vivo to that of C. jejuni grown in vitro. One important issue from these studies is that the relative gene expression level only tells part of the story. The absolute levels of transcript would be more informative by telling how highly a particular gene is expressed. Indeed, a high induction level could just reflect how low the expression is when C. jejuni is grown in vitro, and the absence of induction could just indicate the requirement of this gene under both in vivo and in vitro conditions. Nevertheless, these studies provided novel insights into the transcriptional state of C. jejuni during host colonization. As expected, many of the iron-regulated genes were found to be upregulated under intestinal in vivo conditions compared with in vitro growth (MuellerHinton broth). In the rabbit ileal loop, the genes found to be upregulated or equally expressed be-
tween in vivo and in vitro conditions encode for the ferric enterobactin uptake permease system (CeuBCDE), the TonB-ExbB-ExbD transducing systems, a putative ferric binding transport system (Cj0178 and CjO173c-CjO176c), the ferrous ion transporter FeoB, and the putative iron transporter p19 and Cj16.58. The other genes belonging to iron metabolism found to be upregulated in the rabbit intestine include those encoding CjO236c, CjO722c-Cj0723cy Cj1613c, PanBC, and RpmA. In addition, the genes that encode known or putative oxidative stress defense systems, such as the alkyl hydroxyperoxidase AhpC, the catalase KatA, the thiol peroxidase Tpx, the thioredoxin TrxB, and the ankyrin repeat containing protein Cj 138 6, were also upregulated. Undoubtedly, the in vivo transcriptome of C. jejuni is reflective of an iron-restricted environment. In agreement with this statement, most of the genes with decreased expression level in iron-limited medium (Table 1)were also found to be downregulated in vivo; these genes are CjOO73c, CjO358, nrfA-nrP, napABGHL, and sdhABC. Strikingly, the expression of the iron-regulated genes (Table 1) during C. jejuni chick colonization is significantly different from their expression during the colonization of the rabbit intestine. In fact, most of the genes with decreased expression in ironrestricted conditions were found to be upregulated in vivo in the chick cecum (Woodall et al., 2005). These genes are sdbABC, napABCG, nrfA, and Cj0358. In addition, among all the iron acquisition and oxidative stress defense systems, only chuA and chuB were found to be upregulated in the chick cecum. Furthermore, several of the other iron-repressed genes listed in Table 1 were downregulated in the chick cecum, including Cj0203, panBC, rpmA, CjO982c, and Cj1725c. Thus, and in contrast to the C. jejuni transcriptome in the rabbit intestine, C. jejuni gene expression during colonization of the chick cecum is essentially indicative of an iron-containing environment. Although this discrepancy between the two animal models is surprising, it is important to notice that these in vivo data solely represent a snapshot of the in vivo C. jejuni transcriptome. In the chick colonization experiment, a colonization period of 12 h, compared with 48 h in the rabbit, might not have been sufficient to completely deplete C. jejuni from its stored iron when grown in the iron-replete medium (used to prepare the chick inocula). Furthermore, the absence of gut flora in the chick model will likely affect iron availability because C. jejuni will not have to compete for iron with other microbes. Several other in vitro transcriptome analyses of C. jejuni have now been reported, including the transcriptome profile associated with immobilized growth
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
of C. jejuni (Sampathkumar et al., 2006). Interestingly, compared with growth in broth, growth on agar resulted in the upregulation of iron acquisition and oxidative stress systems. In fact, the pattern of expression of immobilized cells is indicative of an iron-restricted environment with the upregulation of chuA, Cj0178, cfrA, p19, CjO174c-CjO175c, h t A , ahpC, and CjO559 and the downregulation of napABGH, CjOO12c, and a component of the succinate dehydrogenase complex. Recently, C. jejuni has been shown to form biofilms on a variety of surfaces (Kalmokoff et al., 2006). Proteomic analysis of planktonic versus biofilm-grown cells has identified the thiol peroxidase Tpx and the alkyl hydroxyperodase AhpC to be present at a higher level in biofilms (Asakura et al., 2007; Kalmokoff et al., 2006). Altogether, the transcriptome and protein profiles of C. jejuni during immobilized and biofilm growth is indicative of a key role for iron metabolism in the survival and/or maintenance of C. jejuni in biofilms. In P. aeruginosa, biofilm formation is intimately linked to the process of iron acquisition. Similarly, the ability of Carnpylobacter to acquire iron is likely to be a decisive factor for the formation of biofilms (Singh et al., 2002). Given the overwhelming importance of biofilms in bacterial pathogenesis, iron metabolism in C. jejuni could be considered to be a real pathogenic attribute.
CONCLUSIONS In this chapter, we have described the current state of our understanding of the way campylobacters obtain and store iron and how the cell regulates its response to differing iron availability. The importance of elements of both the uptake and response systems in intestinal colonization by Campylobacter highlights the significance of iron metabolism and transport in pathogenicity. Although there have been several significant advances in our understanding, notably in transport systems and the transcriptional response to variation in iron levels, there is still much to do. We are getting closer to understanding the basic biochemistry of the iron transport systems, but the phenotypic consequences of the different complement of transport systems found in Campylobacter stains is unclear; this is also true for the respective roles of the differing TonB/ExbBD systems. In addition, the interplay between the Fur and PerR regulators and the specificity at the level of the promoter is still to be fully elucidated. The transcriptomic analysis of genes involved in iron metabolism has highlighted some important connections between iron limitation and C. jejuni metabolism. Moreover, the
605
differences in transcript profile between animal models and also in biofilms underlines the importance of iron metabolism for C. jejuni. The recent link between the host stress hormone, norepinephrine, and growth in iron-limited conditions is intriguing (Cogan et al., 2007). Clearly, a better understanding of both the role of iron limitation in intestinal colonization and pathogenesis, and the involvement of animal stress levels will have significant implications in addressing the risk of food-borne infection due to Campylobacter. Acknowledgments. Research in the authors’ laboratories has been supported by the Biotechnology and Biological Sciences Research Council (project grants D19661, BBD0127081, and EGA16166; IFR Core Strategic grant), the National Institutes of Health (grant AI055612), and the Canadian Institutes of Health Research.
REFERENCES Abdallah, F. B., and J.-M. El Hage Chahine. 2000. Transferrins: iron release from lactoferrin. J. Mol. Biol. 303:255-266. Adhikari, P., S . A. Berish, A. J. Nowalk, K. L. Veraldi, S. A. Morde, and T. A. Mietzer. 1996. The fipABC locus of Neisseria gonorrhoeae functions in the periplasm-to-cytosol transport of iron. J. Bacteriol. 178:2145-2149. Almiron, M., A. J. Link, D. Furlong, and R. Kolter. 1992. A novel DNA-binding protein with regulatory and protective roles in starved Escherichia coli. Genes Dev. 6:2646-2654. Anderson, B. F., H. M. Baker, E. J. Dodson, G. E. Norris, S. V. Rumball, J. M. Waters, and E. N. Baker. 1987. Structure of human lactoferrin at 3.2-A resolution. Proc. Natl. Acad. Sci. USA 84~1769-1773. Anderson, J. E., P. F. Sparling, and C. N. Cornelissen. 1994. Gonococcal transferrin-binding protein 2 facilitates but is not essential for transferrin utilization. J. Bacteriol. 176:3 162-3 170. Andrews, S. C. 1998. Iron storage in bacteria. Adv. Microb. Physiol. 40:281-351. Andrews, S. C., A. K. Robinson, and F. Rodriguez-Quinones. 2003. Bacterial iron homeostasis. FEMS Microbiol. Rev. 27:215237. Asakura, H., M. Yamasaki, S. Yamamoto, and S. Igimi. 2007. Deletion of peb4 gene impairs cell adhesion and biofilm formation in Campylobacter jejuni. FEMS Microbiol. Lett. 275: 278-285. Baig, B. H., I. K. Wachsmuth, and G. K. Morris. 1986. Utilization of exogenous siderophores by Campylobacter species. J. Clin. Microbiol. 23 :43 1-43 3. Baillon, M. L. A., A. H. M. van Vliet, J. M. Ketley, C. Constantinidou, and C. W. Penn. 1999. An iron-regulated alkyl hydroperoxide reductase (AhpC) confers aerotolerance and oxidative stress resistance to the microaerophilic pathogen Campylobacter jejuni. J. Bacteriol. 181:4798-4804. Bereswill, S., U. Waidner, S. Odenbreit, F. Lichte, F. Fassbinder, G. Bode, and M. Kist. 1998. Structural, functional and mutational analysis of the pfr gene encoding a ferritin from Helicobacter pylori. Microbiology 144:2505-25 16. Bigas, A., M. A. Garrido, I. Badiola, J. Barbe, and M. Llagostera. 2006. Non-viability of Haemophilus parasuis fur-defective mutants. Vet. Microbiol. 118:107-116. Bijlsma, J. J. E., B. Waidner, A. H. M. van Vliet, N. J. Hughes, S. Hag, S. Bereswill, D. J. Kelly, C. M. J. E. VandenbrouckeGrauls, M. Kist, and J. G. Kusters. 2002. The ferric uptake reg-
606
STINT21 ET AL.
ulator (Fur) homologue of Helicobacter pylori is involved in acid resistance. Infect. Immun. 70:606-611. Biswas, G. D., and P. F. Sparling. 1995. Characterization of IbpA, the structural gene for a lactoferrin receptor in Neisseria gonorrhoeae. Infect. Immun. 63:2958-2967. Black, R E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser. 1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis. 157:472-479. Bourke, B., S. T. a1 Rashid, H. L. Bingham, and V. L. Chan. 1996. Characterization of Campylobacter upsaliensis fur and its localization in a highly conserved region of the Campylobacter genome. Gene 183:219-224. Braun, V., and M. Braun. 2002. Active transport of iron and siderophore antibiotics. Cum Opin. Microbiol. 5:194-201. Braun, V., K. Hantke, and W. Koster. 1998. Bacterial iron transport: mechanisms, genetics, and regulation. Met. Ions Biol. Syst. 35~67-145. Bsat, N., A. Herbig, L. Casillas-Martinez, P. Setlow, and J. D. Helmann. 1998. Bacillus subtilis contains multiple Fur homologues: identification of the iron uptake (Fur) and peroxide regulon (PerR) repressors. Mol. Microbiol. 29:189-198. Carniel, E. 2001. The Yersinia high-pathogenicity island: an ironuptake island. Microbes Infect. 3561-569. Cartron, M., S. Maddocks, P. Gillingham, C. Craven, and S. Andrews. 2006. Feo-transport of ferrous iron into bacteria. Biometals 19:143-157. Cawthraw, S., S. F. Park, J. M. Ketley, R. Ayling, and D. G. Newell. 1996. The chick colonization model and its role in molecular biology studies of campylobacters, p. 649-652. In D. G. Newell, J. M. Ketley and R. A. Feldman (ed.), Campylobacter, Helicobacter and Related Organisms, Plenum Press, New York. Chan, A. C. K., B. Lelj-Garolla, F. I. Rosell, K. A. Pedersen, A. G. Mauk, and M. E. P. Murphy. 2006. Cofacial heme binding is linked to dimerization by a bacterial heme transport protein. J. Mol. Biol. 362:1108-1119. Chan, V. L., H. Louie, and H. L. Bingham. 1995. Cloning and transcription regulation of the ferric uptake regulatory gene of Campylobacter jejuni TGH9011. Gene 164:25-31. Chaudhuri, R. R., and M. J. Pallen. 2006. xBASE, a collection of online databases for bacterial comparative genomics. Nucleic Acids Res. 34(Database Issue):D335-337. Chelikani, P., I. Fita, and P. C. Loewen. 2004. Diversity of structures and properties among catalases. Cell. Mol. Life. Sci. 61: 192-208. Chen, C.-Y., S. A. Berish, S. A. Morse, and T. A. Mietzner. 1993. The ferric iron-binding protein of pathogenic Neisseria spp. functions as a periplasmic transport protein in iron acquisition from human transferrin. Mol. Microbiol. 10:3 11-3 18. Cogan, T. A., A. 0. Thomas, L. E. Rees, A. H. Taylor, M. A. Jepson, P. H. Williams, J. Ketley, and T. J. Humphrey. 2007. Norepinephrine increases the pathogenic potential of Campylobacter jejuni. Gut 56:1060-1065. Conrad, M. E., and S. G . Schade. 1968. Ascorbic acid chelates in iron absorption: a role for hydrochloric acid and bile. Gastroenterology 55:35-45. Cooksley, C., P. J. Jenks, A. Green, A. Cockayne, R. P. Logan, and K. R. Hardie. 2003. NapA protects Helicobacter pylori from oxidative stress damage, and its production is influenced by the ferric uptake regulator. J. Med. Microbiol. 52:461-469. Cope, L. D., S. E. Thomas, Z. Hrkal, and E. J. Hansen. 1998. Binding of heme-hemopexin complexes by soluble HxuA protein allows utilization of this complexed heme by Haemophilus influenzae. Infect. Immun. 66:45 11-45 16. Cope, L. D., R. Yogev, U. Muller-Eberhard, and E. J. Hansen. 1995. A gene cluster involved in the utilization of both free
heme and heme:hemopexin by Haemophilus influenzae type b. J. Bacteriol. 177:2644-2653. Cornelissen, C. N., G. D. Biswas, J. Tsai, D. K. Paruchuri, S. A. Thompson, and P. F. Sparling. 1992. Gonococcal transferrinbinding protein 1 is required for transferrin utilization and is homologous to TonB-dependent outer membrane receptors. J. Bacteriol. 1745788-5797. Day, W. A., J. L. Sajecki, T. M. Pitts, and L. A. Joens. 2000. Role of catalase in Carnpylobacterjejuni intracellular survival. Infect. Immun. 68:6337-6345. Deng, K., R. J. Blick, W. Liu, and E. J. Hansen. 2006. Identification of Francisella tularensis genes affected by iron limitation. Infect. Immun. 74:4224-4236. Dhaenens, L., F. Szczebara, and M. 0. Husson. 1997. Identification, characterization, and immunogenicity of the lactoferrinbinding protein from Helicobacter pylori. Infect. Immun. 65: 5 14-5 18. Dhaenens, L., F. Szczebara, S. Van Nieuwenhuyse, and M. 0. Husson. 1999. Comparison of iron uptake in different Helicobacter species. Res. Microbiol. 150:475-48 1. Dorrell, N., M. C. Martino, R. A. Stabler, S. J. Ward, 2. W. Zhang, A. A. McColm, M. J. Farthing, and B. W. Wren. 1999. Characterization of Helicobacter pylori PldA, a phospholipase with a role in colonization of the gastric mucosa. Gastroenterology 117:1098-1104. Dubbels, B. L., A. A. DiSpirito, J. D. Morton, J. D. Semrau, J. N. Neto, and D. A. Bazylinski. 2004. Evidence for a copperdependent iron transport system in the marine, magnetotactic bacterium strain MV-1. Microbiology 150:2931-2945. Dncey, T. F., M. B. Carson, J. Orvis, A. P. Stintzi, and D. W. Dyer. 2005. Identification of the iron-responsive genes of Neisseria gonorrhoeae by microarray analysis in defined medium. J. Bacteriol. 187:4865-4874. Ekins, A., and D. F. Niven. 2002. Identification of fur and fldA homologs and a Pasteurella multocida tbpA homolog in Histophilus ovis and effects of iron availability on their transcription. J. Bacteriol. 184:2539-2542. Elvers, K. T., S. M. Turner, L. M. Wainwright, G. Marsden, J. Hinds, J. A. Cole, R. K. Poole, C. W. Penn, and S. F. Park. 2005. NssR, a member of the Crp-Fnr superfamily from Campylobacter jejuni, regulates a nitrosative stress-responsive reguIon that includes both a single-domain and a truncated haemoglobin. Mol. Microbiol. 57:735-750. Escolar, L., J. Perez-Martin, and V. de Lorenzo. 1999. Opening the iron-box: transcriptional metalloregulation by the Fur protein. ]. Bacteriol. 181:6223-6229. Field, L. H., V. L. Headley, S. M. Payne, and L. J. Berry. 1986. Influence of iron on growth, morphology, outer membrane protein composition, and synthesis of siderophores in Campylobacter jejuni. Infect. Immun. 54: 126-132. Fox, E. M., M. Raftery, A. Goodchild, and G. L. Mendz. 2007. Campylobacter jejuni response to ox-bile stress. FEMS Immunol. Med. Microbiol. 49:165-1 72. Freestone, P. P., R D. Haigh, P. H. Williams, and M. Lyte. 2003. Involvement of enterobactin in norepinephrine-mediated iron supply from transferrin to enterohaemorrhagic Escherichia coli. FEMS Microbiol. Lett. 222:39-43. Fuangthong, M., A. F. Herbig, N. Bsat, and J. D. Helmann. 2002. Regulation of the Bacillus subtilis fur and perR genes by PerR: not all members of the PerR regulon are peroxide inducible. J. Bacteriol. 184:3276-3286. Galindo, M. A., W. A. Day, B. H. Raphael, and L. A. Joens. 2001. Cloning and characterization of a Campylobacter jejuni ironuptake operon. Curr. Microbiol. 42:139-143. Genco, C. A., and D. W. Dixon. 2001. Emerging strategies in microbial haem capture. Mol. Microbiol. 39:l-ll.
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
Ghigo, J. M., S. Letoffe, and C. Wandersman. 1997. A new type of hemophore-dependent heme acquisition system of Serratia marcescens reconstituted in Escherichia coli. J. Bacteriol. 179: 3572-3579. Giansanti, F., Rossi, P., Massucci, M. T., Botti, D., Antonini, G., Valenti, P., and Seganti, L. 2002. Antiviral activity of ovotransferrin discloses an evolutionary strategy for the defensive activities of lactoferrin. Biochem. Cell Biol. 80:125-130. Gomez, J. A., M. T. Criado, and C. M. Ferreiros. 1998. Cooperation between the components of the meningococcal transferrin receptor, TbpA and TbpB, in the uptake of transferrin iron by the 37-kDa ferric-binding protein (FbpA). Res. Microbiol. 149~381-387. Gorringe, A. R., G. Woods, and A. Robinson. 1990. Growth and siderophore production by Bordetella pertussis under ironrestricted conditions. FEMS Microbiol. Lett. 54:101-105. Grant, K. A., I. U. Belandia, N. Dekker, P. T. Richardson, and S. F. Park. 1997. Molecular characterization of pldA, the structural gene for a phospholipase A from Campylobacter coli, and its contribution to cell-associated hemolysis. Infect. Immun. 65: 1172-1 180. Grant, K. A., and S. F. Park. 1995. Molecular characterization of katA from Campylobacter jejuni and generation of a catalasedeficient mutant of Campylobacter coli by interspecific allelic exchange. Microbiology 141: 1369-1376. Gray-Owen, S. D., and A. B. Schryvers. 1996. Bacterial transferrin and lactoferrin receptors. Trends Microbiol. 4:185-191. Grifantini, R., S. Sehastian, E. Frigimelica, M. Draghi, E. Bartolini, A. Muzzi, R. Rappuoli, G. Grandi, and C. A. Genco. 2003. Identification of iron-activated and -repressed Fur-dependent genes by transcriptome analysis of Neisseria meningitidis group B. Proc. Natl. Acad. Sci. USA 100:9542-9547. Guerry, P., J. Perez-Casal, R. Yao, A. McVeigh, and T. J. Trust. 1997. A genetic locus involved in iron utilization unique to some Campylobacter strains. J. Bacteriol. 179:3997-4002. Hannavy, K., G. C. Barr, C. J. Dorman, J. Adamson, L. R. Mazengera, M. P. Gallagher, J. s. Evans, B. A. Levine, I. P. Trayer, and C. F. Higgins. 1990. TonB protein of Salmonella typhimurium. A model for signal transduction between membranes. J. Mol. Biol. 216:897-910. Hantke, K. 1997. Ferrous iron uptake by a magnesium transport system is toxic for Escherichia coli and Salmonella typhimurium. J. Bacteriol. 179:6201-6204. Hantke, K. 2001. Iron and metal regulation in bacteria. Curr. Opin. Microbiol. 4:172-177. Henderson, D., and S. Payne. 1994. Characterization of the Vibrio cholerae outer membrane heme transport protein HutA: sequence of the gene, regulation of expression, and homology to the family of TonB-dependent proteins. J. Bacteriol. 176:32693277. Herbig, A. F., and J. D. Helmann. 2001. Roles of metal ions and hydrogen peroxide in modulating the interaction of the Bacillus subtilis PerR peroxide regulon repressor with operator DNA. Mol. Microbiol. 41:849-859. Herrington, D. A., and P. F. Sparling. 1985. Haemophilw influenzae can use human transferrin as a sole source for required iron. Infect. Immun. 48:248-251. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Dn, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Holmes, K., F. Mulholland, B. M. Pearson, C. Pin, J. McNichollKennedy, J. M. Ketley, and J. M. Wells. 2005. Campylobacter jejuni gene expression in response to iron limitation and the role of Fur. Microbiology 151:243-257.
607
Hornung, J. M., H. A. Jones, and R D. Perry. 1996. The hmu locus of Yersinia pestis is essential for utilization of free haemin and haem-protein complexes as iron sources. Mol. Microbiol. 20:725-739. Hughes, R. 2004. Campylobacter jejuni in Guillain-Barr6 syndrome. Lancet Neurol. 3:644. Husson, M. O., D. Legrand, G. Spik, and H. Leclerc. 1993. Iron acquisition by Helicobacter pylori: importance of human lactoferrin. Infect. Immun. 61:2694-2697. Ishikawa, T., Y. Mizunoe, S. Kawabata, A. Takade, M. Harada, S. N. Wai, and S. Yoshida. 2003. The iron-binding protein Dps confers hydrogen peroxide stress resistance to Campylobacter jejuni. J. Bacteriol. 185:1010-1017. Jakubovics, N. S., and H. F. Jenkinson. 2001. Out of the iron age: new insights into the critical role of manganese homeostasis in bacteria. Microbiology 147:1709-1718. Janvier, B., C. Constantinidou, P. Aucher, Z. V. Marshall, C. W. Penn, and J. L. Fauchere. 1998. Characterization and gene sequencing of a 19-kDa periplasmic protein of Carnpylobacter jejunilcoli. Res. Microbiol. 149:95-107. Kalmokoff, M., P. Lanthier, T. L. Tremblay, M. Foss, P. C. Lau, G. Sanders, J. Austin, J. Kelly, and C. M. Szymanski. 2006. Proteomic analysis of Campylobacter jejuni 11168 biofilms reveals a role for the motility complex in biofilm formation. J. Bacteriol. 188:4312-4320. Kammler, M., C. Schon, and K. Hantke. 1993. Characterization of the ferrous iron uptake system of Escherichia coli. J. Bacteriol. 175:6212-6219. Ketley, J. M. 1997. Pathogenesis of enteric infection by Campylobacter. Microbiology 1435-2 1. Knight, S. A. B., G. Vilaire, E. Lesuisse, and A. Dancis. 2005. Iron acquisition from transferrin by Candida albicans depends on the reductive pathway. Infect. lmmun. 735482-5492. Lee, J. W., and J. D. Helmann. 2007. Functional specialization within the Fur family of metalloregulators. Biometals 20:485499. Lee, J. W., and J. D. Helmann. 2006. The PerR transcription factor senses H,O, by metal-catalysed histidine oxidation. Nature 440: 363-367. Letoffe, S., P. Delepelaire, and C. Wandersman. 2004. Free and hemophore-bound heme acquisitions through the outer membrane receptor HasR have different requirements for the TonBExbB-ExbD complex. J. Bacteriol. 186:4067-4074. Lewis, L. A., M. H. Sung, M. Gipson, K. Hartman, and D. W. Dyer. 1998. Transport of intact porphyrin by HpuAB, the hemoglobin-haptoglobin utilization system of Neisseria meningitidis. J. Bacteriol. 180:6043-6047. Linton, K. J., and C. F. Higgins. 1998. The Escherichia coli ATPbinding cassette (ABC) proteins. Mol. Microbiol. 285-13. Liu, Y., and P. R. Ortiz de Montellano. 2000. Reaction intermediates and single turnover rate constants for the oxidation of heme by human heme oxygenase-1. J. Biol. Chem. 27552975307. Lynch, M., and H. Knramitsu. 2000. Expression and role of superoxide dismutases (SOD) in pathogenic bacteria. Microbes Infect. 2:1245- 1255. Ma, L., W. Kaserer, R. Annamalai, D. C. Scott, B. Jin, X. Jiang, Q. Xiao, H. Maymani, L. M. Massis, L. C. Ferreira, S. M. Newton, and P. E. Klebba. 2007. Evidence of ball-and-chain transport of ferric enterobactin through FepA. J. Biol. Chem. 282: 3 97-406. Madsen, M. L., D. Nettleton, E. L. Thacker, and F. C. Minion. 2006. Transcriptional profiling of Mycoplasma hyopneumoniae during iron depletion using microarrays. Microbiology 152:937944.
608
STINT21 ET AL.
Meimberg, K., N. Fischer, J. D. Rochaix, and U. Muhlenhoff. 1999. Lys35 of PsaC is required for the efficient photoreduction of flavodoxin by photosystem I from Chlamydomonas reinhardtii. Eur. J. Biochem. 263:137-144. Menozzi, F. D., C. Gantiez, and C. Locht. 1991. Identification and purification of transferrin- and lactoferrin-binding proteins of Bordetella pertussis and Bordetella bronchiseptica. Infect. Immun. 59:3982-3988. Merrell, D. S., L. J. Thompson, C. C. Kim, H. Mitchell, L. S. Tompkins, A. Lee, and S. Falkow. 2003. Growth phasedependent response of Helicobacter pylori to iron starvation. Infect. Immun. 71:6510-6525. Miethke, M., H. Westers, E. J. Blom, 0. P. Kuipers, and M. A. Marahiel. 2006. Iron starvation triggers the stringent response and induces amino acid biosynthesis for bacillibactin production in Bacillus subtilis. J. Bacteriol. 188:8655-8657. Miller, C. E., J. D. Rock, K. A. Ridley, P. H. Williams, and J. M. Ketley. 2008. Utilizaton of lactoferrin-bound and transferrinbound iron by Campylobacter jejuni. J. Bacteriol. 190:19001911. Mills, M., and S. Payne. 1997. Identification of shuA, the gene encoding the heme receptor of Shigellu dysenteriae, and analysis of invasion and intracellular multiplication of a shuA mutant. Infect. Immun. 655358-5363. Naikare, H., K. Palyada, R. Panciera, D. Marlow, and A. Stintzi. 2006. Major role for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infect. Immun. 745433-5444. Occhino, D. A., E. E. Wyckoff, D. P. Henderson, T. J. Wrona, and S. M. Payne. 1998. Vibrio cholerae iron transport: haem transport genes are linked to one of two sets of tonB, exbB, exbD genes. Mol. Microbiol. 29:1493-1507. Ochsner, U. A., 2. Johnson, and M. L. Vasil. 2000. Genetics and regulation of two distinct haem-uptake systems, phu and has, in Pseudomonas aeruginosa. Microbiology 146: 185-198. Palyada, K., D. Threadgill, and A. Stintzi. 2004. Iron acquisition and regulation in Campylobacter jejuni. J. Bacteriol. 186:47144729. Park, R. Y., H. Y. Sun, M. H. Choi, Y. H. Bai, and S. H; Shin. 2005. Staphylococcus aureus siderophore-mediated iron-acquisition system plays a dominant and essential role in the utilization of transferrin-bound iron. J. Microbiol. 43:183-190. Park, S . F. 1999. The use of hipO, encoding benzoylglycine amidohydrolase (hippuricase), as a reporter of gene expression in Campylobacter coli. Lett. Appl. Microbiol. 28:285-290. Park, S. F., and P. T. Richardson. 1995. Molecular characterization of a Campylobacter jejuni lipoprotein with homology to periplasmic siderophore-binding proteins. J. Bacteriol. 177: 2259-2264. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. M. van Vliet, S. Whitehead, and B. G . Barrel]. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Pearson, B. M., D. J. Gaskin, R. P. Segers, J. M. Wells, P. J. Nuijten, and A. H. van Vliet. 2007. The complete genome sequence of Campylobacter jejuni strain 811 16 (NCTC11828). J. Bacteri01. 189:8402-8403. Pesci, E. C., D. L. Cottle, and C. L. Pickett. 1994. Genetic, enzymatic, and pathogenic studies of the iron superoxide dismutase of Campylobacter jejuni. Infect. Immun. 62:2687-2694. Pettersson, A., T. Prinz, A. Umar, J. van der Biezen, and J. Tommassen. 1998. Molecular characterization of LbpB, the second
lactoferrin-binding protein of Neisseria meningitidis. Mol. Microbiol. 27599-610. Pfeiffer, J., J. Guhl, B. Waidner, M. Kist, and S. Bereswill. 2002. Magnesium uptake by CorA is essential for viability of the gastric pathogen Helicobacter pylori. Infect. Immun. 70:3930-3934. Pickett, C. L., T. Auffenberg, E. C. Pesci, V. L. Sheen, and S. S. Jusuf. 1992. Iron acquisition and hemolysin production by Campylobacter jejuni. Infect. Immun. 60:3872-3877. Pidcock, K. A., J. A. Wooten, B. A. Daley, and T. L. Stull. 1988. Iron acquisition by Haemophilus influenzae. Infect. Immun. 56: 721-725. Pittman, M. S., K. T. Elvers, L. Lee, M. A. Jones, R. K. Poole, S. F. Park, and D. J. Kelly. 2007. Growth of Campylobacter jejuni on nitrate and nitrite: electron transport to NapA and NrfA via NrfH and distinct roles for NrfA and the globin Cgb in protection against nitrosative stress. Mol. Microbiol. 63575590. Poole, L. B., C. M. Reynolds, Z. A. Wood, P. A. Karplus, H. R. Ellis, and M. Li Calzi. 2000a. AhpF and other NADH:peroxiredoxin oxidoreductases, homologues of low M, thioredoxin reductase. Eur. J. Biochem. 267:6126-6133. Poole, R. K., and M. N. Hughes. 2000b. New functions for the ancient globin family: bacterial responses to nitric oxide and nitrosative stress. Mol. Microbiol. 36:775-783. Postle, K., and R. A. Larsen. 2007. TonB-dependent energy transduction between outer and cytoplasmic membranes. Biometals 20~453-465. Purdy, D., S. Cawthraw, J. H. Dickinson, D. G . Newel!, and S. F. Park. 1999. Generation of a superoxide dismutase (SOD)deficient mutant of Campylobacter coli: evidence for the significance of SOD in Campylobacter survival and colonization. Appl. Environ. Microbiol. 65:2540-2546. Purdy, D., and S. F. Park. 1994. Cloning, nucleotide sequence and characterization of a gene encoding superoxide dismutase from Campylobacter jejuni and Campylobacter coli. Microbiology 140~1203-1208. Raphael, B. H., and L. A. Joens. 2003. FeoB is not required for ferrous iron uptake in Campylobacter jejuni. Can. J. Microbiol. 49~727-731. Redhead, K., T. Hill, and H. Chart. 1987. Interaction of lactoferrin and transferrins with the outer membrane of Bordetella pertussis. J. Gen. Microbiol. 133:891-898. Richardson, P. T., and S. F. Park. 1995. Enterochelin acquisition in Campylobacter coli: characterization of components of a binding-protein-dependent transport system. Microbiology 141: 3 181-3 191. Ridley, K. A., J. D. Rock, Y. Li, and J. M. Ketley. 2006. Heme utilization in Campylobacter jejuni. J. Bacteriol. 188:78627875. Roehrig, S. C., H. Q. Tran, V. Spehr, N. Gunkel, P. M. Selzer, and H. J. Ullrich. 2007. The response of Mannheimia haemolytica to iron limitation: implications for the acquisition of iron in the bovine lung. Vet. Microbiol. 121:316-329. Sampathkumar, B., S. Napper, C. D. Carrillo, P. Willson, E. Taboada, J. H. Nash, A. A. Potter, L. A. Babiuk, and B. J. Allan. 2006. Transcriptional and translational expression patterns associated with immobilized growth of Campylobacter jejuni. Microbiology 152567-577. Schaible, U. E., and S. H. Kaufmann. 2004. Iron and microbial infection. Nut. Rev. Microbiol. 2:946-953. Seliger, S. S., A. R. Mey, A. M. Valle, and S. M. Payne. 2001. The two TonB systems of Vibrio cholerae: redundant and specific functions. Mol. Microbiol. 39:801-812. Singh, P. K., M. R. Parsek, E. P. Greenberg, and M. J. Welsh. 2002. A component of innate immunity prevents bacterial biofilm development. Nature 417552-555.
CHAPTER 33
IRON METABOLISM. TRANSPORT. AND REGULATION
Smith, R. L., and M. E. Maguire. 1998. Microbial magnesium transport: unusual transporters searching for identity. Mol. Microbiol. 28:217-226. Sriyosachati, S., and C. D. Cox. 1986. Siderophore-mediated iron acquisition from transferrin by Pseudomonas aeruginosa. Infect. Immun. 52:885-891. St Maurice, M., N. Cremades, M. A. Croxen, G. Sisson, J. Sancho, and P. S. Hoffman. 2007. F1avodoxin:quinone reductase (FqrB): a redox partner of pyruvate:ferredoxin oxidoreductase that reversibly couples pyruvate oxidation to NADPH production in Helicobacter pylori and Campylobacter jejuni. J. Bacteriol. 189: 4764-4773. Stead, D., and S. F. Park. 2000. Roles of Fe superoxide dismutase and catalase in resistance of Campylobacter coli to freeze-thaw stress. Appl. Environ. Microbiol. 66:3 110-31 12. Stintzi, A., D. Marlow, K. Palyada, H. Naikare, R. Panciera, L. Whitworth, and C. Clarke. 2005. Use of genome-wide expression profiling and mutagenesis to study the intestinal lifestyle of Campylobacter jejuni. Infect. Immun. 73:1797-1810. Stojiljkovic, I., M. Cobeljic, and K. Hantke. 1993. Escherichia coli K-12 ferrous iron uptake mutants are impaired in their ability to colonize the mouse intestine. FEMS Microbiol. Lett. 108:lll115. Stojiljkovic, I., J. Larson, V. Hwa, S. Anic, and M. So. 1996. HmbR outer membrane receptors of pathogenic Neisseria spp.: iron- regulated, hemoglobin-binding proteins with a high level of primary structure conservation. J. Bacteriol. 178:4670-4678. Storz, G., and J. A. Imlay. 1999. Oxidative stress. Curr. Opin. Microbiol. 2:188-194. Suits, M. D., G. P. Pal, K. Nakatsu, A. Matte, M. Cygler, and Z. Jia. 2005. Identification of an Escherichia coli 0157:H7 heme oxygenase with tandem functional repeats. Proc. Natl. Acad. Sci. USA 102~16955-16960. Tannas, T., N. Dekker, G. Bukholm, J. J. E. Bijlsma, and B. J. Appelmelk. 2001. Phase variation in the Helicobacter pylori phospholipase A gene and its role in acid adaptation. Infect. Immun. 69:7334-7340. Tom-Yew, S. A., D. T. Cui, E. G. Bekker, and M. E. Murphy. 2005. Anion-independent iron coordination by the Campylobacter jejuni ferric binding protein. J. Biol. Chem. 280:92839290. Tonello, F., W. G. Dundon, B. Satin, M. Molinari, G. Tognon, G. Grandi, G. Del Giudice, R. Rappuoli, and C. Montecucco. 1999. The Helicobacter pylori neutrophil-activating protein is an iron-binding protein with dodecameric structure. Mol. Microbiol. 34:238-246. Torres, A. G., and S. M. Payne. 1997. Haem iron-transport system in enterohaemorrhagic Escherichia coli 0157:H7. Mol. Microbiol. 23:825-833. Touati, D. 2000. Iron and oxidative stress in bacteria. Arch. Biochem. Biophys. 373:l-6. Tsolis, R. M., A. J. Baumler, F. Heffron, and I. Stojiljkovic. 1996. Contribution of TonB- and Feo-mediated iron uptake to growth of Salmonella typhimurium in the mouse. Infect. Immun. 64: 4549-4556. Tuanyok, A., H. S. Kim, W. C. Nierman, Y. Yu, J. Dunbar, R. A. Moore, P. Baker, M. Tom, J. M. Ling, and D. E. Woods. 2005. Genome-wide expression analysis of iron regulation in Burkholderia pseudomallei and Burkholderia mallei using DNA microarrays. FEMS Microbiol. Lett. 252:327-335. van Vliet, A. H., J. D. Rock, L. N. Madeleine, and J. M. Ketley. 2000. The iron-responsive regulator Fur of Campylobacter jejuni is expressed from two separate promoters. FEMS Microbiol. Lett. 188:115-118. van Vliet, A. H., K. G. Wooldridge, and J. M. Ketley. 1998. Ironresponsive gene regulation in a Campylobacter jejuni fur mutant. J. Bacteriol. 180:5291-5298.
609
van Vliet, A. H. M., M. L. A. Baillon, C. W. Penn, and J. M. Ketley. 1999. Campylobacter jejuni contains two Fur homologs: characterization of iron-responsive regulation of peroxide stress defense genes by the PerR repressor. J. Bacteriol. 181:63716376. van Vliet, A. H. M., M. L. A. Baillon, C. W. Penn, and J. M. Ketley. 2001. The iron-induced ferredoxin FdxA of Campylobacter jejuni is involved in aerotolerance. FEMS Microbiol. Lett. 196~189-193. van Vliet, A. H. M., and J. M. Ketley. 2001. Pathogenesis of enteric Campylobacter infection. J. Appl. Microbiol. 90:45S-56S. van Vliet, A. H. M., J. M. Ketley, S. F. Park, and C. W. Penn. 2002. The role of iron in Campylobacter gene regulation, metabolism and oxidative stress defense. FEMS Microbiol. Rev. 26: 173-186. Velayudhan, J., N. J. Hughes, A. A. McColm, J. Bagshaw, C. L. Clayton, S. C. Andrews, and D. J. Kelly. 2000. Iron acquisition and virulence in Helicobacter pylori: a major role for FeoB, a high-affinity ferrous iron transporter. Mol. Microbiol. 37:274286. Wai, S. N., K. Nakayama, K. Umene, T. Moriya, and K. Amako. 1996. Construction of a ferritin-deficient mutant of Campylobacter jejuni: contribution of ferritin to iron storage and protection against oxidative stress. Mol. Microbiol. 20:1127-1134. Wai, S. N., T. Takata, A. Takade, N. Hamasaki, and K. Amako. 1995. Purification and characterization of ferritin from Campylobacter jejuni. Arch. Microbiol. 164:l-6. Waidner, B., S. Greiner, S. Odenbreit, H. Kavermann, J. Velayudhan, F. Stahler, J. Guhl, E. Bisse, A. H. M. van Vliet, S. C. Andrews, J. G. Kusters, D. J. Kelly, R. Haas, M. Kist, and S. Bereswill. 2002. Essential role of ferritin Pfr in Helicobacter pylori iron metabolism and gastric colonization. Infect. Immun. 70: 3923-3929. Wainwright, L. M., K. T. Elvers, S. F. Park, and R. K. Poole. 2005. A truncated haemoglobin implicated in oxygen metabolism by the microaerophilic food-borne pathogen Campylobacter jejuni. MicroGiology 151:4079-409 1. Wainwright, S. A., J. Velayudhan, and D. J. Kelly. 2001. The magnesium transporter CorA catalyses low-affinity iron uptake in Helicobacter pylori. Int. J. Med. Microbiol. 291:lOO. Wang, G., P. Alamuri, and R. J. Maier. 2006. The diverse antioxidant systems of Helicobacter pylori. Mol. Microbiol. 61:847860. Ward, P. P., and 0. M. Conneely. 2004. Lactoferrin: role in iron homeostasis and host defense against microbial infection. Biometals 17:203-208. Whitby, P. W., T. M. Vanwagoner, T. W. Seale, D. J. Morton, and T. L. Stull. 2006. Transcriptional profile of Haemophilus influenzae: effects of iron and heme. J. Bacteriol. 18856405645. Woodall, C. A., M. A. Jones, P. A. Barrow, J. Hinds, G. L. Marsden, D. J. Kelly, N. Dorrell, B. W. Wren, and D. J. Maskell. 2005. Campylobacter jejuni gene expression in the chick cecum: evidence for adaptation to a low-oxygen environment. Infect. lmmun. 73 5278-5285. Wooldridge, K. G., and A. H. M. van Vliet. 2005. Iron transport and regulation, p. 293-3 10. In J. M. Ketley and M. Konkel (ed.), Campylobacter: Molecular and Cellular Biology. Horizon Scientific Press, Norfolk, United Kingdom. Wooldridge, K. G., and P. H. Williams. 1993. Iron uptake mechanisms of pathogenic bacteria. FEMS Microbiol. Rev. 12:325348. Wooldridge, K. G., P. H. Williams, and J. M. Ketley. 1994. Ironresponsive genetic regulation in Campylobacter jejuni: cloning and characterization of a fur homolog. J. Bacteriol. 17658525856.
610
STINTZI ET AL.
Wyckoff, E. E., M. Schmitt, A. Wilks, and S . M. Payne. 2004. HutZ is required for efficient heme utilization in Vibrio cholerae. J. Bacteriol. 186:4142-4151. Yamasaki, M., S. Igimi, Y. Katayama, S. Yamamoto, and F. Amano. 2004. Identification of an oxidative stress-sensitive protein from Campylobacter jejuni, homologous to rubredoxin oxidoreductase/rubrerythrin.FEMS Microbiol. Lett. 235 57-63. Yuki, N., and M. Koga. 2006. Bacterial infections in GuillainBarre and Fisher syndromes. Cur. Opin. Neurol. 19:451-457.
Zheng, M., B. Doan, T. D. Schneider, and G. Storz. 1999. OxyR and SoxRS regulation of fur. J. Bacteriol. 181:4639-4643. Zhou, D., L. Qin, Y. Han, J. Qiu, Z. Chen, B. Li, Y. Song, J. Wang, Z. Guo, J. Zhai, Z. Du, X. Wang, and R. Yang. 2006. Global analysis of iron assimilation and fur regulation in Yersinia pestis. FEMS Microbiol. Lett. 258:9-17. Zhu, W., A. Wilks, and I. Stojiljkovic. 2000. Degradation of heme in gram-negative bacteria: the product of the hem0 gene of Neisseriae is a heme oxygenase. J. Bacteriol. 182:6783-6790.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
Chapter 34
Regulation of Genes in Campylobacter jejuni MARCM. S. M. WOSTEN,ANDRIESVAN MOURIK,AND Jos P. M.
VAN
PUTTEN
that shows strong similarities among all members of a distinct family (Ramos et al., 2005).
Transcription of DNA in bacteria is catalyzed by a single multisubunit RNA polymerase. This enzyme is recruited to a specific sequence upstream of a gene (the promoter) by a DNA-binding protein, the sigma factor. The sigma factor and the RNA polymerase together constitute the basal transcription apparatus that is required for the baseline transcription of all genes. Besides sigma factors, other specific transcription factors exist that act as switches to enable variable transcription of distinct sets of genes (regulons), often in response to specific environmental stimuli. Examples of regulons are sets of genes encoding components of pathways needed for the utilization of distinct environmental metabolites or components that constitute a developmental pathway. The specific transcription factors can be distinguished into two functional types: those that negatively regulate gene transcription (repressors), and those that positively regulate gene expression (activators). Some transcription factors can act as both activator and repressor, depending on the location of the DNA sequence that is recognized. Transcription factors are grouped on the basis of the presence of conserved motifs and their modes of DNA binding. Structural analyses have revealed that the helix-turn-helix (HTH) signature is the most recurrent DNA-binding motif, present in almost 95% of transcriptional factors in prokaryotes (Aravind et al., 2005; Ramos et al., 2005). Generally, HTH proteins bind as dimers to symmetrical DNA sequences in which each monomer recognizes a half-site, although variations have been found and some proteins recognize direct repeats (Molina-Henares et al., 2006). On the basis of sequence comparisons and phylogenetic, structural, and functional analyses focused on DNA-binding domains, different families of HTH transcriptional regulators have been identified. The HTH structure is usually the only active motif
GENE REGULATION IN CAMPYLOBACTER JEJUNI Genome-wide analysis of several Campylobacter jejuni strains indicates that this species contains between approximately 1,650 and 1,800 genes (Hofreuter et al., 2006; Parker et al., 2006; Parkhill et al., 2000). Transcription profiling on C. jejuni maintained under a variety of growth conditions indicates considerable variation in gene expression (Andersen et al., 2005; Carrillo et al., 2004; Elvers et al., 2005; Gaynor et al., 2005; Hendrixson, 2006; Holmes et al., 2005; MacKichan et al., 2004; Moen et al., 2005; Palyada et al., 2004; Raphael et al., 2005; Stintzi, 2003; Stintzi et al., 2005; Woodall et al., 2005; Wosten et al., 2006), suggesting that gene regulation is essential for the lifestyle of C. jejuni. A search for putative promoters in the C. jejuni genome suggests that the genes are regulated by approximately 650 to 750 promoter elements (Carrillo et al., 2004; Petersen et al., 2003). Computational analysis of sequence similarities with transcriptional regulators in other bacterial species and the presence of conserved DNA binding motifs predict the presence in the C. jejuni genome of at least 37 putative transcription regulators (Table 1). The C. jejuni genome carries only three sigma factors: RpoD, FliA, and RpoN; the remaining 34 regulators belong to the specific transcription factors. In the absence of other major mechanisms of gene regulation, this indicates that all C. jejuni biology, including bacterial replication, adaptation to environments such as cold, nutrient-poor water, and the nutrient-rich habitat of the intestine,
Marc M. S. M. Wosten, Andries van Mourik, and Jos P. M. van Putten University, Yalelaan 1, 3584 CL Utrecht, The Netherlands.
611
Department of Infectious Diseases and Immunology, Utrecht
612
WOSTEN ET AL.
Table 1. C. jejuni transcription factors Transcription factor family rr-factor
NtrC OmpR
Fur
Gene no. CjlOOl Cj0061 Cj0670 Cj1024 Cj1261 Cj0355 Cj0890 Cj 1223 Cj1227 Cj1491 Cj0643 Cjl505 Cj1608 Cj0322 Cj0400
CrpIFnr TetR
Cj466 Cj0368
DeoR LysR
Cj0571 CjlOOO Cjl507 Cj0757 Cj1230 Cj1563 Cj1546 Cj1556 Cj1042 Cj1410 Cj0258 Cjl56l Cj0480 Cj0883 Cj1103 cj394 Cj0422 Cj1036 Cj1387 Cj1533
HrcA MerR MarR AraC/XylS ArsR IcIR Rrf2 CsrA Unknown
Annotation
Reference(s)
phosR dccR
Petersen et al. (2003) Wosten et al. (1998b) Carrillo et al. (2004) Wosten et al. (2004) Bras ct al. (1999) Muller et al. (2007) Wostcn ct al. (2006) MacKichan ct al. (2004)
cbrR
Raphael ct al. (2005)
perR
van Vliet et al. (2002) Holmes et al. (2005); Palyada et al. (2004) Elvers et al. (2005) Lin et al. (2005a); Stintzi et al. (2005)
?OD
fliA rpoN flgR racR
fur
nssR cmeR
modE hrcA hspR
Studholme and Pau (2003) Stintzi et al. (2005) Andersen et al. (2005)
csrA
and bacterial pathogenicity, are largely determined by the controlled activity of a limited number (-2% of the total) of C. jejuni proteins.
C. JEJUNI SIGMA FACTORS RpoD
The main essential sigma factor regulating almost all C. jejuni promoters is RpoD. This protein belongs to the ' a family, which is considered to comprise the primary or housekeeping sigma factors. RpoD recognizes two conserved sequence motifs, known as the -10 and -35 regions, named after their relative positions to transcriptional starting point of +l.Binding of RpoD to the -10 and -35
promoter elements requires two protein regions in the sigma subunit, known as the 2.4 and 4.2 segment (Dombroski et al., 1992). The first segment (2.4) is supposed to confer the contact of the protein with - 10 hexamer region. This segment is very well conserved between Escherichia coli, Bacillus subtilis, and C. jejuni and recognizes the sequence 5'-TATAAT3'. The second segment of RpoD (segment 4.2) recognizes the -35 region of the promoter element. This segment of the protein is much less conserved between E. coli and Campylobacter (Wosten et al., 1998a), which may explain the absence of the typical E. coli -35 nucleotide consensus sequence (TTGACA) in C. jejuni promoters (Wosten et al., 1998b; Petersen et al., 2003). Alignment of the intergenic regions in front of 175 high-confidence genes revealed that the -35 box of the C. jejuni promoters exhibits a strong periodic signal (Petersen et al., 2003). In an effort to identify conserved nucleotide sequences in C. jejuni promoter regions that are recognized by DNA-binding proteins, we aligned 660 putative promoter regions that were all located in intergenic regions and within the first 90-bp region upstream of the predicted start codons with the program Meme (http://meme.nbcr.net/meme/meme. html). By this approach, we identified common promoter elements for 300 of the putative promoters. The identified elements (Fig. 1)were similar to those identified by Petersen et al. (2003). Our comprehensive analysis indicates the consensus sequence of - 10 region of C. jejuni as TAtAAT, with the thymidine at the third position being most variably present. In more than 60% of the investigated putative promoters, this consensus sequence is preceded by a -16 region with the sequence TTTTtGx. In gram-positive bacteria, the -16 region appears to be a basic element of a large portion of the promoters in addition to the -10 and -35 hexamer elements, and the presence of a -16 region may benefit transcription from mainly weak promoters (Voskuil and Chambliss, 1998). The function of the seemingly conserved -16 region of C. jejuni still remains to be defined. Our analysis of C. jejuni promoter regions did not reveal a typical conserved -35 region, but it did indicate the presence an AT-rich region around position -35. In other species, promoters that are activated by specific transcription factors often display a less wellconserved RpoD -35 consensus sequence. We cannot exclude the notion that the presence of this type of promoters in our analysis masked the identification of a conserved -35 hexamer element for C. jejuni. At this point, it should be noted that several E. coli promoters lacking an identifiable -35 region but containing a -16 TG motif are recognized by the conserved segment 3.0 of the sigma subunit (Young
CHAPTER 34
-
REGULATION OF GENES IN C. JETUNI
613
Figure 1. DNA sequence logos to indicate the sequence conservation of the C. jejuni sigma factors RpoD (A), RpoN (B) and FliA (C). Sequence logos were made of 300 RpoD, 17 RpoN, and 10 FliA promoter sequences with the program Weblogo (http: //www.bio.cam.ac.uk/seqlogo/logo.cgi). Sequence conservation is measured in bits, and the degree of conservation of the different nucleotides is indicated by the height of a stack of letters. Numbers under the nucleotides indicate the relative distance to the transcription start site.
et al., 2002). Thus, the presence of a -35 hexamer element is not an absolute requirement for gene transcription. FliA Like RpoD, the alternative sigma factor FliA (a28) belongs to the a 7 0 family of sigma factors. On the basis of transcription profiling of wild-type and FliA-deficient C. jejuni, FliA was found to regulate the activity of at least 10 promoters that direct the transcription of 14 different genes (Carrillo et al., 2004). These genes encode proteins involved in the assembly of the flagellar apparatus, proteins involved in the glycosylation of the major flagellin subunits (Logan et al., 2002), and virulence-associated pro-
teins that are secreted through the flagellum, namely FspA and Cj0977 (Goon et al., 2006; Poly et al., 2007). Consistent with the proposed function of the regulated genes, FliA mutants are nonmotile but still possess a flagellar hook structure. The transcription of fliA is, like that of RpoN, under the control of RpoD. However, the activity of FliA is tightly regulated. FlA is likely kept inactive by the putative antisigma factor FlgM until this protein is secreted through the completed basal-body hook structure that forms the platform of filament formation, reminiscent of the regulation of flagella assembly in other bacterial species. FlgM transcription is thought to be regulated by both RpoN and FliA (Wosten et al., 2004). The consensus sequence that is recognized by FliA is shown in Fig. 1.
614
WOSTEN ET AL.
RpoN In contrast to members of the a 7 0 family, the DNA consensus sequence recognized by RpoN, which belongs to the as4family of sigma factors, is very well conserved throughout the bacterial kingdom (Studholme and Buck, 2000). Genome-wide analysis of C. jejuni that is based on this sequence (TGGCAC-N5-TTGC) indicated the existence of 17 putative RpoN promoters. Transcription profiling confirmed that 15 of these indeed regulate the transcription of the downstream gene or genes. (Carrillo et al., 2004). Comparison of the promoter regions of these genes indicates that the C. jejuni RpoN consensus sequence is larger than the general RpoN consensus sequence and that it prefers an adenine residue instead of a cytosine at position -23 (Fig. 1). The 17 identified RpoN promoters of C. jejuni control the transcription of 23 genes, of which 15 encode proteins that are involved in the assembly of the flagella. Factors regulated by RpoN include components of the basal body, the flagellar hook protein, the filament subunit flagellin Byand the putative antisigma factor FlgM (Carrillo et al., 2004; Wosten et al., 2004). Besides known flagellar genes, RpoN regulates a number of hypothetical genes (Cj0243, Cj0428, Cj1242, and Cj1650), a gene involved in motility but not in assembly (Cj1026) (Sommerlad and Hendrixson, 2007), and a UDP-GlcNAc C6 dehydratase (Cj1293) (Creuzenet, 2004). Consistent with its proposed function, inactivation of RpoN renders the bacteria nonmotile and without flagella. Characteristic for the initiation of transcription of as4-depending genes is that the process depends on at least one other transcription factor that acts as an activator of gene expression.
SPECIFIC TRANSCRIPTION FACTORS Analysis of the C. jejuni genome predicts the presence of approximately 34 transcription factors in addition to the three sigma factors describe above. The specific transcription factors can be classified on the basis of the presence of conserved sequence motifs and homology with related proteins in other species. A description of the more than 15 different transcription factor families in C. jejuni (Table 1)(Ramos et al., 200.5), is given below. Response Regulator Proteins Response regulator proteins are the effector molecules of two-component signal transduction systems that confer the final step in the translation of an
environmental stimulus to altered gene transcription. In general, these systems consist of two proteins, an integral membrane protein, the sensor kinase, and a cytoplasmic protein, the response regulator (Parkinson, 1993). The C-terminal cytoplasmic domain of the sensor protein displays histidine autokinase activity and autophosphorylates itself at a conserved histidine residue upon sensing of a distinct environmental signal (Perraud et al., 1999; Stock et al., 2000). In its simplest form, the phosphorylated protein then serves directly as a phosphoryl donor to the cognate response regulator protein, which usually becomes phosphorylated at a conserved asparate residue on the N-terminal domain (Stock et al., 2000). The response regulator has affinity to distinct DNA sequences present in the promoter region of the genes of a distinct regulon and, depending on its phosphorylation state, activates or represses gene transcription. This ultimately leads to a bacterial phenotype that is optimally adapted to its ecological niche. Analysis of the C. jejuni genome of strain NCTC 11168 predicts the presence of seven putative protein kinase sensor proteins and 10 response regulators. Because a regulon of a two-component system consists of an average of 25 genes (Stock et al., 2000), it is estimated that approximately 20% of the C. jejuni total number of genes is under the control of twocomponent systems. All but one of the response regulators of C. jejuni belong to the OmpR family of DNA binding proteins. The response regulator FlgR represents the NtrC family of transcription factors. OmpR Family of Response Regulators Members of the OmpR family of DNA-binding proteins interact with distinct conserved DNA sequences located in RpoD-dependent promoter regions. The response regulators can act both as a repressor or activator of gene transcription. Of the nine members of this family in C. jejuni (Table l), only five-PhosR, DccR, RacR, Cj1227, and Cj 1491-are located on the chromosome adjacent to their putative cognate sensor kinase. How the remaining orphan response regulators (Cj0355, CbrR, Cj1608, and Cjl50.5) are activated awaits further investigation. Mutants have been obtained for all the response regulators, except for Cj1227, CjO355, and Cjl.505 (Raphael et al., 2005). Although neither a function nor mutants have been described for the Cj0355 response regulator, Cj0355 was able to complement the Helicobucter pylori Cj0355 homolog HP1043, which is required for the normal cell growth (Muller et al., 2007). One of the C. jejuni response regulators that has been studied in more detail is PhosR (Wosten et
CHAPTER 34
al., 2006). PhosR is part of the PhosR/PhosS twocomponent system that regulates the pho regulon of C. jejuni. This systems responds to alterations in the availability of phosphate in the environment. In the presence of low phosphate, PhosS becomes phosphorylated and rapidly transfers its phosphate group to PhosR. This protein binds to three promoter elements that regulate 12 different genes (Fig. 2) (Wosten et al., 2006). Electrophoresis mobility shift assays have demonstrated that PhosR binds to DNA containing the sequence 5'-GTTTCNAAAANGTTTC3'. This so-called pho box is located at the -35 hexamer region of the PhosR regulated promotors. The 12 genes that are regulated by the PhosR/PhosS system encode an alkaline phosphate (PhoA'J), four proteins that are similar to the high-ATP-driven Pi transport system PstSCAB, and seven proteins (including putative ABC transporters) that are all highly conserved among other bacterial species and may function as an unknown uptake system (Wosten et al., 2006). A second relatively well-characterized response regulator is the DccR (diminished capacity to colonize) protein (MacKichan et al., 2004). This regulator protein binds to DNA containing the direct repeat sequence TTCAC-N(6)-TTCAC. This sequence is located in front of six promoter elements and overlaps the putative -35 region. The promoters are predicted to regulate the transcription of eight genes (Fig. 2). The products encoded by the genes are all hypothetical proteins. Inactivation of the DccRS system in strain 81-176 does not affect in vitro growth, but it significantly reduces the colonization of immunocompetent limited flora (I-LF) mice, severe combined immunodeficient limited flora (SCID-LF) mice, and 1-day-old chicken. Disruption of two of the DccSR-regulated genes, Cj200c and Cj1626c, causes a similar attenuated phenotype (MacKichan et al., 2004). On the basis of sequence similarity, three DccR regulated proteins (Cj0606, Cj0607, and Cj0608) may form a type I secretion system as the proteins exhibit homology to HlyD, HlyB, and the TolC protein of the type I secretion systems of E. coli (MacKichan et al., 2004). The putative C. jejuni type I system does not resemble the Sap type I secretion system of Campylobacter fetus (Thompson et al., 1998). The environmental signal that activates the DccRS system is still unknown, although preliminary data indicate that the DccR-regulated promoters are activated when the C. jejuni enters the stationary growth phase (M. M. S. M. Wosten et al., unpublished data). The RacR response regulator is part of the RacRRacS (reduced ability to colonize) two-component signal transduction system-(Bras et al., 1999). The
REGULATION OF GENES IN C. JEJUNI
615
start codon of racS partially overlaps the racR stop codon, suggesting that the genes may be part of one operon. The DNA sequence that is recognized by RacR is unknown. Comparison of protein profiles of a RacR mutant and the parent strain identified at least 11 differentially expressed proteins as members of the racR regulon, including RacR itself. RacR acts as a transcriptional activator of five of these proteins, while two proteins are repressed. The expression of three proteins-RacR, DnaJ, and a cytochrome c peroxidase homolog-responds to changes in growth temperature in a RacR-dependent fashion (Bras et al., 1999). Disruption of RacR results in a reduced growth rate at 42"C, while at 37"C, a slightly lower final cell density is reached compared with the parent strain (Bras et al., 1999). The gene Cj0643 encoding the response regulator designated as CbrR (Campylobacter bile resistance regulator) is not linke2 with a gene &at codes for a cognate sensor kinase (Raphael et al., 2005). The protein lacks an obvious DNA-binding domain but does contain a receiver domain that shows a high degree of similarity with the chemotaxis response regulator CheY that drives flagellar rotation. In addition, the protein has a C-terminal GGDEF domain assumed to be a diguanylate cyclase (Raphael et al., 2005). Inactivation of CbrR renders C. jejuni highly sensitive to bile salts. The mechanism underlying this phenotype is unknown but may be related to the function of CmeR. This transcriptional repressor regulates the multidrug efflux pump CmeABC that confers resistance to antimicrobial agents, including bile salts (Lin et al., 2002). Together, the data suggest that CbrR is not directly involved in gene regulation in C. jejuni, but rather communicates with another response regulator or directly with target proteins. The function of the remaining response regulators (Cj0355, Cj1227, Cj1491, Cjl505c, and Cj1608) that belong to the OmpR family is virtually unknown. The Cj1491 protein is strongly downregulated in vivo in a rabbit ileal loop model (Stintzi et al., 2005), suggesting that this transcription factor might have a role outside the host. Cj1505c has sequence similarity with the Salmonella global regulator SirA (Salmonella invasion regulator), but no functional studies have ceen performed on this protein. Attempts to inactivate the Cj0355 and Cj1227 response regulator genes have failed as yet, suggesting that the corresponding proteins are essential for C. jejuni growth under the conditions used. NtrC Family of Response Regulators The sole response regulator of C. jejuni that does not belong to the OmpR family is the FlgR protein.
616
WOSTEN ET AL.
A
B
5
Phosphate limitation
V DccS
@ j i r/?
Cj0200
phoAcj
Cj0727
Cj0728
Cj0729 Cj0730 Cj073l
Cj0732 Cj0733
Cjl723
Figure 2. Regulons belonging to the (A) FlgS/FlgR, (B) DccS/DccR, and (C) PhosS/PhosR two-component systems. The different two-component systems activate (+) or repress (-) the transcription of the indicated genes and operons. The sizes of the gene arrows correspond to the size of the genes.
CHAPTER 34
FlgR is a member of the NtrC family of response regulators. This group of transcription factors acts in concert with the alternative sigma factor RpoN (Shingler, 1996). Indeed, in C. jejuni, FlgR and RpoN regulate the same promoters and the same sets of genes (Fig. 2) (Hendrixson and DiRita, 2003; Wosten et al., 2004). In other bacterial species, NtrC homologs bind to palindromic or nearly palindromic nucleotide sequences that are located > l o 0 bp upstream of a distinct class of promoters (Kustu et al., 1991). These sequences resemble eukaryotic enhancer elements in that they still facilitate transcription even when moved thousands of nucleotides upstream or downstream of the transcription start site, and regardless of their orientation (Morett and Segovia, 1993). For C. jejuni, an enhancer-like element recognized by FlgR has not been identified. To activate the RpoN-dependent promoters, FlgR needs to be phosphorylated by its cognate sensor FlgS (Wosten et al., 2004). The gene encoding FlgS (Cj0793) is not located adjacent to that encoding FlgR (Cj1024). However, phosphorylation assays with purified recombinant FlgR and FlgS demonstrate that the two protein constitute one two-component system (Wosten et al., 2004). The signal that drives the autophosphorylation of FlgS, and thus the activation of FlgR, is still unknown. Inactivation of both the rpoN and flgR genes results in a nonmotile C. jejuni phenotype that lacks flagella (Jagannathan et al., 2001), as can be expected from the important role of RpoN in the regulation of flagella assembly (see above). Thus, it can be hypothesized that the system responds to a signal that senses needed changes in bacterial motility. Fur Transcription Factor Family Iron is an essential element for bacterial growth. Because the iron bioavailability in an aerobic neutral pH environment (10-l' M) and in mammals M) is much lower than the minimal requirement of bacterial growth ( M), microorganisms contain sophisticated iron acquistion systems. A key factor in the regulation of bacterial iron metabolism is the ferric uptake regulator Fur. This factor acts as a repiessor>f transcription of a distinct subset of genes once complexed to its corepressor Fe(I1). When the intracellular concentration of Fe(I1) exceeds a threshold level, Fur binds Fe(II), and in this active state, the protein binds to a consensus sequence in the DNA, the Fur box. This box is usually located between the -35 and -10 hexamer elements of Fur-repressed promoters. When the intracellular iron concentration becomes low, the Fur repressor dissociates from the DNA, allowing access of RNA polymerase to the
REGULATION OF GENES IN C. TElUhJI
617
promoters that drive the transcription of the ironregulated genes (Andrews et al., 2003; Hantke, 2001). Microarray analyses suggest that the Fur protein of C. jejuni influences the transcription of at least 44 genes involved in iron acquisition and oxidative stress defense (Holmes et al., 2005; Palyada et al., 2004) (Fig. 3). Direct binding of purified Fur to promoter regions has been shown for the elements regulating the p19 (putative iron transporter), c f ~ A (ferricenterobactin receptor), ceuBCDE (ferric-enterobactin uptake permeasee), and chuABCD (heme outer membrane transporter) operons and the Cjl613 gene (encoding a hypothetical protein) (Holmes et al., 2005; Ridley et al., 2006). The consensus sequence recognized by C. jejuni Fur has not been determined experimentally. However, C. jejuni Fur recognizes the E. coli consensus Fur box (Wooldridge et al., 1994), and on the basis of computational analysis of upstream sequences of iron-regulated genes, a C. jejuni consensus Fur-binding site has been proposed (Palyada et al., 2004). Besides influencing the transcription of genes likely involved in iron acquisition and iron homeostasis, Fur also affects genes involved in the oxidative stress defense. Two of these genes, katA (which encodes catalase) and ahpC (which encodes alkyl hydroperoxide reductase), are also regulated by peroxide stress regulator PerR (Fig. 3, van Vliet et al., 1999). Fur might also influence the transcription of perR (Holmes et al., 2005). PerR belongs to the family of the Fur proteins that are all activated in the presence of ferrous iron (van Vliet et al., 2002). For C. jejuni, the PerR protein is assumed to regulate the PerR regulon that comprises, among others, the genes katA, ahpC, and probably also sodB (encoding ironcontaining superoxide dismutase) (Holmes et al., 2005). The apparent linkage of the regulation of iron homeostasis and oxidative stress defense is not unexpected, given that iron-associated oxidative stress may cause major damage to the cell via the generation of reactive oxygen species such as peroxides and hydroxyl radicals (Ratledge and Dover, 2000). Crp/FNr Family Involved in Regulation of Nitrosative Stress Responses In analogy to the oxidative stress response, C. jejuni also exhibits a nitrosative stress (Nss) response (Elvers et al., 2004). In general, nitrosative stress responses serve to protect the bacteria against the toxic effects of nitric oxide and reactive nitrogen species (Poole, 2005). In C. jejuni, the constitutively expressed NfrA protein plays a role in protection against nitrosative stress (Pittman et al., 2007). In ad-
618
WOSTEN ET AL.
Cj1386
katA I
\ I
\
\
I*(-)?
Cjl658
PI9
I
b
%-)* >
ceuB
::
I I
:v
I
+(-)p
Cj0176
CjO175
Cj1660
-
ceuE
Cj0173
> b
chuB
Cj0178 Cj1661
-
CjOl74 chuC
-_ = +
. I
CjOl77
ceuD
-1
-
chuA
(-)P,
ceuC
exbB?
Cjl662
exbBl
chuD
ton57
>
b Cj1663 . Cj1664. Cj1665.
Figure 3. Regulons belonging to Fur and PerR repressor proteins. The response regulators FurR and PerR repress (-) the transcription of the indicated genes and operons. Solid arrows point to genes for which binding of FurR to the promoter regions has been confirmed. Dashed arrows point to genes for direct binding of the transcription factor that still need to be demonstrated. The sizes of the gene arrows correspond to the size of the genes.
dition, the inducible single-domain protein Cgb is involved in the scavenging and detoxification process (Elvers et al., 2004). Expression studies identified the NssR (Cj0466) protein as a regulator of Cgb expression. NssR belongs to the family of Crp-Fnr superfamily of transcriptional regulators. Besides Cgb, NssR also influences the transcription of at least several other genes of unknown function (Fig. 4) (Elvers et al., 2005). The promoter regions of the genes comprising the NssR regulon (cgb, CjO465c, VjOS30, and Cj0761) contain an inverted repeat sequence (TTAAC-N(4)-GTTAA) centered around position -41 relative to the putative transcriptional start site.
This consensus recognition motif for NssR and its location in the genome are consistent with the architecture of a class I1 Fnr-dependent promoter (Korner et al., 2003). Whether NssR directly senses nitrite and nitric oxide, or whether this is conferred by a separate signaling molecule is still unknown. NssR appears to be the sole representative of the Crp-Fnr superfamily of transcription factors in C. jejuni. TetR Family of Transcriptional Repressors One of the systems that contributes to Campylobacter resistance to broad-spectrum antibiotics is
CHAPTER 34
REGULATION OF GENES IN C. TElUNI
619
Other Transcription Factors of C. jejuni
Cj0465
Cj0761
Cj0830
B
cmeA
cmeB
>
b cmeC
Figure 4. Regulons belonging to (A) NssR and (B) CmeR transcription factors. The different two-component systems activate (+) or repress (-) the transcription of the indicated genes and operons. The sizes of the gene arrows correspond to the size of the genes.
the CmeABC efflux pump (Lin et al., 2002; Pumbwe and Piddock, 2002). This membrane channel is composed of three proteins that are encoded by the cmeABC operon. The operon is located directly downstream of the gene cmeR. This gene encodes a transcriptional repressor that interacts with the cmeABC promoter through binding to a unique inverted repeat (5'-TGTAATA-3') (Lin et al., 2005a) (Fig. 4). Both in architecture and function, CmeR of C. jejuni is the only C. jejuni member of the TetR family of transcription repressors. This family of repressors controls gene products involved in multidrug resistance, enzymes implicated in different catabolic pathways, biosynthesis of antibiotics, osmotic stress, and pathogenicity of both gram-negative and gram-positive bacteria (Ramos et al., 2005). The induction of a multidrug efflux pump can be induced by direct interaction of the substrates of the pumps with the repressor molecules. This interaction may cause a conformational change, preventing association of CmeR with the DNA, and consequently derepression of cmeABC transcription. For C. jejuni, the substrate that induces de-repression of the CmeABC efflux pump is not known. On the bases of the inability of CmeR mutants to colonize the animal intestinal tract and the finding that CmeABC is essential for C. jejuni bile resistance, it has been suggested that bile salts may be one of the substrates that activate this system in this species (Lin et al., 2005b).
In addition to the three sigma factors and 14 regulatory proteins describe above, C. jejuni is predicted to contain 20 additional transcription regulators. These factors are discussed briefly because detailed structural and functional knowledge with respect to gene regulation largely awaits future studies. On the basis of sequence characteristics, C. jejuni is predicted to contain one homolog of the DeoR family of transcription factors. The corresponding gene, Cj0.571, is the only transcription factor that is consistently upregulated after infection of a rabbit ileal loop model with C. jejuni (Stintzi et al., 2005). In most cases, DeoR homologs act as transcriptional repressors in sugar metabolism (Engels and Wendisch, 2007). However, the DeoR homolog IgeR has recently been reported to regulate the cytolethal distending toxin of Salmonella enterica serovar Typhi (Haghjoo and GalLn, 2007). In rabbit ileal loop competition assays, inactivation of C. jejuni Cj0571 did not alter bacterial colonization (Stintzi et al., 2005). The signals that activate Cj0571 and the regulon that is influenced by this transcription factor remain to be determined. Campylobacter is predicted to produce two transcription factors belonging to the LysR family of proteins, namely CjlOOO and Cj1507. The LysR family is composed of similar-sized autoregulatory transcriptional regulators with extremely diverse functions (Schell, 1993). To our knowledge, the CjlOOO gene has not been investigated in Campylobacter, and also the function of its homologs in other bacterial species remains to be defined. The C. jejuni gene Cjl.507 likely codes for a molybdate-responsive transcription factor ModE (Studholme and Pau, 2003). In proteobacteria, ModE regulates molybdenum metabolism and homeostasis. In Escherichia coli, ModE is a transcriptional repressor preventing the transcription of the ModABCD operon in the presence of high concentrations of molybdenum. The transition metal molybdenum is essential for life and is found in the catalytic center of more than 30 enzymes involved in the nitrogen, carbon, and sulfur cycles. On the basis of the location of the helix-turn-helix motif in Campylobacter ModE, the protein might be a transcriptional activator rather than a repressor (Studholme and Pau, 2003). Intriguingly, the Cjl507c gene overlaps fdhD (Cj1508c). This gene is predicted to encode a molybdo-enzyme formate dehydrogenase. The overlap in genes may indicate a functional link between the regulatory protein and molybdenum metabolism (Studholme and Pau, 2003). The putative C. jejuni Cj1230 and Cj1563 proteins exhibit similarity with the MerR family of tran-
620
WBSTEN ET AL.
scription factors. This family of proteins comprises metal ion responsive activators (Hobman et al., 2005). Cj1230 codes for the heat shock protein HspR. In other bacterial species, HspR regulates the DnaK operon (Narberhaus, 1999). In H. pylori, HspR represses the transcription of the groESL, hrcAgrpE-dnaK, and cbpA-hspR-orf operons by binding on a conserved HAIR element (Spohn et al., 2004). In Campylobacter, HspR also represses the groELS and hrcA-grpE-dnaK operons as well as a few hypothetical proteins (Andersen et al., 2005). Surprisingly, disruption of HspR also causes a downregulation of some flagellar genes, resulting in less motile bacteria. This may explain why this mutant shows reduced bacterial colonization in the rabbit ileum loop model (Andersen et al., 2005; Stintzi et al., 2005). The signals that activate C. jejuni HspR are unknown. Similarly, no data exist with respect to a second C. jejuni member of the MerR family, Cj1563. Campylobacter also produces a homolog of the HcrA protein (Cj0757). Members of the HcrA transcription factor family also negatively regulate the transcription of heat shock genes including groESL and dnaK (Narberhaus, 1999). HcrA proteins recognize the conserved consensus sequence TTAGCACTC-N9-GAGTGCTAA. This sequence, which also known as the CIRCE element, is located upstream of the groESL and dnaK genes in a variety of bacterial species (Narberhaus, 1999). In H. pylori, HcrA has been demonstrated to influence the transcription of the groESL and dnaK operons, but its activity depends on the presence of HspR (Spohn et al., 2004). This finding, combined with the presence of the CIRCE motif in the promoter region in front of the groELS operon of Campylobacter, suggests that Campylobacter HrcA may have a similar function. In contrast to inactivation of HspR (see above), disruption of HcrA does not alter C. jejuni colonization in the rabbit ileal loop model (Stintzi et al., 2005). The MarR family of prokaryotic transcriptional regulators includes proteins critical for control of virulence factor production, bacterial responses to antibiotics and oxidative stresses, and catabolism of environmental aromatic compounds (Wilkinson and Grove, 2006). The majority of MarR homologs are transcriptional repressors that are autoregulated (Wilkinson and Grove, 2004). Sequence analysis suggests that the Campylobacter proteins Cj1.546 and Cj1556 belong to this transcription factor family, but functional studies are lacking. Homology searches indicate that the C. jejuni proteins Cj1042 and Cj1410 belong to the AraC/ XylS family of transcription factors. Members of this family are transcription activators often involved in
carbon metabolism and responses to environmental stress (Egan, 2002). Transcription analysis suggests that Cj1410 may be induced under conditions of increased oxygen concentration (Moen et al., 2005). The Cj1042 protein has thus far not been investigated. The ArsR family of transcription factors represses the expression of operons linked to stressinducing concentrations of di- and multivalent heavy metal ions (Busenlehner et al., 2003). These repressor proteins bind in most cases to two imperfect inverted repeats in the absence of certain metals such as zinc or cadmium. Metal binding allosterically negatively regulates DNA binding and de-represses transcription (Busenlehner et al., 2003). The C. jejuni proteins CjO258 and Cj1561 appear to belong to this family of transcription repressors. No further information is available on the function of these C. jejuni proteins, except that it has been possible to construct a Cj0258-deficient C. jejuni strain (http://m. lshtm.ac.uk/pmbu/crf/wren-mutants.htm). This indicates that this protein is not essential for C. jejuni survival. The Campylobacter Cj0480 protein appears to be a member of the IcIR family of transcription factors. This family comprises regulators that act as repressors, activators, or both. Members of the IcIR family control genes whose products are involved in the glyoxylate shunt in Enterobacteriaceae, multidrug resistance, degradation of aromatics, inactivation of quorum-sensing signals, and plant pathogenicity and sporulation (Molina-Henares et al., 2006). The function and activation of the C. jejuni IcIR homolog await future investigation. Campylobacter also encodes one protein (Cj0883) that is a like member of the Rrf2 family of regulators. In other bacterial species, these proteins are repressors of genes involved in nitrite, nitric oxide, or iron metabolism (Peres and Harwood, 2006). Inactivation of Cj08 83 causes reduced colonization of the ceca of chicken, although the notion that this is caused by an effect on the downstream located flhA gene (Hendrixson and DiRita, 2004) cannot be excluded. The carbon storage regulator A (CsrA) represses the expression of genes involved in glycogen metabolism, gluconeogenesis, glycolysis, and cell motility. CsrA destabilizes target mRNAs by binding in the vicinity of the ribosome binding site (Baker et al., 2007). The C. jejuni gene Cj1103 likely codes for a CsrA homolog. In the rabbit ileal loop model, csrA mRNA levels of C. jejuni are upregulated consistent with the downregulation of the genes encoding enzymes involved in carbohydrate metabolism in this environment (Stintzi et al., 2005). This suggests that
CHAPTER 34
C. jejuni CsrA may also act as a repressor of carbohydrate metabolism. Apart from various (putative) transcription factors mentioned, C. jejuni has several proteins (Cj0394, Cj0422, Cj1036, Cj1387, and Cj1.533) that possess a HTH domain, but that cannot be linked to a distinct transcription regulator family. Although the HTH domain is primarily found in transcription factors, it can be found in proteins involved DNA repair and replication, RNA metabolism, and proteinprotein interactions in diverse signaling contexts (Aravind et al., 2005). Classification of these c. jejuni proteins as transcription factors awaits further functional analysis of these proteins.
OTHER FACTORS INFLUENCING C. JEJUNI GENE REGULATION Besides transcription factor-mediated regulation of gene transcription, prokaryotes have evolved a number of additional mechanisms to control gene expression. Although data with respect to the role of small noncoding RNAs (Vogel and Wagner, 2007) or DNA methylation (Heusipp et al., 2007) are lacking, slipped-strand mispairing of nucleotides and DNA rearrangements influences gene expression in C. jejuni. Slipped-strand mispairing typically involves the incidental addition or deletion during DNA replication of one or more repetitive units in a homopolymeric stretch of nucleotides. Genome-wide analyses indicates that C. jejuni NCTC 11168 has 23 regions that contain variable poly-GC tracts and at least another 9 regions with potentially variable homopolymeric tracts (Parkhill et al., 2000). Most of these repetitive sequences are located in genes responsible for lipooligosaccharide biosynthesis, capsule biosynthesis, and flagellar modification. In most cases, slippedstrand mispairing will influence the translation of genes into proteins, but when changes occur in the promoter region or in regulatory genes, the event may cause alterations in gene transcription, as evidenced by the mispairing in the flgR gene (Hendrixson, 2006). This results in variable expression of the transcriptional regulator FlgR, and consequently alterations in (flagellar) gene transcription. The biological significance of this type of regulation of flagella synthesis remains to be determined. Genome rearrangements might be an important characteristic of the lifestyle of Campylobacter (Hanninen et al., 1999). Interstrain genetic exchange and intragenomic alterations occur in vivo during C. jejuni infection (Boer et al., 2002). The expression of phase-variable genes is to some extent under environmental control. Iron starvation, for example, in-
REGULATION OF GENES IN c. iEruivr
621
creases the frequency of antigenic and phase variation of Neisseria gonorrhoeae pili and may be correlated with a general change in the rates of DNA recombination and DNA repair (Serkin and Seifert, 2000). Comparative genomic analysis of C. jejuni revealed the existence of seven major plasticity regions (PR), three of which (PR4, PR5, and PR6) contain genes involved in the production and modification of antigenic surface structures. These regions are hypervariable, indicating that antigenic diversity is an important characteristic of the lifestyle of Campylobacter and that animal hosts may be a driving force behind the emergence of new antigenic variants (Pearson et al., 2003). Another mechanism of C. jejuni gene regulation may involve direct modulation of the function of the RNA polymerase. A C. jejuni spoT mutant that has a defect in the stringent response exhibits a strong reduction in the “natural” resistance of C. jejuni to rifampicin (Gaynor et al., 2005). In E. coli, the mediator of the stringent response (ppGpp) binds to the RNA polymerase, close to the rifampicin binding site. The binding of ppGpp alters the promoter specificity, transcription initiation, and elongation (Chatterji and Ojha, 2001). The altered resistance to rifampicin in the C. jejuni SPOTmutant may indicate that C. jejuni ppGpp also interacts with the RNA polymerase. Whether this binding contributes to the altered gene regulation observed during the C. jejuni stringent response (Gaynor et al., 2005) awaits future investigation.
CONCLUDING REMARKS Essentially all of the biological functions encoded in the DNA depend on the interaction of proteins with specific DNA sequences. DNA-binding proteins dictate the correct regulation of gene expression, so that the optimal amount and type of proteins are produced in response to specific internal and external stimuli. Insight in transcriptional regulatory networks is crucial for understanding fundamental processes such as growth control, cell-cycle progression, and development, as well as understanding differentiated cellular functions such as secretion and cell-cell communication. Bacteria are commonly viewed as living independently of each other; however, in their native habitat, most bacterial species colonize biotic and abiotic surfaces in communities through coordinated population behavior. Therefore, understanding how the regulons function together as larger units is important to address the responsive behavior for both single organisms and the population as a whole.
622
WOSTEN ET AL.
Transcription profiling of C. jejuni kept under a variety of environmental conditions indicates that the pathogen has a complex transcription regulation network that fits the enormous adaptation potential of this pathogen. Although considerable progress has been made, the knowledge of the mechanisms that control C. jejuni gene regulation is still fragmentary. The major future challenge is to relate the observed dynamics in gene expression to the function of the numerous transcription factors and other regulatory pathways of C. jejuni. Unraveling the C. jejuni transcription regulation network will provide insight into the basis of complex bacterial behavior, which in turn opens avenues that may permit us to manipulate it to our benefit. Acknowledgment. Our research on Campylobacter jejuni has been supported by grants from the Dutch Organization of Scientific Research.
REFERENCES Andersen, M. T., L. Brondsted, B. M. Pearson, F. Mulholland, M. Parker, C. Pin, J. M. Wells, and H. Ingmer. 2005. Diverse roles for HspR in Campylobacter jejuni revealed by the proteome, transcriptome and phenotypic characterization of an hspR mutant. Microbiology 151:905-915. Andrews, S. C., A. K. Robinson, and F. Rodriguez-Quinones. 2003. Bacterial iron homeostasis. FEMSMicrobiol. Rev. 27:215237. Aravind, L., V. Anantharaman, S. Balaji, M. M. Babn, and L. M. Iyer. 2005. The many faces of the helix-turn-helix domain: transcription regulation and beyond. FEMS Microbiol. Rev. 29:23 1262. Baker, C. S., L. A. Eory, H. Yakhnin, J. Mercante, T. Romeo, and P. Babitzke. 2007. CsrA inhibits translation initiation of Escherichia coli hfq by binding to a single site overlapping the ShineDalgarno sequence. J. Bacteriol. 1895472-548 1. Boer, P., J. A. Wagenaar, R. P. Achterberg, J. P. M. van Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:351-359. Bras, A. M., S. Chatterjee, B. W. Wren, D. G. Newell, and J. M. Ketley. 1999. A novel Campylobacter jejuni two-component regulatory system important for temperature-dependent growth and colonization. J. Bacteriol. 181:3298-3302. Busenlehner, L. S., M. A. Pennella, and D. P. Giedroc. 2003.The SmtB/ArsR family of metalloregulatory transcriptional repressors: structural insights into prokaryotic metal resistance. FEMS Microbiol. Rev. 27:131-143. Carrillo, C. D., E. Taboada, J. H. Nash, P. Lanthier, J. Kelly, P. C. Lau, R. Verhulp, 0. Mykytczuk, J. Sy, W. A. Findlay, K. Amoako, S. Gomis, P. Willson, J. W. Austin, A. Potter, L. Babiuk, B. Allan, and C. M. Szymanski. 2004. Genome-wide expression analyses of Campylobacter jejuni NCTClll68 reveals coordinate regulation of motility and virulence by fZhA. J. Biol. Chem. 279:20327-203 3 8. Chatterji, D., and A. K. Ohja. 2001. Revisiting the stringent response, ppGpp and starvation signaling. Cum. Opin. Microbiol. 4~160-165. Creuzenet, C. 2004. Characterization of CJ1293, a new UDPGlcNAc C6 dehydratase from Campylobacter jejuni. FEBS Lett. 559:136-140.
Dombroski, A. J., W. A. Walter, M. T. Record, Jr., D. A. Siegele, and C. A. Gross. 1992. Polypeptides containing highly conserved regions of transcription initiation factor sigma 70 exhibit specificity of binding to promoter DNA. Cell 70501-512. Egan, S. M. 2002. Growing repertoire of AraC/XylS activators. J. Bacteriol. 1845529-5532. Elvers, K. T., S. M. Turner, L. M. Wainwright, G . Marsden, J. Hinds, J. A. Cole, R. K. Poole, C. W. Penn, and S. F. Park. 2005. NssR, a member of the Crp-Fnr superfamily from Campylobacter jejuni, regulates a nitrosative stress-responsive reguIon that includes both a single-domain and a truncated haemoglobin. Mol. Microbiol. 57:735-750. Elvers, K. T., G. Wu, N. J. Gilberthorpe, R. K. Poole, and S. F. Park. 2004. Role of an inducible single-domain hemoglobin in mediating resistance to nitric oxide and nitrosative stress in Campylobacter jejuni and Campylobacter coli. J. Bacteriol. 186: 5332-534 1. Engels, V., and V. F. Wendisch. 2007. The DeoR-type regulator SugR represses expression of ptsG in Corynebacterium glutamicum. J. Bacteriol. 189:2955-2966. Gaynor, E. C., D. H. Wells, J. K. MacKichan, and S. Falkow. 2005. The Campylobacter jejuni stringent response controls specific stress survival and virulence-associated phenotypes. Mol. Microbiol. 56:8-27. Goon, S., C. P. Ewing, M. Lorenzo, D. Pattarini, G. Majam, and P. Guerry. 2006. A sigma28-regulated nonflagella gene contributes to virulence of Campylobacter jejuni 81-176. Infect. Immun. 74:769-772. Haghjoo, E., and J. E. G a l h . 2007. Identification of a transcriptional regulator that controls intracellular gene expression in Salmonella Typhi. Mol. Microbiol. 64:1549-1561. Hanninen, M. L., M. Hakkinen, and H. Rautelin. 1999. Stability of related human and chicken Campylobacter jejuni genotypes after passage through chick intestine studied by pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 65:2272-2275. Hantke, K. 2001. Iron and metal regulation in bacteria. Cum. Opin. Microbiol. 4:172-177. Hendrixson, D. R. 2006. A phase-variable mechanism controlling the Campylobacter jejuni FlgR response regulator influences commensalism. Mol. Microbiol. 61:1646-1659. Hendrixson, D. R., and V. J. DiRita. 2004. Identification of Campylobacter jejuni genes involved in commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52:471-484. Hendrixson, D. R., and V. J. DiRita. 2003. Transcription of ~ 5 4 dependent but not a28-dependent flagellar genes in Campylobacter jejuni is associated with formation of the flagellar secretory apparatus. Mol. Microbiol. 50:687-702. Heusipp, G., S. Falker, and M. A. Schmidt. 2007. DNA adenine methylation and bacterial pathogenesis. Int. J. Med. Microbiol. 297:1-7. Hobman, J. L., J. Wilkie, and N. L. Brown. 2005. A design for life: prokaryotic metal-binding MerR family regulators. Biometals 18:429-436. Hofreuter, D., J. Tsai, R. 0. Watson, V. Novik, B. Altman, M. Benitez, C. Clark, C. Perbost, T. Jarvie, L. Du, and J. E. Galan. 2006. Unique features of a highly pathogenic Campylobacter jejuni strain. Infect. Immun. 74:4694-4707. Holmes, K., F. Mulholland, B. M. Pearson, C. Pin, J. McNichollKennedy, J. M. Ketley, and J. M. Wells. 2005. Campylobacter jejuni gene expression in response to iron limitation and the role of Fur. Microbiology 151:243-257. Jagannathan, A., C. Constantinidou, and C. W. Penn. 2001. Roles of @ON,fZiA, and fZgR in expression of flagella in Campylobacter jejuni. J. Bacteriol. 183:2937-2942. Korner, H., H. J. Sofia, and W. G . Zumft. 2003. Phylogeny of the bacterial superfamily of Crp-Fnr transcription regulators: ex-
CHAPTER 34
ploiting the metabolic spectrum by controlling alternative gene programs. FEMS Microbiol. Rev. 27559-592. Kustu, S., A. K. North, and D. S. Weiss. 1991. Prokaryotic transcriptional enhancers and enhancer-binding proteins. Trends Biochem. Sci. 16:397-402. Lin, J., M. Akiba, 0. Sahin, and Q. Zhang. 2005a. CmeR functions as a transcriptional repressor for the multidrug efflux pump CmeABC in Campylobacter jejuni. Antimicrob. Agents Chemother. 49: 1067-1 075. Lin, J., C. Cagliero, B. Guo, Y. W. Barton, M. C. Maurel, S. Payot, and Q. Zhang. 2005b. Bile salts modulate expression of the CmeABC multidrug efflux pump in Campylobacter jejuni. J. Bacteriol. 187:74 17-7424. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Chemother. 46:2124-213 1. Logan, S. M., J. F. Kelly, P. Thibault, C. P. Ewing, and P. Guerry. 2002. Structural heterogeneity of carbohydrate modifications affects serospecificity of Campylobacter flagellins. Mol. Microbiol. 46587-597. MacKichan, J. K., E. C. Gaynor, C. Chang, S. Cawthraw, D. G. Newell, J. F. Miller, and S. Falkow. 2004. The Carnpylobacter jejuni dccRS two-component system is required for optimal in vivo colonization but is dispensable for in vitro growth. Mol. Microbiol. 54 :1269- 1286. Moen, B., A. Oust, 0. Langsrud, N. Dorrell, G. L. Marsden, J. Hinds, A. Kohler, B. W. Wren, and K. Rudi. 2005. Explorative multifactor approach for investigating global survival mechanisms of Campylobacter jejuni under environmental conditions. Appl. Environ. Microbiol. 71:2086-2094. Molina-Henares, A. J., T. Krell, M. Eugenia Guazzaroni, A. Segura, and J. L. Ramos. 2006. Members of the IclR family of bacterial transcriptional regulators function as activators and/or repressors. FEMS Microbiol. Rev. 30: 157-186. Morett, E., and L. Segovia. 1993. The sigma 54 bacterial enhancerbinding protein family: mechanism of action and phylogenetic relationship of their functional domains. J. Bacteriol. 175:60676074. Muller, S., M. Pflock, J. Schar, S. Kennard, and D. Beier. 2007. Regulation of expression of atypical orphan response regulators of Helicobacter pylori. Microbiol. Res. 162:l-14. Narberhaus, F. 1999. Negative regulation of bacterial heat shock genes. Mol. Microbiol. 31:l-8. Palyada, K., D. Threadgill, and A. Stintzi. 2004. Iron acquisition and regulation in Campylobacter jejuni. J. Bacteriol. 186:47144729. Parker, C. T., B. Quinones, W. G. Miller, S. T. Horn, and R E. Mandrell. 2006. Comparative genomic analysis of Campylobacter jejuni strains reveals diversity due to genomic elements similar to those present in C. jejuni strain RM1221. J. Clin. Microbiol. 44:4125-4135. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668. Parkinson, J. S. 1993. Signal transduction schemes of bacteria. Cell 73: 857-8 71. Pearson, B. M., C. Pin, J. Wright, K. Anson, T. Humphrey, and J. M. Wells. 2003. Comparative genome analysis of Campylobacter jejuni using whole genome DNA microarrays. FEBS Lett. 554:224-230 Peres, C. M., and C. S. Harwood. 2006. BadM is a transcriptional repressor and one of three regulators that control benzoyl co-
REGULATION OF GENES IN C . IEIUNI
623
enzyme A reductase gene expression in Rhodopseudomonas palustris. ]. Bacteriol. 188:8662-8665. Perraud, A. L., V. Weiss, and R. Gross. 1999. Signalling pathways in two-component phosphorelay systems. Trends Microbiol. 7: 115-120. Petersen, L., T. S. Larsen, D. W. Ussery, S. L. On, and A. Krogh. 2003. RpoD promoters in Campylobacter jejuni exhibit a strong periodic signal instead of a -35 box. J. Mol. Biol. 326:13611372. Pittman, M. S., K. T. Elvers, L. Lee, M. A. Jones, R. K. Poole, S. F. Park, and D. J. Kelly. 2007. Growth of Campylobacter jejuni on nitrate and nitrite: electron transport to NapA and NrfA via NrfH and distinct roles for NrfA and the globin Cgb in protection against nitrosative stress. Mol. Microbiol. 63575590. Poly, F., C . Ewing, S. Goon, T. E. Hickey, D. Rockabrand, G. Majam, L. Lee, J. Phan, N. J. Savarino, and P. Guerry. 2007. Heterogeneity of a Campylobacter jejuni protein that is secreted through the flagella filament. Infect. Immun. 75:3859-3867. Poole, R. K. 2005. Nitric oxide and nitrosative stress tolerance in bacteria. Biochem. Soc. Trans. 33:176-180. Pumbwe, L., and L. J. Piddock. 2002. Identification and molecular characterisation of CmeB, a Campylobacter jejuni multidrug efflux pump. FEMS Microbiol. Lett. 206:185-189. Ramos, J. L., M. Martinez-Bueno, A. J. Molina-Henares, W. Teran, K. Watanabe, X. Zhang, M. T. Gallegos, R. Brennan, and R. Tobes. 2005. The TetR family of transcriptional repressors. Microbiol. Mol. Biol. Rev. 69:326-356. Raphael, B. H., S. Pereira, G. A. Flom, Q. Zhang, J. M. Ketley, and M. E. Konkel. 2005. The Campylobacter jejuni response regulator, CbrR, modulates sodium deoxycholate resistance and chicken colonization. J. Bacteriol. 187:3662-3670. Ratledge, C., and L. G. Dover. 2000. Iron metabolism in pathogenic bacteria. Annu. Rev. Microbiol. 54:881-941. Ridley, K. A., J. D. Rock, Y. Li, and J. M. Ketley. 2006. Heme utilization in Campylobacter jejuni. J. Bacteriol. 188:78627875. Schell, M. A. 1993. Molecular biology of the LysR family of transcriptional regulators. Annu. Rev. Microbiol. 47597-626. Serkin, C. D., and H. S. Seifert. 2000. Iron availability regulates DNA recombination in Neisseria gonorrhoeae. Mol. Microbiol. 37~1075-1086. Shingler, V. 1996. Signal sensing by sigma 54-dependent regulators: derepression as a control mechanism. Mol. Microbiol. 19: 409-416. Sommerlad, S. M., and D. R. Hendrixson. 2007. Analysis of the roles of FlgP and FlgQ in flagellar motility of Campylobacter jejuni. J. Bacteriol. 189:179-186. Spohn, G., A. Danielli, D. Roncarati, I. Delany, R. Rappuoli, and V. Scarlato. 2004. Dual control of Helicobacter pylori heat shock gene transcription by HspR and HrcA. J. Bacteriol. 186: 2956-2965. Stintzi, A. 2003. Gene expression profile of Campylobacter jejuni in response to growth temperature variation. J. Bacteriol. 185: 2009-2016. Stintzi, A., D. Marlow, K. Palyada, H. Naikare, R. Panciera, L. Whitworth, and C. Clarke. 2005. Use of genome-wide expression profiling and mutagenesis to study the intestinal lifestyle of Campylobacter jejuni. Infect. Immun. 73:1797-1810. Stock, A. M., V. L. Robinson, and P. N. Goudreau. 2000. Twocomponent signal transduction. Annu. Rev. Biochem. 69: 183215. Studholme, D. J., and M. Buck. 2000. The biology of enhancerdependent transcriptional regulation in bacteria: insights from genome sequences. FEMS Microbiol. Lett. 186:l-9.
624
WOSTEN ET AL.
Studholme, D. J., and R. N. Pau. 2003. A DNA element recognised by the molybdenum-responsive transcription factor ModE is conserved in Proteobacteria, green sulphur bacteria and Archaea. BMC Microbiol. 3:24. Thompson, S. A., 0. L. Shedd, K. C. Ray, M. H. Beins, J. P. Jorgensen, and M. J. Blaser. 1998. Campylobacter fetus surface layer proteins are transported by a type I secretion system. J. Bacteriol. 180:645 0-645 8. van Vliet, A. H., M. L. Baillon, C. W. Penn, and J. M. Ketley. 1999. Campylobacter jejuni contains two fur homologs: characterization of iron-responsive regulation of peroxide stress defense genes by the PerR repressor. J. Bacteriol. 181:6371-6376. van Vliet, A. H., J. M. Ketley, S. F. Park, and C. W. Penn. 2002. The role of iron in Campylobacter gene regulation, metabolism and oxidative stress defense. FEMS Microbiol. Rev. 26:173-186. Vogel, J., and E. G. Wagner. 2007. Target identification of small noncoding RNAs in bacteria. Curr. Opin. Microbiol. 10:262270. Voskuil, M. I., K. Voepel, and G. H. Chambliss. 1995. The -16 region, a vital sequence for the utilization of a promoter in Bacillus subtilis and Escherichia coli. Mol. Microbiol. 17:271-279. Wilkinson, S . P., and A. Grove. 2004. HucR, a novel uric acidresponsive member of the MarR family of transcriptional regulators from Deinococcus radiodurans.J. Biol. Chem. 279:s 144251450. Wilkinson, S. P., and A. Grove. 2006. Ligand-responsivetranscriptional regulation by members of the MarR family of winged helix proteins. Curr. Issues Mol. Biol. 851-62.
Woodall, C. A,, M. A. Jones, P. A. Barrow, J. Hinds, G. L. Marsden, D. J. Kelly, N. Dorrell, B. W. Wren, and D. J. Maskell. 2005. Campylobacter jejuni gene expression in the chick cecum: evidence for adaptation to a low-oxygen environment. Infect. Immun. 735278-5285. Wooldridge, K. G., P. H. Williams, and J. M. Ketley. 1994. Ironresponsive genetic regulation in Campylobacter jejuni: cloning and characterization of a fur homolog. J. Bacteriol. 17658525856. Wosten, M. M., M. Boeve, W. Gaastra, and B. A. M. van der Zeijst. 1998a. Cloning and characterization of the gene encoding the primary a-factor of Campylobacter jejuni. FEMS Microbiol. Lett. 162:97-103. Wosten, M. M., M. Boeve, M. G. Koot, A. C. van Nuene, and B. A. M. van der Zeijst. 1998b. Identification of Campylobacter jejuni promoter sequences. J. Bacteriol. 180594-599. Wosten, M. M., C. T. Parker, A. van Mourik, M. R. Guilhabert, L. van Dijk, and J. P. M. van Putten. 2006. The Campylobacter jejuni PhosSiPhosR operon represents a non-classical phosphate-sensitive two-component system. Mol. Microbiol. 62:278291. Wosten, M. M., J. A. Wagenaar, and J. P. M. van Putten. 2004. The FlgS/FlgR two-component signal transduction system regulates the flu regulon in Campylobacter jejuni. J. Biol. Chem. 279:16214-16222. Young, B. A., T. M. Gruber, and C. A. Gross. 2002. Views of transcription initiation. Cell 109:417-420.
VI. FOOD SAFETY AND INTERVENTION
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blase1 0 2008 ASM Press, Washington, DC
ChaDter 35
Campylobacter in the Food Supply WILMA JACOBS-REITSMA, ULRIKE
Campylobacter infections in humans are considered to be mainly food-borne, in which foods of animal origin play an important role. Besides the foodrelated cases, close contact with pet animals and activities relating to recreational waters, as well as yet unknown factors contribute to the large number of human illnesses each year due to Campylobacter. Large community outbreaks are relatively rare in Campylobacter epidemiology, but the implicated sources are both epidemiologically and microbiologically identified to be raw milk and untreated surface water. The majority of Campylobacter infections are sporadic (single) cases or small family outbreaks, and the actual source of these types of infection is rarely microbiologically identified. Finding microbiological proof by culturing (the same types of) campylobacters from suspected sources as well as from the patients is not easy to accomplish. The original suspected food item was often either consumed entirely or, by the time of illness, leftovers are no longer available. Moreover, after prolonged storage, it may be difficult to recover any campylobacters from the implicated food items. Epidemiological (case-control) studies have revealed a significant association between Campylobacter infection in humans and the handling and consumption of raw or undercooked poultry. The extent to which poultry consumption is responsible for human campylobacteriosis is still not exactly known, but two new pieces of evidence support the hypothesis of poultry being an important source of human infection. In Belgium, poultry and eggs had to be withdrawn from the market in the summer of 1999 because of dioxin contamination, and this coincided with a reduction of human Campylobacter cases of
LYHS,
AND JAAP WAGENAAR
40% (Vellinga and Van Loock, 2002). In Iceland, chicken sales largely changed from frozen to chilled in 1996, which was followed by a marked increase in human Campylobacter infections. After the introduction of a successful combination of preventive measures, there was a marked decrease in the number of cases again in 2000 (Stern et al., 2003). Still, poultry and poultry products have the doubtful honor of being considered the major cause of human Campylobacter infections, and therefore much Campylobacter research has been related to poultry. Indeed, large percentages of fresh and even frozen poultry products can be found to be contaminated with Campylobacter. Levels of contamination may vary between log 2 and log 6 CFU per carcass. In combination with the relatively low infectious dose of 500 CFU as reported for Campylobacter, poultry products may pose a serious risk to consumers if incorrectly treated during preparation or if insufficiently cooked. Besides poultry, other foods (mainly of animal origin) have to be considered as potential sources of infection because Campylobacter also has been isolated from food items like raw milk, pork, beef, lamb, and seafood. The consumption of untreated water or recreational water, whether intended or not, may also be a significant risk factor for Campylobacter infections. This chapter describes the detection and prevalence of Campylobacter in a wide range of different types of food. Possibilities for minimizing the risks of human disease by this pathogen will be discussed. Because food-borne-related campylobacteriosis still mainly concerns the thermotolerant Campylobacter species (C. jejuni and C. coli, and to a lesser extent C. lari), we will refer to the thermotolerant species.
Wilma Jacobs-Reitsma RIKILT Institute of Food Safety, Bornsesteeg 45, 6708 PD Wageningen, The Netherlands. Ulrike Lyhs * Ruralia Institute, Seinajoki Unit, University of Helsinki, Kampusranta 9C, FI-60320 Seinajoki, Finland. Jaap Wagenaar * Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, P.O. Box 80165,3508 TD Utrecht, The Netherlands.
62 7
628
TACOBS-REITSMA ET AL.
DETECTION OF CAMPYLOBACTER IN FOODS An extensive and detailed description of the developments in Campylobacter detection and the variety of early and more recent isolation media is given by Corry and coworkers (2003).The media used for the isolation of Campylobacter from food or water are derived from the media originally designed for the detection of campylobacters from human stool samples. Fecal samples often contain large numbers of quite viable campylobacters, and detection is possible by direct plating on selective media. Food products, however, may harbor only low numbers of campylobacters, and bacterial cells may be seriously injured by processing procedures such as freezing, cooling, heating, and salting. Therefore, liquid enrichment media have been developed to detect low numbers of Campylobacter cells and to promote the recovery of sublethally damaged cells. Isolation of Campylobacter from the enrichment media is on solid selective agars, often with comparable ingredients to the liquid medium. A variety of selective agents in different combinations and concentrations are incorporated in the various media for Campylobacter, with cefoperazone, amphotericin B, trimethoprim, and vancomycin often used. Several enrichment procedures delay the exposure to some selective agents until cell repair has occurred, in combination with the use of lower incubation temperatures during the first stages of enrichment. Incubation generally is at 42"C, a temperature at which the thermotolerant campylobacters are able to grow well, and which provides an enhanced selectivity. Nowadays, incubation at 413°C is more commonly used because of the practical combination with the stricter Salmonella isolation at that temperature. Sterile sheep's or horse's blood, or charcoal is added to the media to neutralize the toxic effects of oxygen and light. Because of the microaerophilic nature of Campylobacter, it is necessary to incubate in an oxygen-reduced atmosphere, most commonly in jars or in adapted CO, incubators. Several methods are available to achieve the optimal gas mixture of 5 to 7% oxygen, 10% carbon dioxide, and 80% nitrogen and/or hydrogen: atmosphere replacement by the appropriate bottled gas mixture; replacement of two-thirds of the atmosphere with a mixture of either nitrogen or hydrogen plus 5 to 15% carbon dioxide when the appropriate gas-generating envelope is used; and even use of candle jars (Corry et al., 2003; Stern and Kazmi, 1989). Chapter 7 of the U.S. Food and Drug Administration Bacteriological Analytical Manual Online (Hunt et al., 1998) gives a detailed description for the isolation of Campylobacter from food and water.
In general, 25 g of food is combined with a volume of 100 ml of enrichment broth, but different sample preparations are described for various types of food. Whole chicken carcasses are rinsed with peptone water; the liquid is centrifuged, and the resuspended pellet is put into enrichment broth. Water samples, preferably of 2 to 4 liters, are filtered, and the 0.45p-diameter pore size filters are enriched. Immediately after sampling, raw milk is adjusted to a pH of 7.5 0.2 with sterile NaOH to inactivate the lactoperoxidase system, which is toxic to Campylobacter. Preenrichment may either be 3 h at 30°C followed by 2 h at 37°C (frozen samples, water, shellfish) or 4 h at 37°C (all other cases). Finally, for both enrichment procedures, incubation is continued at 42°C for 20 to 44 h. A combination of Bolton enrichment broth base with lysed horse blood and the selective agents vancomycin, trimethoprim lactate, cycloheximide (or amphotericin B), and cefoperazone is used for selective enrichment. Plating for direct counts or after enrichment is on either mCCDA (modified charcoal cefoperazone desoxycholate agar with additional yeast extract) or AHB (Abeyta-Hunt-Bark) agar for 24 to 48 h at 42°C. AHB agar contains sterile horse blood, whereas mCCDA is charcoal based. All incubations are performed under microaerobic conditions, and details are given for the recommended shaking of the enrichment in combination with a bubbler system for continuous gas flow. Identification is by microscopical analysis and a number of biochemical tests, including latex agglutination (Hunt et al., 1998). The International Organization for Standardization (ISO) recently revised International Standard 10272, describing a horizontal method for the detection of thermotolerant Campylobacter in food and animal feeding stuffs (ISO, 2006a). A method for enumeration that uses a spread plate count on mCCDA plates (48 h at 41.5"C7 microaerobically) was added as part 2 (ISO, 2006b). The revised detection method also uses Bolton broth for selective enrichment, starting for 4 h at 37°C and followed by another 44 h at 41.S"C. Plating is on mCCDA agar and on a second solid selective medium of choice. All incubations are in a static microaerobic atmosphere. Sampling specifications are not part of this IS0 method, but a 1 : l O ratio for test porti0n:enrichment broth is given as a general rule (ISO, 2006a). However, fresh poultry meat samples may require modifications to this protocol because many problems were experienced in reading the plates as a result of overgrowth with other flora (Jacobs-Reitsma et al., 2007). International Standard 17995 describes the detection and semiquantitative enumeration of thermotolerant Campylobacter species in water by means
*
CHAPTER 35
of both Preston and Bolton enrichment broth and plating on mCCDA (ISO, 2005). Practical recommendations for large numbers of samples as in baseline studies are provided by the National Advisory Committee on Microbiological Criteria for Foods (2005) and, in Europe, the European Food Safety Authority (EFSA) (2006a). All above-mentioned methods are described and intended for the isolation of thermotolerant Campylobacter. However, some of these “thermotolerant” organisms are reported to be missed occasionally as a result of the antibiotics used in the selective media in combination with the high incubation temperature of 42°C. Membrane filtration and subsequent plating on nonselective (blood) media at 37°C are required to detect these and other, more sensitive members of the Campylobacter family (Corry et al., 2003). Various rapid methods, such as antibody-based methods, immunoassays, enzyme-linked immunosorbent assays, immunomagnetic beads, and electrochemiluminescent detection, are commercially available. These methods still require an initial enrichment culturing step. This is also true for most of the published PCR-based detection methods. Improvements in PCR-related problems like cell concentration, DNA purification, and inhibiting components in food, feces, or media as well as improvements in the sensitivity of detection have led to PCR methods that are directly applicable to food or fecal samples. Developments in real-time PCR may even add a quantitative dimension to detection (EFSA, 2005; Humphrey et al., 2007).
POULTRY MEAT Live poultry, including broilers, laying hens, turkeys, and ducks, are often found to be colonized by large numbers of Campylobacter without the animal showing any signs of clinical illness. Colonization levels in the small intestine, especially in the ceca, range from log 5 to over log 9 CFU/g (Berndtson et al., 1992; Mead et al., 1995; Rosenquist et al., 2006). Colonized birds enter the slaughterhouse with large numbers of Campylobacter on their feathers and skin, as well as in their intestinal tract, and Campylobucter consequently can be found throughout the slaughtering process. This also leads to the contamination of equipment, working surfaces, process water, and air. The large amounts of water used during poultry processing contribute to the spread and survival of Cumpylobacter and complicate in-plant control (Berndtson et al., 1992; EFSA, 2005; NACMCF, 1994). Scalding of the birds, which is necessary to
-
CAMPYLOBACTER IN THE FOOD SUPPLY
629
facilitate defeathering, usually occurs at a water temperature of between 50 and 60°C for several minutes. At higher temperatures, a greater thermal reduction of Campylobacter numbers may occur, but the higher scalding temperature also affects the skin surface in such a way that bacteria attach more firmly to the skin. During defeathering and evisceration, the leakage of intestinal contents almost inevitably contributes to the contamination of carcasses. The next steps of washing and chilling tend to reduce bacteria counts again, but they do not completely eliminate the contamination. Air chilling has been suggested to be more effective than water chilling because of the drying effects, but the structure of the poultry skin, unlike pigskin, is much more difficult to dry completely, even after prolonged chilling (Bryan and Doyle, 1995; EFSA, 2005; Oosterom et al., 1983). Overall, the slaughtering process may reduce the level of contamination about 100 to 1,000 times (Rosenquist et al., 2006). During the processing of broiler flocks colonized with log 6.8 CFU Campylobacter per g cecal content, Mead and coworkers (1995) found a reduction of numbers of Campylobucter on neck skin samples from log 3.7 after exsanguination to log 1.8 after packaging. The distribution and numbers of Campylobacter in 100 freshly slaughtered chickens originating from six different Campylobacter-positive flocks were studied by Berndtson and coworkers (1992). Cecal contents of the birds harbored log 5.8 to over log 9 CFU/g of Campylobacter. During processing, Campylobacter was isolated in 89% of neck skins, 93% of peritoneal cavity swab samples, and 75% of subcutaneous samples. The incidence in 340 muscle samples was low at 3%. Cumpylobacter numbers in neck skin and peritoneal cavities ranged from log 2.4 to log 3.4 CFU/ g or per 4 cm2, whereas numbers in the subcutaneous layer were only log 1.1 to log 1.8 CFU per 4 cm2 (Berndtson et al., 1992). The reported levels of Campylobacter in fresh poultry products vary between less than log 2 to over log 6 CFU per carcass or log 1 to log 4 CFU/100 g (or a fillet) of meat, depending on the different studies and the methodologies used (Atanassova et al., 2007; Berndtson et al., 1992; Bryan and Doyle, 1995; Jsrgensen et al., 2002). Freezing the products diminishes the level of Contamination, but survival at -20°C has been demonstrated for at least 3 months, although at a very low level (EFSA, 2005; NACMCF, 1994). Campylobacter are also frequently found on edible offal, like hearts and livers, more likely to be caused by cross-contamination during processing rather than being the result of infection of the tissue itself (Bryan and Doyle, 1995). It is worth noting that during recent years, the variety of poultry products available in retail markets
630
TACOBS-REITSMAET AL.
has been increasing. In Finland, for example, approximately 80% of poultry products are now sold marinated. High NaCl concentration, low pH, and the addition of different spices to the marinade prevent the growth of spoilage bacteria, thus increasing the shelf life of the meat products. However, the use of marinade for poultry meat did not decrease the survival of Campylobacter, probably as a result of the buffering capability of meat, which quickly neutralized the pH of the acidic marinade (Evans et al., 1998; Perko-Makela et al., 2000). Table 1 and Table 2 summarize examples of the prevalence of Campylobacter in poultry products at the slaughterhouse level and at the retail level, respectively, as reported in various studies in recent years. Earlier studies are summarized by Aho and Hirn (1988), Bryan and Doyle (1995), and JacobsReitsma (2000). The EFSA's yearly updated Community Summary Report on Trends and Sources of Zoonoses, Zoonotic Agents and Antimicrobial resistance in the European Union also presents data from various European countries on Campylobacter in poultry meat (broiler, turkey, duck) at the slaughterhouse, processing, and retail level, as well as in pig and bovine meat and in other foods like milk and fishery products (EFSA, 2006b). Direct correlation of the various studies may be difficult because of the wide variety of methodologies used. Also, the differences in sampling procedures may account for a difference in isolation percentages. Sampling of poultry products may include 25 g of meat without skin; a certain area of skin, either as a whole or surface swabbed; thaw or drip water; or whole carcass rinses with shaking or massage. Besides the detection of Campylobacter in the various stages of poultry processing, much more attention is now paid to quantification of the presence of the bacterium because these quantitative data are required for the risk assessment studies on Campylobacter, which are made most commonly in the poultry meat production chains (Nauta et al., 2005; Rosenquist et al., 2003; Uyttendaele et al., 2006). But also this type of information sometimes lacks useful correlation as a result of differences in the sampling and counting methods used. Besides the described reservoir of Campylobacter in animals for food production, wildlife animals, some of which may be consumed by humans, are frequent sources of Campylobacter. These include waterfowl, hare, wild boar, pheasant, guinea fowl, mallard, gull, pigeon, quail, cockatiel, emu, ostrich, parrot, and domestic duck (Casanovas et al., 1995; Cuomo et al., 2007; Kapperud and Rosef, 1983; Oyarzabal et al., 1995). However, the risk of contracting a Campylobacter infection from any of these
potential sources is probably higher via direct contact.
EGGS In general, commercial table eggs are not likely to be associated with Campylobacter infections. Only one suggested outbreak has been reported. Finch and Blake (1985) found that 26 of 81 individuals developed campylobacteriosis after consuming undercooked eggs. Eggs presumed to be the cause of that outbreak were not available for examination. However, hens on the farm that supplied the eggs yielded a number of Campylobacter serotypes, one of which was identical to the isolate derived from a single patient. Like broilers, laying hens are often found to excrete Campylobacter in high numbers, which may lead to externally contaminated eggs. Doyle (1984a) found 2 of 226 eggshells from a Campylobacterexcreting laying hen flock to be contaminated with Campylobacter. One of the 2 eggs had a dirty shell. Jacobs-Reitsma (1994) isolated Campylobacter from 3 of 179 shells of fecally contaminated eggs (not consumer quality). Survival of Campylobacter on eggshells, however, is considered to be poor because of the sensitivity of the organism to drying. Studies on artificially contaminated eggshells showed that Campylobacter could no longer be detected after storage at room temperature for 48 h (Doyle, 1984a; Shane et al., 1986). Egg penetration studies revealed that Campylobacter did not penetrate into the contents of eggs kept at 25, 37, or 4°C after artificial contamination, but it could occasionally be isolated for a short storage time from the inner shell and membranes of the refrigerated eggs (Doyle, 1984a). At room temperature, Campylobacter was recovered from 3 of 70 shell membranes and 1 of 70 egg contents, but only a short time (2 h) after artificial contamination (Shane et al., 1986). Sahin et al. (2003) inoculated C. jejuni into egg yolk, albumen, and air sac, then stored the eggs at 18°C. C. jejuni survived in the egg yolk longest, up to 14 days. In the same study, Campylobacter was detected by culture and PCR in 3 of 65 pooled samples of eggs laid by Campylobacter-inoculated specific-pathogen-free layers. Campylobacter was not isolated from 500 fresh eggs obtained from commercial broiler-breeder flocks actively shedding Campylobacter in their feces, nor from 1,000 eggs from a commercial hatchery (Sahin et al., 2003). Also, no Campylobacter was isolated from 219 egg yolks (Jacobs-Reitsma, 1994) or 216 egg contents, examined within 12 h of being laid (Doyle, 1984a). Sulonen et al. (2007) isolated Cam-
Table 1. Presence of thermotolerant Campylobucteer at the poultry slaughterhouse level Product
Stage of process
Sample type
It"
% Positiveb
Chicken feathers Chicken breast skin Chicken crop Chicken ceca Chicken colon Chicken carcasses Chicken carcasses Chicken carcasses Chicken intestine Chicken skin Chicken crop Chicken carcasses Chicken offal Chicken neck skin Chicken carcasses Chicken neck skin Chicken carcasses Chicken carcasses Chicken carcasses Chicken carcasses Chicken carcasses Chicken wings, legs, fillet Chicken breasts with skin Chicken breasts without skin Chicken skin, liver, neck Turkey skin Turkey skin Turkey skin Turkey skin Turkey carcasses Turkey skin Turkey skin Turkey breasts Turkey wings Turkey thighs and drumsticks Turkey stock Turkey offal Wild pheasants, various
Before scalding Before scalding Before scalding Before scalding Before scalding After scalding After plucking Before evisceration Before evisceration Before evisceration Before evisceration After evisceration After evisceration Before chilling Before chilling After chilling After chilling After chilling After chilling After chilling After chilling Meat cutting Meat cutting Meat cutting Not specified After killing After scalding After plucking After evisceration Before chilling Chilling 20 min Chilling 24 hours Meat cutting Meat cutting Meat cutting Meat cutting Meat cutting Not specified
1.5 g 6.5 g 5.1 g 7.8 g 3.1 g Rinse Rinse Rinse Swabs Swabs Swabs Rinse 10 g 10 g Rinse 10 g Rinse Rinse Rinse Rinse Rinse 10 g Rinse Rinse Not specified Swabs Swabs Swabs Swabs Rinse Swabs Swabs Not specified Not specified Not specified Not specified Not specified Not specified
18 18 18 18 18 125 15 800 202 202 202 15 21 16 75 16 800 15 213 636 125 12 15 15 111 43 43 43 43 59 43 43 22 22 28 20 52 52
ndL,d ndd 100d
~
"Total number of samples examined. *Detection of Campylobucter, based on enrichment culture unless otherwise stated. "Not determined. dDetection of Campylobucter, based on direct culture. eDetection of C. lelunr by PCR.
loo*
ndd 92 53.3 74.5 94d 78 48d 66.7 19 100 100 100 34.9d 40
lood
16 52 33 33.3 26.7 45.9 76 37.2" 58.1" 72.1 36.9 67.4" 25.6" 4.2 5.6 6.9 2.8 9.7 25.9
Enumeration (log CFU) 5.4lg 3.8lg 4.7lg 7.3lg 7.2/g nd 6.5lsample 2.711111 nd nd nd 6.0lsample nd nd nd nd 0.43 lml 5.41sample 5.21 carcass nd nd nd 4.4isample 4.l/sample nd nd nd nd nd nd nd nd 1.9lg
2.3lg 2.0/g 2.31s 2.5lg nd
Country
Year
Reference
USA USA USA USA USA USA Germany USA USA USA USA Germany France France USA France USA Germany Italy Sweden USA France Germany Germany Germany Germany Germany Germany Germany Canada Germany Germany Germany Germany Germany Germany Germany Germany
1999 1999 1999 1999 1999 2004 2006 2005 2001 2001 2001 2006 1999 1999 2004 1999 2005 2006 2003-2004 2002-2003 2004 1999 2006 2006 1995-1997 2002 2002 2002 2002 2005 2002 2002 2005 2005 2005 2005 2005 1995-1997
Berrang et al. (2000) Berrang et al. (2000) Berrang et al. (2000) Berrang et al. (2000) Berrang et al. (2000) Son et al. (2007) Klein et al. (2007) Berrang et al. (2007) Jeffrey et al. (2001) Jeffrey et al. (2001) Jeffrey et al. (2001) Klein et al. (2007) Denis et al. (2001) Denis et al. (2001) Son et al. (2007) Denis et al. (2001) Berrang et al. (2007) Klein et al. (2007) Manfreda et al. (2006) Lindmark et al. (2006) Son et al. (2007) Denis et al. (2001) Klein et al. (2007) Klein et al. (2007) Atanassova and Ring (1999) Alter et al. (2005) Alter et al. (2005) Alter et al. (2005) Alter et al. (2005) Arsenault et al. (2007) Alter et al. (2005) Alter et al. (2005) Atanassova et al. (2007) Atanassova et al. (2007) Atanassova et al. (2007) Atanassova et al. (2007) Atanassova et al. (2007) Atanassova and Ring (1999)
Table 2. Presence of thermotolerant Cumpylobucter in poultry products at the retail level Product
Product condition
Product type
Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Chicken Turkey Turkey Turkey Turkey Turkey Turkey Turkey Turkey Duck
Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh Fresh and frozen Frozen Cooked Heat-treated Fresh Fresh Fresh Fresh Fresh Fresh and frozen Marinated Heat-treated Fresh
Breasts Breasts, liver Meat, legs Not specified Whole, breasts with skin and pieces Carcasses Wings, thighs, and fillets Meat Meat Thighs and breasts Meat Pieces and organs Legs, skin Breast fillets Deep tissue of muscle Deep tissue of breast fillets Carcasses and cuts Carcasses Sausages, fried breast, wings Not specified Breasts, pieces, organs Breasts Not specified Breast fillets and cutlets Sausages Carcasses and cuts Steaks Breasts, sausages, meat balls, etc. Not specified
“Total number of samples examined. Detection of Cumpylobucter, based on enrichment culture.
‘Not determined.
Sample type Swabs 25 g 10 g 10 g
Rinse Rinse 25 g 25 g
Rinse Rinse 25 g 25 g 25 g
Rinse 10 g 10 g Not specified 25 25 25 25
g g
g g
Rinse 10 g Not specified 25 g
Not specified Not specified Not specified 10 g
12“
% Positiveb
65 133 198 444 300 184 135 155 270 198 23 0 68 140 100 115 55 460 35 5 40 419 172 33 48 75 204 16 36 11
13.8 38.3 49.5 49.9 68.0 70.7 76.3 81.3 81.4 83.3 89.1 67.6 66.0 67.0 27.0 20.0 38.7 35.2 0 0 6.2 14.5 37.5 28.0 0 27.5 6.0 0 45.8
Enumeration (log CFU) nd nd nd nd nd nd nd nd nd nd nd Variable 2.4lg
2.7ifillet <0.3 MPNig 1.3 /fillet nd nd nd nd nd nd nd 2.1Ig
nd nd 2.11g
nd nd
Country
Year
Reference
Germany Germ any Spain Ireland UK USA Bulgaria Italy Korea UK New Zealand Spain Germany Germany Germany Germany Denmark Bulgaria Spain Bulgaria Germany USA Ireland Germany Germany Denmark Germany Germany Ireland
2004 2004 2002 2004 2000 2001 2006 2000-2001 2001-2006 1998 2003-2004 2002 2003-2004 2004 2004 2004 2005 2006 2002 2006 2004 2001 2004 2007 2004 2005 2007 2007 2004
Alter et al. (2004) Alter et al. (2004) Dominguez et al. (2002) Whyte et al. (2004) Harrison et al. (2001) Zhao et al. (2001) Stoyenchev et al. (2007) Pezzotti et al. (2003) Hong et al. (2007) Kramer et al. (2000) Wong et al. (2007) Mateo et al. (2005) Scherer et al. (2006) Luber and Bartelt (2007) Scherer et al. (2006) Luber and Bartelt (2007) Nielsen et al. (2006) Stoyanchev et al. (2007) Mateo et al. (2005) Stoyanchev et al. (2007) Alter et al. (2004) Zhao et al. (2001) Whyte et al. (2004) Atanassova et al. (2007) Alter et al. (2004) Nielsen et al. (2006) Atanassova et al. (2007) Atanassova et al. (2007) Whyte et al. (2004)
CHAPTER 35
pylobacter only from a single eggshell sample out of 360 table eggs originating from organic laying hens. Izat and Gardner (1988) conducted five trials in two commercial egg-processing plants, and no Campylobacter could be detected in the raw eggs or in any of the various processed egg products. So far, the case described by Finch and Blake (1985) is the only egg-associated Campylobacter case, and commercial table eggs remain an uncommon source of infection.
MILK AND DAIRY PRODUCTS Unpasteurized milk is a well-documented cause of a number of outbreaks of campylobacteriosis. The first recognized milk-borne outbreak associated with a Campylobacter-like organism occurred in 1938 (Blaser et al., 1983). A very large outbreak took place in 1979 in the United Kingdom, with more than 2,500 children infected by drinking school milk (Jones et al., 1981). Peterson (2003) described an outbreak of Campylobacter enteritis among people consuming raw milk during a meal. Thirteen of 15 people who drank the milk became ill. C. jejuni was recovered from five of six submitted stool samples cultured. Campylobacter can be isolated from the feces of healthy dairy cows, as well as from the raw milk produced. Doyle and Roman (1982) found 50 out of 78 milk-producing cows on nine farms to excrete Campylobacter in their feces, but the organism was isolated in only 1 out of 108 bulk tank milk samples from these farms. Beumer and coworkers (1988) isolated Campylobacter in 22% of 904 fecal samples and in 4.5% of the individual milk samples from 904 healthy dairy cows on 13 Dutch farms. Ten of 12 English dairy herds (Humphrey and Beckett, 1987) examined by 668 rectal swab samples were Campylobacter positive, with a herd incidence of 10 to 72% of the cows tested. No Campylobacter was detected in 30 bulk milk tanks of the two Campylobacternegative herds, but the organism was isolated from 9 out of 111 samples of five farms with cows positive for C. jejuni. The mean level of contamination was MI“ (most probable number) 16 +. 30 organisms per 100 ml of milk (range, 1 to 100 MPN/100 ml) (Humphrey and Beckett, 1987). Raw milk is considered to become contaminated via cow’s feces (Schildt et al., 2006). However, direct contamination of the milk as a consequence of udder infection by Campylobacter has also been described (Hutchinson et al., 1985; Orr et al., 1995). Besides raw cow’s milk, unpasteurized goat’s milk may also transmit Campylobacter infection from animals to humans (Harris et al., 1987).
CAMPYLOBACTER IN T H E FOOD SUPPLY
633
Table 3 summarizes some larger studies on the presence of Campylobacter in raw milk. Differences in isolation percentages can at least to some extent be attributed to the different sampling procedures and the variety of isolation methods used. Fresh raw milk benefits the availability of the lactoperoxidase system, which is toxic to Campylobacter (Beumer et al., 1988). This system can be neutralized by the addition of NaOH to the milk directly after sampling, which subsequently leads to higher isolation percentages of Campylobacter from the milk (Beumer et al., 1988; Hunt et al., 1998). It was not always specified whether such a neutralization procedure was used during the studies listed in Table 3. Although Campylobacter can be detected in a low percentage of raw milk samples, the actual numbers of campylobacters present are generally low (but enough to cause disease). The common pasteurization process is sufficient to eliminate this risk. Several milk-related outbreaks have been described where the pasteurization process had been inadequate (Blaser et al., 1983; Fahey et al., 1995). From the United Kingdom, outbreaks were reported from properly heattreated milk that was nevertheless contaminated, presumably because birds pecked at the bottles delivered to the doorstep (Hudson et al., 1991). The sensitivity of Campylobacter to low water activity suggests that the organism should not cause problems in foods with high solute concentrations like cheese. In combination with the sensitivity to acid, Campylobacter is not likely to survive in cheese or during its production and ripening. Bachmann and Spahr (1995) examined the ability of potentially pathogenic bacteria to grow and survive during the manufacture and ripening of hard and semihard Swiss cheeses made from raw milk inoculated with log 4 to log 6 CFU/ml of Campylobacter. At the age of commercial ripeness (90 days), both the hard cheese and the semihard cheeses were free of Campylobacter (Bachmann and Spahr, 1995). Butter is considered to be safe from pathogens because of its low water content and relatively high NaCl concentration (EFSA, 2005). However, an outbreak of campylobacteriosis associated with garlic butter in a restaurant was described by Zhao et al. (2000). Inoculation studies of log 4 to log 6/g of C. jejuni into butter stored at 5 or 21°C were performed by the same authors. The results showed that C. jejuni was able to survive for 13 days in refrigerated butter with no garlic stored at 5°C. However, large numbers were killed within a few hours in butter that contained garlic. Fermented milk products like yogurt are not suspected sources of infection because of its low pH and the organisms’ sensitivity to organic acids like lactic acid (Cuk et al., 1987).
634
IACOBS-REITSMA ET AL.
Table 3. Presence of thermotolerant Cumpylobucter in raw milk Product
Description
nu
%I Positiveb
Country
Year
Reference
Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw cow’s milk Raw goat’s milk Raw goat’s milk Raw goat’s milk Raw goat’s milk Raw ewe’s milk Raw ewe’s milk Raw ewe’s milk
Bulk tanks at 15 sites 12 farm bulk tanks 292 farm bulk tanks 48 farm bulk tanks 27 farm bulk tanks 1720 farm bulk tanks Bulk tank samples Bulk tank samples Bulk tank samples Bulk tank samples Dairy bulk tanks Dairy milk supply At retail Various sources Individual samples Bulk tank samples Bulk tank samples Individual samples Bulk tank samples Bulk tank samples
237 153 292 48 69 1,720 131 248 62 127 130 71 985 2,477 1,078 69 344 1,391 56 63
0.4 5.9 12.3 0 1.4 0.5 9.2 2.2 1.6 10.2 1.5 0 5.9 0.04 0 0 0 0 0 0
USA UK USA Poland FR Canada USA USA Ireland Pakistan Poland New Zealand UK UK Germany Germany Switzerland Germany Germany Switzerland
1987 1988 1992 1996 1997 1997 1997 2001-2002 2001-2002 2002-2004 1996 1987 1988 1985 1992 1992 2002 1992 1992 2002
McManus and Lanier (1987) Humphrey and Hart (1988) Rohrbach et al. (1992) Gomolka and Uradzinski (1996) Desmasures et al. (1997) Steele et al. (1997) Jayarao and Henning (2001) Jayarao et al. (2006) Whyte et al. (2004) Hussain et al. (2007) Gomolka and Uradzinski (1996) Stone (1987) Humphrey and Hart (1988) Roberts (1985) Hahn et al. (1992) Hahn et al. (1992) Muehlherr et al. (2003) Hahn et al. (1992) Hahn et al. (1992) Muehlherr et al. (2003)
“Total number of samples examined. bDetection of Cumpylobucter, based on enrichment culture.
In general, milk-borne infection can be adequately controlled by proper pasteurization and the prevention of recontamination after heat treatment.
RED MEAT Like dairy cows, meat-producing cattle may excrete C. jejuni in their feces, with a higher incidence occurring in summer than in winter (Blaser et al., 1983). A carriage rate as high as 89% was reported by Stanley et al. (1998b) in samples of 360 cattle during 1 year. Numbers of thermotolerant Campylobacter were about log 2 MPN/g of fresh feces (Stanley et al., 1998b); this is much lower than the log 6 to 9 commonly found in poultry feces (Berndtson et al., 1992). During a 1-year study, the prevalence of Campylobacter in Finnish beef and dairy cattle at slaughter varied monthly between 18.8 and 44.1%, with beef cattle being more often colonized than dairy cattle. C. jejuni was found mostly in young animals, while C. hyointestinalis subsp. hyointestinalis was detected more frequently in older animals (Hakkinen et al., 2007). Inglis et al. (2004), by means of real-time quantitative PCR, found C. lanienae to be the most prevalent species detected in bovine feces. Swine are also frequently found to harbor Campylobacter, but in this animal reservoir, C. coli is isolated significantly more often than C. jejuni (Fricker and Park, 1989; Mafu et al., 1989; Oporto et al., 2007). Numbers of the organism in intestines are
more comparable to the numbers in cattle than to the numbers found in poultry (Oosterom et al., 1983). Lambs at slaughter were found to carry thermotolerant Campylobacter in 91.7% of 360 samples of the small intestine, which appears to be the most important site of colonization in these animals. Numbers average log 4.0 MPN/g of fresh intestinal content (Stanley et al., 1998a). The level of Campylobacter carriage in sheep and the presence of the bacterium on sheep’s meat at slaughter, however, is reported to be low (Stanley et al., 1998a). Zweifel et al. (2004) collected cecum samples from 653 slaughtered sheep from two Swiss abattoirs and found a Campylobacter prevalence of 17.5%. The slaughtering processes of cattle, swine, and sheep can lead to contamination of the meat with intestinal flora during manual evisceration, but this may occur less frequently than during poultry slaughtering, where the evisceration process is often mechanical and has a high turnover rate. Lammerding and coworkers (1988) isolated high percentages of Campylobacter from cattle, sheep, and pigs just after slaughter and before chilling. Overnight forced-air chilling of the carcasses is supposed to greatly reduce the numbers of Campylobacter on the carcass (EFSA, 2005; Oosterom et al., 1983; Stern and Kazmi, 1989). Ghafir et al. (2007) extensively reports on the prevalence of Campylobacter in poultry, pork, and beef meat at different stages of the production chain in Belgium over several years. Table 4 summarizes the
CHAPTER 35
CAMPYLOBACTER IN T H E FOOD SUPPLY
635
Table 4. Presence of thermotolerant Campylobacter in meat products, excluding poultry Product
Stage of process
nu
% Positiveb
Country
Year
Beef carcasses Beef meat Beef meat Beef Beef Beef Ox liver Unweaned veal Pork freshly ground meat Pork freshly ground sausage Pork carcasses Pork meat Pork Pork Pork Pork meat Pork ground meat and sausage Pork liver Lamb Lamb liver Lamb and mutton
Before chilling Processing plant and retail At retail At retail At retail At retail At retail At retail Processing plant
948 151 250 230 22 1 182 96 90 20
3.5 1.3 1.6 3.5 3.2 0.5 54.2 10.0 0
Finland Italy Korea New Zealand Ireland USA UK New Zealand USA
2003 2000-2001 2001-2006 2003-2004 2001-2002 1999-2000 1998 2003-2004 2000
Hakkinen et al. (2007) Pezzotti et al. (2003) Hong et al. (2007) Wong et al. (2007) Whyte et al. (2004) Zhao et al. (2001) Kramer et al. (2000) Wong et al. (2007) Duffy et al. (2001)
20
0
USA
2000
Duffy et al. (2001)
Processing plant Processing plant and retail At retail At retail At retail At retail At retail
20 175 197 181 230 250 348
62.0 10.3 5.1 1.7 9.1 1.2 1.3
Denmark Italy Ireland USA New Zealand Korea USA
2004 2000-2001 2001-2002 1999-2000 2003-2004 2001-2006 2000
Malakauskas et al. (2006) Pezzotti et al. (2003) Whyte et al. (2004) Zhao et al. (2001) Wong et al. (2007) Hong et al. (2007) Duffy et al. (2001)
At At At At
99 262 96 23 1
71.7 11.8 72.9 6.9
UK Ireland UK New Zealand
1998 2001-2002 1998 2003-2004
Krarner et al. (2000) Whyte et al. (2004) Kramer et al. (2000) Wong et al. (2007)
Processing plant
retail retail retail retail
Reference
“Total number of samples examined. bDetection of Campylobucter, based on enrichment culture.
results of several large and more recent studies on the presence of Campylobacter in various types of meat, excluding poultry. Earlier studies are summarized by Jacobs-Reitsma (2000). As stated for the results on poultry products and milk, these data should be interpreted with care because of the variety of methodologies used. Numbers of Campylobacter on both pork and beef are reported to be low. These contamination levels of about the detection limit of methodologies may contribute significantly to the differences in isolation rates in the various studies. Sufficient heating of red meat products, which are relatively infrequently contaminated with low numbers of Campylobacter, will eliminate this risk of human infection.
WATER AND SEAFOOD Contaminated surface water is recognized as a source of Campylobacter outbreaks in humans; it may also play a role in the contamination of farm animals. The organism could also be isolated from various other water environments, such as streams, seawater, and other recreational waters. These waters are not intended for human consumption, but accidental ingestion of a significant amount of these waters may pose a risk. Municipal water systems, which
are sometimes chlorinated, are regarded to be free of Campylobacter, as long as safety procedures are maintained and mixing with potentially contaminated water sources is avoided. Campylobacter contamination of surface water is likely to originate from fecal contamination by wild birds or domestic animals, or from sewage effluent (Blaser et al., 1983; EFSA, 2005; Koenraad et al., 1997; NACMCF, 1994). Hanninen et al. (2003) described three waterborne outbreaks in Finland caused by C. jejuni. The cause of contamination in two of these cases was probably the runoff of surface water into groundwater wells after heavy precipitation; the cause of the third outbreak remained unresolved. Kuusi et al. (2005) described an outbreak of campylobacteriosis that occurred in 1998 in northern Finland, and drinking nonchlorinated, nonboiled municipal tap water was found to be strongly associated with the illness. Gallay et al. (2006) reported a large waterborne outbreak in a local community in France. C. coli was detected in the stools of patients in combination with group A rotavirus and norovirus. An extensive environmental investigation concluded that a groundwater source had probably been contaminated by agricultural runoff, and a failure in the chlorination system was identified. This was reported to be the first documented waterborne outbreak of C. coli.
636
TACOBS-REITSMA ET AL.
Moore et al. (2001) found untreated surface water from lakes and rivers to be a source of Campylobacter. The contamination here could be due to sewage release, watersheds, runoff from agricultural or residual areas, fecal contamination from wildlife, or floods. Savill et al. (2001) described Campylobacter findings of 29% in 24 tested samples of reticulated drinking water. However, the concentration of Campylobacter in all samples was low, with a maximum MPN of 0.3 per 100 ml. Most of the isolates were identified as C. lari, and differences in survival compared with C. jejuni and C. coli were speculated on. Drinking bottled water was one of the five independent risk factors for Campylobacter infections identified within a case-control study by Evans et al. (2003). Tatchou-Nyamsi-Konig et al. (2007) therefore studied the survival of C. jejuni in bottled filtered natural mineral water at different temperatures. They concluded that in the event of dual contamination (Campylobacter and organic matter), C. jejuni could survive in natural mineral waters at low temperatures (4°C) and in the presence of oxygen. As a consequence of the transient existence of Campylobacter in the marine environment, shellfish may become contaminated by the organism. Consumption of raw clams and raw Pacific oysters has been described as a source of Campylobacter infection in humans (Abeyta et al., 1993). Several investigations on the detection of Campylobacter in different types of seafood have been carried out. The relatively high isolation percentages of C. lari in these products implicate seagulls, which are often found colonized with this particular Campylobacter species, as the primary source of contamination. In a Dutch study (Endtz et al., 1997), Campylobacter was isolated from 41 out of 59 batches of mussels and from 11 out of 41 batches of oysters. High percentages of mussels from Germany, Denmark, and England (65%), as well as oysters from Northern Ireland (25%), were also culture positive for Campylobacter. Further identification of 39 strains isolated from 24 Dutch mussel batches revealed that all but two isolates were C. lari (Endtz et al., 1997). In a study performed in the United Kingdom during 1994 (Wilson and Moore, 1996), 47% of 331 cockles, mussels, and scallops examined shortly after harvesting were found to contain Campylobacter. Three (6%) of 49 samples of depurated and ready-to-eat oysters contained Campylobacter. Most isolations were during the cooler months of the year (November to March). The majority of strains were classified most likely to be C. lari (57% urease-positive thermophilic campylobacters and 24% “classical” C. lari) (Wilson and Moore, 1996). In 2002, three oyster samples from
129 raw shellfish samples examined (2.3%) were found to contain C. jejuni (Whyte et al., 2004). Campylobacter was also isolated from 50 (21%) of 240 samples of fresh hand-picked crab meat from 12 different blue crab processing facilities in the United States (Reinhard et al., 1996). Quantitative levels in all cases were below the detection limit of 0.30 MPN/g. Isolates were phenotypically determined to be 36 C. jejuni and 14 C. coli strains (Reinhard et al., 1996). Shellfish beds located near potentially contaminated sources like sewage effluents, farmland runoff, and waterfowl reservoirs were reported to present a human health risk in the consumption of raw oysters (Abeyta et al., 1993). Moreover, depuration of oysters was shown to be not fully effective in the elimination of potential bacterial pathogens like Cumpylobacter. This indicates that some thermal processing of shellfish is generally advisable (EFSA, 2005; Wilson and Moore, 1996).
OTHER FOODS Evans et al. (2003) compared 213 patients with sporadic Campylobacter infection with patients with negative fecal-samples, and they confirmed that eating chicken was a well-established risk factor. However, they also identified eating salad vegetables other than lettuce (e.g., tomatoes and cucumber) and drinking bottled water as two new potential risk factors for the illness. Several studies are available on the detection of Campylobacter in vegetable food types, although only few of them report on the presence of the organism, and only in low isolation rates of less than 5%. Theoretically, crops may occasionally become contaminated by the application of natural fertilizers, by wild birds and animals, or by contaminated surface water. Also, cross-contamination of fresh produce during harvest, handling, and processing may occur (EFSA, 2005). In a large survey in Canada (Park and Sanders, 1992), a total of 1,564 fresh samples of 10 vegetable types from farmers’ outdoor markets and supermarkets were examined for the presence of thermotolerant campylobacters. No Campylobacter was detected in the 1,031 samples from supermarkets. The detection rates in samples from the outdoor markets ranged from 1.6 to 3.3% and included vegetables like spinach, lettuce, radish, green onions, parsley, and potatoes. No Campylobacter was detected in celery, carrots, cabbage, or cucumber. Campylobacter was also detected in 14 (2.9%) of another 482 outdoor market samples, but all these samples were negative
CHAPTER 35
for Campylobacter after thoroughly washing them with chlorinated water (Park and Sanders, 1992).Another Canadian study (Odumeru et al., 1997) detected no Campylobacter in any of 65 unprocessed or 296 fresh-cut and packaged ready-to-use vegetables like lettuce, carrots, cauliflower, celery, broccoli, or sliced green peppers. In a study of 3,200 uncooked ready-to-eat organic vegetables on retail markets in the United Kingdom, no Campylobacter was detected (Sagoo et al., 2001). McMahon and Wilson (2001) did not detect Campylobacter in 86 organic vegetable samples from large supermarket chains in Northern Ireland. Low isolation rates of 2 (3.6%) of 56 fresh vegetables and 2 (0.5%) of 400 ready-to-use vegetables were reported by Kumar et al. (2001) and FCdCrighi et al. (1999), respectively. Exceptionally high rates were presented by Chai et al. (2007), with average prevalences from 29.4 to 67.7% of Campylobacter in 309 samples of raw vegetables used in ulam, a popular Malaysian salad dish, obtained from three different supermarkets. A study by the Seattle-King County Health Department revealed an increased relative risk of developing campylobacteriosis in individuals that consume mushrooms. This caused Doyle and Schoeni (1986) to examine retail mushrooms for the presence of Campylobacter. C. jejuni was isolated from 3 of 200 twenty-five-gram samples from wrapped packages of fresh mushrooms obtained from local grocery stores. The three positive samples were packaged by the same mushroom producer, but it is unknown how the mushrooms became contaminated. The composted manure used as growth substrate for these mushrooms had been heated to 60°C for 2 h and therefore was not considered a likely source of contamination (Doyle and Schoeni, 1986). Whyte et al. (2004) isolated C. jejuni from 2 out of 217 retail mushroom samples, but McMahon and Wilson (2001) did not isolate the organism from 86 samples of fresh organic mushrooms. Reports on outbreaks of campylobacteriosis attributed to rare sources, such as tuna salad (Roels et al., 1998), are suggested to be in fact a result of crosscontamination during preparation, therefore serving as a vehicle rather than being the original source of infection.
SURVIVAL OF CAMPYLOBACTER O N FOODS The optimum growth temperature of thermotolerant campylobacters lies between 3 7 and 42"C, very close to the body temperature of the chicken of 42°C. Thermotolerant campylobacters are not able to grow below 3 O"C, so actual multiplication during handling
CAMPYLOBACTER IN THE FOOD SUPPLY
637
or storage at room temperature in the kitchen will not readily occur in moderate climates. Especially during storage at room temperature, campylobacters are sensitive to drying. In general, Campylobacter survives better at cooling temperatures than at room temperature. Still, refrigerated storage reduces the level of contamination. In a study by Stern (1995), levels on fresh broiler carcasses as high as log 5.8 were in some cases reduced to nondetectable levels after storage for 10 days at 4°C. Although freezing substantially reduces numbers of Campylobacter cells, survival at -20°C is possible, and Campylobacter could still be recovered at very low levels from frozen chicken drumsticks after 3 months of storage. Also, isolation rates for Campylobacter from frozen meat and poultry products were reported to be five times lower than from the corresponding fresh products (NACMCF, 1994; Stern and Kazmi, 1989). Fernandez and Pison (1996) isolated Campylobacter from a large number of frozen chicken liver samples and typed 22% of the isolates to be C. jejuni; whereas 78% were C. coli. It was suggested that C. coli is better at surviving under frozen storage (and possibly also under other environmental circumstances) than C. jejuni (Fernandez and Pison, 1996). Campylobacters are sensitive to heat and are inactivated by the pasteurization treatments used in practice. D values (decimal reduction values) reported for the organism in skim milk at 48°C ranged from 7.2 to 12.8 min and at 55°C from 0.74 to 1.0 min. High-temperature short-duration pasteurization for 80 s at 60°C also rapidly killed levels of log 6 inoculated in raw milk. D values in meat ranged from 5.9 to 6.3 min for heating at 50°C and less than 1 min for heating at 60°C. D values for C. jejuni in ground chicken heated at 49 or 57°C were about 20 min and 45 s, respectively. Meatballs of ground beef inoculated with more than log 6 Campylobacterlg were cooked to an internal temperature of 60"C, and Campylobacter numbers were reduced to nondetectable levels within 10 rnin (NACMCF, 1994; Stern and Kazmi, 1989). The optimal NaCl concentration for the growth of Campylobacter is 0.5%, but the organism is sensitive to higher concentrations of NaC1, although the extent depends both on the temperature and the food medium. Certain spices, such as oregano, sage, and cloves, were reported to inhibit growth of Campylobacter at 42°C in a 0.5% broth (Stern and Kazmi, 1989). Properly chlorinated drinking water can be regarded as free of Campylobacter. Chlorination of scalding or chilling water during the processing of poultry carcasses is allowed in several countries, but
638
JACOBS-REITSMA ET AL.
the effects are reported to be minimal, most likely as a result of neutralization by organic matter in the chilling tank and the campylobacters being protected by attachment to the skin. In an attempt to reduce contamination of carcasses with Campylobacter during processing, Mead and coworkers (1995) used chlorinated water sprays on equipment and working surfaces, increased chlorine concentrations in processing water, and removed any unnecessary carcass contact surface. These changes significantly reduced the levels of carcass contamination, but in practice, the reduction was considered to have little impact on consumer exposure to Cumpylobucter infection (Mead et al., 1995). A concentration of 0.05% of ascorbic acid inhibits the growth of Campylobacter, and a concentration of 0.09% is bactericidal (NACMCF, 1994). Lactic acid and acetic acid sprays or dips were evaluated for application during the slaughtering process, and Campylobacter contamination levels were found to be reduced (Bolder, 1997; Stern and Kazmi, 1989). Modified atmospheres or vacuum packaging have little effect on the survival of the microaerophilic campylobacters. When stored at 4"C, campylobacters survived just as well on beef or chicken packaged with oxygen-permeable wrap or various modified atmospheres or with vacuum (NACMCF, 1994; Stern and Kazmi, 1989). Campylobacter is quite sensitive to gamma irradiation, with a D value of 32 krad. It was reported that C. jejuni in culture medium and in chicken paste survived an irradiation dose of 0.2 kGy, but not a 1kGy dose. Campylobacter is also more sensitive to ultraviolet irradiation at 254 nm than, e.g., Escherichia coli (NACMCF, 1994). Various disinfecting agents effectively inactivate Campylobacter. Phenolic compounds, iodophors, quaternary ammonium compounds, 70% ethyl alcohol, and glutaraldehyde in commonly applied concentrations were able to inactivate about log 7 Campylobacterlml within 1 min. Levels of 1.25 ppm sodium hypochlorite destroyed log 4 campylobacters per ml within 1 min, and 5 ppm NaOCl killed log 7 CFU per ml within 15 min (Stern and Kazmi, 1989). The majority of Campylobacter studies on growth characteristics and survival were carried out during the early 1 9 8 0 and ~ ~ summarizing reviews can be found in articles by Doyle (1984b) and Stern and Kazmi (1989). The overall results indicate that depending on the Campylobacter strain, the initial number of cells, the type of food, and other environmental conditions, in particular the storage temperature, campylobacters may survive in food products for long periods of time (Doyle, 1984b). In addition to laboratory experiments, there is an urgent need for a
better understanding of the behavior of campylobacters in the food chain and how this affects their ability to survive the various food production processes in practice (Havelaar et al., 2005; Humphrey et al., 2007).
CROSS-CONTAMINATION DURING (DOMESTIC) FOOD PREPARATION The role of cross-contamination has been discussed as an important contributory factor in a number of case-control studies and domestic outbreaks caused by Campylobacter. Contamination pathways can be either direct, from raw meat to products that will not undergo further cooking, or indirect, via work surfaces, hands, or utensils (Cogan et al., 1999; Humphrey et al., 2001; Gorman et al., 2002; Luber et al., 2006; Mattick et al., 2003b). Transfer from unwashed cutting boards to the prepared meal was assumed to be the most important route of cross-contamination. Several factors may influence the number of transferred Campylobacter, such as drip fluid amount, the contact area between raw chicken and the cutting board, and the contact time (Cogan et al., 2002; Rosenquist et al., 2003). It is important to note that external and internal surfaces of poultry packaging may function as a potential vehicle for introducing Campylobacter in retail markets and domestic kitchens. Harrison et al. (2001) isolated Campylobacter from 3% of external and 34% of whole packaging overall. Burgess et al. (2005) reported 3 and 0.8% Campylobacter contamination of the external surface of 895 chicken and 129 turkey meat packages, respectively. In order to estimate the exposure of consumers to Campylobacter during the preparation and consumption of meat, several authors sampled surface (swab samples) and internal parts (deep tissue) of chicken meat separately (Luber and Bartelt, 2007; Scherer et al., 2006). They found low numbers of Campylobacter in the chicken muscle compared and higher numbers on the meat surface (Table 2). Luber and Bartelt (2007) concluded that crosscontamination during the preparation of contaminated poultry is more important for the exposure of consumers to Campylobacter than the consumption of undercooked poultry meat. In contrast to other studies, Luber et al. (2006) used naturally contaminated fresh retail chicken parts and simulated typical situations and handling procedures common in domestic kitchens. The experiments revealed high Campylobacter counts on hands. The importance of unwashed hands compared with
CHAPTER 35
contaminated kitchen utensils as a vehicle for Campylobacter transfer will depend mostly on the consumer behavior in the kitchen. Some authors gave various recommendations for implementing domestic food hygiene in the kitchen. Mattick et al. (2003b) studied the survival of bacteria in food debris on dishes before washing up by air drying up to 72 h at 21°C. Because Campylobacter survived this poorly, they concluded that dirty dishes should be left for some time before washing up. Mattick et al. (2003a) recommended that the water temperature during washing up and rinsing dishes before washing up be should be as high as possible. Combining detergent-based cleaning and rinsing was found to be consistently effective for surfaces contaminated with Campylobacter (Cogan et al., 1999, 2002). Provision of adequate information to all consumers and proper education and training for everyone involved in food preparation and handling are control options for reduced Campylobacter exposure at the consumer level (EFSA, 2005). Although the information campaigns on TV and radio and in the newspapers seem to have contributed to the reduction in Campylobacter infections in Iceland (Stern et al., 2003), it is also considered to be difficult to change people’s behavior permanently, or at least to measure the effectiveness of educational efforts (Havelaar et al., 2005).
PREVENTION OF FOOD-BORNE TRANSMISSION OF HUMAN CAMPYLOBACTERIOSIS Because poultry products are regarded as accounting for the majority of Campylobacter infections in humans, the reduction of the Campylobacter contamination of these products is considered to have the greatest impact on this human health problem. Reduction of the potential risk of contaminated poultry products has to be achieved by the application of good hygienic practices by both the producers of poultry meat products and the consumers of these products. Consumers need professional education programs, starting at primary school levels, in combination with adequate consumer information on proper handling of foods at risk. Product labeling is sometimes used to provide consumers with useful handling and preparation instructions. The National Commission on Microbiological Safety of Foods (NACMCF, 1994) provided a fact sheet with essential basic information about Campylobacter for retail food, food service, and regulatory agency personnel,
-
CAMPYLOBACTER IN THE FOOD SUPPLY
639
as well as a fact sheet for consumers about foodborne illness caused by Campylobacter bacteria. The production of poultry meat products needs hazard analysis critical control point principles to be applied during all stages of production, processing, and distribution (NACMCF, 1994, 1997). As described by the NACMCF (1997), the overall objective of these principles is to ensure that production and processing are conducted to enhance the microbiological safety of the product. This is achieved through effective management of key operations that realistically prevent or control the introduction or growth of pathogens. First, total prevention of Campylobacter colonization of live broiler flocks at the farm level, thereby delivering Campylobacter-free material to the slaughterhouse, would be the neatest way to reduce the level of contaminated poultry products. However, it appears to be complicated to effect this in practice (EFSA, 2005; Havelaar et al., 2005). The impossibilities of intervention strategies at the farm level are discussed in more detail in chapter 37. Second, improvements in processing procedures at the slaughterhouse level and decontamination treatment of end products might be considered. Several procedures are suggested to minimize Campylobacter contamination during poultry processing. These include the following, alone or in some combination: using counterflow water systems during scalding and chilling; rinsing and washing equipment to minimize or reduce cross-contamination; disinfecting carcasses and contact surfaces with chlorine or other bacterial control treatments like trisodium phosphate or lactic acid; or simply freezing the carcasses (Bolder, 1997; EFSA, 2005; Havelaar et al., 2005; Rosenquist et al., 2006). So-called logistic slaughter of known Campylobacter-free flocks first thing in the morning, followed by killing pathogenpositive flocks at the end of the day did relatively little to reduce the risk for consumers. However, additional treatment of the positive-tested flocks only (so they were not used for fresh broiler meat production) reduced this risk substantially (Havelaar et al., 2005). To be applicable in practice, this scheduled treatment requires an accurate and punctual surveillance system for all broiler flocks and a relatively low percentage of Campylobacter infection. In the United States and several other countries, irradiation of packaged fresh or frozen poultry products at 1.5 to 3.0 kGy may be applied as an effective treatment to eliminate Campylobacter from the end product (Bolder, 1997; Havelaar et al., 2005; NACMCF, 1994). Other recognized food-related sources of human campylobacteriosis, in particular raw milk and con-
640
TACOBS-REITSMA ET AL.
taminated water, can be avoided relatively easily: consumers should refrain from drinking any raw milk and should consume only properly heat-treated milk, and they should drink water only from approved sources. These strategies should be undertaken under the realistic assumption that treatment procedures in milk and drinking water production are properly controlled and that recontamination is avoided.
CONCLUSIONS
Campylobacter bacteria are widespread in warmblooded food-producing animals, and during processing, the food products may easily become contaminated. In addition, different types of seafood can be found contaminated with this human pathogen. Campylobacter may survive in food products for long periods of time, depending on the initial number and type of cells, the specific food product, and environmental conditions like storage temperature. Only a small number of Campylobacter ingested via contaminated or cross-contaminated food may lead to infection and disease in humans. The risk of consuming raw foods of animal origin in relation to human campylobacteriosis is evident. This risk can be avoided by only consuming pasteurized milk; thoroughly heating red meat, poultry, and seafood; and drinking water from approved sources. In addition, good food handling practices in the home and in the food-service industry will reduce the risk of illness. REFERENCES Abeyta, C., F. G. Deeter, C. A. Kaysner, R. F. Stott, and M. M. Wekell. 1993. Campylobacter jejuni in a Washington State shellfish growing bed associated with illness. ]. Food Prot. 56:323325. Aho, M., and J. Hirn. 1988. Prevalence of campylobacteria in the Finnish broiler chicken chain from the producer to the consumer. Acta Vet. Scand. 29:451-462. Alter, T., F. Gaul], A. Froeb, and K. Fehlhaber. 2005. Distribution of Campylobacter jejuni strains at different stages of a turkey slaughter line. Food Microbiol. 22:345-351. Alter, T., M. Giirtler, F. G a d , A. Johne, and K. Fehlhaber. 2004. Comparative analysis of the prevalence of Campylobacter spp. in retail turkey and chicken meat. Arch. Lebensmittelhyg. 55:4972. Arsenault, J., A. Letellier, S. Quessy, J. P. Morin, and M. Boulianne. 2007. Prevalence and risk factors for Salmonella and Campylobacter spp. carcass contamination in turkeys slaughtered in Quebec, Canada.]. Food Prot. 70:1350-1359. Atanassova, V., F. Reich, L. Beckmann, and G. Klein. 2007. Prevalence of Campylobacter spp. in turkey meat from a slaughterhouse and in turkey meat retail products. FEMS lmmunol. Med. Microbiol. 49:141-145. Atanassova, V., and C. Ring. 1999. Prevalence of Campylobacter spp. in poultry and poultry meat in Germany. Int. J. Food Microbiol. 5 l:187-190.
Bachmann, H. P., and U. Spahr. 1995. The fate of potentially pathogenic bacteria in Swiss hard and semihard cheeses made from raw milk.]. Dairy Sci. 78:476-483. Berndtson, E., M. Tivemo, and A. Engvall. 1992. Distribution and numbers of Campylobacter in newly slaughtered broiler chickens and hens. lnt. J. Food Microbiol. 15:45-50. Berrang, M. E., J. S. Bailey, S. F. Altekruse, B. Patel, W. K. Shaw, Jr., R. J. Meinersmann, and P. J. Fedorka-Cray. 2007. Prevalence and numbers of Campylobacter on broiler carcasses collected at rehang and postchill in 20 U.S. processing plants. J. Food Prot. 70:1556-1560. Berrang, M. E., R J. Buhr, and J. A. Cason. 2000. Campylobacter recovery from external and internal organs of commercial broiler carcass prior to scalding. Poultry Sci. 79:286-290. Beumer, R R., J. J. Cruysen, and I. R Birtantie. 1988. The occurrence of Campylobacter jejuni in raw cow’s milk. 1. Appl. Bacteriol. 65:93-96. Blaser, M. J., D. N. Taylor, and R. A. Feldman. 1983. Epidemiology of Campylobacter jejuni infections. Epidemiol. Rev. 5: 157-176. Bolder, N. M. 1997. Decontamination of meat and poultry carcasses. Trends Food Sci. Technol. 8:221-227. Bryan, F. L., and M. P. Doyle. 1995. Health risks and consequences of Salmonella and Campylobacter jejuni in raw poultry. J. Food P70t. 58~326-344. Burgess, F., C. L. Little, G. Allen, K. Williamson, and R. T. Mitchelli. 2005. Prevalence of Campylobacter, Salmonella, and Escherichia coli on the external packaging of raw meat. J. Food P70t. 68~469-475. Casanovas, L., M. de Simon, M. D. Ferrer, J. Arques, and G. Monzon. 1995. Intestinal carriage of campylobacters, salmonellas, yersinias and listerias in pigeons in the city of Barcelona.]. Appl. Bacteriol. 78:ll-13. Chai, L. C., T. Robin, U. M. Ragavan, J. W. Gunsalam, F. A. Bakar, F. M. Ghazali, S. Radu, and M. P. Kumar. 2007. Thermophilic Campylobacter spp. in salad vegetables in Malaysia. Int.]. Food Microbiol. 117:106-111. Cogan, T. A., S. F. Bloomfield, and T. J. Humphrey. 1999. The effectiveness of hygiene procedures for prevention of crosscontamination from chicken carcasses in the domestic kitchen. Lett. Appl. Microbiol. 29:354-358. Cogan, T. A., J. Slader, S.F Bloomfield, and T. J. Humphrey. 2002. Achieving hygiene in the domestic kitchen: the effectiveness of commonly used cleaning procedures. ]. Appl. Microbiol. 92~885-892. Corry, J. E. L., H. I. Atabay, S. J. Forsythe, and L. P. Mansfield. 2003. Culture media for the isolation of campylobacters, heliobacters and arcobacters, p. 271-316. In J. E. L. Corry, G. D. W. Curtis, and R. M. Baird (ed.), Handbook of Culture Media for Food Microbiology, 2nd ed. Elsevier, Amsterdam. Cuomo, A., L. Dipineto, A. Santaniello, G. Matteoli, T. Sarli, D. D. Vecchia, A. Fioretti, and L. F. Menna. 2007. Detection of thermotolerant Campylobacter in ostriches (Struthio camelus) in Italy. Vet. J. 174:439-441. Cuk, Z., A. Annan-Prah, M. Janc, and J. Zajc Satler. 1987. Yoghurt: an unlikely source of Campylobacter jejunilcoli. J. Appl. Bacteriol. 63:201-205. Denis, M., M.-J. Refrtgier-Petton, M.-J. Laisney, G. Ermel, and G. Salvat. 2001. Campylobacter contamination in French chicken production from farm to consumers. Use of a PCR assay for detection and identification of Campylobacter jejuni and Campylobacter coli. J. Appl. Microbiol. 91:255-267. Desmasures, N., F. Bazin, and M. Gueguen. 1997. Microbiological composition of raw milk from selected farms in the Camembert region of Normandy. J. Appl. Microbiol. 8353-58.
CHAPTER 35
Dominguez, C., I. Gbmez, and J. Zumalacarregui. 2002. Prevalence of Salmonella and Campylobacter in retail chicken meat in Spain. lnt. 1.Food Microbiol. 72:165-168. Doyle, M. P. 1984a. Association of Campylobacter with laying hens and eggs. Appl. Environm. Microbiol. 47533-536. Doyle, M. P. 1984b. Campylobacter in foods, p. 163-180. In J. P. Butzler (ed.), Campylobacter Infection in Man and Animals. CRC Press, Boca Raton, FL. Doyle, M. P., and D. J. Roman. 1982. Prevalence and survival of Campylobacter jejuni in unpasteurized milk. Appl. Environ. Microbiol. 44:1154-1158. Doyle, M. P., and J. L. Schoeni. 1986. Isolation of Campylobacter jejuni from retail mushrooms. Appl. Environ. Microbiol. 51:449450. Duffy, E. A., K. E. Belk, J. N. Sofos, G. R. Bellinger, A. Pape, and G. C. Smith. 2001. Extent of microbial contamination in United States pork retail products.]. Food Prot. 64:172-178. Endtz, H. P., J. S. Vliegenthart, P. Vandamme, H. W. Weverink, N. P. van den Braak, H. A. Verbrugh, and A. van Belkum. 1997. Genotypic diversity of Campylobacter Iari isolated from mussels and oysters in The Netherlands. Int. 1.Food Microbiol. 34:7988. European Food Safety Authority (EFSA). 2006a. Report of Task Force on Zoonoses Data Collection on proposed technical specifications for a coordinated monitoring programme for Salmonella and Campylobacter in broiler meat in the EU. EFSA]. 92: 1-33. European Food Safety Authority (EFSA). 2005. Scientific report of the scientific panel on biological hazards on the request from the commission related to Campylobacter in animals and foodstuffs. Annex EFSAJ. 173:l-105. European Food Safety Authority (EFSA). 2006b. Trends on sources of zoonoses, zoonotic agents and antimicrobial resistance in the EU in 2004. EFSA]. 2005:310. http://www.efsa.eu.int/ scienceI monitoring-zoonoses/ reports/ 1277-en.htm1. Evans, M. R., W. Lane, J. A. Frost, and G. Nylen. 1998. A Cumpylobacter outbreak associated with stir-fried food. Epidemiol. Infect. 121:275-279. Evans, M. R., C. D. Ribeiro, and R. L. Salmon. 2003. Hazards of healthy living: bottled water and salad vegetables as risk factors for Campylobacter infection. Emerg. Infect. Dis. 9:1219-1225. Fahey, T., D. Morgan, C. Gunneburg, G. K. Adak, F. Majid, and E. Kaczmarski. 1995. An outbreak of Campylobacter jejuni enteritis associated with failed milk pasteurisation. ]. Infect. 31: 137- 143. FtdCrighi, M., C. Magras, M. F. Pilet, D. Woodward, W. Johnson, F. Jugiau, and J. L. Jouve. 1999. Incidence of thermotolerant Campylobacter in foods assessed by NF I S 0 10272 standard: results of a two-year study. Int. 1.Food Microbiol. 16: 195-204. Fernandez, H., and V. Pison. 1996. Isolation of thermotolerant species of Campylobacter from commercial chicken livers. lnt. 1.Food Microbiol. 29:75-80. Finch, M. J., and P. A. Blake. 1985. Foodborne outbreaks of campylobacteriosis: the United States experience, 1980-1982. Am. 1.Epidemiol. 122:262-268. Fricker, C. R., and R. W. A. Park. 1989. A two-year study of the distribution of “thermophilic” campylobacters in human, environmental and food samples from the Reading area with particular reference to toxin production and heat-stable serotype. J. Appl. Bacteriol. 66:477-490. Gallay, A., H. De Valk, M. Cournot, B. Ladeuil, C. Hemery, C. Castor, F. Bon, F. MCgraud, P. Le Cam, J. C. Desenclos; Outbreak Investigation Team. 2006. A large multi-pathogen waterborne community outbreak linked to faecal contamination of a
CAMPYLOBACTER IN THE FOOD SUPPLY
641
groundwater system, France, 2000. Clin. Microbiol. Infect. 12: 561-570. Ghafir, Y., B. China, K. Dierick, L. De Zutter, and G. Daube. 2007. A seven-year survey of Campylobacter contamination in meat at different production stages in Belgium. lnt. 1.Food Microbiol. 116:lll-120. Gorman, R., S. Bloomfield, and C. C. Adley. 2002. A study of cross-contamination of food-borne pathogens in the domestic kitchen in the Republic of Ireland. Int. 1. Food Microbiol. 76: 143-150. Gomolka, M., and J. Uradzinski. 1996. Occurrence of Campylobacter species in raw milk in north eastern Poland. Mikhwissenschafc 51:366-368. Hahn, G., H. Kirchhoff, P. Hammer, E. H. Ubben, and W. H. Heeschen. 1992. Bacteriological findings and their evaluation for milk and dairy products. Arch. Lebensmittelhyg. 43:89-93. Hakkinen, M., H. Heiska, and M.-L. Hanninen. 2007. Prevalence of Campylobacter spp. in cattle in Finland and antimicrobial susceptibilities of bovine Campylobacter jejuni strains. Appl. Environ. Microbiol. 73:3232-323 8 Hanninen, M.-L., H. Haajanen, T. Purnmi, K. Wermundsen, M.-L. Katila, H. Sarkkinen, I. Miettinen, and H. Rautelin. 2003. Detection and typing of Campylobacter jejuni and Campylobacter coli and analysis of indicator organisms in three waterborne outbreaks in Finland. Appl. Environ. Microbiol. 69:1391-1396. Harris, N. V., T. J. Kimball, P. Bennett, Y. Johnson, D. Wakeley, and C. M. Nolan. 1987. Campylobacter jejuni enteritis associated with raw goat’s milk. Am. 1.Epidemiol. 126:179-186. Harrison, W. A., C. J. Griffith, D. Tennant, and A. C. Peters. 2001. Incidence of Campylobacter and Salmonella isolated from retail chicken and associated packaging in South Wales. Lett. Appl. Microbiol. 3 3:45 0-454. Havelaar, A. H., M. J. Nauta, M.-J. J. Mangen, A. G. de Koeijer, M.-J. Bogaardt, E. G. Evers, W. F. Jacobs-Reitsma, W. van Pelt, J. A. Wagenaar, G. A. de Wit, and H. van der Zee. 2005. Costs and Benefits of Controlling Campylobacter in The Netherlands. Integrating Rtsk Analysis, Epidemiology and Economics. RIVM report 250911009/2005. Bilthoven, The Netherlands. Hong, J., J. M. Kim, W. K. Jung, S. H. Kim, W. Bae, H. C. Koo, J. Gil, M. Kim, J. Ser, and Y. H. Park. 2007. Prevalence and antibiotic resistance of Campylobacter spp. isolated from chicken meat, pork, and beef in Korea, from 2001 to 2006.1. Food Prot. 70:860-866. Hudson, S. J., N. F. Lightfoot, J. C. Coulson, K. Russell, P. R. Sisson, and A. 0. Sobo. 1991. Jackdaws and magpies as vectors of milk-borne human Campylobacter infection. Epidemiol. lnfect. 107:363-372. Humphrey, T. J., and P. Beckett. 1987. Campylobacter jejuni in dairy cows and raw milk. Epidemiol. Infect. 98:263-269. Humphrey, T. J., and R. J. Hart. 1988. Campylobacter and Salmonella contamination of unpasteurized cow’s milk on sale to the public. 1.Appl. Bacteriol. 65:463467. Humphrey, T. J., K. W. Martin, J. Slader, and K. Durham. 2001. Campylobacter spp. in the kitchen: spread and persistence. J. Appl. Microbiol. 90:115S-l2OS. Humphrey, T. J., S. O’Brien, and M. Madsen. 2007. Campylobacters as zoonotic pathogens: a food production perspective. Int. 1.Food Microbiol. 117:237-257. Hunt, J. M., C. Abeyta, and T. Tran. 1998. Campylobacter. In FDA Bacteriological Analytical Manual Online, 8th ed. http: I1 www.cfsan.fda.gov/ -ebam/ bam-7.html. Hussain, I., M. Shahid Mahmood, M. Akhtar, and A. Khan. 2007. Prevalence of Campylobacter species in meat, milk and other food commodities in Pakistan. Food Microbiol. 24:219-222. Hutchinson, D. N., F. J. Bolton, P. M. Hinchliffe, H. C. Dawkins, S. D. Horsley, E. G. Jessop, P. A. Robertshaw, and D. E.
642
TACOBS-REITSMA ET AL.
Counter. 1985. Evidence of udder excretion of Campylobacter jejuni as the cause of milk borne Campylobacter outbreak. J. Hyg. (Lond.) 94:205-215. Inglis, G. D., L. D. Kaischuk, and H. W. Busz. 2004. Chronic shedding of Campylobacter species in beef cattle. J. Appl. Microbiol. 97~410-20. IS0 (International Organization for Standardization). 2006a. Microbiology of food and animal feeding stuffs-horizontal method for detection and enumeration of Campylobacter spp. Part 1: detection method. I S 0 10272-1:2006. I S 0 (International Organization for Standardization). 2006b. Microbiology of food and animal feeding stuffs-horizontal method for detection and enumeration of Campylobacter spp. Part 2: enumeration method. ISO/TS 10272-2:2006. I S 0 (International Organization for Standardization). 2005. Water quality-detection and enumeration of thermotolerant Campylobacter species. IS0 17995:2005. Izat, A. L., and F. A. Gardner. 1988. Incidence of Campylobacter jejuni in processed egg products. Poultry Sci. 67:143 1-1435. Jacobs-Reitsma, W. F. 2000. Campylobacter in the food supply. p. 467-481. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM, Washington, DC. Jacobs-Reitsma, W. F. 1994. Epidemiology of Campylobacter in poultry. Thesis. Wageningen Agricultural University, The Netherlands. Jacobs-Reitsma, W. F., W. van Overbeek, S. Oostra-van Dijk, and H. J. Aarts. 2007. Implementation of I S 0 10272:2006 for detection and enumeration of Campylobacter in foods. Zoonoses Public Health 54(Suppl. 1):49. Jayarao, B. M., S. C. Donaldson, B. A. Straley, A. A. Sawant, N. V. Hegde, and J. L. Brown. 2006. A survey of foodborne pathogens in bulk tank milk and raw milk consumption among farm families in Pennsylvania. J. Dairy Sci. 89:2451-2458. Jayarao, B. M., and D. R. Henning. 2001. Prevalence of foodborne pathogens in bulk tank milk. J. Dairy Sci. 84:2157-2162. Jeffrey, J. S., K. H. Tonooka, and J. Lozano. 2001. Prevalence of Campylobacter spp. from skin, crop, and intestine of commercial broiler chicken carcasses at processing. Poultry Sci. 80:13901392. Jones, P. H., A. T. Willis, D. A. Robinson, M. B. Skirrow, and D. S. Josephs. 1981. Campylobacter enteritis associated with the consumption of free school milk. J. Hyg. (Lond.) 87:155-162. Jsrgensen, F., R. Bailey, S. Williams, P. Henderson, D. R. A. Wareing, F. J. Bolton, J. A. Frost, L. Ward, and T. J. Humphrey. 2002. Prevalence and numbers of Salmonella and Campylobacter spp. on raw, whole chickens in relation to sampling methods. Int. J. Food Microbiol. 76:151-164. Kapperud, G., and 0. Rosef. 1983. Avian wildlife reservoir of Campylobacter fetus subsp. jejuni, Yersinia spp., and Salmonlla spp. in Norway. Appl. Environm. Microbiol. 45:375-380. Klein, G., F. Reich, L. Beckmann, and V. Atanassova. 2007. Quantification of thermophilic Campylobacter spp. in broilers during meat processing. Antonie Van Leeuwenhoek, 92:267-273. Koenraad, P. M. F. J., F. M. Rombouts, and S. H. W. Notermans. 1997. Epidemiological aspects of thermophilic Campylobacterin water related environments: a review. Water Environm. Res. 69: 52-63. Kramer, J. M., J. A. Frost, F. J. Bolton, and D. R. A. Wareing. 2000. Carnpylobacter contamination of raw meat and poultry at retail sale: identification of multiple types and comparison with isolates from human infection. J. Food Prot. 63:1654-1659. Kumar, A., R. K. Agarwal, K. N. Bhilegaonkar, B. R. Shome, and V. N. Bachhil. 2001. Occurrence of Campylobacter jejuni in vegetables. Int. J. Food Microbiol. 67: 153-155. Kuusi, M., J. P. Nuorti, M.-L. Hanninen, M. Koskela, V. Jussila, E. Kela, I. Miettinen, and P. Ruutu. 2005. A large outbreak of
campylobacteriosis associated with a municipal water supply in Finland. Epidemiol. Infect. 133593-601. Lammerding, A. M., M. M. Garcia, E. D. Mann, Y. Robinson, W. J. Dorward, R. B. Truscott, and F. Tittiger. 1988. Prevalence of Salmonella and thermophilic Campylobacter in fresh pork, beef, veal and poultry in Canada. J. Food Prot. 51:47-52. Lindmark, H., C. Diedrich, L. Andersson, R. Lindqvist, and E. Olsson Engvall. 2006. Distribution of Campylobacter genotypes on broilers during slaughter. J. Food Prot. 69:2902-2907. Luber, P., and E. Bartelt. 2007. Enumeration of Campylobacter spp. on the surface and within chicken breast fillets. J. Appl. Microbiol. 102:313-318. Luber, P., S. Brynestad, D. Topsch, K. Scherer, and E. Bartelt. 2006. Quantification of Campylobacter species cross-contamination during handling of contaminated fresh chicken parts in kitchens. Appl. Environ. Microbiol. 72:66-70. Mafu, A. A., R. Higgins, M. Nadeau, and G. Cousineau. 1989. The incidence of Salmonella, Campylobacter, and Yersinia enterocolitica in swine carcasses and the slaughterhouse environment. J. Food Prot. 52:642-645. Malakauskas, M., K. Jorgensen, E. M. Nielsen, B. Ojeniyi, and J. E. Olsen. 2006. Isolation of Campylobacter spp. from a pig slaughterhouse and analysis of cross-contamination. Int. J. Food Micro biol. 108:295-3 00. Manfreda, G., A. De Cesare, V. Bondioli, N. J. Stern, and A. Franchini. 2006. Enumeration and identity of Campylobacter spp. in Italian broilers. Poultry Sci. 85556-562. Mateo, E., J. Circamo, M. Urquijo, I. Perales, and A. FernindezAstorga. 2005. Evaluation of a PCR assay for the detection and identification of Campylobacter jejuni and Campylobacter coli in retail poultry products. Res. Microbiol. 1 5 6 5 6 8 4 7 4 . Mattick, K., K. Durham, G. Domingue, F. Jsrgensen, M. Sen, D. W. Schaffner, and T. Humphrey. 2003b. The survival of foodborne pathogens during domestic washing-up and subsequent transfer onto washing-up sponges, kitchen surfaces and food. Int. J. Food Microbiol. 85:213-226. Mattick, K., K. Durham, M. Hendrix, J. Slader, C. Griffith, M. Sen, and T. Humphrey. 2003a. The microbiological quality of washing-up water and the environment in domestic and commercial kitchens. J. Appl. Microbiol. 94:842-848. McMahon, M. A., and I. G. Wilson. 2001. The occurrence of enteric pathogens and Aeromonas species in organic vegetables. Int. J. Food Microbiol. 70:155-162. McManus, C., and J. M. Lanier. 1987. Salmonella, Campylobacter jejuni, and Yersinia enterocolitica in raw milk. J. Food Prot. 50: 51-55. Mead, G. C., W. R. Hudson, and M. H. Hinton. 1995. Effect of changes in processing to improve hygiene control on contamination of poultry carcasses with Campylobacter. Epidemiol. Infect. 115:495-500. Moore, J.EW., P. S. Caldwell, B. C. Millar, and P. G. Murphy. 2001. Occurrence of Campylobacter spp. in water in Northern Ireland: implications for public health. Ulster Med. J. 70:102107. Muehlherr, J. E., C. Zweifel, S. Corti, J. E. Blanco, and R. Stephan. 2003. Microbiological quality of raw goat’s and ewe’s bulk-tank milk in Switzerland. J. Dairy Sci. 86: 849-3856. NACMCF (National Advisory Committee on Microbiological Criteria for Foods). 2005. Analytical utility of Campylobacter methodologies. http://www.fsis.usda.gov/PDF/NACMCFCampylobacter-092805.pdf. NACMCF (National Advisory Committee on Microbiological Criteria for Foods). 1994. Campylobacter jejunilcoli. J. Food Prot. 57:llOl-1121. NACMCF (National Advisory Committee on Microbiological Criteria for Foods). 1997. Generic HACCP application in broiler slaughter and processing. J. Food Prot. 60579-604.
CHAPTER 35
Nauta, M., I. van der Fels-Klerx, and A. Havelaar. 2005. A poultry-processing model for quantitative microbiological risk assessment. Risk Anal. 25:85-98. Nielsen, E. M., V. Fussing, J. Engberg, N. L. Nielsen, and J. Neimann. 2006. Most Campylobacter subtypes from sporadic infections can be found in retail poultry products and food animals. Epidemiol. Infect. 134:758-767. Odumeru, J. A., S. J. Mitchell, D. M. Alves, J. A. Lynch, A. J. Yee, S . L. Wang, S. Styliadis, and J. M. Farber. 1997. Assessment of the microbiological quality of ready to use vegetables for health care food services. J. Food Prot. 60:954-960. Oosterom, J., G. J. A. de Wilde, E. de Boer, L. H. de Blaauw, and H. Karman. 1983. Survival of Campylobacter jejuni during poultry processing and pig slaughtering. 1. Food Prot. 46:702-706, 709. Oporto, B., J. I. Esteban, G. Aduriz, R. A. Juste, and A. Hurtado. 2007. Prevalence and strain diversity of thermophilic campylobacters in cattle, sheep and swine farms. J. Appl. Microbiol. 103: 977-984. Orr, K. E., N. F. Lightfoot, P. R Sisson, B. A. Harkis, J. L. Tweddle, P. Boyd, A. Carroll, C. J. Jackson, D. R A. Wareing, and R. Freeman. 1995. Direct milk excretion of Campylobacter jejuni in a dairy cow causing cases of human enteritis. Epidemiol. Infect. 114:15-24. Oyarzabal., 0.A., D. E. Comer, and F. J. Hoerr. 1995. Incidence of campylobacters in the intestine of avian species in Alabama. Avian Dis. 39:147-151. Park, C. E., and G. W. Sanders. 1992. Occurrence of thermotolerant campylobacters in fresh vegetables sold at farmers’ outdoor markets and supermarkets. Can. J. Microbiol. 38:313-316. Perko-Makela, P., M. Koljonen, M. Miettinen, and M.-L. Hanninen. 2000. Survival of Campylobacter jejuni in marinated and nonmarinated chicken products. J. Food Safety 20:209-216. Peterson, M. C. 2003. Campylobacter jejuni enteritis associated with consumption of raw milk. J. Environ. Health 65:20-21. Pezzotti, G., A. Serafin, I. Luzzi, R. Mioni, M. Milan, and R. Perin. 2003. Occurrence and resistance to antibiotics of Campylobacter jejuni and Campylobacter coli in animals and meat in northeastern Italy. Int. ]. Food Microbiol. 82:281-287. Reinhard, R. G., T. J. McAdam, G. J. Flick, R. E. Croonenberghs, R. F. Wittman, A. A. Diallo, and C. Fernandes. 1996. Analysis of Campylobacter jejuni, Campylobacter coli, Salmonella, Klebsiella pneumoniae, and Escherichia coli 0157:H7 in fresh handpicked blue crab (Callinectes sapidus) meat. J. Food Prot. 59: 803-807. Roberts, D. 1985. Microbiological aspects of goat’s milk-a public health laboratory service survey. Goat Vet. SOC.J. 6:lO-13. Roels, T. H., B. Wickus, H. H. Bostrom, J. J. Kazmierczak, M. A. Nicholson, T. A. Kurzynski, and J. I?. Davis. 1998. A foodborne outbreak of Campylobacter jejuni (0:33) infection associated with tuna salad: a rare strain in an unusual vehicle. Epidemiol. Infect. 121:281-287. Rohrbach, B. W., F. A. Draughon, P. M. Davidson, and S . P. Oliver. 1992. Prevalence of Listeria monocytogenes, Campylobacter jejuni, Yersinia enterocolitica, and Salmonella in bulk tank milk: risk factors and risk of human exposure.]. Food Prot. 55: 93-97. Rosenquist, H., N. L. Nielsen, H. M. Sommer, B. Norrungand, and B. B. Christensen. 2003. Quantitative risk assessment of human campylobacteriosis associated with thermophilic Campylobacter species in chickens. lnt. ]. Food Microbiol. 83:87103. Rosenquist H., H. M. Sommer, N. L. Nielsen, and B. B. Christensen. 2006. The effect of slaughter operations on the contami-
CAMPYLOBACTER IN THE FOOD SUPPLY
643
nation of chicken carcasses with thermotolerant Campylobacter. Int. 1.Food Microbiol. 108:226-232. Sagoo, S. K., C. L. Little, and R. T. Mitchell. 2001. The microbiological examination of ready-to-eat organic vegetables from retail establishments in the United Kingdom. Lett. Appl. Microbiol. 33:434-439. Sahin, O., P. Kobalka, and Q. Zhang. 2003. Detection of Campylobacter in chicken eggs. J. Appl. Microbiol. 95:1070-1079. Savill, M. G., J. A. Hudson, A. Ball, J. D. Klena, P. Scholes, R. J. Whyte, R. E. McCormick, and D. Jankovic. 2001. Enumeration of Campylobacter in New Zealand recreational and drinking waters.]. Appl. Microbiol. 91:38-46. Scherer K., E. Bartelt, C. Sommerfeld, and G. Hildebrandt. 2006. Quantification of Campylobacter on the surface and in the muscle of chicken legs at retail.]. Food Prot. 69:757-761. Schildt, M., S . Savolainen, and M. L. Hanninen. 2006. Longlasting Campylobacter jejuni contamination of milk associated with gastrointestinal illness in a farming family. Epidemiol. Infect. 134:401-405 Shane, S. M., D. H. Gifford, and K. Yogasundram. 1986. Campylobacter jejuni contamination of eggs. Vet. Res. Comm. 10: 487-492. Son, I., M. D. Englen, M. E. Berrang, P. J. Fedorka-Cray, and M. A. Harrison. 2007. Prevalence of Archobacter and Campylobacter on broiler carcasses during processing. Int. /. Food Microbiol. 113~16-22. Stanley, K. N., J. S. Wallace, J. E. Currie, P. J. Diggle, and K. Jones. 1998a. Seasonal variation of thermophilic campylobacters in lambs at slaughter.]. Appl. Microbiol. 84:1111-1116. Stanley, K. N., J. S. Wallace, J. E. Currie, P. J. Diggle, and K. Jones. 1998b. The seasonal variation of thermophilic campylobacters in beef cattle, dairy cattle, and calves. J. Appl. Microbiol. 85~472-480. Steele, M. L., W. B. McNab, C. Poppe, M. W. Griffiths, Chen Shu, S. A. Degrandis, L. C. Fruhner, C. A. Larkin, J. A. Lynch, and J. A. Odumeru. 1997. Survey of Ontario bulk tank raw milk for food-borne pathogens.]. Food Prot. 60:1341-1346. Stern, N. J. 1995. Influence of season and refrigerated storage on Campylobacter spp. contamination of broiler carcasses. 1.Appl. Poultry Res. 4:235-23 8. Stern, N. J., K. L. Hiett, G. A. Alfredsson, K. G. Kristinsson, J. Reiersen, H. Hardardottir, H. Briem, E. Gunnarsson, F. Georgsson, R Lowman, E. Berndtson, A. M. Lammerding, G. M. Paoli, and M. T. Musgrove. 2003. Campylobacter spp. in Icelandic poultry operations and human disease. Epidemiol. Infect. 130: 23-32. Stern, N. J., and S. U. Kazmi. 1989. Campylobacter jejuni, p. 71110. In M. P. Doyle (ed.), Food-borne Bacterial Pathogens. Marcel Dekker, New York. Stone, D. L. 1987. A survey of raw whole milk for Campylobacter jejuni, Listeria monocytogenes and Yersinia enterocolitica. N. Z. J. Dairy Sci. Technol. 22:257-264. Stoyanchev, T., I. Vashin, C. Ring, and V. Atanassova. 2007. Prevalence of Campylobacter spp. in poultry and poultry products for sale on the Bulgarian retail market. Antonie Van Leeuwenhoek 92:285-288. Sulonen, J., R. Karenlampi, U. Holma, and M.-L. Hanninen. 2007. Campylobacter in Finnish organic laying hens in autumn 2003 and spring 2004. Poult. Sci. 86:1223-1228. Tatchou-Nyamsi-Konig, J.-A., A. Moreau, M. FCdCrighi, and J.-C. Block. 2007. Behavior of Campylobacter jejuni in experimentally contaminated bottled natural mineral water. J. Appl. Microbiol. 103:280-288. Uyttendaele, M., K. Baert, Y. Ghafir, G. Daube, L. De Zutter, L.
644
TACOBS-REITSMA ET AL.
Herman, K. Dierick, D. Pierard, J. J. B. Horion, and J. Debevere. 2006. Quantitative risk assessment of Campylobacter spp. in poultry based meat preparations as one of the factors to support the development of risk-based microbiological criteria in Belgium. Int. J. Food Microbiol. 111:149-63. Vellinga, A., and F. Van Loock. 2002. The dioxin crisis as experiment to determine poultry-related Campylobacter enteritis. Epidemiol. Infect. Dis. 8:19-22. Whyte, P., K. McGill, D. Cowley, R. H. Madden, L. Moran, P. Scates, C. Carroll, A. O’Leary, S. Fanning, J. D. Collins, E. McNamara, J. E. Moore, and M. Cormican. 2004. Occurrence of Campylobacter in retail foods in Ireland. Int. J. Food Microbiol. 95: 111-1 18. Wilson, I. G., and J. E. Moore. 1996. Presence of Salmonella spp. and Campylobacter spp. in shellfish. Epidemiol. Infect. 116:147153.
Wong, T. L., L. Hollis, A. Cornelius, C. Nicol, R. Cook, and J. A. Hudson. 2007. Prevalence, numbers, and subtypes of Campylobacter jejuni and Campylobacter coli in uncooked retail meat samples. 1.Food Prot. 70566-573. Zhao, T., M. P. Doyle, and D. E. Berg. 2000. Fate of Cumpylobacter jejuni in butter. J. Food Prot. 63:120-122. Zhao, C., G . Beilei, J. De Villena, R. Sudler, E. Yeh, S. Zhao, D. G. White, D. Wagner, and J. Meng. 2001. Prevalence of Campylobacter spp., Escherichia coli, and Salmonella Serovars in retail chicken, turkey, pork, and beef from the greater Washington, DC, area. Appl. Environ. Microbiol. 675431-5436. Zweifel, C., M. A. Zychowska, and R. Stephen. 2004. Prevalence and characteristics of Shiga toxin-producing Escherichia coli, Salmonella spp. and Campylobacter spp. isolated from slaughtered sheep in Switzerland. Int. J. Food Microbiol. 92:45-53.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser 0 2008 ASM Press, Washington, DC
ChaDter 36
Transmission of Antibiotic Resistance from Food Animals to Humans FRANKM. AARESTRUP, PATRICKF. MCDERMOTT, AND HENRIK C. WEGENER
Campylobacter jejuni and Campylobacter coli are among the most common causes of bacterial diarrhea in humans worldwide (Nachamkin et al., 2000). Of these, C. jejuni accounts for the vast majority (>95%) of diagnosed infections. Patients usually recover without antimicrobial therapy, but in some patients with severe, prolonged, or relapsing illness, therapy may be indicated. Macrolides are normally considered the drug of choice, but fluoroquinolones (FQs) are also recommended, and they are the drugs of choice for empirical treatment of the undiagnosed diarrheal condition (Blaser, 1990; Goodman et al., 1990; Petruccelli et al., 1992; Salazar-Lindo et al., 1986; Skirrow and Blaser, 2000). However, increases in the occurrence of Campylobacter causing infections in humans that are resistant to macrolides and FQs have been reported in several countries (Endtz et al., 1991; Gaudreau and Gilbert, 1998; Hoge et al., 1998; Rautelin et al., 1991; Reina et al., 1994; Sanchez et al., 1994; Sjogren et al., 1997; Smith et al., 1999). Food of animal origin is considered the most important sources of Campylobacter causing infections in humans. Consequently, the development of antimicrobial resistance in Campylobacter spp. due to the use of antimicrobial agents in food animals is a matter of concern. Several studies have reported a frequent or increasing occurrence of resistance to macrolides and FQs among Campylobacter from food animals (Cabrita et al., 1992; Endtz et al., 1991; Hariharan et al., 1990; Kaneuchi et al., 1988; Saenz et al., 2000; Wang et al., 1984). This chapter reviews the current knowledge on antimicrobial susceptibility, occurrence of antimicrobial resistance, and transmission of antimicrobial resistant Campylobacter from food animals.
ANTIMICROBIAL SUSCEPTIBILITY TESTING Methods In general, there are two different methods for testing the susceptibility of a bacterial isolate to antimicrobial agents, dilution and diffusion methods. Several variations of both methods have been used worldwide for susceptibility testing of Campylobacter. This includes a number of diffusion (disk, tablets, Etest) and dilution (macro- and broth microdilution, and agar dilution) methods. Campylobacter requires microaerobic conditions and supplemented media for growth. Large variations in media and incubation conditions have been used. Minimal inhibitory concentrations (MICs) and inhibition zones can vary greatly depending on the method and medium used. Thus, it is important that results are generated by means of validated testing methods to ensure appropriate categorization of the isolates. Reference methods for in vitro antimicrobial susceptibility testing are established in multilaboratory studies, with testing conducted over several days. Ideally, the method should be easy to perform, flexible, and inexpensive. Even though the testing itself is fairly easy to perform, consistency and strict adherence to protocol are essential to ensure reliable data. Each method requires the use of an appropriate quality-control organism or organisms along with specific quality-control ranges for each drug of interest, a standardized test medium that supports the growth of clinical isolates, an appropriate inoculum size, and standardized growth parameters (atmosphere, incubation temperature, and time). Although many bacteria of clinical interest may be tested by means of the same in vitro testing conditions, bacteria
Frank M. Aarestrup and Henrik C. Wegener National Food Institute, Bulowsvej 27, DK-1790 Copenhagen V, Denmark. Patrick F. McDermott * National Antimicrobial Resistance Monitoring System, Center for Veterinary Medicine, U.S.Food & Drug Administration, 8401 Muirkirk Rd., Mod 2, Laurel, MD 20708.
645
646
AARESTRUP ET AL.
that require increased incubation times, different incubation temperatures or atmospheric conditions, or a different growth medium may require the use of genus-specific testing conditions. Thus, although Cumpylobucter jejuni may inhabit the same in vivo environment as Escherichiu coli, its in vitro cultivation and susceptibility testing require growth media and atmospheric conditions appropriate to the fastidious nature of the genus. For this same reason, it is necessary to validate genus-specific quality-control organisms. Because in vitro antimicrobial susceptibility methods are relatively simple, it is not uncommon for laboratories to generate their own variations of the standardized testing methods for Cumpylobucter, then interpret the results on the basis of breakpoints recognized for the standardized method or for other pathogens. It should be emphasized that this is not a valid way to generate or report susceptibility data and could lead to unreliable and misleading data. For example, Campylobacter jejuni and C. coli can be screened for ciprofloxacin resistance by displaying no zone of inhibition when a 5-pg disk mass is used (CLSI M45 document). This no-zone-of-inhibition phenotype correlates >99% with isolates displaying ciprofloxacin MIC values >2 mg/L. It is not valid, however, to use this endpoint when applying disks containing other than 5 pg of ciprofloxacin. Only data generated by means of standardized methods can be used to categorize a bacterium as susceptible, intermediate, or resistant to a given antimicrobial agent. Methods that have not been independently validated in a multilaboratory format must demonstrate an acceptable level of agreement with the reference testing method. The goal of the method validation process is to ensure intra- and interlaboratory reproducibility of the testing method. Once a standardized method is established, reliable in vitro susceptibility data can be generated for clinical isolates for use in setting interpretive criteria predictive of clinical response. A number of different volunteer consensus organizations publish standards for antimicrobial susceptibility testing and interpretive criteria (clinical breakpoints). The most widely used are those published by the Clinical and Laboratory Standards Institute (CLSI; http: //www.clsi.org), the European Committee for Antimicrobial Susceptibility Testing (EUCAST; http: //www.eucast.org), and the International Organization for Standardization (ISO; http://www.iso.org). The goal of these organizations is to enhance the clinical use of antimicrobial agents and increase the likelihood of therapeutic success. An agar dilution method for susceptibility testing of Cumpylobucter has been developed and recognized
by CLSI (McDermott et al., 2004). This method was developed with quality-control ranges for five antimicrobials (ciprofloxacin, doxycycline, gentamicin, erythromycin, and meropenem) with C. jejuni ATCC 33560 as the quality-control organism. Subsequently, a broth microdilution method was developed with quality-control ranges for 14 different antimicrobials established (McDermott et al., 2005). The guidelines give detailed information on media and incubation conditions, including temperature and time. Currently the agar and broth microdilution methods are the only internationally recognized methods for susceptibility testing of Cumpylobucter. A number of studies have compared different methods for susceptibility testing of Cumpylobucter. In general, these studies have shown good agreement between test methods (Engberg et al., 1999; Gaudreau and Gilbert, 1997; Luber et al., 2003; Oncul et al., 2003). However, it is currently not possible to correlate the exact MIC values obtained from, e.g., Etest and agar dilution (Ge et al., 2002) or MIC and disk diffusion (Luangtongkum et al., 2007). Thus, further standardization of diffusion methods are needed before these can be recommended for susceptibility testing of Cumpylobucter. Clinical Breakpoints and Epidemiological Cutoff Values Because adequate, well-controlled clinical outcome trials have not been performed for the treatment of campylobacteriosis, clinical breakpoints have not been formally established for any antimicrobial agent with in vitro activity against Cumpylobucter. This has also been hampered by a general lack of data on the pharmacokinetics of antimicrobials in the intestinal tract. Regardless, determination of the antimicrobial susceptibility of Cumpylobucter can still be used to guide the most appropriate therapy for clinical infections by identifying strains with acquired resistance traits. Because Cumpylobucter is the most common food-borne bacterial pathogen in many countries, surveillance systems to monitor this bacterium in the food supply have become more common. Thus, susceptibility testing is also valuable for generating public health monitoring data on the occurrence of acquired resistance among bacteria in Cumpylobucter from different reservoirs. Semiquantitative methods for determining the MIC of an antimicrobial agent for a given bacterial pathogen is the gold standard for susceptibility testing. However, because clinical efficacy data are sparse to nonexistent, different interpretive criteria are used in different laboratories. In addition, programs designed to monitor antimicrobial susceptibility over time may use break-
CHAPTER 36
points that might presage impending clinical resistance. Thus, MIC threshold values appropriate for predicting clinical efficacy might differ from those used for monitoring purposes. An isolate might, through mutations or horizontal gene transfer, develop reduced susceptibility to a given drug but still have a sufficiently low MIC to enable a successful clinical outcome from therapy. Thus, for monitoring purposes, this isolate may signify an emerging public health problem, yet remain categorized for clinical purposes as susceptible. It is therefore important to differentiate between MIC thresholds used for clinical purposes (clinical breakpoints) from those used for monitoring (epidemiological cutoff values). Clinical breakpoints The development of clinical breakpoints requires microbiological MIC data generated by means of standardized in vitro testing methods, pharmacokinetic and pharmacodynamic information, and, most importantly, outcome data from clinical efficacy trials (Watts and Lindemann, 2006). These three types of data, taken together, usually are sufficient to establish threshold antimicrobial concentration levels for infections that are likely to respond when treated at the approved dosages (susceptible organisms), and those that are expected to fail to respond when treated with the approved dosage (resistant organisms). The intermediate category is used as a buffer zone to account for day-to-day variability in in vitro antimicrobial susceptibility testing methods, to provide flexibility for sites of infection where the agent is concentrated, or for agents where increased dosage ranges are defined. Epidemiological cutoff values Epidemiological cutoff values focus on separating isolates in the normal wild-type population from isolates with some type of acquired mechanisms that reduces the normal susceptibility of these isolates. Thus, this threshold value does not take into consideration any data about doses or clinical efficacy, but is only based on microbiological data aimed at optimum detection of acquired decreases in susceptibility. When working with epidemiological cutoff values, there is no intermediate category; isolates are recorded as wild type or non-wild type. Epidemiological cutoff values are normally recommended for monitoring changes in susceptibility. However, it should be emphasized that an epidemiological cutoff value is not a clinical breakpoint. Sources of clinical breakpoints and epidemiological cutoff values A number of different national and international committees publish tables for clinical breakpoints.
TRANSMISSION OF ANTIBIOTIC RESISTANCE
647
The most widely used globally are those provided by CLSI, which publishes methods for susceptibility testing and also provides tables with clinical breakpoints approved by the U.S. Food and Drug Administration (FDA) in the United States. Clinical breakpoints are available for both MIC determination and disk diffusion testing for a large number of different antimicrobial agents and bacterial pathogens. In Europe, EUCAST provides epidemiological cutoff values and clinical breakpoints. Data are currently only available for MIC determinations. For Campylobacter, only epidemiological cutoff values are available. CLSI includes an intermediate category that is based on the presence or absence of known resistance determinants, whereas EUCAST only has a cutoff value for the wild-type population. In addition, CLSI recommend the same cutoff values for both C. coli and C. jejuni, whereas EUCAST has found it advantageous to establish species-specific values. Thus, some differences in the suggested epidemiological cutoff values are apparent (Table l), which clearly shows a need for international harmonization.
MECHANISM OF ANTIMICROBIAL RESISTANCE IN CAMPYLOBACTER Resistance to antimicrobial agents can be either intrinsic or acquired. Intrinsic resistance describes a general insensitivity of all isolates of a given bacterial species to an antimicrobial agent or class. Acquired resistance refers to qualitative genetic differences present in some strains of a species as the result of either mutations or the acquisition of foreign DNA. Knowledge on the genetic basis of resistance is important to understand the ecological dynamics of resistance and to identify potential reservoirs of resistance for human infections. Knowing whether a mechanism is transferable between strains, or whether the resistance only is transmitted with the resistant clone is important for understanding the relative risks associated with various types of resistance reservoirs. The most common resistance mechanisms identified in Campylobacter are outlined below and in Table 2. Aminoglycoside Resistance Resistance to aminoglycosides is normally mediated by enzymes that modify the drugs. These enzymes are divided into three different groups on the basis of the reaction they mediate (Shaw et al., 1993): aminoglycoside phosphotransferases (APH), aminoglycoside adenyltransferases (AAD or ANT), and aminoglycoside acetyltransferases (AAC). Many enzymes
648
AARESTRUP ET AL.
Table 1. Epidemiological cutoff values recommended by CLSI (http: //www.clsi.org) and EUCAST (http:I/www.eucast.org) Organism
Antimicrobial
Campylobacter jejuni
Erythromycin Ciprofloxacin Azithromycin Clindamycin Nalidixic acid Telithromycin Tetracycline Streptomycin Gentamicin Erythromycin Ciprofloxacin Azithromycin Clindamycin Nalidixic acid Telitromycin Tetracycline Streptomycin Gentamicin
Campylobacter coli
mediate aminoglycoside resistance, and the nomenclature for both genes and enzymes is complex (Shaw et al., 1993). In Campylobacter, kanamycin resistance is mediated by APH (3') types I, 111, IV, and VII and by APH (3") (Rivera et al., 1986; Taylor et al., 1988; Tenover and Elvrum, 1988; Tenover et al., 1992). The ant(3')-la and ant(6)-la genes have been found to encode streptomycin resistance in Campylobacter (Pinto-Alphandary et al., 1990; Shaw et al., 1993). The ant(3')-la gene (ANT(3')-I) is commonly found in gram-negative bacterial species, often as part of class I integrons, whereas the ant(6)-lagene (ANT(6)I) mainly has been found in staphylococci (PintoAlphandary et al., 1990; Shaw et al., 1993). Campylobacter have been shown to contain class I integrons, with the ant(3')-Ia gene identical to gene cassettes in Salmonella and E. coli, suggesting a transfer between these species (O'Halloran et al., 200). Gentamicin has been recommended for the treatment of campylobacteriosis, particularly in patients with systemic infection (Skirrow and Blaser, 2002). Resistance to gentamicin is rare in Campylobacter and has not been characterized at the genetic level. The sat4 gene encoding resistance to streptothricins has been observed in C. coli of different animal and clinical sources in Germany (Bischoff et al., 1996; Bottcher and Jacob, 1992). The different aminoglycoside resistance genes have also been found in other bacterial species, mainly those that are gram positive. This could indicate that Campylobacter mainly acquires horizontally transferred genes from gram-positive bacteria.
EUCAST cutoff value (mgiliter), R >
CLSI cutoff value (mg/liter), R >
4 1 16 2 2 1 16 1 -
16 2 4 4 32 8 8 4 16 2 4 4
32 2
4 2
32 8 8
4
However, the presence of ant(3')-la and APH(3 ') of type I indicates that genes also may be acquired from gram-negative bacterial species. P-Lactam Resistance With the exception of carbapenems, the majority of C. jejuni and C. coli strains are resistant to plactam agents, principally penicillins and cephalosporins. For this reason, p-lactams are often incorporated into Campylobacter selective growth media used in primary isolation protocols. However, Campylobacter are moderately sensitive to cefotaxime, ceftazidime, and cefpirome (van der Auwera et al., 1985). Resistance to p-lactam antibiotics is in most bacterial species caused by the production of plactamases that break the p-lactam ring of the antibiotics. However, in some bacteria, changes in the penicillin-binding proteins or lack of penetration of the drug into the bacteria are the main mechanisms of resistance. A large proportion of C. coli and C. jejuni produce P-lactamases (Lachance et al., 1991, 1993; Tajada et al., 1996). However, the P-lactamase of C. jejuni and C. coli seems to play a role only in resistance to amoxicillin, ampicillin, and ticarcillin. With penicillin G, piperacillin, and cephalosporins, the mechanism of resistance in C. jejuni and C. coli is primarily considered to result from their lowaffinity penicillin-binding proteins and on low permeability (Lachance et al., 1991; Tajada et al., 1996). Because of this intrinsic resistance in many isolates, p-lactam susceptibility is normally not determined for Campylobacter.
Table 2. Antimicrobial resistance mechanism in Campylobacter Antimicrobial
Resistance mechanisms
Aminoglycosidesa Kanamycin, neomycin, gentamicin B Kanamycin, neomycin, gentamicin B Kanamycin, neomycin Kanamycin, neomycin Streptomycin
Inactivation of the drug by chemical modification
Streptomycin, spectinomycin Streptothricins p-lactam Chloramphenicol Fluoroquinolones
Resistance genes or mutations
Transmissible
Host-range
Reference
aph(3’)-Ia
+
Enterobacteriaceae, gram-positive cocci
Ouellette et al. (1987), Pinto-Alphandary et al. (1990), Shaw et al. (1993)
aph(3‘)-1II
Gram-positive cocci
Derbise et al. (1997), Shaw et al. (1993)
ant(6)-Ia
+ + + +
Gram-positive cocci Campylo bacter Gram-positive cocci
ant(3 I)-Ia
+
Gram-negative bacteria
Shaw et al. (1993) Tenover et al. (1992) Derbise et al. (1997), Pinto-Alphandary et al. (1990), Shaw et al. (1993) Pinto-Alphandary et al. (1990), Shaw et al. Derbise et al. (1997), Moon et al. (1996) Lachance et al. (1991, 1993) Wang and Taylor (1990)
aph(3 ’)-U aph(3 ‘)-VlI
(1993)
ND
ND
Gram-positive cocci ND
cat
ND
ND
sat4
Production of p-lactamses or low permeability Inactivation of the drug by chemical modification Mutations in the drugsensitive target
gyrA; Ala-70 to Thr; Thr-86 to Ile, Lys, Ala, Val; Asp90 to Ala, Asn, Tyr
-
+
Macrolides
Mutations in the drugsensitive target
23s rDNA; A to G at position 2075; A to C at position 2074
Sulfonamides
Mutations in the drugsensitive target Protection of the target
Mutations in folP gene
-
tet(0)
+
Replacement of the sensitive target by a new enzyme
dfrl
Zirnstein et al. (1999), Ruiz et al. (1998), Wang et al. (1993), Zirnstein et al. (1999), Bachoual et al. (2001), Cooper et al. (2002), Piddock et al. (2003), Luo et al. (2003)
Campylobacter
Jensen and Aarestrup (2001), Niwa et al. (2001), Vacher et al. (2003), Payot et al. (2004), Gibreel et al. (ZOOS), Kim et al. (2006)
Tetracyclines Trimethoprim
“Only the most commonly used for therapy.
dfr9
ND ND
Gibreel and Skold (1999) Gram-positive cocci
Enterobacteriaceae
Buu-Hoi et al. (1989), Manavathu et al. (1990), Zilhao et al. (1988) Gibreel and Skold (1998, 2000)
650
AARESTRUP ET AL.
evidence is available, mutation in gyrA is still the most important mechanism (Payot et al., 2006).
Chloramphenicol Resistance Chloramphenicol resistance is mainly due to the production of chloramphenicol acetyltransferase enzymes (encoded by the cut genes) that acetylate chloramphenicol, thereby preventing binding to the ribosome (Shaw, 1992). In C. coli a single cut gene has been identified (Wang and Taylor, 1990). Resistance to chloramphenicol has not become widespread among Campylobucter. Fluoroquinolone Resistance Fluoroquinolones inhibit the activity of DNA gyrase, and in most bacterial species, resistance is due to mutations in the gyrase or topoisomerase genes. Although in other enteric pathogens such as E. coli and Salmonella, two topoisomerase gene mutations (typically in gyrA and purC) are needed to confer high-level resistance to FQs, only a single mutation is needed in Cumpylobucter. Resistance appears after mutations in the gyrA gene encoding part of the A subunits of DNA gyrase. Cloning and sequencing of the C. jejuni gyrA gene have demonstrated that mutations in gyrA at positions Thr-86, Asp-90, and Ala70 are often observed (Wang et al., 1993), with mutation at position threonine-8 6-isoleucine (ACT ATT) being most common (Bachoual et al., 2001; Cooper et al., 2002; Piddock et al., 2003; Ruiz et al., 1998; Wang et al., 1993; Zirnstein et al., 1999). Other substitutions of the gyrA gene have also been observed, including Ala-70-Thr (Wang et al., 1993), Thr-86-Lys (Luo et al., 2003; Ruiz et al., 1998), Thr86-Ala (Bachoual et al., 2001), Thr-86-Val (Piddock et al., 2003), Asp-90-Asn (Bachoual et al., 2001; Luo et al., 2003; Piddock et al., 2003; Wang et al., 1993), and Asp-90-Tyr (Bachoual et al., 2001). In rare isolates of Cumpylobacter that are nalidixic acid resistant and ciprofloxacin susceptible, the Thr86Ala substitution was identified (Jesse et al., 2006). Double mutations resulting in two amino acid modifications have also been reported, including Thr-86-Ile and Pro-104-Ser (Piddock et al., 2003) and Thr-86-Ile and Asp-90-Asn (Ruiz et al., 2005). The secondary FQ target in other gram-negative enteric organisms, purC, is absent in Cumpylobacter (Payot et al., 2006). A number of different efflux systems might also contribute to the emergence of FQ resistance (Ge et al., 2005; Lin et al., 2002; Pumbwe and Piddock, 2002; Pumbwe et al., 2005). Thus, efflux systems might mediate low-level resistance so that Cumpylobucter more easily under selective pressure might survive and develop mutations mediating full FQ resistance (Yan et al., 2006). However, until further
-
Macrolide Resistance Macrolides are bacteriostatic agents that inhibit peptide chain elongation by binding to the 50s subunit of the bacterial ribosome, encoded by the 23s rRNA genes. Acquired resistance to macrolides can be based on different mechanisms: target modification by point mutation or methylation of 23s rRNA, hydrolysis of the macrolide lactone ring, and the activity of energy-dependent efflux pumps (Schwarz et al., 2006). In Campylobucter, it has been shown that acquired resistance is not consistent with the presence of an rRNA methylase, with modification of the antibiotic, or with efflux (Yan and Taylor, 1991). Sequencing of 23s rRNA genes from erythromycin resistant and susceptible C. coli and C. jejuni have mainly identified mutations at positions A2074 and A2075 (corresponding in E. coli to positions 2058 and 2059), of the 23s rRNA at the erythromycin biding site (Gibreel et al., 2005; Jensen and Aarestrup, 2001; Niwa et al., 2001; Payot et al., 2004; Vacher et al., 2003). Campylobucter contains three copies of the rRNA gene; evidence suggests that at least two copies must be mutated to cause resistance (Gibreel et al., 2005). An A2075G transition is the most frequent mutation observed in clinical strains. It is usually present in all three copies of the 23s rRNA gene (Gibreel et al., 2005) and can confer high MICs (>128 pg/ml). Resistance acquired through mutations in the chromosome is normally not considered to be transferable. However, Gibreel et al. (2005) were able to transfer a PCR fragment of the 23s region containing the macrolide resistance mutations into a C. jejuni strain by natural transformation. Furthermore, Kim et al. (2006) succeeded in transferring erythromycin resistance by using whole DNA extracts from different donor strains into different C. coli recipients from turkey and swine. Mutations affecting macrolide binding have also been identified in the ribosomal proteins L4 and L22 (Corcoran et al., 2006). The role of these mutations in macrolide resistance in Campylobacter requires further investigations. As with FQ resistance, efflux pumps are known to play a role in the intrinsic reduced susceptibility to macrolides (Cagliero et al., 2005; Lin et al., 2002; Payot et al., 2004; Pumpwe and Piddock, 2002; Pumpwe et al., 2004) and might contribute to the selection of full macrolide resistance. Gibreel et al. (2007) found that inactivation of the cmeB gene had no effect on resistance when all copies of 23s carried an A2074C mutation, whereas it led to a major in-
CHAPTER 36
crease in susceptibility when two or three copies had an A2075G transition. However, the A2074C isolates in this study all had an initial MIC that were out of the MIC test range, and no comparison was done to erythromycin-susceptible isolates. Thus, it could be possible that the CmeABC pumps simply gives a basic reduction in the MIC, which is independent of acquired mechanisms. As with FQ resistance, the importance of these efflux pumps in acquired resistance requires further investigations. Sulfonamide Resistance Sulfonamides are structural analogues of paminobenzoic acid (PABA) and compete with PABA for the enzyme dihydropteroate synthease (DHPS), thereby preventing PABA from incorporating into folic acid. In gram-negative bacteria, resistance to sulfonamides is normally due to the acquisition of horizontally transferable drug resistant variants of DHPS (Schwarz et al., 2006). In gram-positive bacteria, the most common mechanisms are mutations in the gene encoding DHPS (Schwarz et al., 2006). Gibreel and Skold (1999) found that sulfonamide resistance in C. jejuni was associated with the mutational substitution of four amino acid residues in DHPS, resulting in reduced affinity for sulfonamides. No other mechanisms of sulfonamide resistance have been found in Campylobacter. Tetracycline Resistance Tetracycline is considered as a second-line treatment for Campylobacter. It is used mainly in developing regions because of its low cost and low toxicity. Bacterial resistance to tetracycline resistance is mediated by four different mechanisms: efflux, modification of the tetracycline molecule, protection of the ribosomal binding site for tetracycline, and mutations in 16s rDNA (Schwarz et al., 2006). In C. coli and C. jejuni, tetracycline resistance is mainly due to ribosomal protection mediated by the t e t ( 0 ) gene (Manavathu et al., 1988). The t e t ( 0 ) gene is widespread in Campylobacter, and it can be located on either plasmids or the chromosome (Dasti et al., 2007; Pratt and Korolik, 2005; Taylor et al., 1983; Tenover et al., 1983, 1985). The conjugative plasmid that carries t e t ( 0 ) in C. jejuni transfers only between Campylobacter species (Taylor et al., 1983; Tenover et al., 1983, 1985). Some studies indicate that the gene mainly plasmid located in C. jejuni and chromosome located in C. coli (Dasti et al., 2007; Pratt and Korolik, 2005). The gene has also been found in different gram-positive bacterial species, including enterococci and streptococci (Schwarz et al., 2006),
TRANSMISSION OF ANTIBIOTIC RESISTANCE
651
suggesting a gram-positive origin of the gene. Studies have also shown that inactivation of the CmeABC efflux pump in C. jejuni can increase tetracycline susceptibility in tet(0)-positive isolates (Gibreel et al., 2007). Trimethoprim Resistance Trimethoprim acts by binding to and inhibiting the activity of the enzyme dihydrofolate reductase (dfr). Resistance is due to the acquisition of horizontally transferred dfr genes that are not inhibited by trimethoprim. In Campylobacter two different genes (dfrl and dfr9) have been found to mediate resistance (Gribeeel and Skold, 1998, 2000). The genes have been found on the chromosome in transposons or integrons (Gribeel and Skold, 1998, 2000). These two dihydrofolate reductases are also found in gramnegative bacterial species, mainly Enterobacteriaceae, indicating that Campylobacter also may acquire genes from this group (Gribeel and Skold, 1998, 2000). Both genes are present in approximately 10% of all Campylobacter isolates, but dfrl is present in almost all isolates (Gribeel and Skold, 1998; Jansson et al., 1992). Thus, Campylobacter is normally considered intrinsically resistant to trimethoprim. EMERGENCE OF RESISTANCE AND ASSOCIATION WITH ANTIMICROBIAL USE Surveillance of antimicrobial resistance in Campylobacter is important for assessing the association between antimicrobial use and occurrence of antibiotic resistance. Monitoring systems suggest that macrolide resistance is uncommon in C. jejuni. In the United States, macrolide resistance (MICs 2 8 pg/ml) is observed in approximately 1% of human isolates, with higher rates reported in other countries. Resistance to macrolides (and other antimicrobials) is usually higher in C. coli; resistance to erythromycin resistance ranges from 4 to 6% in the United States and up to 14% in other regions. Among food animals, resistance is generally higher in isolates from swine and poultry production environments, where macrolides are used routinely. In contrast to the relatively stable low incidence of macrolide resistance, FQ-resistant C. jejuni has emerged in many regions over the past two decades. This increase has been attributed in part to the use of FQs (sarafloxacin and enrofloxacin) in poultry medicine, prompting the FDA to withdraw approval of FQs in poultry. Experimental Studies Several animal experimental studies have shown that the use of standard doses of antimicrobials will
652
AARESTRUP ET AL.
select for antimicrobial resistance in the gut flora of food animals. For Campylobacter, such studies have particularly focused on FQs and macrolides because of the clinical importance of these antimicrobials in the treatment of human campylobacteriosis. A summary of selected experiments is provided in Table 3. Several experiments with both chickens and pigs administered FQs have shown a rapid emergence of FQ-resistant Campylobacter, and the resistant isolates have in most cases persisted throughout the experimental period (Delsol et al., 2004; McDermott et al., 2002; Takahashi et al., 2005; van Boven et al., 2003). In a single study, the treatment eradicated the colonization with C. jejuni (Takahashi et al., 2005). This was most likely because the inoculation dose was so low that resistant mutants did not arise in the gut of the chickens. In the single report on use of tylosin and selection for erythromycin resistance, resistant variants did not emerge during routine treatment, but only when tylosin was fed continuously to the chickens for 17 or 31 days (Lin et al., 2007). The authors attribute this to the low frequency of emergence of erythromycin-resistant mutants that they observed during in vitro experiments. Thus, there was insufficient time for erythromycin-resistant mutants to arise during a 3-day treatment course, but this time was sufficient when tylosin was used as a growth promoter. Epidemiological Field Studies A few studies have compared the occurrence of antimicrobial resistance among Campylobacter from farms with a low or no use of antimicrobial agents with farms that use antimicrobials. Luangtongkum et al. (2006) compared the antimicrobial susceptibility of 694 Campylobacter from 5 organic and 10 conventional broiler farms and 5 organic and 10 conventional turkey farms. The organic farms had never used antimicrobials, whereas the conventional farms routinely used antimicrobials. A major difference in the occurrence of resistance was observed. Thus, the conventional farms had significantly higher frequencies of resistance to ampicillin, tetracycline, kanamycin, macrolides, and quinolones. This was the case for both C. jejuni and C. coli, as well as other Campylobacter species. Avrain et al. (2003) studied the association between production type, antimicrobial use and antimicrobial resistance among Campylobacter from broilers. They observed a significant difference in the distribution of C. jejuni and C. coli as well as antimicrobial resistance according to production type (standard or export and free range) and antimicrobial administra-
tion. A difference between production types was only observed for tetracycline. Thus, 81 and 90% of the C. coli isolates from standard and export production were tetracycline resistant compared with 5 1% in free-range production. Associations between use of antimicrobial agents for growth promotion or therapy were observed, but not necessarily between use of a specific agent and resistance to that agent. Sat0 et al. (2004) compared the antimicrobial susceptibilities of 332 Campylobacter isolates isolated from 30 organic and 30 conventional dairy farms. Most of the isolates were susceptible to the antimicrobials tested (ciprofloxacin, gentamicin, and erythromycin), except tetracycline, against which 45% of all isolates were resistant. No differences were observed between organic and conventional farms. Gebreyes et al. (2005) compared the antimicrobial susceptibilities of Campylobacter coli with extensive and intensive antimicrobial-free swine farms. This study did not compare these farms to farms that used antimicrobials. They found a higher frequency of resistance for erythromycin, nalidixic acid, and tetracycline in the intensive compared with extensive production. A difference in occurrence of resistance to ciprofloxacin and nalidixic acid was apparent in this study, but this could perhaps be because these authors used the CLSI guidelines for susceptibility testing of Enterobacteriaceae. Cui et al. (2005) compared the occurrence of antimicrobial resistance among C. coli and C. jejuni isolated from organic and conventional chickens. They found a statistically significant higher prevalence of resistance to ciprofloxacin among C. jejuni from conventional compared with organic chickens. Erythromycin resistance was also more common among conventional chickens, but not significantly so, whereas tetracycline resistance was more common among organic chickens. In contrast, resistance to both erythromycin (significant) and tetracycline was more common among C. coli isolated from organic chickens. Ecological Observations
A few studies have assessed the effect of major changes in the use of antimicrobial agents in primary food production on the occurrence of resistance among Campylobacter on the farm, at retail, and in human isolates. Endtz et al. (1991) screened 883 strains of Cumpylobacter spp. isolated between 1982 and 1989 from human stools and poultry products for quinolone resistance. In this period, the prevalence of resistant strains isolated from poultry products increased from 0 to 14%. During the same period, the
Table 3. Experimental studies on the emergence of antimicrobial resistance in Cumpylobucter after treatment Reference Takahashi et al. (2005)
Animal species Chickens
Van Boven et al. (2003)
Chickens
McDermott et al. (2002)
Broiler chickens
Delsol et al. (2004)
Piglets
Lin et al. (2007)
Broiler chickens
No. of experiments and no. of animals in each group Study 1: Two groups of 15 chickens inoculated with lo6 C. jejuni ATCC 33560 at day 17
Study 2 : Two groups of 15 chickens inoculated with lo7 and 10' C. jejuni ATCC 33560 at day 18 and 23, respectively Sixteen individually housed chickens colonized with FQ-susceptible C. jejuni at day 8; from day 21, eight chosen for further study Study 1: Two groups (treatment and control) of 25 chickens each. Colonized with a mixture of five C. jejuni strains Study 2: Two groups (treatment and control) of 50 chickens each. Colonized with a mixture of five C. jejuni strains. Two groups of six piglets Study 1: Three experiments with 10 to 15 chickens in each group (control and treatment group). Experiment A chickens were inoculated with a mixture of two C. jejuni strains. Experiment B chickens were inoculated with a mixture of two C. coli strains and for experiment C with a C. jejuni strain. Study 2 : Two experiments with 9 to 11 chickens in each group, all inoculated with a C. jejuni strain.
Treatment
0utcome
50 ppm in drinking water from day 21 to 30 to six of the eight chickens
Study 1: C. jejuni was isolated from approximately 50% of the control group and disappeared from the administration group during treatment. No resistant isolates were found. Study 2: Isolation of C. jejuni from most chickens and 100% resistance found in the administration group In five chickens of the treatment group, an emergence of FQ-resistant isolates was observed.
Study 1: Enrofloxacin at 40 ppm in drinking water for 5 days
Rapid and persistent emergence of ciprofloxacin resistance in C. jejuni
SO ppm enrofloxacin in drinking water for 3 days to one group
Study 2: Sarafloxacin at 40 ppm in drinking water for 5 days One group given 15 mg enrofloxacin 1 pigld for 5 days Study 1: Tylosin at 0.53 g/liter in drinking water for 3 days for experiment A and B; three times treatment with tylosin at 0.53 glliter in drinking water for 3 days
FQ-resistant C. coli found at levels of 40 to 80% in the treated group Study 1: No emergence of erythromycin-resistant iso1ates
Study 2: Tylosin at 50 mgikg feed for 41 days
Study 2 : Erythromycin resistance emerged at 17 and 3 1 days after inoculation, respectively.
654
AARESTRUP ET AL.
prevalence in humans increased from 0 to 11%. The increase of quinolone resistance coincided with an increasing use of FQs in human and veterinary medicine. The extensive use of enrofloxacin in poultry and the almost exclusive transmission route of Campylobacter from chicken to humans in The Netherlands suggest that the resistance observed was mainly due to the use of enrofloxacin in the poultry industry. Smith et al. (1999) evaluated resistance to quinolones among Campylobacter isolates from Minnesota residents during the period from 1992 through 1998. A total of 4,953 isolates from humans received by the Minnesota Department of Health were tested for resistance to nalidixic acid. They conducted a case-comparison study of patients with ciprofloxacinresistant Campylobacter jejuni isolated during 1996 and 1997. Domestic chicken was evaluated as a potential source of quinolone-resistant Campylobacter. They found that the proportion of quinoloneresistant C. jejuni isolates from humans increased from 1.3% in 1992 to 10.2% in 1998. During 1996 and 1997, infection with quinolone-resistant C. jejuni was associated with foreign travel and with the use of a quinolone before the collection of stool specimens. However, quinolone use could not account for more than 15% of the cases from 1996 through 1998. The number of quinolone-resistant infections that were acquired domestically also increased during the period from 1996 through 1998. Ciprofloxacinresistant C. jejuni was also isolated from 14% of 91 domestic chicken products obtained from retail markets in 1997 and molecular subtyping showed an association between resistant C. jejuni strains from chicken products and domestically acquired infections in Minnesota residents. They concluded from these studies that the increase in quinolone-resistant C. jejuni infections in Minnesota mainly was due to infections acquired during foreign travel. However, the number of quinolone-resistant infections acquired domestically had also increased, largely because of the acquisition of resistant strains from poultry, and thus, the use of FQs in poultry, which began in the United States in 1995, had created a reservoir of resistant C. jejuni in the food supply. Several other researchers have also observed an association between introduction of FQs for veterinary use and emergent and increased resistance to FQs among Campylobacter isolated from food animals and humans (Engberg et al., 2004; Luber et al., 2003). Because FQ use in humans can lead to resistance in Campylobacter during therapy, it can be difficult to determine whether increased resistance in human strains is mainly caused by veterinary or human drug use. However, humans do not normally carry Campylobacter. Considering that poultry are a
major source of Campylobacter infection, that FQ use in poultry rapidly leads to FQ resistance in native strains, the close temporal association between the emergence/increase of resistance among isolates from food animals and the occurrence of resistance among isolates from humans, and the absence of a similar association when FQs were initially licensed for humans, poultry FQ use can be considered the major driving force for the increase of resistance among human isolates. This was the conclusion reached by the FDA in withdrawing the approval of FQs for use in poultry (Federal Register, 2005).
OCCURRENCE OF RESISTANCE IN DIFFERENT FOOD ANIMAL RESERVOIRS Because of differences in interpretive criteria and the failure of many laboratories to implement standardized testing procedures, great care has to be taken when comparing data from different sources and over time. The occurrence of antimicrobial resistance among C. coli and C. jejuni isolated from broilers in Denmark is shown in Table 4 and among broilers in the United States in Fig. 1. There is a general difference in the occurrence of resistance between C. coli and C. jejuni even when isolates originate from the same reservoirs, with C. coli generally being more resistant. Especially for macrolides and related compounds, and for FQs (as well as streptomycin and tetracycline in Denmark), a higher frequency of resistance is observed among C. coli. In the United States, a higher frequency of tetracycline resistance is observed among C. jejuni, which is not the case in Denmark. There is currently no explanation for the apparent difference in the tendency of C. coli and C. jejuni to develop resistance to macrolides and streptomycin. Both C. coli and C. jejuni have been reported to contain three copies of rRNA (Kim et al., 1992; Taylor et al., 1992). Possible explanations include species differences in the stability of the mutations, or different capacities to exchange DNA by natural transformation. However, this will have to await further studies. Recent data on the occurrences of resistance to selected antimicrobial agents among C. coli and C. jejuni isolated from various animal and retail sources are given in Tables 5 to 8. Major differences in the prevalences of resistance are apparent among countries. The most likely explanation for these differences is differences in use of antimicrobial agents. However, antimicrobial use data are only available from a few countries, making such comparisons difficult. In the countries were data are available from '
CHAPTER 36
-
TRANSMISSION OF ANTIBIOTIC RESISTANCE
655
Table 4. Occurrence of antimicrobial resistance among Cumpylobucter coli and Campylobucter jejuni isolated from broiler chickens in Denmark from 1997 to 2001 Occurrence of resistance among isolates from broilers Antimicrobial agent
Campylobacter coli (n = 86)
Campylobacter jejuni ( n = 370)
1.2 37.2 9.3 18.6 1.2
0.0 1.6 4.6 5.1 1.4
Chloramphenicol Erythromycin Fluoroquinolone Streptomycin Tetracycline
both the primary food production and retail, a reasonable correlation between the prevalences of resistance in animals and food is found. However, because an increasing part of the food products consumed in one country originates from another country, such comparisons are difficult and in the future may be impossible as trade is expected to increase. Antimicrobial resistance seems to evolve independently of the strain type. Thus, no correlation between Campylobacter clonal type and antimicrobial resistance has been detected so far when studying wild-type strains (Cullen et al., 2007; Wittwer et al., 2005). Antimicrobial use is important, but it is probably not the only factor that determines the rate of the emergence, spread, and persistence of antimicrobial in the resistance in Campylobacter. The association is complex because resistant clones occur naturally at low levels in the absence of selective pressure, and resistant clones can adapt to a nonselective environment and compete favorably with sensitive clones. The association between use of a given antimicrobial agent and emergence of resistance to other classes is common as a result of the simultaneous occurrence of several resistance genes in the bacterial strains. The
difference in antimicrobial resistance among extensive and intensive antimicrobial-free production could indicate that stress factors in general might render Campylobacter more prone to become resistant.
TRANSMISSION OF RESISTANT STRAINS ALONG THE FOOD CHAIN
A general overview of transmission along the food chain is provided in chapter 35. There are no known biological reasons why resistant Campylobacter should not transmit from animals to humans as well as susceptible Campylobacter. Thus, when resistance emerges in Campylobacter in animals, resistant Campylobacter transmits to humans. This concept has been supported by several investigations showing a temporal association between resistance emergence and increase in animals and humans as mentioned previously (Endtz et al., 1991; Smith et al., 1999). In addition, as has already been mentioned, a recent study has even indicated that that certain types gain increased fitness when acquiring FQ resistance mutations (Luo et al., 200563). It has recently been speculated that isolates with acquired resistance to anti-
50.0% 45.0% 40.0% 35.0%
30.0%
1 s
25.0%
20.0% 15.0% 10.0%
5.0% 0.0%
TEL
CLI
AZI
ERY
CIP
NAL
TET
Figure 1. Occurrence of antimicrobial resistance among Cumpylobucter coli and Cumpylobucter jejuni from U.S. chickens during 2002 to 2006.
656
AARESTRUP ET AL.
Table 5 . Recent data on the occurrence of antimicrobial resistance among Campylobucter jejuni from broilers in different countries Percentage resistant to: Country Austria Denmark Italy Japan The Netherlands Spain United States a 'I
Year 2005 2005 2005 2001-2003 2005 2005 1998-2006
No. of isolates 195 76 36 39 78 16 3,548
Reference Ciprofloxacin
Erythromycin
Streptomycin
Tetracycline
49.7 7.9 66.7 17.9" 43.6 93.8 8.8-20.36
3.1 0 0 0 0 6.3 0.2-3. lb
2.1 1.3
29.2 5.3 52.8 41.0 42.3 43.8 34-59b
_.
2.6 0 6.3 -
EFSA (2006) EFSA (2006) EFSA (2006) Ishihara et al. (2006) EFSA (2006) EFSA (2006) U.S. Department of Agriculture (2007)
Enrofloxacin. Range indicated highest and lowest rates for the 9 years of testing.
microbial agents may also survive better along the food chain because of cross-resistance to disinfectants. This, however, needs further study.
SOURCE ATTRIBUTION STUDIES Case-control studies have provided valuable information on the sources of human campylobacteriosis. Most studies, conducted in several parts of the world, have identified consumption of poultry (especially undercooked chicken) as the main risk factor for sporadic campylobacteriosis. Additional risk factors include traveling, drinking water or milk, barbecuing, swimming in water, and contact with pets. A number of case-control studies have specifically addressed risk factors for quinolone-resistant Campylobacter (Table 9). Five of the six case-control studies identified foreign travel as a risk factor for acquisition of an infection with a quinolone-resistant Cumpylobucter. The study by Smith et al. (1999) specifically identified traveling to Mexico, Caribbean countries or Central America, Asia, or Spain as significant risk factors for patients in Minnesota. A Campylobacter sentinel study in the United Kingdom identified traveling to Portugal, Cyprus, and Spain as
significant risk factors for patients in England and Wales (2002). Several studies have also indicated that imported food products are a risk for infections with antimicrobial-resistant Campylobacter. In the future, trade of food products is expected to increase. Thus, the selection of antimicrobial-resistant Campylobacter in one country can be a problem for trading partners.
INFECTION IN HUMANS AND ITS CONSEQUENCES Similar to what has been reported for Salmonella infections, there is evidence that antimicrobial resistance per se causes adverse health outcomes in patients with campylobacteriosis (chapter 6 ) . Several studies have shown that infections with quinoloneresistant Campylobacter in humans are associated with adverse effects for human health, mainly measured by prolonged diarrhea (Campylobacter Sentinel, 2002; Engberg et al., 2004; Helms et al., 2005; Nelson et al., 2004; Smith et al., 1999). Although the results of these studies are not all statistically significant, they point in the same direction, and taken to-
Table 6 . Recent data on the occurrence of antimicrobial resistance among Cumpylobucter coli from pigs in different countries Percentage resistant to: Country Austria Denmark Italy Japan The Netherlands Spain Sweden a
Enrofloxacin.
Year 2005 2005 2001-2003 2005 2005
No. of isolates 219 105 40 68 153 143 97
Reference Ciprofloxacin
Erythromycin
Streptomycin
Tetracycline
29.2 14.3 35.0 30.9" 4.6 87.9
19.2 20.0 37.5 48.5 9.2 69.5 0
78.1 47.6 55.9 86.3 90.1 -
76.7 5.7 87.5 82.4 86.3 98.6 4.1
-
EFSA EFSA EFSA EFSA EFSA EFSA EFSA
(2006) (2006) (2006) (2006) (2006) (2006) (2006)
TRANSMISSION OF ANTIBIOTIC RESISTANCE
CHAPTER 36
657
Table 7 . Recent data on the occurrence of antimicrobial resistance among Campylobacter jejuni from cattle in different countries Percentage resistant to: Country
Year
Austria Denmark Italy Japan The Netherlands I?
No. of isolates
Reference Ciprofloxacin
Erythromycin
Streptomycin
Tetracycline
29.8 31.7 13.0 17.1" 34.0
2.8 2.4 0 0 0
5.7 0 2.9 9.1
29.0 0 22.2 34.3 63.6
141 41 54 35 44
2005 2005 2005 2001-03 2005
EFSA (2006) EFSA (2006) EFSA (2006) Ishihara et al. (2006) EFSA (2006)
Enrofloxacin.
gether, they suggest that there is a longer duration of illness in patients infected with FQ-resistant strains compared with susceptible strains. In addition it has been shown that there is an excess risk of death or invasive illness after an infection with a quinoloneor macrolide-resistant Campylobacter compared with susceptible strains (Helms et al., 2005). The potential of antibiotic exposure to coselect for virulence traits has not been examined systematically in Campylobacter. Although the relationship between virulence and antimicrobial resistance is complex and unclear, the clinical evidence suggests that the health burden posed by Campylobacter may be potentiated by resistant strains.
RISK ASSESSMENT OF ANTIMICROBIAL RESISTANT CAMPYLOBACTER Risk assessments are mainly performed to provide a systematic organization of scientific evidence about an acknowledged health hazard; the magnitude of exposure to that hazard; the estimation of the likelihood of adverse outcomes in populations as a result of exposures to the hazardous agents; the existence of remaining gaps in scientific knowledge about the risk in question; and the overall uncertainty due to data gaps. The risk assessment process helps to provide risk managers with tools for choosing between options designed to eliminate or control the hazard. Risk assessments can take many forms, from quali-
tative discussions of potential hazards, to exposures to the hazards and the attendant risk of adverse health effects, to highly sophisticated and rigorous quantitative estimates of probabilities of occurrence for each step in the process, from hazard through exposure and finally to risk. To our knowledge, the latter has never been completed for antimicrobial resistance, although the FDA has an open quantitative risk assessment document for streptogramin-resistant Enterococcus faecium (http://www.fda.gov/cvm/ Documents/ SREF-RA-FinalDraft. pdf). Antimicrobial resistance risk assessment (ARRA) is an emerging area of human health risk assessment. ARRA is closely related to microbiological risk assessment, which is the analytic process used to assess the risks of illness from food-borne microbial pathogens. Two international organizations have recommended risk analysis models for assessing the risks from either microbiological agents or veterinary antimicrobial drugs in food animal uses: the Codex Alimentarius Commission and the Office International des Epizooties. A distinct attribute of ARRA is the focus on the biological determinants of antimicrobial resistance carried by bacterial strains of interest. Thus, an A R M might include descriptions on the role of commensal organisms that might serve as carriers of the resistance determinant or determinants. Another distinguishing feature of ARRA is that the adverse health effect of concern (impaired treatment of illness) may not be observed until treatment is attempted with the antimicrobial drug of interest. In
Table 8. Recent data on the occurrence of antimicrobial resistance among Campylobacter jejuni from broiler meat in different countries Percentage resistant to: Country Denmark Latvia Korea Norway UK USA
Year 2005 2005 2004 2005 2005 2004
No. of isolates 76 30 116 595 35 510
Reference Ciprofloxacin
Erythromycin
Streptomycin
Tetracycline
5.3 16 92.2 2 15.1
1.3 7.1 0 1.8 0 0.8
3.9 -
2.6 23.1 99.1 0 50.2
-
-
EFSA (2006) EFSA (2006) EFSA (2006) EFSA (2006) EFSA (2006) NARMS (2004)
Table 9. Case-control studies evaluating risk factors for quinolone resistant Cumpylobucter infections Reference
Study population and
No. patients with
YWS)
Smith et al. (1999)
Minnesota, 1992-1998
The Campylobacter Sentinel Surveillance Scheme Collaborators (2002)
England and Wales, 1 April 2000 to 31 May 2001, N = 3,489
Engberg et al. (2004)
Denmark, 1 May 2001 to 10 June 2002, N = 126
Kassenborg et al. (2004) Nelson et al. (2004) Johnson et al. (2007)
USA, 1998-1999, N = 858 Seven states in USA (19981999), N = 740 Southern Alberta, Canada (1February 2004 to 29 January 2005), N = 205
Only risk factors associated with increased risk are included. bUnivariateanalysis.
a
Multivariate analysis, matched odds ratio (95% CI)
P-value
47 (36) 14 (11)
26.0 (8.6-78.6) 45.5 (9.7-214)
<0.001 <0.001
23 (18) 7 (5) 26 (20)
40.7 (10.2-163.0) 48.6 (4.1-570.0) 7.5 (2.6-21.3)
<0.001 0.002 <0.001
8 5 48 92 90 80 30 14
22.4 (4.4-115.0) 11.7 (1.3-108.0) 6.9 (3.5-13.4) 5.0 (2.1-11.6) 3.7 (1.69-8.1) 2.1 (1.4-3.1) 16.8 (3.4-82.2) 19.1 (2.2-167.3)
<0.001 0.03 <0.001 <0.001 0.001 <0.001 <0.001 0.008
Potential risk factor" Resistant isolates Foreign travel to: Mexico Caribbean countries or Central America (not Mexico) Asia Spain Use of a quinolone before collection of stool specimen Travel-related infections (Portugal, Cyprus, Spain) Consumption of chicken Consumption of bottled water Domestically acquired infections: Consumption of cold meat (precooked) Foreign travel Consumption of fresh poultry other than chicken and turkey Swimming (pool, ocean, lake, or other places) Consumption of chicken or turkey cooked at a commercial establishment Foreign travel Foreign travel Possession of nonprescribed antibiotics
(2) (1) (14) (27) (26) (27) (71) (33)
Susceptible isolates
20 (48)
10.0 5.0 (1.14-22.0)
0.003
18 (55)
(1.3-78 .O)
0.03
29 (41)
<0.01 93.2 (29.6-292.9) 13.3 (2.2-80.9)
<0.001 0.005
CHAPTER 36
general, the overall purpose of an A R M is to give a quantitative estimate of the impact of antimicrobial use in a given reservoir in relation to human health. This helps give the policy makers a solid basis for making decisions on targeted interventions aimed at achieving the most human health benefit. A major impediment to risk assessment modelling, and a cause of much uncertainty, is the scarcity of data on the use of antimicrobial agents in food animal species down to the individual herd or group level. A sound causal understanding of the association between use of a given antimicrobial agent and the emergence and spread of resistance is seldom established. Even though numerous studies have shown that use is the main driving factor for the emergence of resistance, the association is not linear but also determined by the way the antimicrobial agent is used. Thus, it has been suggested that increased dosages might help to avoid the selection for quinolones resistance in Salmonella (Wiuff et al., 2003), and that continuous feeding of tylosin supplemented feed to chickens might select for macrolide resistant C. jejuni, whereas single treatments are less likely to do so. Exposure is also difficult to measure. Quantitative data on the load of antimicrobial resistant pathogens in different food products, as well as data about human consumption habits, are available in some cases. In addition, outcome data on the human health consequences are often also missing or incomplete. Only a few risk assessments on antimicrobialresistant Campylobacter have been performed. The FDA did a risk assessment on the use of FQs in poultry for the treatment of respiratory diseases (Bartholomew et al., 2003; Vose et al., 2001). The model was linear and did not take into account the differences in pathogen load caused by handling along the food chain. The model used available case-control studies to estimate the consequences for human health and the fraction of infections with quinoloneresistant Campylobacter that could be attributed to poultry. The model assumed that use of FQs in the primary production was the only factor selecting for resistance, and that removal of the selective pressure would lead to a major reduction in numbers of resistant bacteria. The major weakness of the model is the need for a precise estimate of the relationship between the exposure and the expected number of human cases (K). Cox and Popken (2006) adopted the basic principles from the FDA risk assessment in quantifying the human health risks associated with the use of FQs in poultry. The main purpose of the study was to develop a risk-benefit model, weighing the risks to human health with an expected benefit arising from a smaller Campylobacter load resulting from the use of FQs. Singer et al. (2007) developed
TRANSMISSION OF ANTIBIOTIC RESISTANCE
659
a similar model on use of macrolides, and the risks and benefits thereof. A risk model developed by Alban et al. (2007) estimated that of the 186 infections caused by macrolide-resistant Campylobacter recorded in Denmark in 2004, a total of 85 could be ascribed to imported pork (95% CI, 19 to 142), 72 could be ascribed to imported poultry meat (95% CI, 21 to 132), 19 could be ascribed to Danish poultry meat (95% CI), and seven could be ascribed to Danish pork (95% CI, 2 to 20). This risk assessment did not determine the association between the use of antimicrobial agents in the primary production and selection for resistance. It stresses, however, the relative risks of imported versus domestically produced meat and meat types. INTERVENTIONS AIMED AT REDUCING THE OCCURRENCE OF ANTIMICROBIALRESISTANT CAMPYLOBACTER As outlined in the previous sections, several studies have shown that the use of antimicrobials will select for resistance. The association is not always linear, and it is especially difficult to estimate under field conditions, where other factors can contribute to the emergence and spread of resistance. In contrast, only a few studies have described the effect of different interventions aimed at reducing the occurrence of antimicrobial resistant Campylobacter. In Denmark, major changes in the consumption of the macrolide tylosin have taken place in the production of pigs from 1995 to 2006. During the same time period, continuous monitoring of macrolide resistance among C. coli from pigs was performed. The findings are summarized in Fig. 2. The consumption of macrolides increased from 1995 to 1996 to a maximum of 76 tonnes. This was followed by a decrease in consumption to the lowest level of 11 tonnes in 1999 because of the voluntary stop in the use of antimicrobial growth promoters in 1999. Tylosin consumption has since increased to just above 20 tonnes annually, which is still considerably lower than its use before the use of antimicrobial growth promoters was ended in Denmark. The prevalence of resistance has closely followed these changes. The occurrence of resistance increased to a peak at 71% in 1997, one year after the peak consumption. Resistance rates subsequently decreased and as of this writing are at 13%. It is difficult to determine the potential consequences for human health of this reduced resistance. Most infections in humans are caused by C. jejuni, and only a small fraction are caused by C. coli. In addition, in Denmark, only a limited number of Campylobacter cases were routinely tested for antimicrobial susceptibility, and until 1999, no species identification was
660
AARESTRUP ET AL.
120
olo
1
100
Decreased usage for growth promotion
60
0C fr
.-a?
60
s9 .I
8
60
Y
m
I
40
a
1 If
40
20 20
0
0 1995 1996 1997 1996
1999 2000 2001 2002 2003 2004 2005 2006
Figure 2. Macrolide resistance among Campylobacter coli from pigs and consumption of tylosin for growth promotion and therapy in Denmark, 1995 to 2006.
performed. Regardless, these monitoring data provide strong evidence that reduction in tylosin use in pigs has led to a reduction in the occurrence of resistance among C. coli. Similarly in the United States, the FQ ban in poultry was implemented in September 2005, when ciprofloxacin resistance rates peaked at 19.6% in Campylobacter from retail meats and 15% in broilers. Data from 2006 isolates show a slight decrease in meat isolates to 18.6% and a 41% decrease to 8.8% in broiler isolates (Fig. 3). These data support the value of limiting animal drug exposure on a national scale as a means to reverse antimicrobial resistance in Campylobacter.
It is normally expected that antimicrobial resistance imposes a fitness cost for the bacteria and that resistance will disappear after the selective pressure is removed. This traditional viewpoint has been questioned for, among others, Salmonella by the observation of compensatory mutations in resistant mutants restoring the original fitness (Bjorkman et al., 1998, 2000; Maisnier-Patin and Anderson, 2004). For Campylobacter, only limited information is available on the ecological fitness of resistant variants. Recently, Luo et al. (2005) examined the biological fitness of a FQ-susceptible C. jejuni strain and its in vivo selected FQ-resistant mutant in chickens. When
25
20
1 l5 ae I0
5
0 1998
1999
2M)o
2001
2002
2003
2004
2005
Zoo6
* September2005 Figure 3. Occurrence of antimicrobial resistance among Campylobacter jeiuni from U.S. chickens, 1998 to 2006.
CHAPTER 36
monoinoculated into the chickens, both the FQsusceptible and -resistant strains were able to colonize the chickens equally efficiently in the absence of antibiotic use. When coinoculated, the FQ-resistant mutant out competed the susceptible wild type. When a series of clonally related FQ-resistant and -susceptible strains derived from chickens (Luo et al., 2005) was used in pairwise competitions, the FQ-resistant strains outcompeted the susceptible strains in six of the eight cases. By use of isogenic strains generated by natural transformation) it was shown that the increase in fitness was directly linked to the C257T mutation in gyrA, which confers high-level FQ resistance in Campylobacter. It was also shown that this mutation could confer a fitness cost in a small number of isolates. Thus, increases in or reductions of fitness might depend on the genetic background of the mutation. How this mutation interacts with the other genes in Campylobacter remains to be elucidated. Field studies of tetracycline-resistant strains on organic poultry farms where no antibiotics are used have also revealed an interesting dynamic (Luangtongkum et al., 2006). It was found that tetracyclineresistant Campylobacter coexisted with or replaced tetracycline-susceptible isolates. Horizontal transfer of tetracycline resistance plasmids in the absence of selection pressure has also been observed in the intestinal tract of chickens (Avrain et al., 2004).
FUTURE ASPECTS Campylobacteriosis is one of the most common causes of diarrhea in humans worldwide. Even though only a small fraction of people with infections require antimicrobial treatment) the availability of effective antimicrobial agents is necessary. Thus, measures should be taken to ensure that the use of antimicrobial agents for animals does not adversely affect human health. This requires continuously monitoring antimicrobial resistance, creating models to analyze risks associated with different selection pressures, and assessing the public health impact of resistant Campylobacter infections. The magnitude of the global trade in food products points to a need to do this on a worldwide basis. Thus, global monitoring of the occurrence of Campylobacter and its antimicrobial resistances by means of standardized methodologies should be implemented. Studies aimed at the human health consequences of infections with antimicrobialresistant strains should help guide interventions aimed at limiting the spread of the most important types of resistant strains. The effects of successful intervention measures can be used to direct similar in-
TRANSMISSION OF ANTIBIOTIC RESISTANCE
661
terventions in other countries and have the potential to considerably improve food safety worldwide.
REFERENCES Alban, L., E. 0. Nielsen, and J. Dahl. 2008. A human health risk assessment for macrolide-resistant Campylobacter associated with the use of macrolides in Danish pig production. Prev. Vet. Med. 83:115-129. Avrain, L., F. Humbert, R. L’Hospitalier, P. Sanders, C. VernozyRozand, and I. Kempf. 2003. Antimicrobial resistance in Campylobacter from broilers: association with production type and antimicrobial use. Vet. Microbiol. 96:267-276. Avrain, L., C. Vernozy-Rozand, and I. Kempf. 2004. Evidence for natural horizontal transfer of tetO gene between Campylobacter jejuni strains in chickens. J. Appl. Microbiol. 97:134-140. Bachoual R., S. Ouabdesselam, F. Mory, C. Lascols, C. J. Soussy, and J. Tankovic. 2001. Single or double mutational alterations of gyrA associated with fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli. Microb. Drug Resist. 7: 257-26 1. Bartholomew, M. J., D. J. Vose, L. R. Tollefson, and C. C. Travis. 2005. A linear model for managing the risk of antimicrobial resistance originating in food animals. Risk Anal. 25:99-108. Bischoff K., and J. Jacob. 1996. The sat4 streptothricin acetyltransferase gene of Campylobacter coli: its distribution in the environment and use as epidemiological marker. Zentrulbl. Hyg. Umweltmed. 198:241-257. Bjorkman, J., D. Hughes, and D. I. Anderson. 1998. Virulence of antibiotic-resistant Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 95:3949-3953. Bjorkman, J., I. Nagaev, 0. G. Berg, D. Hughes, and D. I. Andersson. 2000. Effects of environment on compensatory mutations to ameliorate costs of antibiotic resistance. Science 287: 1479-1482. Blaser, M. J. 1995. Campylobacter species, p. 1948-1956. In G. L. Mandell, J. E. Bennett, and R. D o h (eds.), Principles and Practice of Infectious Diseases, 4th ed. Churchill Livingstone, New York. Bottcher, I., and J. Jacob. 1992. The occurrence of high-level streptothricin resistance in thermotolerant campylobacters isolated from the slurry of swine and the environment, Zentralbl. Bakteriol. 277:467-473. van Boven, M., K. T. Veldman, M. C. de Jong, and D. J. Mevius. 2003. Rapid selection of quinolone resistance in Campylobacter jejuni but not in Escherichia coli in individually housed broilers. J. Antimicrob. Chemother. 52:719-723. Buu-Hoi, A., C. Le Bougutnec, and T. Horaud. 1989. Genetic basis of antibiotic resistance in Aerococcus viridans. Antimicrob. Agents Chemother. 3 3 :529-5 34. Cagliero C., C. Mouline, S. Payot, and A. Cloeckaert. 2005. Involvement of the CmeABC efflux pump in the macrolide resistance of Campylobacter coli. J. Antimicrob. Chemother. 56:948950. Cumpylobuctev Sentinel Surveillance Scheme Collaborators. 2002. Ciprofloxacin resistance in Campylobacter jejuni: case-case analysis as a tool for elucidating risks at home and abroad. J. Antimicrob. Chemother. 50561-568. Clinical Institute Laboratory Standards. 2006. Methods for antimicrobial dilution and disk susceptibility testing of infrequently isolated or fastidious bacteria. Approved guideline. CLSI M45A. Clinical Institute Laboratory Standards, Wayne, PA. Cooper R., H. Segal, A. J. Lastovica, and B. G. Elisha. 2002. Genetic basis of quinolone resistance and epidemiology of resistant
662
AARESTRUP ET AL.
and susceptible isolates of porcine Campylobacter coli strains. J. Appl. Microbiol. 93:241-249. Corcoran D., T. Quinn, L. Cotter, and S. Fanning. 2005. Relative contribution of target gene mutation and efflux to varying quinolone resistance in Irish Campylobacter isolates. FEMS Microbiol. Lett. 253:39-46. Corcoran D., T. Quinn, L. Cotter, and S. Fanning. 2006. An investigation of the molecular mechanisms contributing to highlevel erythromycin resistance in Campylobacter. Int. J. Antimicrob. Agents 27:40-45. Cox, L. A. Jr., and D. A. Popken. 2006. Quantifying potential human health impacts of animal antibiotic use: enrofloxacin and macrolides in chickens. Risk Anal. 26:135-146. Cui, S., B. Ge, J. Zheng, and J. Meng. 2005. Prevalence and antimicrobial resistance of Campylobacter spp. And salmonella serovars in organic chickens from Maryland retail stores. Appl. Environ. Microbiol. 71:4108-4111. Cullen, P., J. Abbott, S. L. Friedman, L. L. English, E. Hall Robinson, P. F. McDermott, D. G. White, s. Zhao, and the NARMS Working Group. 2007. Antimicrobial susceptibility profiles and molecular characterization of Campylobacter spp. recovered from NARMS Retail Meats-2005. 11th Annu. PulseNet Update Mtng., Providence, RI, 16 to 19 April 2007. Dasti, J. I., U. Gross, S. Pohl, R. Lugert, M. Weig, and R. SchmidtOtt. 2007. Role of the plasmid-encoded tet(0) gene in tetracycline-resistant clinical isolates of Campylobacter jejuni and Campylobacter coli. J. Med. Microbiol. 56:833-837. Delsol, A. A., J. Sunderland, M. J. Woodward, L. Pumbwe, L. J. Piddock, and J. M. Roe. 2004. Emergence of fluoroquinolone resistance in the native Campylobacter coli population of pigs exposed to enrofloxacin. J. Antimicrob. Chemother. 53:872874. Derbise, A., G. de Cespedes, and N. el Solh. 1997. Nucleotide sequence of the Staphylococcus aureus transposon, Tn5405, carrying aminoglycosides resistance genes. J. Basic Microbiol. 37: 379-384. Endtz, H. P., G. J. Ruijs, B. van Klingeren, W. H. Jansen, T. van der Reyden, and R. P. Mouton. 1991. Quinolone resistance in Campylobacter isolated from man and poultry following the introduction of fluoroquinolones in veterinary medicine. J. Antimicrob. Chemother. 27:199-208. Engberg, J., S. Andersen, R. Skov, F. M. Aarestrup, and P. GernerSmidt. 1999. Comparison of two agar dilution methods and three agar diffusion methods, including the Etest, for antibiotic susceptibility testing of thermophilic Campylobacter species. Clin. Microbiol. Infect. 5580-584. Engberg, J., J. Neimann, E. M. Nielsen, F. M. Aarestrup, and V. Fussing. 2004. Quinolone-resistant Campylobacter infections: risk factors and clinical consequences. Emerg. Infect. Dis. 10: 1056-1063. European Food Safety Authority (EFSA). 2006. The community summary report on trands and sources of zoonoses, zoonotic agents, antimicrobial resistance and foodborne outbreaks in the European Union in 2005. Federal Register. 2005. Enrofloxacin for poultry; final decision on withdrawal of new animal drug application following formal evidentiary public hearing. Department of Health and Human Services, U.S. Food and Drug Administration. Fed. Regist. 70: 44105. Gaudreau, C., and H. Gilbert. 1997. Comparison of disc diffusion and agar dilution methods for antibiotic susceptibility testing of Campylobacter jejuni subsp. jejuni and Campylobacter coli. J. Antimicrob. Chemother. 39:707-712. Gaudreau, C., and H. Gilbert. 1998. Antimicrobial resistance of clinical strains of Campylobacter jejuni subsp. jejuni isolated
from 1985 to 1997 in Quebec, Canada. Antimicrob. Agents Chemother. 42:2106-2108. Ge, B., S. Bodeis, R. D. Walker, D. G. White, S. Zhao, P. F. McDermott, and J. Meng. 2002. Comparison of the Etest and agar dilution for in vitro antimicrobial susceptibility testing of Campylobacter. J. Antimicrob. Chemother. 50:487-494. Ge, B., P. F. McDermott, D. G. White, and J. Meng. 2005. Role of efflux pumps and topoisomerase mutations in fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 49:3347-3354. Gebreyes, W. A., S. Thakur, and W. E. Morrow. 2005. Campylobacter coli: prevalence and antimicrobial resistance in antimicrobial-free (ABF) swine production systems. J. Antimicrob. Chemother. 56:765-768. Gibreel, A., and 0. Skold. 1998. High-level resistance to trimethoprim in clinical isolates of Campylobacter jejuni by acquisition of foreign genes (dfrl and dfr9) expressing drug-insensitive dihydrofolate reductases. Antimicrob. Agents Chemother. 42: 3059-3064. Gibreel, A., and 0. Skold. 1999. Sulfonamide resistance in clinical isolates of Campylobacter jejuni: mutational changes in the chromosomal dihydropteroate synthase. Antimicrob. Agents Chemother. 43 :2156-2160. Gibreel, A., and 0. Skold. 2000. An integron cassette carrying dfil with 90-bp repeat sequences located on the chromosome of trimethoprim-resistant isolates of Campylobacter jejuni. Microb. Drug Resist. 6:91-98. Gibreel, A., V. N. Kos, M. Keelan, C. A. Trieber, S. Levesque, S. Michaud, and D. E. Taylor. 2005. Macrolide resistance in Campylobacter jejuni and Campylobacter coli: molecular mechanism and stability of the resistance phenotype. Antimicrob. Agents Chemother. 49:2753-2759. Gibreel, A., N. M. Wetsch, and D. E. Taylor. 2007. Contribution of the CmeABC efflux pump to macrolide and tetracycline resistance in Campylobacter jejuni. Antimicrob. Agents Chemother. 51~3212-3216. Goodman, L. J., G. M. Trenholme, R. L. Kaplan, J. Segreti, D. Hines, R. Petrak, J. A. Nelson, K. W. Mayer, W. Landau, and G. W. Parkhurst. 1990. Empiric antimicrobial therapy of domestically acquired acute diarrhea in urban adults. Arch. Intern. Med. 150541-546. Han, K., S. S. Jang, E. Choo, S. Heu, and S. Ryu. 2007. Prevalence, genetic diversity, and antibiotic resistance patterns of Campylobacter jejuni from retail raw chickens in Korea. Int. /. Food Microbiol. 11450-59. Hariharan, H., T. Wright, and J. R. Long. 1990. Isolation and antimicrobial susceptibility of Campylobacter coli and Campylobacter jejuni from slaughter hogs. Microbiologica 13:l-6. Helms, M., J. Simonsen, K. E. Olsen, and K. Mdbak. 2005. Adverse health events associated with antimicrobial drug resistance in Campylobacter species: a registry-based cohort study. J. Infect. Dis. 191:1050-1055. Hoge, C. W., J. M. Gambel, A. Srijan, C. Pitarangsi, and P. Echeverria. 1998. Trends in antibiotic resistance among diarrheal pathogens isolated in Thailand over 15 years, Clin. Infect. Dis. 26~341-345. Ishihara, K., T. Yamamoto, S. Satake, S. Takayama, S. Kubota, H. Negishi, A. Kojima, T. Asai, T. Sawada, T. Takahashi, and Y. Tamura. 2006. Comparison of Campylobacter isolated from humans and food-producing animals in japan. J. Appl. Microbiol. 100:153-160. Jensen, L. B., and F. M. Aarestrup. 2001. Macrolide resistance in Campylobacter coli of animal origin in Denmark. Antimicrob. Agents Chemother. 45 :371-3 72. Jesse, T. W., M. D. Englen, L. G. Pittenger-Alley, and P. J. Fedorka-Cray. 2006. Two distinct mutations in gyrA lead to cip-
CHAPTER 36
rofloxacin and nalidixic acid resistance in Campylobacter coli and Campylobacter jejuni isolated from chickens and beef cattle. J. Appl. Microbiol. 100:682-688. Johnson, J. Y., L. M. McMullen, P. Hasselback, M. Louie, G. Jhangri, and L. D. Saunders. 2007. Risk factors for ciprofloxacin resistance in reported Campylobacter infections in southern Alberta. Epidemiol. Infect. 3:l-10. Kaneuchi, C., M. Ashihara, Y. Sugiyama, and T. Imaizumi. 1988. Antimicrobial susceptibility of Campylobacter jejuni, Campylobacter coli, and Campylobacter laridis from cats, dogs, pigs, and seagulls. ]pn. 1. Vet. Sci. 50:685-691. Kassenborg, H. D., K. E. Smith, D. J. Vugia, T. Rabatsky-Ehr, M. R. Bates, M. A. Carter, N. B. Dumas, M. P. Cassidy, N. Marano, R. V. Tauxe, and F. J. Angulo; Emerging Infections Program FoodNet Working Group. 2004. Fluoroquinoloneresistant Campylobacter infections: eating poultry outside of the home and foreign travel are risk factors. Clin. Infect. Dis. 38(S~ppl.3):279-284. Kim, N. W., H. Bingham, R. Khawaja, H. Louie, E. Hani, K. Neote, and V. L. Chan. 1992. Physical map of Campylobacter jejuni TGH9011 and localization of 10 genetic markers by use of pulsed-field gel electrophoresis. J. Bacteriol. 174:3494-3498. Kim, J.-S., D. K. Carver, and S. Kathariou. 2006. Natural transformation-mediated transfer of erythromycin resistance in Campylobacter coli strains from turkeys and swine. Appl. Environ. Microbiol. 72: 1316-132 1. Lachance, N., C. Gaudreau, F. Lamothe, and L. A. Lariviere. 1991. Role of the beta-lactamase of Campylobacter jejuni in resistance to beta-lactam agents. Antimicrob. Agents Chemother. 35:813-818. Lachance, N., C. Gaudreau, F. Lamothe, and F. Turgeon. 1993. Susceptibilitiesof beta-lactamase-positiveand -negative strains of Campylobacter coli to beta-lactam agents. Antimicrob. Agents Chemother. 37:1174-1176. Lin, J., L. 0. Michel, and Q. Zhang. 2002. CmeABC functions as a multidrug efflux system in Campylobacter jejuni. Antimicrob. Agents Chemother. 46:2124-213 1. Lin, J., M. Yan, 0. Sahin, S. Pereira, Y. J. Chang, and Q. Zhang. 2007. Effect of macrolide usage on emergence of erythromycinresistant Campylobacter isolates in chickens. Antimicrob. Agents Chemother. 51:1678-1686. Luangtongkum, T., T. Y. Morishita, A. J. Ison, S. Huang, P. F. McDermott, and Q. Zhang. 2006. Effect of conventional and organic production practices on the prevalence and antimicrobial resistance of Campylobacter spp. in poultry. Appl. Environ. Microbiol. 72:3600-3607. Luangtongkum, T., T. Y. Morishita, A. B. El-Tayeb, A. J. Ison, and Q. Zhang. 2007. Comparison of antimicrobial susceptibility testing of Campylobacter spp. by the agar dilution and the agar disk diffusion methods. J. Clin. Microbiol. 45590-594. Luber, P., E. Bartelt, E. Genschow, J. Wagner, and H. Hahn. 2003. Comparison of broth microdilution, E Test, and agar dilution methods for antibiotic susceptibility testing of Campylobacter jejuni and Campylobacter coli. J. Clin. Microbiol. 41: 1062-1068. Luber, P., J. Wagner, H. Hahn, and E. Bartelt. 2003. Antimicrobial resistance in Campylobacter jejuni and Campylobacter coli strains isolated in 1991 and 2001-2002 from poultry and humans in Berlin, Germany. Antimicrob. Agents Chemother. 47: 3825-3830. Luo, N., 0. Sahin, J. Lin, L. 0. Michel, and Q. Zhang. 2003. In vivo selection of Campylobacter isolates with high levels of fluoroquinolone resistance associated with gyrA mutations and the function of the CmeABC efflux pump. Antimicrob. Agents Chemother. 47:390-394.
TRANSMISSION OF ANTIBIOTIC RESISTANCE
663
Luo, N., S. Pereira, 0. Sahin, J. Lin, S. Huang, L. Michel, and Q. Zhang. 2005. Enhanced in vivo fitness of fluoroquinoloneresistant Campylobacter jejuni in the absence of antibiotic selection pressure. Proc. Natl. Acad. Sci. USA 102541-546. Maisnier-Patin, S., and D. I. Andersson. 2004. Adaptation to the deleterious effects of antimicrobial drug resistance mutations by compensatory evolution. Res. Microbiol. 155:360-369. Manavathu, E. K., C. L. Fernandez, B. S. Cooperman, and D. E. Taylor. 1990. Molecular studies on the mechanism of tetracycline resistance mediated by Tet(0). Antimicrob. Agents Chemother. 34:71-77. McDermott, P. F., S. M. Bodeis, L. L. English, D. G. White, R. D. Walker, S. Zhao, S. Simjee, and D. D. Wagner. 2002. Ciprofloxacin resistance in Campylobacter jejuni evolves rapidly in chickens treated with fluoroquinolones. J. Infect. Dis. 185:837840. McDermott, P. F., S. M. Bodeis, F. M. Aarestrup, S. Brown, M. Traczewski, P. Fedorka-Cray, M. Wallace, I. A. Critchley, C. Thornsberry, S. Graff, R Flamm, J. Beyer, D. Shortridge, L. J. Piddock, V. Ricci, M. M. Johnson, R N. Jones, B. Reller, S. Mirrett, J. Aldrobi, R. Rennie, C. Brosnikoff, L. Turnbull, G. Stein, S. Schooley, R. A. Hanson, and R. D. Walker. 2004. Development of a standardized susceptibility test for Campylobacter with quality-control ranges for ciprofloxacin, doxycycline, erythromycin, gentamicin, and meropenem. Microb. Drug Resist. 10:124-13 1. McDermott, P. F., S. M. Bodeis-Jones, T. R. Fritsche, R N. Jones, and R. D. Walker. 2005. Broth microdilution susceptibilitytesting of Campylobacter jejuni and the determination of quality control ranges for fourteen antimicrobial agents. J. Clin. Microbiol. 43:6 136-613 8. Miflin, J. K., J. M. Templeton, and P. J. Blackall. 2007. Antibiotic resistance in Campylobacter jejuni and Campylobacter coli isolated from poultry in the South-East Queensland region. ,J. Antimicrob. Chemother. 59:775-778. Nachamkin, I., J. Engberg, and F. M. Aarestrup. 2000. Diagnosis and antimicrobial susceptibility of Campylobacter spp., p. 4566. In I. Nachamkin and M. J. Blaser (eds.), Campylobacter, 2nd ed. ASM Press, Washington DC. NARMS. 2004. NARMS retail meat annual report, 2004. http:// www.fda.gov/cvm/NARMSReport2004.htm. Nelson, J. M., K. E. Smith, D. J. Vugia, T. Rabatsky-Ehr, S. D. Segler, H. D. Kassenborg, S. M. Zansky, K. Joyce, N. Marano, R M. Hoekstra, and F. J. Angulo. 2004. Prolonged diarrhea due to ciprofloxacin-resistant Campylobacter infection. J. Infect. Dis. 190:1150-1157. Niwa, H., T. Chuma, K. Okamoto, and K. Itoh. 2001. Rapid detection of mutations associated with resistance to erythromycin in Carnpylobacterjejunilcoli by PCR and line probe assay. Int. 1.Antimicrob. Agents 18:359-364. O’Halloran, F., B. Lucey, B. Cryan, T. Buckley, and S. Fanning. 2004. Molecular characterization of class 1 integrons from Irish thermophilic Campylobacter spp. J. Antimicrob. Chemother. 53: 952-957. Oncul, O., P. Zarakolu, 0. Oncul, and D. Gur. 2003. Antimicrobial susceptibility testing of Campylobacter jejuni: a comparison between Etest and agar dilution method. Diagn. Microbiol. Infect. Dis. 45:69-71. Ouellette, M., G. Gerbaud, T. Lambert, and P. Courvalin. 1987. Acquisition by a Campylobacter-like strain of aphA-1, a kanamycin resistance determinant from members of the family Enterobacteriaceae.Antimicrob. Agents Chemother. 3 1:1021-1026. Paulander, W., S. Maisnier-Patin, and D. I. Andersson. 2007. Multiple mechanisms to ameliorate the fitness burden of mupirocin resistance in Salmonella typhimurium. Mol. Microbiol. 64: 10381048.
664
AARESTRUP ET AL.
Payot S., L. Avrain, C. Magras, K. Praud, A. Cloeckaert, and E. Chaslus-Dancla. 2004. Relative contribution of target gene mutation and efflux to fluoroquinolone and erythromycin resistance, in French poultry and pig isolates of Campylobacter coli. Int. J. Antimicrob. Agents 23:468-472. Payot, S., J. M. Bolla, D. Corcoran, S. Fanning, F. Megraud, and Q. Zhang. 2006. Mechanisms of fluoroquinolone and macrolide resistance in Campylobacter spp. Microbes Infect. 8:1967-1971. Petruccelli, B. P., G. S. Murphy, J. L. Sanchez, S. Walz, R. DeFraites, J. Gelnett, R. L. Haberberger, P. Echeverria, and D. E. Taylor. 1992. Treatment of traveler's diarrhea with ciprofloxacin and loperamide. J. Infect. Dis. 165557-560. Piddock, L. J., V. Ricci, L. Pumbwe, M. J. Everett, and D. J. Griggs. 2003. Fluoroquinolone resistance in Campylobacterspecies from man and animals: detection of mutations in topoisomerase genes. J. Antimicrob. Chemother. 51:19-26. Pinto-Alphandary, H., C. Mabilat, and P. Courvalin. 1990. Emergence of aminoglycoside resistance genes aadA and aadE in the genus Campylobacter.Antimicrob. Agents Chemother. 34: 12941296. Pratt, A., and V. Korolik. 2005. Tetracycline resistance of Australian Campylobacterjejuni and Campylobacter coli isolates. J. Antimicrob. Chemother. 55:452-460. Pumbwe, L., and L. J. Piddock. 2002. Identification and molecular characterisation of CmeB, a Campylobacter jejuni multidrug efflux pump. FEMS Microbiol. Lett. 206:185-189. Pumbwe L., L. P. Randall, M. J. Woodward, and L. J. Piddock. 2004. Expression of the efflux pump genes cmeB, cmeF and the porin gene porA in multiple-antibiotic-resistant Campylobacter jejuni. J. Antimicrob. Chemother. 54:341-347. Pumbwe L., L. P. Randall, M. J. Woodward, and L. J. Piddock. 2005. Evidence for multiple-antibiotic resistance in Campylobacter jejuni not mediated by CmeB or CmeF. Antimicrob. Agents Chemother. 49:1289-1293. Rautelin, H., 0.-V. Renkonen, and T. U. Kosunen. 1991. Emergence of fluoroquinolone resistance in Campylobacterjejuni and Campylobacter coli in subjects from Finland. Antimicrob. Agents Chemother. 35:2065-2069. Reina, J., M. J. Ros, and A. Serra. 1994. Susceptibilities to 10 antimicrobial agents of 1,220 Campylobacter strains isolated from 1987 to 1993 from feces of pediatric patients.Antimicrob. Agents Chemother. 3 8:29 17-2920. Rivera, M. J., J. Castillo, C. Martin, M. Navarro, and R. GomezLus. 1986. Aminoglycoside-phosphotransferases APH(3')-IV and APH(") synthesized by a strain of Campylobacter coli. J. Antimicrob. Chemother. 18 :153-1 5 8, Ruiz, J., P. Goni, F. Marco, F. Gallardo, B. Mirelis, T. Jimenez De Anta, and J. Vila. 1998. Increased resistance to quinolones in Campylobacter jejuni: a genetic analysis of gyrA gene mutations in quinolone-resistant clinical isolates. Microbiol. lmmunol. 42:223-226. Ruiz J., A. Moreno, M. T. Jimenez de Anta, and J. Vila. 2005. A double mutation in the gyrA gene is necessary to produce high levels of resistance to moxifloxacin in Campylobacter spp. clinical isolates. Int. J. Antimicrob. Agents 25542-545. Saenz, Y., M. Zarazaga, M. Lantero, M. J. Gastanares, F. Baquero, and C. Torres. 2000. Antibiotic resistance in Campylobacter strains isolated from animals, foods, and humans in Spain in 1997-1998. Antimicrob. Agents Chemother. 44:267-271. Salazar-Lindo, E., R. B. Sack, E. Chea-Woo, B. A. Kay, I. Piscoya, and R. Y. Leon-Barua. 1986. Early treatment with erythromycin of Campylobacterjejuni associated dysentery in children. J. Pediatr. 109:3555-3560. SLnchez, R., V. FernLndez-Baca, M. D. Diaz, P. Munoz, M. Rodriguez-Crbems, and E. Bouza. 1994. Evolution of susceptibili-
ties of Campylobacter spp. to quinolones and macrolides. Antimicrob. Agents Chemother. 38: 1879-1 882. Sato, K., P. C. Bartlett, J. B. Kaneene, and F. P. Downes. 2004. Comparison of prevalence and antimicrobial susceptibilities of Campylobacter spp. isolates from organic and conventional dairy herds in Wisconsin. Appl. Environ. Microbiol. 70:1442-1447. Schwarz, S., A. Cloeckaert, and M. C. Roberts. 2006. Mechanisms and spread of bacterial resistance to antimicrobial agents, p. 8398. In F. M. Aarestrup (ed.), Antimicrobial Resistance in Bacteria of Animal Origin. ASM Press, Washington, DC. Shaw, K. J., P. N. Rather, R S. Hare, and G. H. Miller. 1993. Molecular genetics of aminoglycoside resistance genes and familial relationships of the aminoglycoside-modifying enzymes. Microbiol. Rev. 57:138-163. Singer, R. S., L. A. Jr. Cox, J. S. Dickson, H. S. Hurd, I. Phillips, and G. Y. Miller. 2007. Modeling the relationship between food animal health and human foodborne illness. Prev. Vet. Med. 79: 186-203. Sjsgren, E., G. B. Lindblom, and B. Kaijser. 1997. Norfloxacin resistance in Campylobacter jejuni and Campylobacter coli isolates from Swedish patients. J. Antimicrob. Chemother. 40:257261. Shaw, W. V. 1997. Chemical anatomy of antibiotic resistance: chloramphenicol acetyltransferase. Sci. Prog. 76565-580. Skirrow, M. B., and M. J. Blaser. 2002. Campylobacterjejuni, p. 719-739. In M. J. Blaser, P. D. Smith, J. I. Ravdin, H. B. Greenberg, and R. L. Guerrant (ed.), Infections of the Gastrointestinal Tract. Lippincott Williams and Wilkins, Baltimore. Smith, K. E., J. M. Besser, C. W. Hedberg, F. T. Leano, J. B. Bender, J. H. Wicklund, B. P. Johnson, K. A. Moore, and M. T. Osterholm. 1999. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992-1998. Investigation Team. N. Engl. J. Med. 340:1525-1532. Tajada, P., J. L. Gomez-Graces, J. I. Alos, D. Balas, and R. Cogollos. 1996. Antimicrobial susceptibilities of Campylobacterjejuni and Campylobacter coli to 12 beta-lactam agents and combinations with beta-lactamase inhibitors. Antimicrob. Agents Chemother. 40:1924-1925. Takahashi, T., K. Ishihara, A. Kojima, T. Asai, K. Harada, and Y. Tamura. 2005. Emergence of fluoroquinolone resistance in Campylobacter jejuni in chickens exposed to enrofloxacin treatment at the inherent dosage licensed in Japan. J. Vet. Med. B Infect. Dis. Vet. Public Health 52:460-464. Taylor, D. E., R. S. Garner, and B. J. Allan. 1983. Characterization of tetracycline resistance plasmids from Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 24: 930-935. Taylor D. E., W. Yan, L. K. Ng, E. K. Manavathu, and P. Courvalin. 1988. Genetic characterization of kanamycin resistance in Campylobacter coli. Ann. Inst. Pasteur Microbiol. 139:665-676. Taylor, D. E., M. Eaton, W. Yan, and N. Chang. 1992. Genome maps of Campylobacter jejuni and Carnpylobactercoli. 1.Bacteriol. 174:23 32-233 7. Tenover, F. C., and P. M. Elvrum. 1988. Detection of two different kanamycin resistance genes in naturally occurring isolates of Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 32:1170-1173. Tenover, F. C., M. A. Bronsdon, K. P. Gordon, and J. J. Plorde. 1983. Isolation of plasmids encoding tetracycline resistance from Campylobacter jejuni strains isolated from simians. Antimicrob. Agents Chemother. 23:320-322. Tenover, F. C., S. Williams, K. P. Gordon, C. Nolan, and J. J. Plorde. 1985. Survey of plasmids and resistance factors in Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 27:37-4 1.
CHAPTER 36
Tenover, F. C., C. L. Fennell, L. Lee, and D. J. LeBlanc. 1992. Characterization of two plasmids from Campylobacterjejuni isolates that carry the aphA-7 kanamycin resistance determinant. Antimicrob. Agents Chemother. 3 6: 7 12-7 16. U.S. Department of Agriculture, Agricultural Research Service. 2007. http://www.ars.usda.gov/SP2UserFiles/Place/66120508 / NARMS / percent - resistance / CampyResistSummary98 - 06 updated041307.pdf. Vacher S., A. Menard, E. Bernard, and F. Megraud. 2003. PCRrestriction fragment length polymorphism analysis for detection of point mutations associated with macrolide resistance in Campylobacter spp. Antimicrob. Agents Chemother. 47:1125-1128. Van der Auwera, P., and B. Scorneaux. 1985. In vitro susceptibility of Campylobacter jejuni to 27 antimicrobial agents and various combinations of beta-lactams with clavulanic acid or sulbactam. Antimicrob. Agents Chemother. 28:37-40. Vose, D.,K. Hollinger, M. J. Bartholomew, S. Thompson, C. Andres, D. Brooks, and K. Hemming. 2001. Risk assessment on the human health impact of fluoroquinolone resistant Campylobacter associated with the consumption of chicken. http: // www.fda.gov/ cvm/ antimicrobiallRisk-asses.htm. Watts J. L., and C.J. Lindeman. 2006. Antimicrobial susceptibility testing, p. 29-35. In F. M.Aarestrup (ed.), Antimicrobial Resistance in Bacteria ofAnimal Origin. ASM Press, Washington, DC. Wang, W. L., L. B. Reller, and M. J. Blaser. 1984. Comparison of antimicrobial susceptibility patterns of Campylobacterjejuni and Campylobacter coli. Antimicrob. Agents Chemother. 26235 1353. Wang, Y.,and D. E. Taylor. 1990. Chloramphenicol resistance in Campylobacter coli: nucleotide sequence, expression, and cloning vetor construction. Gene 94:23-28.
TRANSMISSION OF ANTIBIOTIC RESISTANCE
665
Wang Y., W. M. Huang, and D. E. Taylor. 1993. Cloning and nucleotide sequence of the Campylobacter jejuni gyrA gene and characterization of quinolone resistance mutations, Antimicrob. Agents Chemother. 37:457-463. Wittwer, M., J. Keller, T. M. Wassenaar, R Stephan, D. Howald, G . Regula, and B. Bissig-Choisat. 2005. Genetic diversity and antibiotic resistance patterns in a Campylobacterpopulation isolated from poultry farms in Switzerland. Appl. Enuiron. Microbiol. 71:2840-2847. Wiuff, C., J. Lykkesfeldt, 0. Svendsen, and F. M. Aarestrup. 2003. The effects of oral and intramuscular administration and dose escalation of enrofloxacin on the selection of quinolone resistance among Salmonella and coliforms in pigs. Res. Vet. Sci. 75: 185-193. Yan, W., and D. E. Taylor. 1991. Characterization of erythromycin resistance in Campylobacter jejuni and Campylobacter coli. Antimicrob. Agents Chemother. 35:1989-1996. Yan, M., 0. Sahin, J. Lin, and Q. Zhang. 2006. Role of the CmeABC efflux pump in the emergence of fluoroquinoloneresistant Campylobacter under selection pressure. J. Antimicrob. Chemother. 58:1154-1159. Zilhao, R.,B. Papadopoulou, and P. Courvalin. 1988. Occurrence of the Campylobacter resistance gene tetO in Enterococcus and Streptococcus spp. Antimicrob. Agents Chernother. 32:17931796. Zirnstein, G., Y. Li, B. Swaminathan, and F. J. Angulo. 1999. Ciprofloxacin resistance in Campylobacter jejuni isolates: detection of gyrA resistance mutations by mismatch amplification mutation assay PCR and DNA sequence analysis./. Clin. Microbiol. 37~3276-3280.
Cumpylobacter, 3rd ed. Edited by I. Nachamkin, C. M. Szyrnanski, and M. J. Blaser Q 2008 ASM Press, Washington, DC
Chapter 37
Poultry Colonization with Campylobacter and Its Control at the Primary Production Level JAAD A.
WAGENAAR, WILMAJACOBS-REITSMA, MERETE HOFSHAGEN, AND DIANENEWELL
meat as an important source for human campylobacteriosis (Friedman et al., 2000; Studahl and Andersson, 2000). However, accurate source attribution of human campylobacteriosis has proved difficult. The role of protective immunity in humans is poorly understood, but it is likely that immunity confounds many case-control studies. This may explain why, in some studies, poultry meat is identified as a protective factor (Adak et al., 1995; Friedman et al., 2004). Moreover, thus far, molecular epidemiology has failed to contribute substantially to the debate on source attribution and has not revealed intrinsic differences between chicken and human isolates. Nevertheless, quantitative attribution studies of different potential sources estimate poultry meat to be responsible for 20 to 40% of all cases in Northern Europe (Vellinga and Van Look, 2002; Wingstrand et al., 2006).
The handling and consumption of poultry is an important source of human campylobacteriosis. It is widely assumed that the control of Campylobacter in meat-producing poultry, with the aim of reducing the numbers of Campylobacter on poultry meat at the retail level, will reduce the public health burden of human campylobacteriosis. There are four levels at which Campylobacter contamination control might be targeted and implemented: on the poultry farm during rearing; at the interface between poultry farm and slaughterhouse during catching and transportation; at the slaughterhouse during processing; and in kitchens during meat handling and cooking. The first two levels we define as primary production. In a previous review (Newel1 and Wagenaar, 2000), the then current knowledge of poultry infections and their control at the farm level was discussed. In this chapter, we update this knowledge and extend it to include the interface between the farm and the slaughterhouse. The focus of the review remains on the control of Campylobacter in primary production because in this food chain, the gut of living poultry is the only amplification point for Campylobacter. Therefore, the control of this pathogen would have the greatest impact if the colonization of live animals could be prevented or reduced, thus avoiding the introduction of high numbers of Campylobacter into the subsequent phases of the production process.
COLONIZATION OF BROILERS Several types of poultry (broilers, layers, turkeys, ducks, fowl, quails, and ostriches) can become colonized with Campylobacter (Borck and Pedersen, 2005; Yogasundram et al., 1989). However, research to date has focused primarily on colonization in intensively reared broilers because this is the most easily (and largest) market sector to study and perhaps to address. Many longitudinal studies have indicated that broilers are free of Campylobacter at day of hatch but intensively reared flocks become detectably positive generally at 2 to 3 weeks of age (this is called
ROLE OF POULTRY MEAT IN HUMAN CAMPYLOBACTERIOSIS Epidemiological and exposure studies have implicated the handling and consumption of poultry
Jaap A. Wagenaar Department of Infectious Diseases and Immunology, OIE Reference Laboratory for Campylobacteriosis and WHO Collaboration Centre for Campylobacter, Faculty of Veterinary Medicine, Utrecht University, P.O. Box 80.165, 3508 TD Utrecht, The Netherlands; and Animal Sciences Group, OIE Reference Laboratory for Carnpylobacteriosis and WHO Collaboration Centre for Campylobacter, P.O. Box 65, 8200 AB Lelystad, The Netherlands. Wilma F. Jacobs-Reitsma * RIKILT Institute of Food Safety, Bornsesteeg 45, 6708 PD Wageningen, The Netherlands. Merete Hofshagen * National Veterinary Institute, P.O. Box 8156, N-0033 Diane G . Newell Veterinary Laboratories Agency (Weybridge), New Haw, Addlestone, Surrey KT15 3NB, United Oslo, Norway. Kingdom.
667
668
WAGENAAR ET AL.
the lag phase). In experimental studies, day-old chickens and older animals are highly susceptible to colonization with Campylobacter. Shedding starts 2 to 3 days after challenge. There are differences in colonization capacity of strains, and in vivo passage increases the colonization potential of individual strains (Cawthraw et al., 1996; Chen et al., 2006; Ringoir and Korolik, 2003). Cecal colonization reaches the highest levels after 5 days. A slight decline of colonization occurs after about 4 weeks, and negative birds may then occur. In one experimental study, the prevalence of birds shedding Campylobacter, from first challenge at 1 day of age, declined from more than 80% to <40% over the 43 days investigated (Achen et al., 1998). Several reports describe mixed infections in flocks (Jacobs-Reitsma et al., 1995; Van de Giessen et al., 1992). During this cocolonization, strains with a high colonization potential are able to replace others (Korolik et al., 1998). Such mixed colonization may lead to a greater variety in Cumpylobacter genotypes in the intestine because strains may exchange DNA leading to chimeras. In this way, the genomic diversity of the colonizing strains can increase (De Boer et al., 2002). Such factors may be important when subtyping methods are used for strain tracing, though this may occur relatively rarely (A. M. Ridley et al., personal communication). Campylobacter is a commensal organism in broilers, and a recent report confirms the absence of associated clinical signs and pathology (Dhillon et al., 2006). Colonization is largely in the ceca and is primarily confined to the intestinal mucous layer over the intestinal crypts of the villi (Beery et al., 1988). There is no detectable adhesion or invasion to intestinal epithelial cells. However, Campylobacter is recoverable from extraintestinal sites, including liver and spleen, suggesting translocation across the intestinal epithelium (Knudsen et al., 2006). Strainspecific differences in the potential to translocate may relate to invasive properties. It has been well established that after colonization, mucosal and systemic antibody responses are induced and increase over time (Cawthraw et al., 1994; Rice et al., 1997). Such protective immunity may be important in the dynamics of flock infection. Because most of the broiler parent flocks are colonized with Cumpylobucter, chicks will have high levels of antiCampylobacter-specific maternal immunoglobulin (Ig) G at day of hatch. These maternal antibodies decrease over the subsequent 2 to 3 weeks. Experimental infections have shown that 3-day old chicks from C. jejuni-colonized hens exhibit a 2- to 4-day delay in colonization when compared with chicks from non-Cumpylobacter colonized hens, by both a homologous and heterologous strain (Sahin et al.,
2003). This indicates a protective role for maternal antibodies, which may be reflected in the 2- to 3week lag phase of infection.
EPIDEMIOLOGY OF POULTRY FLOCK COLONIZATION
An age-related increase in prevalence of flock positivity is well established, with most flocks investigated becoming detectably colonized from 2 to 3 weeks of age. However, mathematical modeling of the spread of flock colonization has estimated that without factors that would delay transmission, a flock will test positive when routine surveillance methods are used only at 7 days after exposure (Katsma et al., 2007; Van Gerwe et al., 2005). Thus, during the early days of infection, routine surveillance programs might not detect the few colonized birds. This effect may explain, at least in part, the observed lag phase. However, other factors, such as feed composition and the previously indicated maternal immunity may also contribute to the lag phase. Once the first birds in a flock become colonized, fecal shedding (at about >lo6 per g of feces) and coprophagy promote the rapid transmission of infection through the flock. National surveys indicate that many intensively reared broiler flocks are Campylobacter-positive at slaughter. Depending on the country, the prevalence of positive flocks may vary. However, comparison of prevalence rates between countries is fraught with difficulty because of unharmonized sampling strategies, transport conditions and analytical methods. Within the European Union (EU), 3 to 91% of broiler flocks tested were reported as Campylobacter-positive from different countries in 2004 (European Food Safety Authority, 2005). The monitoring of Campylobacter has recently become obligatory within the EU (Zoonoses Directive (2003/99/EC) and harmonized procedures, for sampling, transport, culture and reporting, have now been produced for application across all member states to provide a first comparative baseline survey of Cumpylobacter in broiler flocks in 2008. There appears to be a consistently reported lower prevalence of positive flocks in the north of Europe compared with more southern countries like the United Kingdom and The Netherlands (Saleha et al., 1998). The reason for this remains unknown. Zootechnical parameters (number of animals per farm, climatic conditions, and the distance between farms) may all have an influence on the colonization rate. It is also possible that management practices in the poultry industries of these Northern European countries are the reason, i.e., the industry is more
CHAPTER 37
closely regulated than elsewhere in Europe. Several countries have claimed reductions in Campylobacter prevalence in broiler flocks as a result of implementation of control and prevention measures. The quantitative effect of these control measures is hard to estimate as controls are missing and natural variation in prevalence occurs over time. Surveys in several countries have indicated a seasonal variation in the prevalence of poultry flock colonization. This seasonality generally demonstrates a higher rate of infection in summertime compared with the winter period (Anonymous, 2006; JacobsReitsma et al., 1994). The reason for this seasonal variation is unknown but may reflect levels of environmental contamination (Nylen et al., 2002). Certainly, poultry houses have more ventilation in the summer, potentially increasing the contact with the outside environment. Colonization is primarily with C. jejuni in all studies of intensively reared broilers, with a lower proportion of C. coli strains and rarely other species. The overall relative distribution of C. jejuni to C. coli strains recovered varies between countries. For example in Denmark, the United Kingdom and the Netherlands the proportion of C. jejuni is 85,95,and 65%, respectively (Evans, 1997; Nielsen and Nielsen, 1999; W. Jacobs-Reitsma, personal communication). Factors that influence this are unknown but in older animals, e.g., from organic production and laying hens, there is a shift toward C. coli (El-Shibiny et al., 2005) suggesting that C. coli strains can outcompete C. jejuni strains over time.
RISK FACTORS AND SOURCES OF COLONIZATION FOR POULTRY Reducing human campylobacteriosis is an important public health goal in most industrialized countries. In recent years, several quantitative risk assessment models for Campylobacter in poultry have been developed (Hartnett et al., 2001; Nauta et al., 2007; Rosenquist et al., 2003). Such models have been used to estimate that reducing the average number of Campylobacter on a broiler carcass by 2 log would reduce the number of campylobacteriosis cases by 30% (Rosenquist et al., 2003). Thus, producing Campylobacter-negative flocks would have a significant public health benefit. For intensively reared flocks, this is feasible, at least in some countries. However, Campylobacter-negative free-range flocks rarely occur (Heuer et al., 2001; Huneau-Salaun et al., 2007; Van Overbeke et al., 2006) presumably as a consequence of substantial environmental exposure.
POULTRY COLONIZATION WITH CAMPYLOBACTER
669
Campylobacter is widespread in the farm environment, and horizontal transmission into the flock clearly occurs. The role of vertical transmission as a source is still debated. Campylobacter is recoverable, albeit at a low incidence and level, from semen (Buhr et al., 2005) and ovarian follicles (Cox et al., 2005), and most recently from paper pad tray liners in hatcheries (Byrd et al., 2007). To date, there is no direct evidence that these Campylobacter derived from parents are a significant source of colonization for descending flocks. Extensive molecular epidemiological studies on grandparent flocks showed that vertical transmission into parent flocks was not detectable (Callicott et al., 2006). This is supported by the delayed colonization in intensively reared flocks and negativity in experimental chicks maintained under containment conditions. Thus, all the evidence suggests that vertical transmission occurs rarely, if at all. Certainly, in terms of infection control, environmental sources will be significantly more important than a possible risk of vertical transmission. Once the first bird in a flock becomes colonized, upregulation of colonization potential, high levels of fecal shedding and extensive coprophagy ensure the rapid dissemination of infection throughout the flock. From risk assessment modelling studies the primary measures to control Campylobacter need to be implemented at the farm level. However, before targeted intervention strategies can be implemented, the sources and routes of infection for broiler flocks must be identified (Newel1 and Fearnley, 2003; Shreeve et al., 2002) and for effective implementation of intervention strategies such sources need to be prioritorized. There are three main approaches to the identification of risk factors for, and sources of, flock colonization. One approach is to use epidemiological case control studies relating flock positivity to environmental factors and management practices collated by questionnaire (e.g., Huneau-Salaun et al., 2007). This approach provides indirect indicators of possible factors, but the conclusions are generally broad, may depend on personal perception, and may depend on the observer or recorder. A more targeted approach is the use of molecular epidemiology to track specific strain types, prospectively recovered from the farm environment, for comparison with those which cause colonization in the flock. This approach has been applied over the last few years, but there are several problems. Firstly the culture and subsequent maintenance of environmentally stressed Campylobacter is difficult, so the recovery of organisms around the farm may be very insensitive. Thus, the absence of culturable organisms may misleadingly indicate Campylobacter-negative
670
WAGENAAR E T AL.
environments. The second problem is the selection and application of appropriate molecular typing methods for Campylobacter. As previously indicated (Newel1 et al., 2000) there are many typing methods available with various properties and the typing technique should be selected to best address the question asked. For the study of source attribution in humans, molecular typing has been remarkably unhelpful, largely as a consequence of the susceptibility of Campylobacter to genomic variation. However, in the case of a poultry flock exposed to environmental Campylobacter, the epidemiology is as for any single point acute outbreak situation. Because the geographical and temporal lines are limited, the risk of genomic variation is limited, and therefore, any rapid, cost-effective typing method providing a reasonable level of discrimination will be appropriate. Some of these molecular typing techniques can be adapted to generate novel methods to track specific strains around the farm (Ridley et al., in press). The final approach adopted has been to determine the effects on prevalence of the use of intervention procedures selected to target sources considered to be of high risk. However, in practice such experiments are difficult to effectively control. Given the diversity of farm environments and the variation in poultry farming management practices world wide, it is not surprising that an extensive list of risk factors and sources for Campylobacter in poultry flocks has been generated. These were reviewed in detail by Newell and Fearnley (2003). Systematic reviews are an approach to the unbiased analysis of such data collections. Such reviews have been extensively used in medical research but are only recently applied to the agricultural sector. In the only systematic review to date, Adkin et a1 (2006) analyzed all primary research on the presence or absence of Campylobacter in broilers on the farm. Publications (n = 159) were selected for review from an initial list of 1476. Data were collected from 74 papers reporting studies investigating over 1,200 flocks: 25% from the United States and 58% from Europe. For analysis, sources or risk factors were grouped; for example, depopulation schedule equalled staggered slaughter or slaughter in multiple batches. There were 14 sources and 37 contributing factors considered. The major factors identified as associated with an increased risk included age of broilers, staggered slaughter, multiple houses on the farm, on-farm staff, and other animals on the farm, while those factors identified as associated with a decreased risk included use of a hygiene barrier, the parent company and the season of rearing (winter). Some of these consistent sources and risk factors require elucidation. The increased risk associated
with age of broilers, and the presence of other animals on the farm is clearly related to opportunity for exposure from the environment. The effect of multiple houses on the farm reflects the chance of infection being trafficked from Campylobacter-positive to -negative houses. The association of positivity with partial depopulation is thought to result from breakdown of biosecurity associated with the entrance of catching crews into the house. Interestingly a decreased risk is frequently associated with the parent company, which probably reflects levels of management and general biosecurity practices. The final association with season of rearing is a very interesting and generally consistent factor, as yet unexplained. As previously indicated the timing and duration of this seasonal peak in flock positivity varies between countries and can even vary among years (Nylen et al., 2002). Undoubtedly the sources and risk factors for Campylobacter in poultry flocks are different during this seasonal peak than they are at other times of the year. The advantage of the systematic study is that it identifies sources and risk factors that are consistently and commonly reported worldwide. These are clearly the starting point for targeted intervention. The disadvantage of such a study is that important local sources and risk factors may not be identified. For example in studies in Nordic countries nonpotable water supplies are frequently identified as a risk factor (Kapperud et al., 1993). In these countries, well water is commonly used to supply poultry houses. Moreover, Nordic countries frequently report waterassociated human outbreaks of campylobacteriosis, so water also seems a feasible source for poultry. However, elsewhere in Europe, water supply is rarely reported as a risk factor. Such differences may provide research opportunities to investigate geographically related variations in the epidemiology of Campylobacter in broiler flocks. The role of flies in transmitting Campylobacter was discussed more than 20 years ago and has been reidentified recently. Flies appear to act as vectors for Campylobacter, and the fly traffic in and out of broiler houses is huge (Hald et al., 2004; Shane et al., 1985). In addition, flies have a seasonal prevalence comparable to that seen in broiler flocks and human cases. In some studies, controlling flies showed both delayed and reduced Campylobacter colonization in poultry flocks (Hald et al., 2007).
STRATEGIES TO PREVENT INTRODUCTION OF CAMPYLOBACTER INTO A FLOCK Campylobacter does not grow outside of the host; therefore Campylobacter numbers on a contam-
CHAPTER 37
inated carcass from a Campylobacter-colonized bird will tend to decline during processing. As a consequence, risk assessment models clearly indicate that reduction of Campylobacter at the end of the food chain is best achieved if the colonization of the live animal can be prevented. Identifying risk factors for the introduction of Campylobacter means that targeted intervention strategies can be implemented. Such strategies have to be separated into those currently available, which at the moment means related to biosecurity, and those under development such as vaccination, competitive exclusion, genetic resistance, and the use of probiotics. General Biosecurity Measures All the evidence indicates that Campylobacter is transported into a poultry house from the external environment. Many potential vehicles have been identified, including farm staff, rodents, insects, and wild birds. Theoretically, a high level of biosecurity on the farm should protect against Campylobacter. Certainly when raised under category 2 containment, experimental layer flocks can remain Campylobacter free (S. Cawthraw, personal communication). Under commercial conditions, some correlation has been found between high biosecurity levels and absence of Campylobacter (Van de Giessen et al., 1998), but even a very high level of biosecurity, as in grandparent stock raising, does not guarantee a Campylobacter-free flock (W. Jacobs-Reitsma, personal communication). Modern poultry houses can be considered closed environments, and as such, they should be relatively biosecure with restricted entrances protected by hygiene barriers. Nevertheless, biosecurity breeches occur. Every broiler house on a farm is entered by production staff at least once, and usually three or more times per day. In addition, maintenance workers and other authorized visitors will break this barrier frequently, and the risk of infection increases with the number of staff involved (RefrCgier-Petton et al., 2001). Thus, during the life of the birds (approximately 42 days), this means that there are approximately 50 to 150 occasions on which the barrier is broken by human traffic, each of which provides an opportunity for the entrance of Campylobacter. Thus, the chance of Campylobacter exposure increases with bird age, which is consistent with observed data. The risk of barrier breakdown involved on each occasion is a reflection of the level of personal biosecurity precautions used by the poultry production staff and visitors, and the level of environ-
POULTRY COLONIZATION WITH CAMPYLOBACTER
671
mental load in the external environment. The education of poultry farmers to improve general hygiene and enhance disease prevention has the advantage of addressing biosecurity against all infectious poultry diseases, not just Campylobacter. Because of the sensitivity of intensively reared flocks to Campylobacter, this colonization could be viewed as an indicator of effective biosecurity, and Campylobacter negativity therefore could have a perceived commercial benefit to the farmer as an added incentive. However, conflicting reports come from two Scandinavian countries undertaking poultry farmer education programs: Norway reported a positive effect from its education program, but Iceland failed to observe any effect (Hofshagen and Kruse, 2005; Reiersen et al., 2005). In the United Kingdom, education was targeted at the chicken catchers as well as the poultry farmers, but whether this has been effective remains subjective. Clearly there are behavioral components associated with biosecurity that require greater investigation and understanding. It has been difficult to assess the effectiveness of improving general biosecurity in the reduction of Campylobacter positivity. Mathematical models ( C A W ; http://www.rivm.nl/carma/index-eng. html) predict that such approaches should be effective but without clear recommendations on which measures should be taken and the cost of such measures, then feasibility and practicality are impossible to determine. Specific Biosecurity Measures Although no single biosecurity measure can be reliably implemented to prevent the ingress of Campylobacter into a poultry house, as indicated by the systematic review, some farming management practices must be considered as highly important. These include multispecies farming and thinning or partial depopulation. Multispecies farming Multispecies farming is a clear risk factor for Campylobacter-positive poultry flocks on the same site. This is supported by increasingly sound evidence that Campylobacter strains initially found in other livestock like pigs and calves, at the time of chick placement, are subsequently found in the poultry flock. Single-species farming is therefore advisable. However, such changes in farming practice places economic burdens on the farmers who tend to respond by increasing the numbers of broiler houses and possibly the stocking density. This may, in return, intro-
672
WAGENAAR ET AL.
duce other important risk factors (Katsma et al., 2007) and negate the benefit. Thinning Thinning is the process of partially depopulating broiler houses to provide more space for the remaining birds on ethical and economic grounds. In some instances partial depopulation also occurs as a consequence of a limited capacity in transportation, lairage, and abattoir processing. The practicalities of thinning may vary from country to country and this can be reflected in the associated risk. Nevertheless, as the systematic review indicates, thinning is a consistent risk factor in most studies (Adkin et al., 2006). However, in some studies, the increased risk of thinning could be entirely accounted for by the increased age of the broilers (Russa et al., 2005). The period of time between thinning and final depopulation of the flock is crucial. Differences in this interval, as well as varying flock sizes, probably account for the range of opinions on thinning. The common perception is that thinning causes an increased risk of introducing Campylobacter into a flock through inadequate disinfection of equipment and inadequate hygiene practices of catching crews. Epidemiological studies suggest that 50% of Campylobacter-negative flocks become positive at final depopulation after thinning. This may be an underestimate as a result of the low sensitivity of routine surveillance methods at low levels of in-flock prevalence.
INTERVENTION METHODS UNDER DEVELOPMENT Biosecurity alone is unlikely to always protect flocks from colonization with Campylobacter. Additional and complementary interventions will also be required particularly for extensively reared (free range) broilers. To date, no such measures are available, but research on competitive exclusion, vaccination, and genetic resistance is ongoing. Competitive Exclusion and Use of Probiotics The principles and possible applications of competitive exclusion with defined or undefined agents to prevent Campylobacter colonization in poultry have been previously reviewed (Mead, 2000; Newel1 and Wagenaar, 2000). Despite the success of such approaches for organisms like Salmonella the overviews suggest that the efficacy for Carnpylobacter is
unpredictable. Some preparations can reduce the level of colonization (Mead et al., 1996; Stern, 1992), and apparently even commercially available competitive exclusion agents can achieve substantial reductions in both the level and prevalence of Campylobacter colonization in some experiments (Hakkinen and Schneitz, 1999; Wieliczko, 1995). However, in other experiments (Ah0 et al., 1992), similar agents had no effect despite efficacy against Salmonella. The reasons for these inconsistencies remain unclear but may reflect the variable nature of the competitive exclusion agent and of the susceptibility of the Campylobacter strain. It is now well recognized that effective agents may produce antiCampylobacter metabolites (Mead et al., 1996; Schoeni and Wong, 1994). Recent studies from U.S.Russian collaborations have defined such metabolites from several microbial agents, including Bacillus circulans, Paenibacillus polymyxa and Lactobacillus salivarius (Stern et al., 2005, 2006). The bacteriocins can be isolated and purified. The mechanisms of action are as yet unclear but observations in turkeys indicate the induction of changes in gut morphology (Cole et al., 2006). Like the bacteriocins, other chemical agents may also be effective in reducing colonization. High-molecular-weight carbohydratesfor example, glucuronic acid-enriched polysaccharides-are also active against Campylobacter in vitro (Wittschier et al., 2007) but to date have been ineffective in vivo. Such agents could be used as therapeutics to reduce the levels of Campylobacter carried in the avian gut just before slaughter. Campylobacter appears to occupy a unique ecological niche in the avian gut. As a natural extension of this, an alternative intervention approach is to use Campylobacter strains, which have a high colonizing potential but are nonpathogenic, as competitive excluders of potentially pathogenic environmental strains (homologous competitive exclusion) (Wassenaar et al., 1994). Further recent extensive research (Pope et al., 2007) that used in vivo models confirms differences in colonization potential between strains and further supports differences in strain virulence for humans. By means of oral models of chick colonization, highly colonizing strains have been identified (Cawthraw et al., 1996; Korolik et al., 1996). However, the propensity for genomic instability in some strains and a growing awareness of the correlation between genomic content and colonization potential (A. M. Ridley et al., personal communication) may preclude the development of such approaches. This clearly indicates the need for more fundamental research on the colonization mechanisms of Campylobacter in the avian gut.
CHAPTER 37
Vaccination Chickens colonized with C. jejuni induce both systemic and mucosal humoral responses (Cawthraw et al., 1994; Widders et al., 1996) of IgY, IgA and IgM isotypes. Campylobacter-specific antibodies are detectable between 10 days and 3 weeks after experimental inoculation depending on the chicken line (W. Ang, personal communication). These antibodies are directed against a variety of C. jejuni antigens. Flagellin is the first antigen to be recognized by all antibody isotypes, followed by antibodies directed against other bacterial antigens (Cawthraw et al., 1994; Rice et al., 1997). Interestingly this specificity may be compartmentalized such that a narrower spectrum of antigens is detected by mucosal rather than circulating antibodies (Cawthraw et al., 1994). The increasing levels of antibodies are paralleled by declining levels of colonization, suggesting a protective nature for these antibodies (Cawthraw et al., 1994; Myszewski and Stern, 1990; Rice et al., 1997). Such antibody responses appear to at least partly protect from rechallenge with the homologous strain (Cawthraw et al., 1998) after clearance of the primary colonization with antibiotics. Effective vaccine strategies directed against Campylobacter in broiler chickens have yet to be developed. Such strategies must either provide protection from exposure, presumably from day 1 of age to the point of slaughter; eliminate colonization by the time of slaughter; or reduce the Campylobacter numbers in the ceca. Given that the average life span of broiler chickens is only 6 weeks, this is a very short period in which to induce effective antibody responses, especially in an immunologically immature animal. The three main challenges in development of a vaccine in poultry are the identification of cross-protective antigens, the induction of a rapid and effective immune response, and the development of novel adjuvants to further stimulate immunity against Campylobacter antigens (reviewed by De Zoete et al., 2007). An alternative or complementary strategy is passive immunization. Orally administered antiCampylobacter antibodies have both therapeutic and prophylactic properties in chickens (Stern et al., 1990b; Tsubokura et al., 1997). So one approach may be to immunize parent flocks to produce passively protected chicks. Detectable circulating maternal IgG antibodies showing bactericidal effects are present in young chicks (Cawthraw et al., 1994; Sahin et al., 2001). Three-day-old chicks from c.jejunicolonized hens exhibit a 2- to 4-day delay in colonization compared with chicks derived from noncolonized hens, by both homologous and heterologous strains (Sahin et al., 2003). This observation
POULTRY COLONIZATION WITH CAMPYLOBACTER
673
suggests a role of these antibodies in the lag phase in colonization. The improvement and extension of the protective capacity of any such maternally derived immune protection requires further investigation. The choice of vaccine candidates and delivery systems will dictate the outcome of any immunization. Parenteral immunization, though feasible for breeder flocks, would be too costly for broilers. Oral delivery systems would be less expensive and easier to administer. Orally delivered killed vaccines have little, if any, protective capacity, even when delivered with mucosal adjuvants (Cawthraw et al., 1994; Rice et al., 1997) unless coupled together (Khoury and Meinersmann, 1995). Nevertheless, in ovo vaccination may overcome some of these problems and can certainly induce immune responses, though the protective capacity of such responses has yet to be determined (Noor et al., 1995). Oral live vaccines currently appear to be the candidates of choice. Live Campylobacter, which would colonize chickens but be nonpathogenic to humans, would potentially also act as homologous competitive exclusion agents, but their use will be hampered by the genetic instability of the Campylobacter genome as indicated previously. The approach of live vectors like attenuated Salmonella genetically engineered to express appropriate Campylobacter antigens is more promising. However, the antigens inducing protective immune responses are as yet unidentified. The flaA gene has been expressed in Salmonella vectors (Ellen-Vercoe et al., 1996). The engineered product can confer partial protection. This protection is more significant against the challenge with the homologous strain than with a heterologous strain (Kauc and Nachamkin, 1998). Expression of the potential vaccine candidate Pebl in an attenuated Salmonella failed to reduce intestinal colonization (Sizemore et al., 2006). However, oral immunization of chickens with an attenuated Salmonella expressing C. coli CjaA reduced the colonization by 6 logs in the majority of birds when challenged with C. jejuni (Wyszynska et al., 2004). These observations show the potential of the use of an interspecies cross-protective vaccine. Genetic Resistance Differences in the susceptibility of chicken lines for Campylobacter colonization have been reported (Stern et al., 1990a; Boyd et al., 2005). However, experimental observations also indicate that such differences also occur between individual chickens from the same time and batch and housed under identical conditions. Thus, this approach needs more investigation before it can be potentially exploited.
674
WAGENAAR ET AL.
STRATEGIES TO ELIMINATE CAMPYLOBACTER FROM COLONIZED FLOCKS There are currently two experimental approaches to reduce the shedding of Campylobacter in colonized chickens. The first is the administration of lytic bacteriophages (chapter 3 8). Under experimental conditions, this approach has been shown to give a 2 to 3 log reduction in Campylobacter shedding (LOCCarrillo et al., 2005; Wagenaar et al., 2005). The use of lytic phages in primary production will, however, face several problems in implementation. In particular, there are concerns for the amplification of bacteriophage-resistant Campylobacter in the environment, preventing repeated use of phages on specific farms. In an alternative strategy, phages may be used as a treatment to decontaminate the end product (Atterbury et al., 2003). The second approach to reduce shedding in Campylobacter-colonized chickens is treatment with bacteriocins. Experimental bacteriocin treatment of Campylobacter-colonized birds shows a consistent reduction in colonization of at least 1 million-fold compared with levels found in the untreated group (Stern et al., 2006). The application of bacteriocins is promising, but efficacy will need to be assessed in field trials and a risk assessment of the safety of treated products will be required. SCHEDULED SLAUGHTERINGAN APPROACH AT THE FARM-SLAUGHTERHOUSE INTERFACE The separation of Campylobacter-positive and -negative flocks and decontamination of the meat from positive flocks is an alternative strategy to reducing the risk of campylobacteriosis in humans. If the Campylobacter status of a flock could be identified at the farm level, this would offer the possibility to direct any Campylobacter-negative flocks to a Campylobacter-free slaughter line. This approach has been implemented in Iceland, Denmark, and Norway. One issue is the sensitivity of the on-farm test. For example, any misidentified positive flock (i.e., a false-negative result) would be treated as safe meat. Therefore, the incidence of false-negative tests needs to be as low as possible. Because flock positivity increases with age and after thinning, such tests should be performed as close as possible to the slaughter time. Unfortunately, for conventional culture techniques, samples need to be taken several days before slaughter in order to obtain a culture-positive result and this time delay may not enable the detection of those flocks that become positive between sampling and slaughter (Hofshagen and Bruheim, 2004).
There are two alternatives to conventional culture approaches: either the use of a rapid on-site test which could be done at the moment of depopulation (such a test is currently under evaluation in the Netherlands) or the use of a rapid PCR on samples taken just before depopulation. This approach is currently in use in Denmark (Lund et al., 2003). As indicated, this practical approach is already implemented in several countries. A good example is provided by the Norwegian experience developed in an Action Plan initiated in 2001. The Norwegian surveillance program includes virtually all broiler flocks, which are sampled twice throughout the growing period. The first sample is taken before slaughter by the farmer. From the start in 2001, this sampling was performed approximately 8 days before slaughter and from 2005 onward, this sample has been taken a maximum of 4 days before slaughter. The sample consists of a pool of ten swabs from fresh fecal droppings. All flocks are sampled again on arrival at the slaughter plant. From May 2004 onward, the samples have been intact ceca from 10 birds per flock. The cecal contents are pooled to one sample in the laboratory. Currently, samples taken at the farm are analyzed by real-time PCR (Lund et al., 2003), while samples taken in the slaughterhouse are analyzed by culture that uses the Nordisk Metodikkkomite for Levnetsmidler (1990) method without enrichment. Action is taken on all carcasses from flocks that are identified as positive for Campylobacter by the preslaughter sample. All such carcasses are either heat-treated or frozen for a minimum of 3 weeks before retail. Currently, no action is taken on the carcasses from flocks, which are only positive at the slaughterhouse. The important aspect of this action plan is to identify as many positive flocks as possible at the preslaughter sample so that these flocks can be subject to interventions to reduce Campylobacter contamination levels at retail. The feasibility of strategies like scheduled slaughtering remains an issue to the poultry industry. Obviously countries with a relatively small and wellstructured poultry industry like Norway and Iceland can introduce such procedures. Whether the highthroughput industries of countries like the United States or other parts of Europe could is questionable. More importantly, the Nordic countries have an overall low level of positive flocks (
CHAPTER 37
*
retail products, then the risk to human public health should be significantly reduced. However, there are unexplained results from countries where this approach was implemented. Even in countries with apparently effective intervention strategies to reduce Campylobacter contamination on poultry meat, the incidence of campylobacteriosis has remained the same or even increased (A. Engvall, personal communication). The reasons for this are debatable but may reflect increased nondomestic cases or even increased susceptibility in the general population as a result of lower exposure and consequently lower levels of immunity. The efficacy of a single intervention can be confounded by the adoption of multiple approaches at the same time, such as the freezing of positive flocks and the introduction of an extensive advertising campaign on kitchen hygiene and is therefore hard to determine after implementation of an intervention.
CONCLUSION In conclusion, campylobacteriosis in humans is a serious and apparently increasing problem with huge social and economic consequences worldwide. The contamination of meat products, especially poultry, by Campylobacter appears to be a significant risk factor. The problem of Campylobacter in the poultry food chain needs to be urgently addressed, but there are no immediate solutions available. The introduction of measures to improve biosecurity are practical but may be only partly effective. The costs and benefits of such measures remain difficult to estimate, and the practicality of implementation appears to vary significantly across the industry. Other potential preventions and therapeutic strategies are as yet not commercially available and remain under development. One practical control strategy that can be implemented at the farm-slaughterhouse interface is to separate colonized and noncolonized flocks during processing, and to subsequently treat the meat from Campylobacter-positive flocks. This approach is clearly feasible in countries with an overall low flock positivity rate, but in most countries where positivity is high, this would place a significant burden on the poultry industry. There remains an urgent need for reliable and practical measures of intervention in primary production. Acknowledgments. We acknowledge the Ministry of Agriculture, Fisheries and Foods, Great Britain, and the European Unionfunded Network of Excellence Med-Vet-Net for support for the writing of this review (D.N.).
POULTRY COLONIZATION WITH CAMPYLOBACTER
675
REFERENCES Achen, M., T. Y. Morishita, and E. C. Ley. 1998. Shedding and colonization of Campylobacter jejuni in broilers from day-ofhatch to slaughter age. Avian Dis. 42:732-737. Adak, G. K., J. M. Cowden, S. Nicholas, and H. S. Evans. 1995. The Public Health Laboratory Service national case-control study of primary indigenous sporadic cases of Campylobacter infection. Epidemiol. Infect. 115:15-22. Adkin, A., E. Hartnett, L. Jordan, D. Newell, and H. Davison. 2006. Use of a systematic review to assist the development of Campylobacter control strategies in broilers. J. Appl. Microbiol. 100:306-3 15. Aho, M., L. Nuotio, E. Nurmi, and T. Kiiskinen. 1992. Competitive exclusion of Campylobacter from poultry with K-bacteria and Broilact. fnt. J. Food Microbiol. 15:265-275. Anonymous. 2006. Annual report on zoonoses in Denmark 2006. Technical University of Denmark, Copenhagen. Atterbury, R. J., P. L. Connerton, C. E. Dodd, C. E. Rees, and I. F. Connerton. 2003. Application of host-specific bacteriophages to the surface of chicken skin leads to a reduction in recovery of Campylobacter jejuni. Appl. Environ. Microbiol. 69: 6302-6306. Beery, J. T., M. B. Hugdahl, and M. P. Doyle. 1988. Colonization of gastrointestinal tracts of chicks by Campylobacter jejuni. Appl. Environ. Microbiol. 54:2365-2370. Borck, B., and K. Pedersen. 2005. Pulsed-field gel electrophoresis types of Campylobacter spp. in Danish turkeys before and after slaughter. Int. J. Food Microbiol. 101:63-72. Boyd, Y., E. G. Herbert, K. L. Marston, M. A. Jones, and P. A. Barrow. 2005. Host genes affect intestinal colonization of newly hatched chickens by Campylobacter jejuni. fmmunogenetics 57: 248-253. Buhr, R. J., M. T. Musgrove, L. J. Richardson, N. A. Cox, J. L. Wilson, J. S. Bailey, D. E. Cosby, and D. V. Bourassa. 2005. Recovery of Campylobacter jejuni in feces and semen of caged broiler breeder roosters following three routes of inoculation. Avian Dis.49577-581. Byrd, J., R H. Bailey, R. Wills, and D. Nisbet. 2007. Recovery of Campylobacter from commercial broiler hatchery trayliners. Poult. Sci. 86:26-29. Callicott, K. A., V. Friethriksdbttir, J. Reiersen, R. Lowman, J. R. Bisaillon, E. Gunnarsson, E. Berndtson, K. L. Hiett, D. S. Needleman, and N. J. Stern. 2006. Lack of evidence for vertical transmission of Campylobacter spp. in chickens. Appl. Environ. Microbiol. 725794-5 79 8. Cawthraw, S., R. Ayling, P. Nuijten, T. Wassenaar, D. G. and Newell. 1994. Isotype, specificity, and kinetics of systemic and mucosal antibodies to Campylobacter jejuni antigens, including flagellin, during experimental oral infections of chickens. Avian Dis. 38:341-349. Cawthraw, S. A., C. Gorringe, and D. G. Newell. 1998. Prior infection, but not a killed vaccine, reduces colonization of chickens by Campylobacter jejuni, p. 364-367. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. University of Cape Town, Cape Town. Cawthraw, S. A., T. M. Wassenaar, R. Ayling, and D. G . Newell. 1996. Increased colonization potential of Campylobacter jejuni strain 8 1116 after passage through chickens and its implication on the rate of transmission within flocks. Epidemiol. Infect. 117: 213-215. Chen, L., H. Geys, S. Cawthraw, A. Havelaar, and P. Teunis. 2006. Dose response for infectivity of several strains of Campylobacter jejuni in chickens. Risk Anal. 26:1613-1621.
676
WAGENAAR ET AL.
Cole, K., M. B. Farnell, A. M. Donoghue, N. J. Stern, E. A. Svetoch, B. N. Eruslanov, L. I. Volodina, Y. N. Kovalev, V. V. Perelygin, E. V. Mitsevich, I. P. Mitsevich, V. P. Levchuk, V. D. Pokhilenko, V. N. Borzenkov, 0. E. Svetoch, T. Y. Kudryavtseva, I. Reyes-Herrera, P. J. Blore, F. Solis de 10s Santos, and D. J. Donoghue. 2006. Bacteriocins reduce Cumpylobacter colonization and alter gut morphology in turkey poults. Poult. Sci. 85:1570-1575. Cox, N. A., J. S. Bailey, L. J. Richardson, R. J. Buhr, D. E. Cosby, J. L. Wilson, K. L. Hiett, G. R. Siragusa, and D. V. Bourassa. 2005. Presence of naturally occurring Campylobacter and Salmonella in the mature and immature ovarian follicles of late-life broiler breeder hens. Avian Dis. 49:285-287. De Boer, P., J. A. Wagenaar, R. P. Achterberg, J. P. M. van Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:351-359. De Zoete, M. R., J. P. van Putten, and J. A. Wagenaar. 2007. Vaccination of chickens against Campylobacter. Vaccine 25: 5548-5557. Dhillon, A. S., H. L. Shivaprasad, D. Schaberg, F. Wier, S. Weber, and D. Bandli. 2006. Campylobacter jejuni infection in broiler chickens. Avian Dis. 5055-58. Ellen-Vercoe, E., S . Cawthraw, D. G . Newell, and M. J. Woodward. 1996. Expression of Campylobacter jejuni fl.A epitopes within a modified salmonella flagellin in Salmonella enteritidis, p. 667-672. In D. G. Newell, J. M. Ketley, and R. A. Feldman (ed.), Campylobacter, Helicobacter and Related Organisms. Plenum Press, New York. El-Shibiny, A., P. L. Connerton, and I. F. Connerton. 2005. Enumeration and diversity of Campylobacter and bacteriophages isolated during the rearing cycles of free-range and organic chickens. Appl. Environ. Microbiol. 71:1259-1266. European Food Safety Authority. 2005. The community summary report on trends and sources of zoonoses, zoonotic agents and antimicrobial resistance in the European Union in 2004. 21 December 2005. http: // www.efsa.europa.eu/ EFSA/ efsa-locale1178620753812~1178620772157.htm. Evans, S. A. 1997. Epidemiological studies of Salmonella and Campylobacter in poultry. Ph.D. Thesis. University of London, London. Friedman, C. R., R. M. Hoekstra, M. Samuel, R. Marcus, J. Bender, B. Shiferaw, S. Reddy, S. D. Ahuja, D. L. Helfrick, F. Hardnett, M. Carter, B. Anderson, and R. V. Tauxe. 2004. Risk factors for sporadic Cumpylobacter infection in the United States: a case-control study in FoodNet sites. Emerging Infections Program FoodNet Working Group. Clin. Infect. Dis. 38(S~ppl.3):S285-S296. Friedman, C. R., J. Neimann, H. C. Wegener, and R. V. Tauxe. 2000. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations, p. 121-139. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Hakkinen, M., and C. Schneitz. 1999. Efficacy of a commercial competitive exclusion product against Cumpylobacter jejuni. Br. Poult. Sci. 40: 6 19-62 1. Hald, B., H. Skovgird, D. D. Bang, K. Pedersen, J. Dybdahl, J. B. Jespersen, and M. Madsen. 2004. Flies and Campylobacter infection of broiler flocks. Emerg. Infect. Dis. 10:1490-1492. Hald, B., H. Skovgird, and H. Sommer. 2007. Screen out insect vectors to significantly reduce Campylobacter prevalence in broilers. Zoonoses Public Health 54(Suppl. 1):14. Hartnett, E., L. Kelly, D. Newell, M. Wooldridge, and G. Gettinby. 2001. A quantitative risk assessment for the occurrence of Campylobacter in chickens at the point of slaughter. Epidemiol. Infect. 127:195-206.
Heuer, 0. E., K. Pedersen, J. S. Andersen, and M. Madsen. 2001. Prevalence and antimicrobial susceptibility of thermophilic Campylobacter in organic and conventional broiler flocks. Lett. Appl. Microbiol. 33:269-274. Hofshagen, M., and T. Bruheim. 2004. The surveillance and control program for Campylobacter in broiler flocks in Norway: annual report 2004. http://www.vetinst.no/Arkiv/Pdf-filer/ NOK-2004/ 19-Carnpylobucter-in-broiler-flockspdf. Hofshagen, M., and H. Kruse. 2005. Learning from experiences: the Norwegian action plan against Campylobacter spp. in broilers. In V. Korolik, A. Lee, H. Mitchell, G. Mendz, B. Fry, and P. Coloe (ed.), 13th International Workshop on Campylobacter, Helicobacter and Related Organisms. Griffith University Publications, Cold Coast, Australia. Gold Coast, Queensland, Australia, 4-8 September. Huneau-Salaiin, A., M. Denis, L. Balaine, and G . Salvat. 2007. Risk factors for Campylobacter spp. colonization in French freerange broiler-chicken flocks at the end of the indoor rearing period. Prev. Vet. Med. 80:34-48. Jacobs-Reitsma, W. F., N. M. Bolder, and R. W. Mulder. 1994. Cecal carriage of Campylobacter and Salmonella in Dutch broiler flocks at slaughter: a one-year study. Poult. Sci. 73:12601266. Jacobs-Reitsma, W. F., A. W. Van de Giessen, N. M. Bolder, and R W. Mulder. 1995. Epidemiology of Cumpylobacter spp. at two Dutch broiler farms. Epidemiol. Infect. 114:413-421. Kapperud, G., E. Skjerve, L. Vik, K. Hauge, A. Lysaker, I. Aalmen, S. M. Ostroff, and M. Potter. 1993. Epidemiological investigation of risk factors for Cumpylobacter colonization in Norwegian broiler flocks. Epidemiol. Infect. 111:245-255. Katsma, E., A. A. De Koeijer, W. F. Jacobs-Reitsma, M. J. Mangen, and J. A. Wagenaar. 2007. Assessing interventions to reduce the risk of Campylobacter prevalence in broilers. Risk Anal. 27~863-876. Kauc, L., and I. Nachamkin. 1998. A recombinant Salmonella typhimurium strain expressing Campylobacter jejuni flagellin confers partial colonization protection in chicks, p. 166-170. In A. J. Lastovica, D. G. Newell, and E. E. Lastovica (ed.), Campylobacter, Helicobacter and Related Organisms. University of Cape Town, Cape Town. Khoury, C. A., and R. J. Meinersmann. 1995. A genetic hybrid of the Campylobacter jejuni flaA gene with LT-B of Escherichia coli and assessment of the efficacy of the hybrid protein as an oral chicken vaccine. Avian Dis. 39:812-820. Knudsen, K. N., D. D. Bang, L. 0. Andresen, and M. Madsen. 2006. Campylobacter jejuni strains of human and chicken origin are invasive in chickens after oral challenge. Avian Dis. 5O:lO14. Korolik, V., M. R Alderton, S. C. Smith, J. Chang, and P. J. Coloe. 1998. Isolation and molecular analysis of colonising and noncolonising strains of Carnpylobacter jejuni and Campylobacter coli following experimental infection of young chickens. Vet. Microbiol. 60:239-249. Korolik, V., J. Chang, N. Stern, and P. J. Coloe. 1996. Differentiation of Campylobacter strains from chickens in the USA using a DNA probe, p. 203-207. In D. G. Newell, J. M. Ketley, and R. A. Feldman (ed.), Campylobacter, Helicobacter and Related Organisms. Plenum Press, New York. LOCCarrillo, C., R. J. Atterbury, A. El-Shibiny, P. L. Connerton, E. Dillon, A. Scott, and I. F. Connerton. 2005. Bacteriophage therapy to reduce Carnpylobacter jejuni colonization of broiler chickens. Appl. Environ. Microbiol. 71:6554-6563. Lund, M., A. Wedderkopp, M. Waino, S. Nordentoft, D. D. Bang, K. Pedersen, and M. Madsen. 2003. Evaluation of PCR for detection of Campylobacter in a national broiler surveillance programme in Denmark. ]. Appl. Microbiol. 94:929-935.
CHAPTER 37
Mead, G. C. 2000. Prospects for “competitive exclusion’’ treatment to control salmonellas and other foodborne pathogens in poultry. Vet. J. 159:lll-123. Mead, G. C., M. J. Scott, T. J. Humphrey, and K. McAlpine. 1996. Observations on the control of Campylobacter jejuni infection of poultry by “competitive exclusion.” Avian Pathol. 25: 69-79. Myszewski, M. A., and N. J. Stern. 1990. Influence of Campylobacter jejuni cecal colonization on immunoglobulin response in chickens. Avian Dis. 34588-594. Nauta, M. J., W. F. Jacobs-Reitsma, and A. H. Havelaar. 2007. A risk assessment model for Campylobacter in broiler meat. Risk Anal. 27:845-861. Newell, D. G., and C. Fearnley. 2003. Sources of Campylobacter colonization in broiler chickens. Appl. Environ. Microbiol. 69: 4343-4351. Newell, D. G., J. A. Frost, B. Duim, J. A. Wagenaar, R. H. Madden, J. Van der Plas, and s. L. W. On. 2000. New developments in the subtyping of Campylobacter species, p. 27-44. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Newell, D. G., and J. A. Wagenaar. 2000. Poultry infections and their control at the farm level, p. 497-509. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Nielsen, E. M., and N. L. Nielsen. 1999. Serotypes and typability of Campylobacter jejuni and Campylobacter coli isolated from poultry products. lnt. J. Food Microbiol. 46:199-205. Noor, S. M., A. J. Husband, and P. R. Widders. 1995. In ovo oral vaccination with Campylobacter jejuni establishes early development of intestinal immunity in chickens. Br. Poult. Sci. 36: 563-573. Nordisk Metodikkkomite for Levnetsmidler. 1990. Campylobacter jejunilcoli detection in foods, 2nd ed. Method 119. Nordisk Metodikkkomite for Levnetsmidler, Oslo, Norway. http: // www.nmkl.org/. Nylen, G., F. Dunstan, S. R. Palmer, Y. Andersson, F. Bager, J. Cowden, G. Feierl, Y. Galloway, G. Kapperud, F. MCgraud, K. Mdbak, L. R. Petersen, and P. Ruutu. 2002. The seasonal distribution of Campylobacter infection in nine European countries and New Zealand. Epidemiol. Infect. 128:383-390. Pope, C., J. Wilson, E. N. Taboada, J. MacKinnon, C. A. F. Alves, J. H. E. Nash, K. Rahn, and G. W. Tannock. 2007. Epidemiology, relative invasive ability, molecular characterization and competitive performance in the chicken gut of Campylobacter jejuni strains. Appl Environ Microbiol. 73:7959-7966. RefrCgier-Petton, J., N. Rose, M. Denis, and G. Salvat. 2001. Risk factors for Campylobacter spp. contamination in French broilerchicken flocks at the end of the rearing period. Prev. Vet. Med. 50~89-100. Reiersen, J., H. Hardardottir, E. Gunnarsson, V. Fridriksdottir, G. Sigmundsdottir, and K. Kristinsson. 2005. In V. Korolik, A. Lee, H. Mitchell, G. Mendz, B. Fry, and P. Coloe (ed.), 23th International Workshop on Campylobacter, Helicobacter and Related Organisms. Griffith University Publications, Cold Coast, Australia. Gold Coast, Queensland, Australia, 4-8 September. Rice, B. E., D. M. Rollins, E. T. Mallinson, L. Carr, and S. W. Joseph. 1997. Campylobacter jejuni in broiler chickens: colonization and humoral immunity following oral vaccination and experimental infection. Vaccine 15:1922-1932. Ridley, A. M., V. M. Allen, M. Sharma, J. A. Harris, and D. G. Newell. A real time polymerase chain reaction approach for the detection of environmental sources of Campylobacter strains colonizing broiler flocks. Appl. Environ. Microbiol, in press. Ringoir, D. D., and V. Korolik. 2003. Colonisation phenotype and colonisation potential differences in Campylobacter jejuni strains
POULTRY COLONIZATION WITH CAMPYLOBACTER
677
in chickens before and after passage in vivo. Vet. Microbiol. 92: 225-235. Rosenquist, H., N. L. Nielsen, H. M. Sommer, B. Norrung, and B. B. Christensen. 2003. Quantitative risk assessment of human campylobacteriosis associated with thermophilic Campylobacter species in chickens. Int. J. Food Microbiol. 83:87-103. Russa, A. D., A. Bouma, J. C. M. Vernooij, W. Jacobs-Reitsma, and J. A. Stegeman. 2005. No association between partial depopulation and Campylobacter spp. colonization of Dutch broiler flocks. Lett. Appl. Microbiol 41:280-285. Sahin, O., N. Luo, S. Huang, and Q. Zhang. 2003. Effect of Campylobacter-specific maternal antibodies on Campylobacter jejuni colonization in young chickens. Appl. Environ. Microbiol. 695372-5379. Sahin, O., Q. Zhang, J. C. Meitzler, B. S. Harr, T. Y. Morishita, and R. Mohan. 2001. Prevalence, antigenic specificity, and bactericidal activity of poultry anti-Campylobacter maternal antibodies. Appl. Environ. Microbiol. 67:3951-3957. Saleha, A. A., G. C. Mead, and A. L. Ibrahim. 1998. Campylobacter jejuni in poultry production and processing in relation to public health. World Poult. Sci. J. 54:49-57. Schoeni, J. L., and A. C. Wong. 1994. Inhibition of Campylobacter jejuni colonization in chicks by defined competitive exclusion bacteria. Appl. Environ. Microbiol. 60:1191-1197. Shane, S. M., M. S. Montrose, and K. S. Harrington. 1985. Transmission of Campylobacter jejuni by the housefly (Musca domestics). Avian Dis. 29:384-391. Shreeve, J. E., M. Toszeghy, A. Ridley, and D. G . Newell. 2002. The carry-over of Campylobacter isolates between sequential poultry flocks. Avian Dis. 46:378-85. Sizemore, D. R., B. Warner, J. Lawrence, A. Jones, and K. P. Kilken. 2006. Live, attenuated Salmonella typhimurium vectoring Campylobacter antigens. Vaccine 24:3793-3 803. Stern, N. J. 1992. Reservoirs for Campylobacter jejuni and approaches for intervention in poultry, p. 49-60. In I. Nachamkin, M. J. Blaser, and L. Tompkins (ed.), Campylobacter jejuni: Current Topics and Future Trends. ASM Press, Washington, DC. Stern, N. J., R. J. Meinersmann, N. A. Cox, J. S. Bailey, and L. C. Blankenship. 1990a. Influence of host lineage on cecal colonization by Campylobacter jejuni in chickens. Avian Dis. 34:602606. Stern, N. J., R. J. Meinersmann, and H. W. Dickerson. 1990b. Influence of antibody treatment of Campylobacter jejuni on the dose required to colonize chicks. Avian Dis. 34595-601. Stern, N. J., E. A. Svetoch, B. V. Eruslanov, Y. N. Kovalev, L. I. Volodina, V. V. Perelygin, E. V. Mitsevich, I. P. Mitsevich, and V. P. Levchuk. 2005. Paenibacillus polymyxa purified bacteriocin to control Campylobacter jejuni in chickens. J. Food Prot. 68: 1450-1453. Stern, N. J., E. A. Svetoch, B. V. Eruslanov, V. V. Perelygin, E. V. Mitsevich, I. P. Mitsevich, V. D. Pokhilenko, V. P. Levchuk, 0. E. Svetoch, and B. S. Seal. 2006. Isolation of a Lactobacillus salivarius strain and purification of its bacteriocin, which is inhibitory to Campylobacter jejuni in the chicken gastrointestinal system. Antimicrob. Agents Chemother. 50:3111-3 116. Studahl, A., and Y. Andersson. 2000. Risk factors for indigenous Campylobacter infection: a Swedish case-control study. Epidemiol. Infect. 125:269-275. Tsubokura, K., E. Berndtson, A. Bogstedt, B. Kaijser, M. Kim, M. Ozeki, and L. Hammarstrom. 1997. Oral administration of antibodies as prophylaxis and therapy in Campylobacter jejuniinfected chickens. Clin. Exp. Immunol. 108:451-455. Van de Giessen, A., S. I. Mazurier, W. Jacobs-Reitsma, W. Jansen, P. Berkers, W. Ritmeester, and K. Wernars. 1992. Study on the epidemiology and control of Campylobacter jejuni in poultry broiler flocks. Appl. Environ. Microbiol. 58:1913-1917.
678
WAGENAAR ET AL.
Van de Giessen, A. W., J. J. Tilburg, W. S. Ritmeester, and J. Van der Plas. 1998. Reduction of Campylobacter infections in broiler flocks by application of hygiene measures. Epidemiol Infect. 121: 57-66. Van Genve, T. J. W. M., A. Bouma, W. F. Jacobs-Reitsma, J. van den Broek, D. Klinkenberg, J. A. Stegeman, and J. A. P. Heesterbeek. 2005. Quantifying transmission of Campylobacter spp. among broilers. Appl. Environ. Microbiol. 715765-5770. Van Overbeke, I., L. Duchateau, L. De Zutter, G . Albers, and R. Ducatelle. 2006. A comparison survey of organic and conventional broiler chickens for infectious agents affecting health and food safety. Avian Dis. 50:196-200. Vellinga, A., and F. Van Loock. 2002. The dioxin crisis as experiment to determine poultry-related Campylobacter enteritis. Emerg. Infect. Dis. 8:19-22. Wagenaar, J. A., M. A. P. Van Bergen, M. A. Mueller, T. M. Wassenaar, and R. M. Carlton. 2005. Phage therapy reduces Campylobacter jejuni colonization in broilers. Vet. Microbiol. 109~275-283. Wassenaar, T. M., D. G. Newell, B. A. M. and Van der Zeijst. 1994. Towards the use of attenuated Campylobacter mutants as a probioticum in poultry, p. 85-90. In J. F. Jensen, M. M. Hinton, and R. W. A. W. Mulder (ed.), Prevention and Control of Potenially Pathogenic Microorganisms in Poultry and Poultry Meat Processing. FLAIR no. 6, vol. 12, Probiotics and Pathogenicity. COW-DLO Het Spelderholt.
Widders, P. R, R. Perry, W. I. Muir, A. J. Husband, and K. A. Long. 1996. Immunisation of chickens to reduce intestinal colonisation with Campylobacter jejuni. Br. Poult. Sci. 37:765-778. Wieliczko, A. 1995. The role of Campylobacter spp. in the pathology of poultry. I11 Effects of selected feed supplements on the colonization of the alimentary tract of chickens by C. jejuni. Med. Weteruynaryjna 5 1~693-696. Wingstrand, A., J. Neimann, J. Engberg, E. M. Nielsen, P. GernerSmidt, H. C. Wegener, and K. Melbak. 2006. Fresh chicken as main risk factor for campylobacteriosis, Denmark. Emerg. Infect. Dis. 12:280-285. Wittschier, N., C. Lengsfeld, S. Vorthems, U. Stratmann, J. F. Ernst, E. J. Verspohl, and A. Hensel. 2007. Large molecules as anti-adhesive compounds against pathogens. J. Pbam. PbamzaC O ~ .59~777-786. Wyszynska, A., A. Raczko, M. Lis, and E. K. Jagusztyn-Krynicka. 2004. Oral immunization of chickens with avirulent Salmonella vaccine strain carrying C. jejuni 72Dz/92 cjaA gene elicits specific humoral immune response associated with protection against challenge with wild-type Campylobacter. Vaccine 22: 1379-13 89. Yogasundram, K., S. M. Shane, and K. S . Harrington. 1989. Prevalence of Campylobacter jejuni in selected domestic and wild birds in Louisiana. Avian Dis. 33:664-667.
Campylobacter, 3rd ed. Edited by 1. Nachamkin, C. M. Szymanski, and M. J. Blaser Q 2008 ASM Press, Washington, DC
Chapter 38
Bacteriophage Therapy and Campylobacter LAN
F. CONNERTON, PHILLIPPA L. CONNERTON, PAUL BARROW,BRUCEs. SEAL, ROBERTJ. ATTERBURY
AND
al., 1999). Bacteriophages are classified into 12 families, based principally on their morphological characteristics and nucleic acid content (Ackermann, 2001). The specificity of bacteriophages can be exploited by humans in many ways, for example phage typing schemes or for the rapid identification of bacteria. However, over the last few years, the main focus of interest had been on the use of bacteriophages as treatments for bacterial disease or to control pathogens in food. This has become fueled particularly by the increase in antibiotic resistance of bacteria but also in the case of food, for a general public desire to reduce the use of chemicals in food production. The intense interest in bacteriophages has led to the publication of several reviews on all aspects of the use of bacteriophage (Kutter and Sulakvelidze, 2005). Control of Campylobacter is an obvious target for phage therapy because a large proportion of poultry reared for meat harbor these organisms as a part of their intestinal flora with few practical alternatives for reduction (Corry and Atabay, 2001; Newel1 and Wagenaar, 2000). Campylobacter are present in the intestines of poultry at very high densities, ranging between log 4 and log 8 CFU/g (Rudi et al., 2004). This is advantageous to the potential success of phage treatments because phages can easily locate their prey and quickly multiply with plentiful supplies of their host. Simply introducing bacteriophages is unlikely to result in the elimination of the target bacteria because like most other predators, they seldom eliminate their prey in nature (Alexander, 1981; van den Ende, 1973). Mathematical models of the risk of Campylobacter infection acquired from eating contaminated
This chapter will discuss efforts to exploit Campylobacter-specific bacteriophages to reduce the numbers of Campylobacter jejuni and C. coli colonizing poultry and contaminating poultry meat products. Controlling campylobacters in poultry represents one of the greatest challenges to the agriculture and food industries if they are to achieve consumer and governmental demands to reduce human food-borne disease. These studies have sought to investigate the sustainable use of bacteriophages in these industries, and in doing so, they have revealed some important aspects of Campylobacter ecology in response to phage predation. Essentially these studies highlight key differences between the biological outcomes of phage infection of Campylobacter in laboratory experiments compared with those occurring within the avian intestinal tract, which represents the natural habitat of C. jejuni and C. coli. This review will also document the history, isolation, essential characteristics, types, and sources of bacteriophages that infect Campylobacter. Practical considerations and potential shortcomings associated with the therapeutic uses of bacteriophages will be reviewed from specific application studies. Bacteriophages are defined as viruses that can infect, multiply, and kill susceptible bacteria. Often simply referred to as phage, bacteriophages are ubiquitous in the environment and are found often in large numbers wherever suitable bacterial hosts are located, for example in seawater, soil, or the intestines of animals. It has been estimated that the total number of phage in the biosphere is approximately 1031 phage particles, making them the most abundant biological entity on the planet (Hendrix et
Ian F. Connerton and Phillippa L. Connerton School of Biosciences, Division of Food Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough Leics, LE12 SRD, United Kingdom. Paul Barrow School of Veterinary Medicine & Science, University of Nottingham, Sutton Bonington Campus, Loughborough Leics, LE12 SRD, United Kingdom. Bruce S. Seal Microbiologist & Research Leader, Poultry Microbiological Safety Research Unit, Russell Research Center, Agricultural Research Service, USDA, 950 College Station Road, Athens, GA 30605. Robert J. Atterbury Department of Clinical Veterinary Science, University of Bristol, Langford, Bristol, BS40 SDU, United Kingdom.
-
679
680
CONNERTON ET AL.
chicken have indicated that reductions of Campylobacter numbers by 100-fold or more could result in a significant reduction (30 times less) in the incidence of campylobacteriosis (Rosenquist et al., 2003). Treatments that do not eliminate but reduce the numbers of campylobacters below critical thresholds may therefore result in significant benefit to public health. It is also possible that combining other strategies, such as physical and hygiene-control measures, with bacteriophage treatments could reduce the numbers of human campylobacteriosis cases.
HISTORY Bacteriophages were first reported in 1915 by Fredrick William Twort when he described a transmissible “glassy transformation” of micrococcus cultures that resulted in dissolution of the bacteria. Subsequently Felix Hubert d’HCrelle (1917) reported a microscopic organism that was capable of lysing Shigella cultures on plates that resulted in clear spaces in the bacterial lawn that he called plaques. The term bacteriophage was introduced by d’HCrelle (1917) as he attributed the replicate nature of this phenomenon to bacterial viruses, and he was the first to formally propose that bacteriophages could be used to combat bacterial infections because they were a natural form of resisting bacterial disease. During 1919, d’HCrelle utilized phages isolated from poultry feces as a therapy to treat chicken typhus and further utilized this approach to successfully treat dysentery among humans (Summers, 2001). During 1921, Bruynoghe and Maisin (1921) used bacteriophages to eliminate skin disease caused by Staphylococcus infections. Before the discovery and widespread use of antibiotics, bacterial infections were treated by administering bacteriophages and were marketed by L‘Oreal in France. Although Eli Lilly Co. sold phage products for human use up until the 1940s, early clinical studies with bacteriophages were not extensively undertaken in the United States and Western Europe after the 1930s and 1940s. Bacteriophages were and continue to be sold in the Russian Federation and Eastern Europe as treatments for bacterial infections (Sulakvelidze et al., 2001). In veterinary practice, experimental alimentary Escherichia coli infections in mice and cattle were controlled by bacteriophage therapy (Smith and Huggins, 1982; Smith et al., 1987a, 1987b), while Barrow et al. (1998) reported the use of lytic bacteriophage against E. coli septicemia and meningitis in chickens and young cattle. Huff et al. (2002a, 2002b, 2003) reported the use of lytic bacteriophage to reduce effects of E. coli respiratory illness in chickens,
and bacteriophages have also been proposed as a strategy for control of food-borne pathogens (Hudson et al., 2005). For further insights into the application of bacteriophage to control specific bacteria in poultry, see the review by Joerger (2003).
BACTERIOPHAGE STRUCTURE AND TAXONOMY Bacteriophages have been identified in a variety of forms and may contain RNA or DNA genomes of varying sizes that can be single- or double-stranded nucleic acid (Ackermann, 2003, 2006, 2007). Of all the bacteriophages examined by transmission electron microscopy, 95% of those reported are tailed, with only 3.7% being filamentous, polyhedral, or pleomorphic (Ackermann, 2007). The tailed bacteriophages contain a linear, double-stranded DNA genome that can vary from 17 to 500 kb in the order Caudovirales, which is further divided into three families on the basis of tail morphology (Ackermann, 2003, 2006). These bacterial viruses have icosahedral heads, while those phages with contractile tails are placed in the Myoviridae, those with a long noncontractile tails are placed in the Siphoviridae, and phages with short tails are members of the Podoviridae. Although bacteriophages of the Caudovirales (tailed phages) may be physically similar, it has been difficult to classify them by DNA or protein sequences as a result of the tremendous diversity caused by horizontal gene transfer (Casjens, 2005) and the mosaic nature of bacteriophage genomes (Hendrix et al., 1999). There appears not to be even one candidate conserved gene that can be utilized to categorize all phages for a suitable classification scheme (Nelson, 2004). One approach has been to construct a socalled phage proteomic tree on the basis of predicted protein sequences of a bacterial virus (Rohwer and Edwards, 2002), another approach is to divide bacteriophages on the basis of genome type (for example single-stranded RNA or DNA) with a further demarcation by physical characteristics such as tailed or filamentous types (Lawrence et al., 2002), and a genomic modular approach has also been suggested to classify bacteriophages (Proux et al., 2002). The most frequently encountered Campylobacter bacteriophage are the double-stranded DNA, tailed phage, with icosahedral heads, belonging to the family Myoviridae. In common with the prototype phage T4 of Escherichia coli, Campylobacter phage of the Myoviridae have DNA base modifications that make them difficult to clone and sequence, nevertheless the first genomic sequence has recently been completed at the Sanger Institute (http://www.sanger.ac.uk).
CHAPTER 38
Preliminary annotation of this sequence identifies the structural proteins that make up the capsid and tail fibers that can be confirmed from protein sequences. The T4-type hallmark protein gp23 represents the major capsid protein and is clearly recognizable in the Campylobucter phage proteome, but strangely, it shows the highest degree of protein sequence similarity with translations of T4-type bacteriophage metagenomic sequences recovered from marine environments. However, the genome contains a number of new reading frames encoding hitherto unrecognized proteins, and needless to say, these do not assist in the categorization process (A. E. Scott et al., unpublished observations).
BACTERIOPHAGE THAT INFECT CAMPYLOBACTER The first reports of the isolation of Cumpylobacter bacteriophages were in 1960s, where phage that infected Vibrio coli (now known as C. coli) or V. fetus (now known as C. fetus) were isolated from cattle and pigs (Firehammer and Border, 1968; Fletcher, 1968; Fletcher and Bertschinger, 1964). In the early 1980s, "C. jejuni" (perhaps C. fetus using the current nomenclature) isolated from aborted sheep fetuses were found to induce lysogenic phages when treated with mitomycin C (Bryner et al., 1982). Cumpylobacter phages were also reported to play a role in the autoagglutination of cells, which interfered with attempts to serotype Campylobacter isolates (Ritchie et al., 1983).
PHAGE TYPING Phage typing for Campylobacter spp. has been developed (Frost et al., 1999; Grajewski et al., 1985; Khakhria and Lior, 1992; Sails et al., 1998; Salama et al., 1990) and has been compared with other classification schemes (Gibson et al., 1995; Hopkins et al., 2004). Grajewski et al. (1985) collected samples from poultry manure and isolated 45 bacteriophages by spot-testing on Campylobacter isolates that were considered suitably divergent for typing purposes. From these phages, 14 clusters were identified as appropriate for typing assays. Although 88% of the campylobacters examined were typed by this method, not all bacterial isolates could be categorized, and none of the bacteriophages utilized in the scheme were characterized by electron microscopy. Phagetyping has been utilized to differentiate Campylobacter spp. strains from animal and environmental sources isolated during outbreak investigations. In
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
681
some outbreaks phage-typing proved potentially more discriminatory than serotyping or biotyping (Salama et al., 1990). Khakhria and Lior (1992) developed an extended phage-typing scheme for C. jejuni and C. coli from 46 different phage types by using 19 typing phages. These investigators examined Campylobacter isolates from 17 different countries and were able to demonstrate that contaminated cattle and poultry appeared to be the most common sources of infection for humans. Frost et al. (1999) isolated 57 C. jejuni phage types, with the ten most prevalent phage types accounting for 60% of the bacterial isolates tested, but 16% of the isolates could not be typed by this method. Only 12 phage types were identified among C. coli. Although phage-typing schemes have been utilized by investigators, Gibson et al. (1995) concluded that pulsed-field gel electrophoresis (PFGE) was a more discriminatory subtyping method. Hopkins et al. (2004) extended the comparisons by examining fluorescent amplified fragment length polymorphism analysis of C. jejuni and C. coli with phage types. These investigators demonstrated a lack of congruity between some serotypes andlor phage types with this genotype method.
CAMPYLOBACTER PHAGE CHARACTERISTICS The phages that make up the phage typing system of Frost et al. (1999) were characterized by Sails et al. (1998). All the phages reported by these investigators had icosahedral heads and long contractile tails that were classified as members of the Myoviridae. The phages were further subdivided into three groups according to genome size and head diameter. Two phages with head diameters of 140.6 and 143.8 nm and large genome sizes of 320 kb were classified as group I. Five phages were classified into group I1 had average head diameters of 99 nm and average genome sizes of 184 kb. Group I11 contained nine phages with average head sizes of 100 nm and average genome sizes of 138 kb. The sixteen phages used in the United Kingdom typing scheme were also categorized into four groups on the basis of their patterns of activity against spontaneous, transposon and defined mutants of C. jejuni (Coward et al., 2006). Interaction of C. jejuni bacteriophage with the host bacterium was dependent on a functional capsular polysaccharide, whereas two of the other three groups were dependent on the ability of the bacteria to be motile. Consequently, the extracellular polysaccharide and flagella may serve as receptors for C. jejuni phages. This last finding is in-
682
CONNERTON ET AL.
teresting in that a C. jejuni flagellum-specific phage PV22 from Proteus vulgaris was identified as interacting with C. jejuni by attachment to the flagella followed by translocation of the phage to the polar region of the bacterium up to the point of DNA injection (Zhilenkov et al., 2006). Phage accumulated primarily on the surface of cells at sites where flagella originated; PV22 did not inject DNA into C. jejuni and did not produce lytic plaques on medium containing C. jejuni cells. Recently, phages were isolated that exhibited differential lytic activities to various C. jejuni strains examined for viral infection from the Russian Federation. Two bacteriophages had contractile tails considered to be morphotype A1 of the family Myoviridae, while a third had a long noncontractile tail of morphotype B 1 in the family Siphoviridae. A fourth phage had an icosahedral head that was classified as morphotype B1 of the Siphoviridae, while a fifth phage had an icosahedral head with a short tail of morphotype C1 in the Podoviridae. Transmission electron micrographs of a few representative Campylobacter phages are shown in Fig. 1. Recent genome sequence data has revealed the presence of Mu-like phage sequences in C. jejuni RM1221 (Fouts et al., 2005), and that these sequences are present in other C. jejuni strains but not in the prototype genomic sequence of C. jejuni NCTC 11168 (Parkhill et al., 2000). The majority of the open reading frames within the prophage regions are hypothetical, and as yet have not been ascribed any function.
SOURCES OF CAMPYLOBACTER BACTERIOPHAGE Campylobacter phage have been isolated from various sources including the feces of pigs, cattle, and sheep (Bryner et al., 1970, 1973; Firehammer and Border, 1968); abattoir effluent, sewage, manure and the excreta of both broiler and layer chickens (Grajewski et al., 1985; Khakhria and Lior, 1992; LOC Carrillo et al., 2007; Sails et al., 1998; Salama et al., 1989). However, isolations of C. jejuni phage from sources not related to poultry were rare (Zhilenkov et al., 2004). As campylobacters are so pervasive in commercial broiler chickens, the majority of studies examining Campylobacter phage have concentrated on broiler farms. Campylobacter-specific bacteriophages can be readily isolated from poultry excreta (Atterbury et al., 2005; Connerton et al., 2004; El-Shibiny et al., 2005; LOCCarrillo et al., 2007). Additionally, at least a proportion of the phages associated with broiler chickens remain viable in processing plants
and can be isolated from retail chicken portions (Atterbury et al., 2003b; Tsuei et al., 2007). One report suggested that large populations of host bacteria (2.5 log CFU/g) on the surface of food are required in order to successfully isolate phages from foods (Greer, 2005). The numbers of Campylobacter present on the surface of broiler chickens at retail have been reported to reach levels of l o 9 CFU per carcass (Jorgensen et al., 2002). Campylobacter phages on the skin of retail chicken portions have been recovered at levels of 2 X lo3 PFU/10 cm2 (Atterbury et al., 2003b), which is similar to the levels observed for coliphages in previous studies (Kennedy and Bitton, 1987). This study found that Campylobacter phage could be isolated from chicken skin only when detectable levels of their host were also present.
BACTERIOPHAGE THERAPY Selection of appropriate phage is paramount to the success of phage therapy. It is essential that the selected phages demonstrate high virulence against the target bacteria. Virulent phages always follow a course of infection that results in a burst of phage through cellular lysis, after a relatively fixed interval. In contrast, lysogenic or temperate phages can integrate their DNA into the host DNA and render the host bacterium immune to further infection and are therefore unsuitable for phage therapy. Lysogenic phages are also prone to transduction, the phage mediated process where genetic material is transferred from one bacterial host to another, which can result in the dissemination of pathogenic traits amongst their hosts (Boyd and Brussow, 2002; Cheetham and Katz, 1995). Once appropriate virulent phages have been selected, it is generally recognized that phage therapy confers several advantages over conventional antimicrobial therapies. These are summarized in Table 1 together with some potential shortcomings. Adoption of phage therapy has been held back by a lack of consistent proof of efficacy. Previous failures may have been due to a general lack of understanding of the kinetics of phage replication, which is a density-dependent process to which mathematical models can be applied (Bull et al., 2002; Levin and Bull, 1996; Payne and Jansen, 2001). Phage replication is critically dependent on the density of bacteria present. A consequence of these kinetics is that there will be a threshold above which phage numbers increase and below which they decrease; this theoretical value has been called the phage proliferation threshold (Payne and Jansen, 2003; Wiggins and Alexander, 1985). The threshold value presumes that
CHAPTER 38
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
683
Figure 1. Transmission electron micrographs of Cumpylobucter bacteriophage. (A) Bacteriophage CP8 used for phage therapy trials (LOCCarrillo et al., 2005). (B) Bacteriophage CP220 empty capsid after DNA insertion. (C) Bacteriophage NCTC 12677, which is one of the large phage: used for phage typing (Frost et al., 1999). (D) Bacteriophage CampMu observed by Scott et al. (2007a).
all the bacteria and phage present are available for infection to occur, but this may not may be true in vivo, where greater rate of phage losses may be encountered and the bacteria may become portioned or sequestered from the main population. The latter
consideration is of particular importance to phage applications against bacteria colonizing the intestinal tracts of animals because the adherence of the target bacteria to the large surface area available and the nature of the mucosa to which they adhere could
Table 1. Advantages and disadvantages of bacteriophage therapy over conventional antimicrobial treatments Advantages
Disadvantages
Bacteriophages are self-replicating and self-limiting and will only multiply when sensitive bacteria are present. The specificity of bacteriophages means that resident gut flora will not be affected by their action. Bacteriophages can be selected that target surface receptors that are either important for the fitness of the host bacteria or have a key role in pathogenesis of the bacteria. Bacteriophages are isolated from the animals themselves so the likelihood of side effects such as an allergic response is remote. Phages are generally inexpensive to prepare.
Potential to transfer pathogenic traits but this can be overcome by careful selection of only virulent phage. Careful selection of the phages to be applied is required in order to retain species specificity but exhibit virulence against a wide range of different strains of the target bacterial species. The target bacteria may develop phage resistance to escape phage therapy. However, the careful selection of phage that target key surface receptors as outlined in the adjacent column can help prevent this.
684
CONNERTON ET AL.
limit their interaction with phage. For these reasons, the effective proliferation threshold may be greater in vivo. The outcome of phage therapy will also depend up on various life-history parameters, including the inoculum size, inoculum timing, phage absorptionrate, and burst size (Levin and Bull, 1996; Payne and Jansen, 2001,2002; Weld et al., 2004). It is therefore difficult to translate information gained from interaction of homogeneous bacteria and phage populations in a well-mixed and controlled environment to the situation in vivo. The intestinal lumen is a complex environment, where various physical (for example, constant flow of material), physiological (for example, levels of oxygen), host defenses and biochemical factors (such as pH) all influence the populations of colonizing bacteria. In addition, the kinetics of phage absorption in the intestine may be quite different from those in laboratory media as a result of the viscosity of the mucous layer (Weld et al., 2004). When phage are mixed with bacteria at ratios at which they greatly outnumber the bacteria (high multiplicity of infection), then the bacteria may become lysed from without. This occurs because the phage particles interact with the bacteria in large numbers, and together destabilize bacterial membranes. This process may lead to an initial drop in bacterial counts and phage titer, with yet more phage losses occurring as they adhere to bacterial-cell debris rather than to growing bacteria (Rabinovitch et al., 2003). The selection of resistant bacteria has always been perceived as a potential obstacle to phage therapy (Barrow, 2001), and has been reported after experimental phage treatments (Sklar and Joerger, 2001; Smith and Huggins, 1982; Smith et al., 1987a). However, as argued above, if careful consideration is given to the selection of the phage from an environment to which they are to be deployed, then phage resistance can have unfavorable consequences for the bacteria, such as a reduction in the colonization potential or loss of virulence (Adams, 1959; Scott et al., 2007b). Evidence against the dominance of phage-resistant populations in Campylobacter can be gained from the examination of natural phage infections. In a longitudinal study of a broiler chicken house naturally infected with Campylobacter and phage over three successive rearing cycles, it was observed that occasional phage-resistant Campylobacter strains could be isolated, but these did not become the dominant type or outgrow cocolonizing sensitive types (Connerton et al., 2004). Instead, they coexisted with the phage-sensitive types. The phageresistant strains were found to be genotypically unrelated to the phage-sensitive strains and not resistant mutants of the same genotype. The observation that
phage-resistant campylobacters do not emerge as dominant populations, despite their obvious advantage in the presence of phage, and the observation that the majority of infected flocks do not lead to Campylobacter strain carry-over (Petersen and Wedderkop, 2001; Shreeve et al., 2002), indicates that phage treatment would be unlikely to select for the persistence of specific resistant types in the broilerhouse environment.
EXAMPLES OF THE APPLICATION OF BACTERIOPHAGE TO REDUCE CAMPYLOBACTER IN POULTRY Phage Treatment of Chickens There are a number of general practical considerations that need to be addressed before any application of phage therapy. How these are related to the phage therapy of broiler chickens to reduce Campylobacter colonization is outlined in Table 2, but of these, the choice of phage is probably the most critical to the success of phage treatment. Phage treatment of chickens was first reported by Wagenaar et al. (2005). In a series of experiments, the authors compared the effects of both therapeutic and preventative treatment of broiler chickens by using two of the phage used in the phage typing scheme of Frost et al. (1999). The administration of phage NCTC 12671 (dose approximately log 10 PFU) on each of six consecutive days to birds at 15 days of age that had been infected 5 days earlier with C. jejuni C356 resulted in a 3 log CFU/g decline in cecal counts of C. jejuni within 48 h, compared with nonphage-treated controls. After 5 days the Campylobacter counts in the treated birds had increased but remained 1 log CFU/g below controls. Preventative phage treatment using phage NCTC 12671, where doses between log 9.7 and log 10.3 PFU were given each day to chicks from 7 days of age, delayed but did not prevent the onset of C. jejuni colonization in young birds. The peak titers remained 2 log lower than the controls. In both applications, the colony forming units and phage forming units rose and fell over time, and were out of phase with each other, which is typical of predator-prey populations in nature. In a third experiment, birds were infected at 32 days of age and treated with both phage NCTC 12669 and phage NCTC 12671 at 39 days of age at doses between log 9.7 and log 11.6 PFU, for four consecutive days. Birds of this age were used in order to mimic the age at which broiler chickens are approaching the optimum size for slaughter because this would be the ideal time to administer the treatment to naturally infected birds. Campylobacter counts in
CHAPTER 38
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
685
Table 2. Practical considerations for phage therapy Aspect under consideration Choice of phage
Method of application
Timing of intervention
Quality control
Cocktails
Rationale Not all phages that are able to lyse the target bacteria in the laboratory are suitable for practical application (Reynaud et al., 1992). Minimally, they must be able to reach the target in sufficient titers to effect changes in the target population. Phages have the advantage of being fairly robust, and therefore they can be simply added to drinking water and feed, provided that the intended targets are intestinal bacteria. However, some phages may be sensitive to the low pH encountered in the stomach or proventriculus of avian species (Leverentz et al., 2001). This problem can be overcome through the use of antacid or by selection of appropriate low-pH-tolerant phage. Antacids, such as Maalox (aluminum and magnesium hydroxide) or calcium carbonate, have been used to improve the ability of phage to survive acidity in digestive systems (Smith et al., 1987b; Koo et al., 2001). The administration of phage 2 to 3 days before slaughter would limit exposure to the phage, while allowing the phage to replicate before removing chickens from the broiler house, and any surviving phage-resistant bacteria in their intestinal tract. The lead time must account for phage absorption rates, phage replication rates, inherent dilution factors associated with the intestinal contents, and transit time of the gut. It is necessary to ensure that the treatment phage can be distinguished from wild-type phage, and that the phages recoverable from treated birds are of the same type as those administered. Phage stocks should be quality controlled by using genotypic methods alongside their lysis patterns on stock Campylobacter strains. Because of the highly specific nature of phages, it has been suggested that they be applied as a mixture or “cocktail” to cover a broader range of hosts (Kudva et al., 1999; Sklar and Joerger, 2001). This tactic will assist in the efficacy of the phage preparation but will require that the individual components be produced and tested individually, to ensure their contribution to the host range of the target bacterium.
this experiment decreased by 1.5 log CFU/g cecal contents compared with controls within 24 h of administration, but rose slightly after four days remaining approximately 1 log CFU/g lower than the counts in the control birds. No adverse effects of phage treatment on the birds were observed. LOCCarrillo et al. (2005) performed phage therapy experiments by using bacteriophage and C. jejuni host strains isolated from broiler chickens. A large number of phages were characterized, but only those that showed a broad lytic spectrum were selected for the phage therapy experiments. The kinetics of bacteriophage replication and host growth were first examined in vitro in order to gauge the course of infection for subsequent in vivo experiments. To evaluate the efficacy of phage therapy, low passage C. jejuni isolates from broiler chickens were selected on the basis of their ability to reproducibly colonize 20- to 22-day-old broilers. The age of the birds was selected to parallel the first observation of colonization often observed in commercial broiler chickens (Newel1 and Fearnley, 2003). By using phagesensitive C. jejuni isolates HPC5 and GIIC8 at a dose of log 8 CFU, maximal colonization could be achieved within 48 h, and these levels could be shown to be maintained over a 9-day period (chickens 22 to 30 days of age) without any evidence for the invasion of non-gastrointestinal tract tissues (liver, pancreas, heart, and kidney).
Phage therapy experiments were carried out by using bacteriophages administered as a single treatment to 25-day-old birds precolonized with Campylobacter, in an antacid suspension. Phages were administered at three different doses, log 9, log 7, and log 5 PFU. All the experimental bacteriophage treatments of C. jejuni colonized birds resulted in the phage persisting in the chicken intestinal tract, implying that the phages administered were delivered to the intestinal sites colonized by C. jejuni and were able to replicate within this environment. It became apparent that the optimum dose for phage therapy was log 7 PFU with the higher (log 9 PFU) and lower doses (log 5 PFU) of phage being generally less effective (LOCCarrillo et al., 2005). The reason for the highest dose being less effective was unclear, but it may have been because higher phage densities are prone to phage aggregation and nonspecific association with digesta or nonhost bacteria (Rabinovitch et al., 2003). For the lower dose (log 5 PFU) the numbers surviving ingestion may be simply insufficient to cause an effect within the experimental time frame. The success of phage treatment in reducing Campylobacter numbers depended on the bacteriophage and on the colonizing strain. Some phage, for example CP8, were very effective with one strain (GIIC8) but ineffective with another (HPC5). The ability of different bacteriophage to reduce the numbers of campylobacters in the cecal contents of chick-
686
CONNERTON ET AL.
ens colonized with several different host strains after a 48-h period is shown in Fig. 2. The reductions observed in Cumpylobucter colonization levels after phage administration were between log 1.5 and log 5 CFU/g of intestinal contents. The numbers of campylobacters in the cecal contents, the upper and lower intestines of the birds, were sampled over a five day period after administration of phage and the results summarized in Table 3 in terms of log reductions in Cumpylobucter counts compared with chickens administered with a placebo. The greatest reductions recorded in the cecal Cumpylobacter counts were observed between 24 and 72 h postphage treatment in all three sample sites, after which the Cumpylobacter counts began to recover. The single treatment regime appeared to be effective over 48 h, which would give a window of opportunity to catch and transport the birds for slaughter. The selection of resistant bacteria has always been perceived as a potential drawback to phage therapy and has been reported after phage treatments of experimental animals colonized by Escherichiu coli (Sklar and Joerger, 2001; Smith and Huggins, 1982; Smith et al., 1987a). However, for C. jejuni colonizing broiler chickens, phage resistance can be correlated with a reduced ability to compete against phagesensitive types in the same environment (Scott et al., 2007a, 2007b). In the phage therapy experiments of
LOCCarrillo et al. (2005) the incidence of phageresistant phenotypes recovered from the intestinal contents of birds postphage challenge was 4%. Notably, the phage-resistant types did not become the dominant population in phage-treated chickens despite the continued presence and predation of their sensitive counterparts. Chickens subject to environmental challenge will encounter diverse Cumpylobacter genotypes; under these conditions, it is therefore more likely that succession by phage insensitive genotypes will occur rather than through the selection of phage-resistant strains (Connerton et al., 2004). To ensure the effectiveness of phage therapy and prevent the selection of phage-resistant types or insensitive types in mixed populations, it would be prudent to use more than one phage. In practice cocktails of phages bearing different receptor specificities are preferred. Clearly chickens reared outdoors will be exposed to a wider set of Campylobucter genotypes and possibly bacteriophages than those reared in barns, where biosecurity measures are used to limit the exposure of the birds to biological agents that will affect their welfare or market value. For these reasons limited phage cocktails could be used in these situations that could easily be refreshed with new phages, and with the possibility cycling their use to retain their efficacy.
Controls
1
Campylobacter host: Bacteriophage :
GllC8 CP8
i
T
CP30 F3
-7-
c10
HPCS
HPCS
CP30
CP34
CP8
Figure 2. Comparison of different phage/host combinations in cecal contents, 48 h after phage was administered to precolonized chickens (n 2 5 birds per sample point) together with controls. A single log 7 PFU dose was administered to the treatment group at 25 days of age. Adapted from LOCCarrillo et al. (2005) and Scott et al. (2007a).
CHAPTER 38
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
687
Table 3. Comparison of the effect of different phage/host combinations“
Canzpylobacter host
Days after intervention
CP8
GIIC8
034
HPC5
CP8
HPC5
1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Phage
Log reductionb in count in: Upper intestine
Ceca
Lower intestine
4.1 3.9 3.6 2.5
5.4 5.4 4.2 2.6 2.1 2.7 1.5 2.9 2.8
4 4 4.2 3.2 1.8 2.2 2.1 3 2.6 2.5 1.7 0.6 1.4 1.2 1.1
NSD 1.6 1.1 2.6 2.1 1.9 1.4 1.2
NSD NSD NSD
NSD NSD NSD NSD NSD NSD
“Adapted from data in LOCCarrillo et al. (2005). Results are expressed over 5 days, in three parts of the intestine, after phage were administered to precolonized chickens (n 2 5 birds per sample point), expressed as log Cumpy106ucter count reduction compared with controls. *Compared with untreated Cumpylo6ucter colonized controls. ‘NSD, no significant difference (P 2 0.05).
Use of Bacteriophage as a Disinfection Agent for Poultry Meat Using bacteriophage to reduce contamination of foods with zoonotic pathogens such as Cumpylobucter spp. requires an in-depth understanding of the epidemiology of the pathogen against which the phage preparation is to be used. It also requires the identification of critical intervention points in the processing cycle where phage application will be most beneficial (Stone, 2002). Phages may be applied directly onto raw produce or onto environmental surfaces in processing facilities, to reduce numbers of food-borne pathogens (Sulakvelidze and Barrow, 2005). Phages could be sprayed onto the chicken carcasses after postchill processing (e.g., after the chlorine wash in chiller tanks in the United States, or after processing through air chillers in Europe). Unlike in vivo phage therapy, the aim of carcass treatment is not to reduce bacterial numbers through viral replication, but to kill them rapidly through passive inundation or “lysis from without.” This phenomenon occurs when a large number of phage adsorb to the host bacterium en masse. This compromises the integrity of the cell wall and causes the bacterium to swell and burst (Delbruck, 1940). The advantage of targeting phage treatment on processed carcasses is that the campylobacters are relatively immobile and unable to grow under refrigeration conditions. As such, there is no growth of phage-resistant or insensitive subpopulations on the surface of the product. Additionally, there is lim-
ited recycling of phage in the environment, reducing the possibility of resistance developing. Several studies have investigated the use of phages to reduce Cumpylobucter numbers on the surface of experimentally contaminated chicken skin. Goode et al. (2003) demonstrated that spraying chicken skin contaminated with lo4 CFU of C. jejuni C222 with lo6 PFU of phage 12673 reduced Cumpylobacter recoveries by 95%. A 1log CFU reduction of C. jejuni NCTC 12662 was recorded by Atterbury et al. (2003a) using phage NCTC 12674. This study also found that a combination of phage application and freezing led to a greater reduction in Cumpylobucter numbers (2.5 log CFU) than the sum of reductions when these methods were applied separately. As such, phage treatment could be particularly useful in countries such as Iceland, which routinely freeze Cumpylobacter-positive birds (Stern et al., 2003). Atterbury et al. (2003a) suggested that the observed reduction in Cumpylobucter contamination was most likely to result from the phage adsorbing to their hosts and then initiating replication during the recovery phase on agar plates. Demonstrating efficacy in laboratory trials is only the starting point in the process of making phage treatment of food surfaces viable, and many additional important issues will have to be addressed before phages can be used to improve food safety in real-life settings (Greer, 2005). For example, phage preparations added to foods must meet stringent requirements for purity, which requires the develop-
688
CONNERTON ET AL.
ment of commercially viable protocols for large-scale phage production (Greer, 2005).
PUBLIC ACCEPTABILITY OF THE TECHNOLOGY The reaction of the public to the use of bacteriophage in food is largely untested. To address this issue, a focus-group study to gauge public acceptability of phage intervention was initiated (http: 1lwww.defra.gov.uk 1 Science 1 LINK I publications 1 newsletters/foodlink/FoodLINK~Issue47.pdf). In general, the idea of phage intervention therapy was well received by volunteers drawn from the public. There was a general desire for further research, but, at the end of the proceedings, only one of 23 participants retained concerns over the use of bacteriophage for reducing the incidence of campylobacters in chickens. Key findings were the public preference for labeling the treated product and the opinion that treated foods may be worth a premium. The results of this focus-group study could be of value in relation to other bacteriophage applications involving food production.
REGULATORY ISSUES There are many restrictions on the types of additives and processing aids that can be used in food production. Therefore, it was important to establish that Campylobacter-specific phages could be found naturally associated with poultry meat (Rees and Dodd, 2006). However, government regulatory bodies may yet require that phages used for biocontrol purposes be removed from the final product or inactivated (Rees and Dodd, 2006). Atterbury et al. (2003a) found that freezing chicken skin inoculated with Campylobacter phage led to an appreciable loss of phage recovery, which suggests that physical methods could be used after treatment to inactive the phages if required. One possible method to do this would be crust freezing, which minimizes damage to the appearance of the carcass (James et al., 2007). The U.S. Food and Drug Administration has approved the use of a specific bacteriophage preparation (LMP-102) on ready-to-eat foods and meat/ poultry for the control of Listeria spp. (Lang, 2006). There are several companies in the United States and Europe working toward the commercial use of bacteriophage to control food-borne pathogens. However, the use of bacteriophage in the European Union would be subject to directive 891107/EEC (Food Additives and Processing Aids), which states that the use
of any chemical or substance in food preparation or processing is banned unless it is explicitly authorized by the European Union (http://ec.europa.eu/food/fs/ sfp/addit-flavor/flavO7-en.pdf). An additional problem is that the use of phage to decontaminate poultry could fit the definition of both a food additive and a processing aid. Although the use of phage treatments is being debated among the European Union member states, the bureaucratic nature of the organization means that approval for such treatments is unlikely to materialize in the immediate future.
IMPLICATIONS OF PHAGE INFECTION ON THE ECOLOGY AND EVOLUTION OF CAMPYLOBACTER Bacteriophage Influence the Strains of Campylobacter that Populate Chickens Campylobacters are fastidious bacteria, sensitive to atmospheric oxygen, and they have little or no global stress response mechanisms to counter adverse environmental conditions (Murphy et al., 2006). This coupled with the compactness of the genome means that campylobacters must generate diversity within a population at a genomic level to enable some strains to survive potential challenges. Recombination of genetic material between C. jejuni genotypes in vivo is a frequent event that gives rise to heterogenic populations (Avrain et al., 2004; Fearnhead et al., 2005; Schouls et al., 2003). It has also been established that some strains are better able to colonize and persist in chickens for much longer periods of time than others (Gaynor et al., 2004; Jones et al., 2004; McCrea et al., 2006). When chickens are exposed to multiple genotypes of Carnpylobacter, the dominant genotype, and therefore, the strain most likely to be recovered at that time is the one that is able to exploit the environment to the greatest degree under the prevailing conditions. However, the dominant Carnpylobacter type and species may change several times over the rearing cycle of chickens exposed to multiple sources of Campylobacter such as free-range or organic flocks (El-Shibiny et al., 2005, 2007). One of the environmental pressures encountered by many bacterial populations is predation by bacteriophage. When bacteriophages are present, succession of one genotype over another may be due to incursion of new genotypes with innate phage insensitivity, rather than de novo development of resistance (Connerton et al., 2004). Further to this work, Scott et al. (2007a) examined the C. jejuni strains F2E1, F2E3, and F2C10 that were all isolated from chickens during the study of Connerton et al. (2004). Strains F2E1 and F2C10 were, respectively, resistant and sensitive to all 25
CHAPTER 38
bacteriophage isolated from the same flock. Strain F2E3 was also isolated from this flock but exhibited sensitivity to 6 of the 25 bacteriophage isolated from the same flock. Characterization of these strains by PFGE, heat-stable antigen agglutination, and antibiotic tolerance profiling indicated that F2E1 and F2E3 were likely to have had a shared ancestry because they shared a common PFGE macrorestriction profile, heat-stable serotype and similar antibiotic tolerances or sensitivity profiles. Strain F2C10 had a different PFGE macrorestriction profile, heat-stable serotype and antibiotic profile, indicating it to be of a different clonal lineage. However, F2E3 differed from F2E1 at a single multilocus sequencing type allele, which it shared with F2C10. A strategy to PCR walkout from this locus established that F2E3 was a recombination product of F2E1 and F2C10, in which a 112-kb region of DNA had been substituted from F2C10. The relationships between these three strains thus allowed an opportunity to examine competitive success in vitro and in vivo between C. jejuni strains that were inherently sensitive and insensitive to bacteriophage CP30 from the same source (F2C10 and F2E1, respectively) and strains of altered resistance acquired through horizontal DNA transfer (F2E3). Administration of strains F2E1, F2E3, and F2C10 individually to broiler chickens indicated that all were efficient colonizers of chickens, although strain F2C10 colonized to a significantly higher level, indicating that there was no significant fitness cost associated with differential phage sensitivity, but rather that the strain lineage was more important in determining its ability to colonize. However, on competitive colonization of broiler chickens the resistant strain F2E1 was unable to compete successfully with either of the bacteriophage-sensitive strains and became a minor subpopulation. In the presence of bacteriophage this situation was completely reversed. It was now the resistant strain F2E1 that dominated and the sensitive strains reduced to subpopulations as a result of bacteriophage predation of F2E3 or F2C10. In this model system the presence of bacteriophage was the key determinant for the success or otherwise of C. jejuni strains in broiler chickens. That F2E1 was outcompeted to such a degree in the absence of bacteriophage suggests that its selection and subsequent survival owes much to the presence of bacteriophage. Colonization in the absence of competitors indicated that all of these strains could colonize well, but it was clear that the strain insensitive to the bacteriophage was associated with a competitive fitness disadvantage. This competitive difference was not evident in vitro. Cocultures of these strains showed little or no difference in their growth rates or yields. These findings have implications regarding
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
689
the types of strains isolated from different sources because the presence of bacteriophage may bias the isolation rates of different strains colonizing the same intestinal environment. Genomic Rearrangement of Carnpylobacter jejuni in Response to Bacteriophage Predation The selective pressure exerted by bacteriophage predation in the poultry intestine can also influence the evolution Carnpylobacter genome. Scott et al. (2007b) reported intragenomic inversions of up to 590 kb about the origin of replication of the C. jejuni HPCS chromosome in response to exposure to virulent bacteriophage in broiler chickens. These strains were recovered as a minor phage-resistant population (4%) from chickens after phage therapy with bacteriophage CP34 (LOCCarrillo et al., 2005). The recombination breakpoints of the inversions were identified and found to lie between a central copy and either of two flanking copies of Mu-like prophage DNA sequences located in the genome of C. jejuni HPCS. Upon laboratory culture, these strains retained their genotype and remained resistant to the virulent bacteriophage that affected their selection. However, reintroduction of these strains into chickens revealed they were compromised in their ability to colonize chickens when administered at low doses as compared with their progenitor HPCS, and all the C. jejuni recovered from these birds had reverted to phage sensitivity and had undergone a further round of genome rearrangement involving the Mu-like prophage elements. The reversion to phage sensitivity was accompanied by the recovery of the ability to efficiently colonize chickens in these strains. Recombination between the Mu-like prophage elements leading to resistance to the virulent phage also lead to reactivation of the lysogen and the production of infectious CampMu (Fig. 1). The CampMu arising from the alternative recombination events exhibited different host ranges that could be propagated on independent C. jejuni strains, implying variation in the gene contents of the CampMu. The spontaneous production of CampMu bacteriophages after bacteriophage therapy is of concern because Mu bacteriophages are potential agents of mutation. However, recent evidence suggests that Campylobacter populations are already exposed to CampMu bacteriophages (Barton et al., 2006) as a result of the widespread presence of the prophages in Campylobacter populations (Parker et al., 2006). There was a striking difference between the responses of C. jejuni HPCS to virulent bacteriophage in vitro compared with that observed in the intestinal tracts of broiler chickens. Infection of C. jejuni HPC5
690
CONNERTON ET AL.
by bacteriophage CP34 in laboratory culture resulted in over 90% of the bacterial survivors being phage resistant and nonmotile with no evidence of any genomic rearrangement. The phage-resistant C. jejuni recovered from chickens were as motile as C. jejuni HPCS. Functional flagella have been reported as necessary for the efficient colonization of chickens by C. jejuni (Nachamkin et al., 1993; Wassenaar et al., 1993). Therefore, any mutants escaping phage predation through alterations in flagella function will not compete with their motile counterparts in chickens. It is likely that bacteriophage CP34 has selected the flagella as part of its infection process because it is an essential component of successful C. jejuni that are capable of densely populating the chicken intestine. C. jejuni flagellin is known to be polymorphic and variably 0-glycosylated leading to differences in serospecificity (Logan et al., 2002; Thibault et al., 200 1).This variation could enable campylobacters to evade bacteriophage predation without any loss in motility, although in the arms race between predator and prey this will in turn select bacteriophage that can overcome any such host recognition problems. Virulent bacteriophage can provoke host recombination with the potential to activate dormant prophages. However, although pathogen evolution can be rapid, resistance to the therapeutic bacteriophage is associated with a fitness cost rendering the resistant strain noncompetitive in the absence of the bacteriophage (Scott et al., 2007b).
CONCLUSIONS Attempts to utilize bacteriophage, initially for typing purposes and more recently for their biocontrol potential, have led to a greater awareness of the role that phage play in the complex ecology of Cumpylobacter. Although the bacteriophages of Cumpylobacter were once considered merely an interesting adjunct to mainline epidemiological and pathogenesis studies, we now believe they not only hold the potential to reduce the burden of campylobacters in poultry, but are also of key importance to generally understanding the dynamics of Campylobacter populations in avian species. Bacteriophage predation can create subpopulations in the intestinal tracts of domestic poultry that are unlikely to recovered by means of conventional culture methods. Although these types may constitute a minor part of the total population and be practically invisible to surveillance methods, they can still be present in more than sufficient numbers to be transferred through the food chain and cause human disease. Finally, it should be noted that bacteriophage can shape the evolution of
Campylobacter genomes as they do in other bacterial genera. REFERENCES Ackermann, H.-W. 2003. Bacteriophage observations and evolution. Res. Microbiol. 154:245-251. Ackermann, H.-W. 2006. Classification of bacteriophages, p. 816. In R. Calendar and S. T. Abedon (ed.), The Bacteriophages. Oxford University Press, Oxford. Ackermann, H.-W. 2007. 5500 phages examined in the electron microscope. Arch. Virol. 152:227-243. Ackermann, H.-W. 2001. Frequency of morphological phage descriptions. Arch. Virol. 1465343-857. Adams, M. H. (ed.). 1959. Bacteriophage. Interscience, New York. Alexander, M. 1981. Why microbial predators and parasites do not eliminate their prey and hosts. Annu. Rev. Microbiol. 35: 113-133. Atterbury, R. J., P. L. Connerton, C. E. Dodd, C. E. Rees, and I. F. Connerton. 2003a. Application of host-specific bacteriophages to the surface of chicken skin leads to a reduction in recovery of Campylobacter jejuni. Appl. Environ. Microbiol. 69: 6302-6306. Atterbury, R. J., P. L. Connerton, C. E. Dodd, C. E. Rees, and I. F. Connerton. 2003b. Isolation and characterization of Campylobacter bacteriophages from retail poultry. Appl. Environ. Microbiol. 69:4511-4518. Atterbury, R, E. Dillon, C. Swift, P. Connerton, J. Frost, C. Dodd, C. Rees, and I. Connerton. 2005. Correlation of Campylobacter bacteriophage with reduced presence of hosts in broiler chicken caeca. Appl. Environ. Microbiol. 71:4885-4887. Avrain, L., C. Vernozy-Rozand, and I. Kempf. 2004. Evidence for natural horizontal transfer of tetO gene between Campylobacter jejuni strains in chickens. J. Appl. Microbiol. 97:134-140. Barrow, P. 2001. The use of bacteriophages for treatment and prevention of bacterial disease in animals and animal models of human infection. J. Chem. Techol. Biotechnol. 76:677-682. Barrow, P., M. Lovell, and A. Berchieri, Jr. 1998. Use of lytic bacteriophage for control of experimental Escherichia coli septicemia and meningitis in chickens and calves. Clin. Diugn. Lab. Immunol. 5:294-298. Barton, C , L. K. Ng, S. D. Tyler, and C. G. Clark. 2006. Temperate bacteriophages affect pulsed-field gel electrophoresis patterns of Campylobacter jejuni. ]. Clin. Microbiol. 45:386-391. Boyd, E. F., and H. Briissow. 2002. Common themes among bacteriophage-encoded virulence factors and diversity among the bacteriophages involved. Trends Microbiol. 10521-529. Bruynoghe, R., and J. Maisin. 1921. Essais de thhapeutique au moyen du bacteriophage. C. R. Soc. Biol. 85:1120-1121. Bryner, J. H., A. E. Ritchie, G . D. Booth, and J. W. Foley. 1973. Lytic activity of vibrio phages on strains of Vibrio fetus isolated from man and animals. Appl. Microbiol. 26:404-409. Bryner, J. H., A. E. Ritchie, J. W. Foley, and D. T. Berman. 1970. Isolation and characterization of a bacteriophage for Vibriofetus. J. Virol. 6:94-99. Bryner, J. H., A. E. Ritchie, and J. W. Foley. 1982. Techniques for phage typing Campylobacterjejuni. p. 52-56. In D. G. Newell (ed.), Campylobacter: Epidemiology, Puthogenisis and Biochemistry, MTP Press, Lancaster, United Kingdom. Bull, J. J., B. R. Levin, T. Derouin, N. Walker, and C. A. Bloch. 2002. Dynamics of success and failure in phage and antibiotic therapy in experimental infections. BMC Microbiol. 2:35. Casjens, S. R. 2005. Comparative genomics and evolution of the tailed-bacteriophages. Cttrr. Opin. Microbiol. 8:45 1-458.
CHAPTER 38
Cheetham, B. F., and M. E. Katz. 1995. A role for bacteriophages in the evolution and transfer of bacterial virulence determinants. Mol. Microbiol. 18:201-208. Connerton, P. L., C. M. LOCCarrillo, C. Swift, E. Dillon, A. Scott, C. E. Rees, C. E. Dodd, J. Frost, and I. F. Connerton. 2004. Longitudinal study of Campylobacter jejuni bacteriophages and their hosts from broiler chickens. Appl. Environ. Microbiol. 70: 3877-38 83. Corry, J. E. L., and H. I. Atabay. 2001. Poultry as a source of Campylobacter and related organisms. J. Appl. Microbiol. 90: 96s-114s. Coward, C., A. J. Grant, C. Swift, J. Philp, R. Towler, M. Heydarian, J. A. Frost, and D. J. Maskell. 2006. Phase-variable surface structures are required for infection of Campylobacterjejuni by bacteriophages. Appl. Environ. Microbiol. 72:4638-4647. Delbruck, M. 1940. The growth of bacteriophage and lysis of the host. J. Gen. Physiol. 23:643-660. d’HCrelle, F. 1917. Sur un microbe invisible antagoniste des bac. dysentkriques. C. R. Acad. Sci. Paris 165:373-375. El-Shibiny, A., P. L. Connerton, and I. F. Connerton. 2007. Campylobacter succession in broiler chickens. Vet. Microbiol. 125: 323-332. El-Shibiny, A., P. L. Connerton, and I. F. Connerton. 2005. Enumeration and diversity of campylobacters and bacteriophages isolated during the rearing cycles of free-range and organic chickens. Appl. Environ. Microbiol. 71:1259-1266. Fearnhead, P., N. Smith, M. Barrigas, A. Fox, and N. French. 2005. Analysis of recombination in Campylobacter jejuni from MLST population data. J. Mol. Evol. 61:333-340. Firehammer, B. D., and M. Border. 1968. Isolation of temperate bacteriophages from Vibrio fetus. Am. J. Vet. Res. 29:22292235. Fletcher, R., and H. Bertschinger. 1964. A method of isolation of Vibrio coli from swine faecal material by selective filtration. Zentralbl. Veterinaeromed.B 11:169-174. Fletcher, R. D. 1968. Activity and morphology of Vibrio coli phage. Am. J. Vet. Res. 26:361-364. Fouts, D. E, E. F. Mongodin, R E. Mandrell, W. G. Miller, D. A. Rasko, J. Ravel, L. M. Brinkac, R. T. DeBoy, C. T. Parker, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, S. A. Sullivan, J. U. Shetty, M. A. Ayodeji, A. Shvartsbeyn, M. C. Schatz, J. H. Badger, C. M. Fraser, and K. E. Nelson. 2005. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol. 3:e15. Frost, J. A., J. M. Kramer, and S. A. Gillanders. 1999. Phage typing of Campylobacter jejuni and Campylobacter coli and its use as an adjunct to serotyping. Epidemiol. Infect. 123:47-55. Gaynor, E., S. Cawthraw, G. Manning, J. MacKichan, S. Falkow, and D. Newell. 2004. The genome-sequenced variant of Campylobacter jejuni NCTC 11168 and the original clonal clinical isolate differ markedly in colonization, gene expression, and virulence-associated phenotypes. J. Bacteriol. 186503-517. Gibson, J., C. Fitzgerald, and R. Owen. 1995. Comparison of PFGE, ribotyping and phage-typing in the epidemiological analysis of Campylobacter jejuni serotype HS2 infections. Epidemiol. Infect. 115:215-225. Goode, A., V. Allen, and P. Barrow. 2003. Reduction of experimental Salmonella and Campylobacter contamination of chicken skin by application of lytic bacteriophages. Appl. Environ. Microbiol. 69:s 032-5 036. Grajewski, B. A., J. W. Kusek, and H. M. Gelfand. 1985. Development of a bacteriophage typing system for Campylobacter jejuni and Campylobacter coli. J. Clin. Microbiol. 22:13-18. Greer, G. 2005. Bacteriophage control of food-borne bacteria. J. Food Prot. 68:1102-1111.
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
691
Hendrix, R. W., M. C. Smith, R. N. Burns, M. E. Ford, and G. F. Hatfull. 1999. Evolutionary relationships among diverse bacteriophages and prophages: all the world’s a phage. Proc. Natl. Acad. Sci. USA 96:2192-2197. Hopkins, K. L., M. Desai, J. A. Frost, J. Stanley, and J. M. Logan. 2004. Fluorescent amplified fragment length polymorphism genotyping of Campylobacter jejuni and Campylobacter coli strains and its relationship with host specificity, serotyping, and phage typing J. Clin. Microbiol. 42:229-35. Hudson, J., C. Billington, G. Carey-Smith, and G. Greening. 2005. Bacteriophages as biocontrol agents in food. J. Food Prot. 68: 426-437. Huff, W. E., G. R. Huff, N. C. Rath, J. M. Balog, and A. M. Donoghue. 2002a. Prevention of Escherichia coli infection in broiler chickens with a bacteriophage aerosol spray. Poult. Sci. 81: 1486-9 1. Huff, W. E., G. R Huff, N. C. Rath, J. M. Balog, and A. M. Donoghue. 2003. Bacteriophage treatment of a severe Escherichia coli respiratory infection in broiler chickens. Avian Dis. 47: 1399-1405. Huff, W. E., G. R Huff, N. C. Rath, J. M. Balog, H. Xie, P. A. Moore, Jr., and A. M. Donoghue. 2002b. Prevention of Escherichia coli respiratory infection in broiler chickens with bacteriophage (SPR02). Poult. Sci. 81:437-441. James, C., S. J. James, N. Hannay, G. Purnell, C. Barbedo-Pinto, H. Yaman, M. Araujo, M. L. Gonzalez, J. Calvo, M. Howell, and J. E. Corry. 2007. Decontamination of poultry carcasses using steam or hot water in combination with rapid cooling, chilling or freezing of carcass surfaces. Int. J. Food Microbiol. 114:195-203. Joerger, R. D. 2003. Alternatives to antibiotics: bacteriocins, antimicrobial peptides and bacteriophages. Poult. Sci. 82:640-647. Jones, M. A., K. L. Marston, C. Woodall, D. Maskell, D. Linton, B. Wren, and P. A. Barrow. 2004. Adaptation of Campylobacter jejuni NCTC11168 to high level colonization of the avian gastro-intestinal tract; capsule production is required for colonization. Infect. Immun. 72:3769-76. Jorgensen, F., R. Bailey, S. Williams, P. Henderson, D. Wareing, F. Bolton, J. Frost, L. Ward, and T. Humphrey. 2002. Prevalence and numbers of Salmonella and Campylobacter spp. On raw, whole chickens in relation to sampling methods. Int. J. Food Microbiol. 76:151-164. Kennedy, J., and G. Bitton. 1987. Bacteriophages in foods, p. 289316. In G. Bitton, G. Gerba, and S. Goyal, (ed.) Phage Ecology. John Wiley and Sons, New York. Khakhria, R., and H. Lior. 1992. Extended phage-typing scheme for Campylobacter jejuni and Campylobacter coli.Epidemiol. Infect. 108:403-14. Koo, J., D. L. Marshall, and A. DePaola. 2001. Antacid increases survival of Vibrio vulnificus and Vibrio vulnificus phage in a gastrointestinal model. Appl. Environ. Microbiol. 67:2895-2902. Kudva, I. T., S. Jelacic, P. I. Tarr, P. Youderian, and C. J. Hovde. 1999. Biocontrol of Escherichia coli 0 1 5 7 with 0157-specific bacteriophages. Appl. Environ. Microbiol. 65:3767-3773. Kutter, E., and A. Sulakvelidze (ed.). 2005. Bacteriophages:Biology and Applications. CRC Press, Boca Raton, FL. Lang, L. H. 2006. FDA approves use of bacteriophages to be added to meat and poultry products. Gastroenterology 131:1370. Lawrence, J. G., G. F. Hatfull, and R. W. Hendrix. 2002. Imbroglios of viral taxonomy: genetic exchange and failings of phenetic approaches. J. Bacteriol. 184:4891-4905. Leverentz, B., W. Conway, Z. Alavidze, W. Janisiewicz, Y. Fuchs, M. Camp, E. Chighkadze, and A. Sulakvelidze. 2001. Examination of bacteriophage as a biocontrol method for Salmonella on fresh-cut fruit: a model study. J. Food Prot. 64:1116-1121.
692
CONNERTON ET AL.
Levin, B., and J. Bull. 1996. Phage therapy revisited: the population biology of a bacterial infection and its treatment with bacteriophage and antibiotics. Am. Nut. 147:88 1-898. LOCCarrillo, C. M., R. J. Atterbury, A. El-Shibiny, P. L. Connerton, E. Dillon, A. Scott, and I. F. Connerton. 2005. Bacteriophage therapy to reduce Cumpylobucter jejuni colonization of broiler chickens. Appl. Environ. Microbiol. 71:6554-6563. LOCCarrillo, C. M., P. L. Connerton, T. Pearson, and I. F. Connerton. 2007. Free-range layer chickens as a source of Campylobucter bacteriophage. Antonie Van Leeuwenhoek 92:275-284. Logan, S., J. Kelly, P. Thibault, C. Ewing, and P. Guerry. 2002. Structural heterogeneity of carbohydrate modifications affects serospecificity of Campylobacter flagellins. Mol. Microbiol. 2: 587-597. McCrea, B., K. Tonooka, C. Van Worth, E. Atwill, and J. Schrader. 2006. Colonizing capability of Campylobacter jejuni genotypes from low-prevalence avian species in broiler chickens. J. Food h o t . 69~417-420. Murphy, C., C. Carroll, and K. Jordan. 2006. Environmental survival mechanisms of the foodborne pathogen Cumpylobucter jejuni. J. Appl. Microbiol. 100:623-632. Nachamkin, I., X. H. Yang, and N. J. Stern. 1993. Role of Cumpylobucter jejuni flagella as colonization factors for three-dayold chicks: analysis with flagellar mutants. Appl. Environ. Microbiol. 59:1269-1273. Nelson, D. 2004. Phage taxonomy: we agree to disagree. J. Bacteriol. 186:7029-7031. Newell, D., and C. Fearnley. 2003. Sources of Campylobacter colonisation in broiler chickens. Appl. Environ. Microbiol. 69: 4343-4351. Newell, D., and J. Wagenaar. 2000. Poultry infections and their control at farm level, p. 497-509 In I. Nachamkin, and M. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, DC. Parker, C. T., B. Quinones, W. G . Miller, S. T. Horn, and R. E. Mandrel]. 2006. Comparative genomic analysis of Cumpylobucter jejuni strains reveals diversity due to genomic elements similar to those present in C. jejuni strain RM1221. J. Clin. Microbiol. 44:4125-4135. Parkhill, J., B. Wren, K. Mungall, J. Ketley, C. Churcher, D. Basham, T. Chillingworth, R Davies, T. Feltwell, s. Holroyd, K. Jagels, A. Karlyshev, S. Moule, M. Pallen, C. Penn, M. Quail, M. Rajandream, K. Rutherford, A. van Vliet, S. Whitehead, and B. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobucter jejuni reveals hypervariable sequences. Nuture 403:665-668. Payne, R., and V. Jansen. 2001. Understanding bacteriophage therapy as a density-dependent kinetic process. J. Theor. Biol. 208: 37-48. Payne, R. J., and V. A. Jansen. 2002. Evidence for a phage proliferation threshold? J. Virol. 76:13123. Payne, R. J., and V. A. Jansen. 2003. Pharmacokinetic principles of bacteriophage therapy. Clin. Phumzucokinet. 42:3 15-325. Petersen, L., and A. Wedderkopp. 2001. Evidence that certain clones of Campylobacter jejuni persist during successive broiler flock rotations. Appl. Environ. Microbiol. 67:2739-2745. Proux, C., D. van Sinderen, J. Suarez, P. Garcia, V. Ladero, G. F. Fitzgerald, F. Desiere, and H. Brussow. 2002. The dilemma of phage taxonomy illustrated by comparative genomics of Sfi21like Siphoviridae in lactic acid bacteria. J. Bucteriol. 184:60266036. Rabinovitch, A., I. Aviram, and A. Zaritsky. 2003. Bacterial debris-an ecological mechanism for coexistence of bacteria and their viruses. J. Theor. Biol. 224:377-383. Rees, C. E., and C. E. Dodd. 2006. Phage for rapid detection and control of bacterial pathogens in food. Adv. Appl. Microbiol. 59: 159-186.
Reynaud, A., L. Cloastre, J. Bernard, H. Laveran, H. W. Ackermann, D. Licois, and B. Joly. 1992. Characteristics and diffusion in the rabbit of a phage for Escherichia coli 0103. Attempts to use this phage for therapy. Vet. Microbiol. 30:203-212. Ritchie, A. E., J. H. Bryner, and J. W. Foley. 1983. Role of DNA bacteriophage in Cumpylobucter auto-agglutination. J. Med. Microbiol. 16:333-340. Rohwer, F., and R. Edwards. 2002. The Phage Proteomic Tree: a genome-based taxonomy for phage. J. Bucteriol. 184:4529-35. Rosenquist, H., N. L. Nielsen, H. M. Sommer, B. Norrung, and B. B. Christensen. 2003. Quantitative risk assessment of human campylobacteriosis associated with thermophilic Campylobacter species in chickens. Int. J. Food Microbiol. 83:87-103. Rudi, K., H. K. Hoidal, T. Katla, B. K. Johansen, J. Nordal, and K. S. Jakobsen. 2004. Direct real-time PCR quantification of Campylobacter jejuni in chicken fecal and cecal samples by integrated cell concentration and DNA purification. Appl. Environ. Microbiol. 70:790-797. Sails, A. D., D. R. A. Wareing, F. J. Bolton, A. J. Fox, and A. Curry. 1998. Characterisation of 16 Cumpylobucter jejuni and C. coli typing bacteriophages. J. Med. Microbiol. 47:123-128. Salama, S., F. Bolton, and D. Hutchinson. 1990. Application of a new phage typing scheme to campylobacters isolated during outbreaks. Epidemiol. Infect. 104:405-411. Salama, S., F. J. Bolton, and D. N. Hutchinson. 1989. Improved method for the isolation of Campylobucter jejuni and Cumpylobacter coli bacteriophages. Lett. Appl. Microbiol. 85-7. Schouls, L., S. Reulen, B. Duim, J. Wagenaar, R. Willems, K. Dingle, F. Colles, and J. van Embden. 2003. Comparative genotyping of Cumpylobacter jejuni by amplified fragment length polymorphism, multilocus sequence typing and short repeat sequencing: strain diversity, host range and recombination.], Clin. Microbiol. 41:l.S-26. Scott, A. E., A. R. Timms, P. L. Connerton, A. El-Shibiny, and I. F. Connerton. 2007a. Bacteriophage influence Cumpylobacter jejuni types populating broiler chickens. Environ. Microbiol. 9: 2341-2353. Scott, A. E., A. R. Timms, P. L. Connerton, C. LOCCarrillo, K. A. Radzum, and I. F. Connerton. 2007b. Genome dynamics of Campylobacter jejuni in response to bacteriophage predation. PLoS Pathog. 3x119. Shreeve, J. E., M. Toszeghy, A. Ridley, and D. G. Newell. 2002. The carry-over of Campylobacter isolates between sequential poultry flocks. Avian Dis. 46:378-385. Sklar, I. B., and R. D. Joerger. 2001. Attempts to utilize bacteriophage to combat Salmonella entericu serovar Enteritidis infection in chickens. J. Food Sufi 21:15-30. Smith, H. W., and M. B. Huggins. 1982. Successful treatment of experimental Escherichia coli infections in mice using phage: its general superiority over antibiotics. J. Gen. Microbiol. 128:307318. Smith, H. W., M. B. Huggins, and K. M. Shaw. 1987a. The control of experimental Escherichia coli diarrhoea in calves by means of bacteri0phages.J. Gen. Microbiol. 133:llll-1126. Smith, H. W., M. B. Huggins, and K. M. Shaw. 1987b. Factors influencing the survival and multiplication of bacteriophages in calves and in their environment. J. Gen. Microbiol. 133:11271135. Stern, N. J., K. L. Hiett, G. A. Alfredsson, K. G. Kristinsson, J. Reiersen, H. Hardardottir, H. Briem, E. Gunnarsson, F. Georgsson, R. Lowman, E. Berndtson, A. M Lammerding., G. M. Paoli, and M. T. Musgrove. 2003. Cumpylobucter spp. in Icelandic poultry operations and human disease. Epidemiol. Infect. 130: 23-32. Stone, R. 2002. Stalin’s forgotten cure. Science 298:728-73 1.
CHAPTER 38
Sulakvelidze, A., Z. Alavidze, and J. G. Morris, Jr. 2001. Bacteriophage therapy. Antimicrob. Agents Chemother. 45:649-659. Sulakvelidze, A., and P. Barrow. 2005. Phage therapy in animals and agribusiness, p. 335-380. In E. Kutter, and A. Sulakvelidze (eds.), Bacteriophages: Biology and Applications. CRC Press, Boca Raton, FL. Summers, W. C. 2001. Bacteriophage therapy. Annu. Rev. Microbiol. 55:437-45 1. Thibault, P., S. M. Logan, J. F. Kelly, J-R Brisson, C. P. Ewing, T. J. Trust, and P. Guerry. 2001. Identification of the carbohydrate moieties and glycosylation motifs in Campylobacter jejuni flagellin. ]. Biol. Chem. 276:34862-34870. Tsuei, A. C., G. V. Carey-Smith, J. A. Hudson, C. Billington, and J. A. Heinemann. 2007. Prevalence and numbers of coliphages and Campylobacter jejuni bacteriophages in New Zealand foods. Int. J. Food Microbiol. 116:121-5. Twort, F. W. 1915. An investigation on the nature of ultramicroscopic viruses. Lancet ii:1241-1243. van den Ende, P. 1973. Predator prey interactions in continuous culture. Science 181:562-564.
BACTERIOPHAGE THERAPY AND CAMPYLOBACTER
693
Wagenaar, J., M. van Bergen, M. Mueller, T. Wassenaar, and R. Carlton. 2005. Phage therapy reduces Campylobacter jejuni colonization in broilers. Vet. Microbiol. 109275-283. Wassenaar, T. M., B. A. van der Zeijst, R. Ayling, and D. G. Newell. 1993. Colonization of chicks by motility mutants of Campylobacter jejuni demonstrates the importance of Flagellin A expression. J. Gen. Microbiol. 139:1171-1175. Weld, R J., C. Butts, and J. A Heinemann. 2004. Models of phage growth and their applicability to phage therapy. ]. Theor. Biol. 227: 1-1 1. Wiggins, B. A., and M. Alexander. 1985. Minimum bacterial density for bacteriophage replication: implications for significance of bacteriophages in natural ecosystems. Appl. Environ. Microbiol. 49:19-23. Zhilenkov, E. L., V. M. Popova, D. V. Popov, L. Y. Zavalsky, E. A. Svetoch, N. J. Stern, and B. S. Seal. 2006. The ability of flagellum-specific Proteus vulgaris bacteriophage PV22 to interact with Campylobacter jejuni flagella in culture. Virol.]. 3 5 0 . Zhilenkov, E. L., V. M. Popova, M. E. Zhilenkov, E. A. Svetoch, N. J. Stern, and B. S. Seal. 2004. Isolation and preliminary characterization of bacteriophage that infect Campylobacter jejuni. ASM Conf. New Phage Biol. Key Biscayne, FL.
INDEX
Index Terms
Links
A ABC transporters Abdominal pain, Campylobacter enteritis
42
451
583
594
123
126
213
401
403
423
107
127
101
Abortion farm animals
human Acetyl phosphate
585
N-Acetylglucosamine-6-phosphate2-epimerase
488
Acetyltransferase
494
538
Acid stress, C. jejuni
572
576
AcrA protein
450
Actin filaments, Campylobacter invasion of intestinal epithelium
300
Acute-phase response
342
320
Adherence, bacterial C. jejuni to intestinal epithelial cells
298
capsular polysaccharide in
512
N-linked protein glycosylation in
456
Adhesins
91
317
319
298
317
513 ADP-L,D-Hep
492
Aer receptors
356
Aerobic stress C. jejuni
572
long-term aerobic adaptation
574
short-term
574
short-term aerobic tolerance
574
AFLP (amplified fragment length polymorphism)
574
192
196
Arcobacter
19
203
C. concisus
204
C. fetus
203
C. helveticus
204
216
222
This page has been reformatted by Knovel to provide easier navigation.
319
Index Terms
Links
AFLP (amplified fragment length polymorphism) (Cont.) C. jejuni
251
C. lari
203
C. upsaliensis
203
differentiation of Campylobacter species
16
epidemiologic typing
232
fluorescent
197
national subtyping network
283
non-jejuni, non-coli Campylobacter
203
Agar dilution method, antimicrobial susceptibility testing
646
Age distribution, Campylobacter infections
166
170
ahpC gene
340
572
583
603
605
617
583
598
601
604
246
249
602
604
AhpC protein
599
AIDP (acute inflammatory demyelinating polyneuropathy) Alcian Blue, stabilization of capsular polysaccharide
510
Alkaline phosphatase
615
Alkyl hydroxyperoxidase
573
599
617 Alkyl peroxides
598
Allelic profiles
29
Alpha 1-antitrypsin
343
AMAN (acute motor axonal neuropathy)
247
immune attack on nerve root axons
387
rabbit model
381
treatment
395
amiA gene
249
579
Amino acid biosynthesis involvement of Pgl proteins
87 458
Amino acid catabolism
47
Amino acid identities, between Campylobacter proteomes
76
Amino acid transport
42
583
involvement of Pgl proteins
458
Aminoglycoside acetyltransferase
268
647
Aminoglycoside adenyltransferase
268
647
Aminoglycoside phosphotransferase
268
647
Aminoglycoside resistance
267
647
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
AMLP1 element
79
Amoxicillin resistance
234
Amoxicillin-clavulanic acid, clinical indications
114
648
Amplified fragment length polymorphism, see AFLP Ampicillin, clinical indications
128
Ampicillin resistance
234
Amyloid A, serum
343
Amyloid P, serum
343
Anaplerotic reactions
236
648
44
Animal models, see also specific animals C. fetus infections
218
423
C. jejuni infections
367
435
colonization modes
369
defective for adaptive immunity or with altered flora
372
disease models
370
future goals
375
major uses and desirable characteristics
368
reagents and resources
369
specific-pathogen-free status
368
strain variations
369
C. jejuni vaccine development
373
435
GBS
254
256
ant(6)-Ia gene
648
ant(3’)-Ia gene
648
Antiganglioside antibodies
381
GBS
248
serum
252
Antigen [a]
381
252
409
Antigen sequence typing, C. jejuni
36
Antimicrobial agents, see also specific drugs C. fetus infections
128
C. hyointestinalis infections
133
C. lari infections
134
C. upsaliensis infections
131
Campylobacter enteritis
111
enteric helicobacters
140
Antimicrobial peptides/proteins, produced by neutrophils
340
This page has been reformatted by Knovel to provide easier navigation.
652
Index Terms
Links
Antimicrobial resistance
316
618
see also specific drugs association to antimicrobial use
651
bacteria within biofilms
581
consequences in human infections
656
in different food animal reservoirs
654
emergence
651
ecological observations
652
epidemiological field studies
652
experimental studies
651
fitness cost for bacteria
660
future aspects
661
interventions to reduce occurrence
659
mechanisms
263
plasmid-borne
659
447
76
source attribution studies
656
surveillance
182
transmission along food chain
655
transmission from food animals to humans
645
Antimicrobial resistance risk assessment
657
Antimicrobial susceptibility testing
230
clinical breakpoints
646
epidemiological cutoff values
646
methods
645
in vitro susceptibility profiles
232
658
232
aphC gene
618
Apoptosis, C. jejuni-induced
308
Appendicitis
104
Aquatic environment, survival of campylobacters
580
AraC protein
612
620
19
203
318
Arcobacter AFLP antimicrobial susceptibility
235
characteristics of genus
11
differentiation of species
19
isolation
17
17
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Arcobacter (Cont.) novel species
18
species of genus
17
taxonomy
5
Arcobacter butzleri AMLP1
17 81
antimicrobial susceptibility taxonomy
235 5
Arcobacter butzleri infections
125
Arcobacter cibarius
18
Arcobacter cryaerophilus
17
antimicrobial susceptibility Arcobacter cryaerophilus infections
130
137
235 125
Arcobacter halophilus
18
Arcobacter infections
137
137
characterization of isolates
138
clinical features
126
138
epidemiology
126
137
microbiology
125
137
pathogenesis
138
Arcobacter nitrofigilis
17
Arcobacter skirrowii
5
17
Arcobacter skirrowii infections
126
138
ArsR protein
612
620
Arthritis reactive burden of illness
154
after Campylobacter enteritis
109
vaccine-related
432
437
septic
108
127
Asparaginase
47
Aspartate transport
42
Aspartate/glutamate transaminase
47
Aspartate:ammonia lyase
47
astA gene
551
atpA gene
206
404
459
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
ATPase F-type
42
P-type
42
Autoagglutination (AAG phenotype)
477
Autoregulation, pgl genes
457
Average nucleotide identity
583
7
Azithromycin, clinical indications
111
Azithromycin resistance
265
B Bacillosamine
459
Bacillosamine phosphate
450
Bacteremia C. fetus
124
C. jejuni
164
C. jejuni subsp. doylei
135
C. lari
133
C. upsaliensis
128
in Campylobacter enteritis
105
enteric helicobacters
139
Bacteriocins
672
Bacterioferritin
598
Bacteriophage
65
Campylobacter-specific C. jejuni response to phage predation non-jejuni Campylobacter
126
131
600
681 689 81
phage characteristics
681
resistant bacteria
684
source
682
strains of Campylobacter that populate chickens
688
historical aspects
680
structure
680
taxonomy
680
Bacteriophage Mu
79
Bacteriophage receptor
514
Bacteriophage resistance
684
682
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Bacteriophage therapy advantages and disadvantages
682
elimination of Campylobacter from colonized flock
679
elimination of Campylobacter from poultry disinfection of meat
687
proof of efficacy
682
treatment of live chickens
684
public acceptability
688
regulatory issues
688
Bacteriophage typing
222
Bacteroides gracilis
4
Bacteroides ureolyticus
4
Barrett’s esophagus
136
Basal body
546
Beef
634
Beta-lactamases
234
Bile
272
Bile acids, defense against enteric pathogens Bile resistance
Bile stress, C. jejuni
232
681
20
137
548
269
648
334
336
344
271
316
336
576
615
619
572
576
514
581
Biofilms adaptation to environments encountered in vivo
581
antimicrobial resistance of bacteria within
581
autoagglutination
583
C. jejuni
580
formation
478
505
584 AI-2-mediated quorum sensing
585
conditions promoting
584
cyclic di-GMP
585
regulatory factors
584
stringent control
584
two-component regulatory systems
585
iron metabolism and
605
molecular themes underlying physiology
583
structure and function
581
survival in aquatic environments
580
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Biosecurity, poultry house
670
general measures
671
multispecies farming
671
thinning
672
Biotyping
232
Blood, innate immune defenses
334
Blood samples
228
Bottled water
636
Bovine venereal campylobacteriosis
401
Bovine vibriosis
123
340
Breast milk colostral antibodies
429
oligosaccharides
334
344
157
629
667
298
309
Broilers Broth microdilution method, antimicrobial susceptibility testing
646
Buerger’s disease
136
Burden of illness definition
151
measurement
151
campylobacteriosis
151
acute infections vs. postinfective complications
157
ascertainment of outcomes
154
disability weights
156
immunity and
155
industrialized countries
171
integrating with risk assessment and economics
157
perspectives and future research
160
survival pyramid
152
United States
167
Bursitis
108
Butter
633
C C4-dicarboxylate transport
44
CadF protein
91 407
This page has been reformatted by Knovel to provide easier navigation.
319
Index Terms
Links
Calcium ions Campylobacter invasion of intestinal epithelium
302
intracellular, toxin effects
307
Campylobacter characteristics of genus
11
chemotaxis pathway
351
chickens
157
colonization without disease
432
culture
229
differentiation of species
15
evolution
84
phage effects identification
177
180
688 230
isolation
4
12
enrichment culture
229
628
filtration
229
from food or water
628
metabolomics phenotypic characteristics phylogenic tree population studies
523 15 5
84
27
PulseNet
280
salt sensitivity
575
species of genus
30
637
12
species tree
5
taxonomy
3
227
thermophilic
230
403
thermotolerant
627
629
637
typing systems epidemiologic methods
232
molecular methods
232
phenotypic methods
232
whole-genome taxonomy Campylobacter bubulus Campylobacter canadensis
5 4 15
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Campylobacter coli, see also Non-jejuni Campylobacter antimicrobial susceptibility
232
characteristics of species
14
clonal complexes
36
comparative genomics
73
epidemiology
37
evolution
201
host species
200
isolation
229
MLST
29
O-linked flagellar glycosylation
530
PFGE
282
population studies
36
sequence types
36
taxonomy
36
204
4
transformation
562
VBNC state
577
Campylobacter coli infections
99
182
see also Enteritis, Campylobacter clinical features
127
diagnosis
227
extraintestinal
105
microbiology
125
Campylobacter concisus, see also Non-jejuni Campylobacter AFLP
204
antimicrobial susceptibility
236
characteristics of species
13
comparative genomics
73
PFGE
202
RAPD
204
taxonomy
5
Campylobacter concisus infections
135
clinical features
126
epidemiology
126
microbiology
125
Campylobacter cyaerophila
130
5
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Campylobacter curvus, see also Non-jejuni Campylobacter characteristics of species
13
comparative genomics
73
taxonomy
5
Campylobacter curvus infections
135
clinical features
127
microbiology
125
Campylobacter fetus
213
AFLP
203
cell envelope proteins
405
characteristics of species
216
222
126
214
29
37
204
218
222
12
epidemiology
403
genome plasticity
405
genome sequence
406
genome size
404
genomic islands
405
historic classification
213
host associations
214
isolation
124
lipopolysaccharide
407
metabolism
404
MLST
molecular identification
216
PCR methods
216
218
PFGE
202
222
phenotypic identification
215
404
population studies
37
RAPD
218
reptile strains
221
16S rRNA
222
sap genes
415
sequence types
222
37
serum-resistant strains
409
S-layer
409
sap genes
411
S-layer proteins
411
antigenic diversity in ovine immune responses
404
505
414
423
This page has been reformatted by Knovel to provide easier navigation.
216
Index Terms
Links
Campylobacter fetus (Cont.) antigenic variation
415
in ovine abortion
423
regulation of production
422
secretion
419
structure
404
subspecies differentiation justification
219
molecular
218
phenotypic
216
using animal models
218
surface carbohydrate structures taxonomy
407 4
213
401
typing molecular methods
222
phenotypic methods
222
Campylobacter fetus infections
123
animal models
410
animals
401
clinical features
124
diagnosis
214
epidemiology
124
microbiology
124
pathogenesis
401
treatment
128
Campylobacter fetus subsp. fetus
12
423
130
214
403
37
123
213
37
123
213
126
401 see also Non-jejuni Campylobacter antimicrobial susceptibility comparative genomics subspecies differentiation Campylobacter fetus subsp. venerealis
235 73 216 12 401
subspecies differentiation Campylobacter fetus subsp. venerealis biovar intermedius Campylobacter gracilis Campylobacter gracilis infections
216 216
220
5
12
135
This page has been reformatted by Knovel to provide easier navigation.
14
Index Terms
Links
Campylobacter helveticus
15
AFLP
204
MLST
29
Campylobacter hominis
128
37
204
130
132
141
see also Non-jejuni Campylobacter characteristics of species
14
comparative genomics
73
genomic reduction
88
Campylobacter hyointestinalis characteristics of species PFGE
12 203
taxonomy
5
Campylobacter hyointestinalis infections
131
clinical features
126
diagnosis
131
epidemiology
126
132
microbiology
125
131
treatment
132
Campylobacter hyointestinalis subsp. hyointestinalis
12
131
Campylobacter hyointestinalis subsp. lawsonii
12
131
Campylobacter infections animal models
367
ascertainment of outcomes
154
burden of illness
151
children
177
common-source outbreaks
172
developing countries
156
diagnosis
227
culture and isolation of campylobacters
229
direct detection in stool samples
228
identification of campylobacters
230
specimen considerations
227
disability weights
156
epidemiology
429
429
161
430
182
318
GBS, see Guillain-Barré syndrome immune response
155
immunocompromised patients
163
incidence
429 This page has been reformatted by Knovel to provide easier navigation.
432
Index Terms
Links
Campylobacter infections (Cont.) industrialized countries
156
innate immunity
333
molecular epidemiology
191
161
163
see also Molecular typing morbidity
152
protective factors
182
serology
232
age-related
429
sporadic
174
surveillance pyramid
152
United States
164
Campylobacter insulaenigrae characteristics of species MLST
430
88
167
141
15 204
Campylobacter jejuni acid stress
572
576
aerobic stress
572
574
AFLP
251
amino acid catabolism
47
amino acid transport
42
anaplerotic reactions
44
antigen sequence typing
36
antimicrobial susceptibility
232
bile stress
572
biofilms
580
capsular polysaccharide
66
central carbon metabolism
44
characteristics of species
14
chickens
264
clonal complexes cold stress
576
483
505
298
580
32 571
colonization
41
comparative genomics
63
comparative phylogenomics
68
competence
49
563
electron transport
48
epidemiology
36
50
This page has been reformatted by Knovel to provide easier navigation.
430
Index Terms
Links
Campylobacter jejuni (Cont.) evolution
201
flagellar genes
323
flagellar type III secretion system
315
550
GBS-related gene characteristics
251
gene-specific variation
251
rarity of disease
254
strains that cannot trigger GBS
255
whole-genome polymorphisms
250
gene gazing
64
gene regulation
611
gene sequencing
65
genome diversity
63
67
genome sequences
41
63
genomic rearrangements in response to phage
689
genomics to phenomics
66
gluconeogenesis
44
heat stress
571
homoplasy ratio
560
host species
199
isolation
229
lipooligosaccharide
560
66
251
381
483
32
38
204
572
576
617
N-linked protein glycosylation
66
448
537
O-linked flagellar glycosylation
66
524
organic acid transport
44
507 lipopolysaccharide
339
microaerophily
49
MLST
29
nitrogen assimilation
47
nitrosative stress
oxidative stress
572
pathogenesis
64
pathogenicity islands
65
PFGE
195
phase variation
65
physiology and metabolism
41
280
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Campylobacter jejuni (Cont.) plasmids
64
plasticity regions
621
polysaccharide capsule
439
polysaccharides
514
population studies
32
pseudogenes
65
pVir
565
RAPD
204
restriction/modification system
64
sequence types
32
sigma factors
559
36
612
solute transport
41
starvation
572
574
stationary phase stress
572
574
stress response
515
571
surface glycostructures taxonomy
66 4
under oxygen limitation
54
VBNC state
577
virulence factors/virulence phenotypes
316
Campylobacter jejuni infections
99
see also Enteritis, Campylobacter adherence to intestinal epithelium
298
317
animal models
367
435
clinical features
127
130
diagnosis
227
extraintestinal
105
host cell entry
297
host damage
307
innate immunity
333
invasion of intestinal epithelium
299
microbiology
125
pathogenesis
307
translocation across intestinal mucosa
305
Campylobacter jejuni subsp. doylei
7
comparative genomics
73
metabolism
85
319
315
14
This page has been reformatted by Knovel to provide easier navigation.
34
Index Terms
Links
Campylobacter jejuni subsp. doylei infections
134
Campylobacter jejuni subsp. jejuni Campylobacter lanienae
7
14
13
141
34
Campylobacter lari, see also Non-jejuni Campylobacter AFLP
203
antimicrobial susceptibility
235
characteristics of species
15
comparative genomics
73
genomic reduction
87
MLST
29
taxonomy
37
204
5
UPTC variants Campylobacter lari infections
133 133
clinical features
126
diagnosis
133
epidemiology
126
microbiology
125
treatment
134
Campylobacter mucosalis Campylobacter mucosalis infections
5
133
133
13
125
Campylobacter nitrofigilis
5
Campylobacter rectus
5
14
125
127
12
14
125
135
4
13
125
137
Campylobacter sputorum subsp. bubulus
4
13
Campylobacter sputorum subsp. sputorum
4
13
Campylobacter rectus infections Campylobacter showae Campylobacter showae infections Campylobacter sputorum Campylobacter sputorum infections
135
135
203
Campylobacter upsaliensis, see also Non-jejuni Campylobacter AFLP
203
antimicrobial susceptibility
235
characteristics of species
15
comparative genomics
73
methyltransferases
90
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Campylobacter upsaliensis, see also Non-jejuni (Cont.) MLST
29
PFGE
203
taxonomy
37
204
5
Campylobacter upsaliensis infections
128
characterization of isolates
130
clinical features
126
130
detection and Cape Town protocol
128
205
epidemiology
126
129
microbiology
125
128
pathogenesis
131
treatment of
131
Campylobacter vaccine bovine venereal campylobacteriosis
402
development
429
animal models human
373
435
429
517
candidate selection process
433
epidemiological issues and strain diversity
431
killed whole cell vaccines
437
517
live attenuated vaccines
437
517
outer membrane proteins and secreted proteins
433
safety concerns
432
517
subunit vaccines
439
517
target populations
430
veterinary
517
Campylobacteraceae, characteristics of family Campylobacter-like organisms “Candidatus Arcobacter sulfidicus”
11 4
108
135
514
18
capA gene
320
CapA protein
319
321
Cape Town protocol
128
205
Capsular polysaccharide
298
in adhesion
512
bacterium-host interaction
516
biochemistry
507
biological role
505
512
This page has been reformatted by Knovel to provide easier navigation.
139
Index Terms
Links
Capsular polysaccharide (Cont.) biosynthesis
491
influence of growth conditions C. jejuni
506
515 66
483
comparative genomics
509
detached
514
discovery of CPS production in campylobacters
507
evidence from genome sequencing of C. jejuni
508
gastroenteritis and
516
genetics
506
group II
506
group III
506
immunogenicity
513
in vitro models of infection
512
phase variation
515
phospholipid anchor
510
512
phosphoramidate modification
510
516
sorbofuranose modification
510
stabilization with Alcian Blue
510
structural analyses
510
subunit vaccines
439
types
506
Carbapenems, clinical indications
128
Carbon dioxide, effect on transformation
562
CARMA project
157
Cary-Blair medium, modified
228
Case-control method
627
sporadic infections
505
515
175
Caspases
309
cat genes
649
Catalase
573
599
602
334
338
344
C. fetus infections
214
401
C. jejuni infections
199
drug-resistant Campylobacter
657
617 Cathelicidins Cattle
This page has been reformatted by Knovel to provide easier navigation.
604
Index Terms
Links
Caveolae, Campylobacter invasion of intestinal epithelium
303
cbrR gene
336
572
CbrR protein
577
615
ccoNOPQ genes
87
cdt genes
90
Cefoperazone resistance
270
Cell wall, VBNC state
578
Cellulitis
404
Central carbon metabolism
44
Central genotypes
32 127
Cephalosporin(s), clinical indications
128
Cephalosporin resistance
234
236
Cet proteins
356
548
ceu genes
595 594
ceuBCDE genes
617
ceuC gene
594
ceuD gene
594
ceuE gene
231
Ceu proteins
612
269
648
308
Central nervous system infections
ceuB gene
585
594
603
604
CeuBCD protein
594
CeuE protein
596
602
cfrA gene
595
603
CfrA protein
594
602
cft gene
598
603
Cft protein
598
600
602
cgb gene
572
576
619
Cgb protein
576
618
Cgp proteins
448
452
493
cgtA gene
488
491
494
cgtB gene
488
494
CgtA protein
491
493
497
CgtB protein
491
494
497
605
cgt genes
Cgt proteins
che genes
355 This page has been reformatted by Knovel to provide easier navigation.
617
Index Terms
Links
Che proteins CheA protein
351
355
357
548
583 CheB protein
359
548
CheR protein
359
548
CheV protein
358
548
Chew protein
351
355
358
548
CheY protein
351
355
357
360
580
548 Che signaling cascade
546
548
Cheese
633
Chemoattractants
351
Chemokines, Campylobacter-induced
308
342
351
546
Chemosensory pathway, see Chemotaxis pathway Chemotaxis pathway adaptation
359
Campylobacter adherence to intestinal epithelium
298
CheA histidine kinase
357
CheR protein
359
Chew and CheV proteins
358
CheY protein
360
E. coli
352
genome content of components
355
group A receptors
355
groups B, C, and Aer receptors
356
model
363
organization
355
receptor specificity
357
sensory receptor complex
357
variations
354
Chest wall abscess
108
Chickens, see also Poultry C. jejuni
199
264
298
Campylobacter
157
177
180
Campylobacter-specific phages
682
colonization
513
role of N-linked protein glycosylation
639
457
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Chickens, see also Poultry (Cont.) drug-resistant Campylobacter
654
vaccination
673
Children, Campylobacter infections
100
Chloramphenicol, clinical indications
114
Chloramphenicol acetyltransferase
650
Chloramphenicol resistance
76
food-animal reservoirs
655
mechanisms
649
Chlorine disinfection
581
Cholangitis
127
Cholecystitis
106
Chorioamnionitis
127
ChpA protein
595
177
429
236
270
637
127
336
chu genes chuA gene
596
604
chuABCD genes
603
617
chuB gene
604
chuBCD genes
596
Chu proteins ChuA protein
594
ChuABCD protein
602
ChuZ protein
594
596
329
434
91
309
336
407
Cia proteins CiaB protein
recognition and export by type III secretion system
598
320
326
323
Ciprofloxacin, clinical indications
111
Ciprofloxacin resistance
183
233
235
263
561
650
652
656
85
406
CIRCE element
620
Citrate synthase
87
Citric acid cycle
44
with characteristics of anaerobes Cj1496c protein
46 453
Cja proteins CjaA protein
43
CjaC protein
452
458
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
CJIE3 element
82
CJIE4 element
82
Clarithromycin, clinical indications
111
Clarithromycin resistance
265
CLIE1 element
82
Clindamycin resistance
233
Clinical breakpoints
646
Clonal complexes
29
C. coli
36
C. jejuni
32
Clonal population
31
ClonalFrame algorithm
34
clp genes
235
198
560
572
cme genes cmeABC genes
619
cmeB gene
650
cmeDEF genes
565
cmeR gene
572
612
619
263
270
336
619
651
Cme proteins CmeABC efflux pump
CmeDEF efflux pump
270
CmeR protein
271
Cmg proteins
77
CMLP1 element
79
615
CMP sugars
475
CMP-Kdo synthetase
506
CMP-Leg5Ac7Ac
532
CMP-Leg5Am7Ac
533
CMP-Leg5AmNMe7Aq
533
CMP-NeuSAc
493
CMP-Neu5Ac synthase
488
CMP-Pse5Ac7AC
526
533
CMP-Pse5Ac7Am
530
532
CMP-PseAm
526
530
619
525
525
533
see also CMP-LegSAm7Ac Cold stress, C. jejuni
571
Colitis
104 This page has been reformatted by Knovel to provide easier navigation.
576
Index Terms
Links
Collectins
334
343
Colonization biofilm formation C. jejuni
581 41
capsular polysaccharide in
513
without disease
432
poultry
513
bacteriophage therapy
679
eliminating Campylobacter from colonized flock
674
epidemiology of flock colonization
668
lag phase
668
phage effects
688
prevalence rates between countries
668
49
629
639
679
prevention of introduction of Campylobacter into flock
670
risk factors
669
role of N-linked protein glycosylation
457
seasonal variation
669
sources
669
transcriptional profile of iron metabolism genes
604
Colonization models
369
Colonization potential
668
Colony morphology
230
672
com genes comABC genes
572
comB3 gene
566
Common-source outbreaks
172
evolution and
201
industrialized countries
173
molecular typing studies
201
United States
172
Comparative genomics C. jejuni
7
35
63
capsular polysaccharide
509
DNA microarray analysis
66
non-jejuni Campylobacter
73
Comparative phylogenomics C. jejuni
627
67 68
This page has been reformatted by Knovel to provide easier navigation.
667
Index Terms
Links
Competence DNA discrimination
562
evolution
559
genetics
563
molecular biology
563
natural
559
Competitive exclusion, prevention of colonization of poultry Complement inhibitors
672 343
Complement system complement factor C3
410
complement-mediated nerve injury in GBS
394
innate immune defenses
334
resistance of C. fetus
410
342
Complex I, see NDH-1 complex Conjugate vaccines
439
Contingency (phase-variable) genes, non-jejuni Campylobacter
78
CorA protein
593
Cost of illness methodology
152
cps genes
509
C-reactive protein
342
157
CRISPR elements HRM analysis non-jejuni Campylobacter Cross-contamination, during food preparation
194 82 180
638
Cross-reactive antibodies, to lipooligosaccharide and gangliosides
253
Crp protein
612
617
Csp proteins
323
csrA gene
620
CsrA protein
612
620
cst-II/III gene
251
394
Cst-II/III protein
493
497
cts genes
563
566
488
This page has been reformatted by Knovel to provide easier navigation.
494
Index Terms
Links
Cts proteins CtsD protein
560
563
CtsE protein
560
563
CtsF protein
560
563
CtsG protein
560
CtsP protein
560
CtsT protein
560
CtsW protein
565
CtsX protein
564
CURIE2 element Cutting boards
82 180
Cyanide-insensitive oxidase
53
Cyanide-sensitive oxidase
53
Cyclic di-GMP
563
638
585
cydAB operon 53
87
Cysteine transport
43
cytC gene
66
Cytochrome(s)
50
Cytochrome c
52
Cytochrome c nitrite reductase
56
Cytochrome c oxidase
87
Cytochrome c peroxidase
53
Cytochrome cp-oxidase
53
Cytokines, Campylobacter-induced Cytolethal distending toxin vaccine target
458
65
406
308
318
342
90
131
307
434
Cytoskeleton, Campylobacter invasion of intestinal epithelium Cytotoxin
299
320
307
D Dairy cows
633
Dairy products
633
DALY (disability-adjusted life-year)
151
Dcc proteins
614
dccR gene
612
dcu genes
44
572
574
Defensins
334
338
344
This page has been reformatted by Knovel to provide easier navigation.
406
Index Terms
Links
Denaturing gradient gel electrophoresis
194
Dendritic cells
308
318
334
342
612
619
Developing countries, Campylobacter infections
100
156
161
dfr genes
270
649
651
2,4-Diacetamide-bacillosamine biosynthetic pathway
530
Diarrheal disease
127
innate immune defenses DeoR protein
344
Developed countries, see Industrialized countries
429
see also Enteritis, Campylobacter Arcobacter
138
C. fetus
124
C. jejuni subsp. doylei
135
C. lari
133
C. upsaliensis
128
enteric helicobacters
139
hydrogen-requiring campylobacters
136
toxins and
307
Dicarboxylate/amino acid:cation symporter family
403
130
44
Diffusion method, antimicrobial susceptibility testing
645
Dihydrofolate reductase
270
Dihydroorotase
127
651
84
Dihydropteroate synthase
651
Dilution method, antimicrobial susceptibility testing
645
Disability weights
156
Dish washing
639
Disinfecting agents
638
dmhA gene
509
dmsA gene
66
DMSO reductase
56
65
DNA methylation
562
transformation, see Transformation DNA gyrase
263
DNA microarray
194
comparative genomics
650
66
species-specific
231
DNA uptake sequence
562
This page has been reformatted by Knovel to provide easier navigation.
87
430
Index Terms
Links
dna genes dnaJ gene
572
dnaK gene
572
620
DnaJ protein
572
615
Doc proteins
548
Domestic animals
126
dprA gene
566
DprA protein
560
566
dps gene
340
572
Dps protein
598
600
Drinking water
635
Drug efflux pumps
263
269
336
270
336
650
406
multidrug pumps Dry Spot Campylobacter Test Kit
574
231
E Eggs, Campylobacter in food supply
630
Electron acceptors
50
86
alternative
50
54
fumarate
55
hydrogen peroxide
53
nitrate and nitrite
55
S- or N-oxides
56
Electron donors
50
hydrogen and formate
51
organic acids
52
sulfite
52
Electron transport
48
C. fetus
406
iron in
591
organization and assembly of electron transport chain
50
oxygen-dependent respiration
53
under oxygen limitation
54
Electroporation
406
50
562
Embden-Meyerhof-Parnas pathway
44
85
Empyema
108
127
Endocarditis
127
Endoscopy, Campylobacter enteritis
104
110
This page has been reformatted by Knovel to provide easier navigation.
86
598
Index Terms
Links
Endosomes, intracellular movement of Campylobacter
307
Endotoxin
483
Enrichment culture
229
628
Enrofloxacin
651
653
Enrofloxacin resistance
263
Enteritis, Campylobacter
99
abdominal pain
101
bacteremia
105
capsular polysaccharide and
516
children
100
clinical pathology
99
developing countries
100
diagnosis
110
diarrheal stage
101
early infections
316
extraintestinal infections
105
hematology and biochemistry
110
hepatobiliary infections
105
immune response
126
227
99
immunodeficient patient
103
incubation period
101
infective dose
101
intestinal complications
103
late infection
318
late-onset complications
109
microbiology
110
model
315
morbidity
102
neonatal infections
103
onset and prodrome
101
pathogenesis
315
rashes
103
recovery stage
101
renal and urinary tract disease
107
treatment
111
Enterobactin
594
Enterobactin uptake permease
604
Enterotoxin
307
316
602
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Entner-Doudoroff pathway
85
406
Epidemiological cutoff values
646
ERIC-PCR
194
erm genes
265
Erythema nodosum
103
Erythromycin, clinical indications
111
128
Erythromycin resistance
111
183
233
265
650
656
Escherichiu coli, chemotaxis pathway
352
Evolution Campylobacter, phage effects
688
competence
559
flagella
327
outbreaks and
201
type III secretion system
327
exb genes
597
exbB gene
603
exbB1 gene
618
exbB2 gene
618
exbD gene
603
Exb proteins ExbB protein
594
602
604
ExbD protein
594
602
604
305
308
Extracellular-signal-regulatedkinase, Campylobacter infections
F Farm animals, sporadic infections and contact with animals
181
Farm environment
669
fdh genes
51
fdxA gene
572
FdxA protein
600
Fecal sample
227
feo genes
593
574
This page has been reformatted by Knovel to provide easier navigation.
235
Index Terms
Links
Feo proteins FeoA protein
592
FeoB protein
458
FeoC protein
592
592
594
601
Ferredoxin
574
Ferret model, C. jejuni infections
436
Ferrichrome
594
597
Ferritin
592
598
600
602
Fibronectin receptors
298
309
318
320
Filtration method, Campylobacter isolation
229
fla genes
323 323
328
459
550
616
flaA gene
283
600
550 regulation by sigma
552
fiaB gene
459
fiaC gene
584
fla typing
192
fla-RFLP
192
fla-SVR sequencing
192
recombination
561
222
Fla proteins FlaA protein
317
326
546
550
552
583
glycosylation
471
subunit vaccine
439
517
317
326
546
548
583
FlaB protein
583 glycosylation FlaC protein
471 326
vaccine target FlaG protein
434 549
Flagella assembly in campylobacters
545
comparisons to other bacteria
554
flagellin glycosylation
555
polar assembly
554
regulation of flagellar number
554
Campylobacter adherence to intestinal epithelium
298
614
319
This page has been reformatted by Knovel to provide easier navigation.
548
548
Index Terms
Links
Flagella (Cont.) Campylobacter invasion of intestinal epithelium
305
chemosensory signal transduction pathway
351
evolution
327
323
identification of flagellar proteins from genomic sequences
546
motility components involved
545
phase variation
553
requirements revealed by mutagenesis screens
546
mutants
323
O-linked glycosylation, see O-linked flagellar glycosylation Flagellar export apparatus
551
Flagellar genes C. jejuni
550
expression of early genes
554
Pseudomonas
550
regulation
547
fluA by sigma
552
Salmonella
547
transcriptional regulatory cascade
550
Vibrio
550
Flagellar number
551
regulation
554
Flagellin
317
biosynthesis
339
546
459
glycosylation, see O-linked flagellar glycosylation structure of glycoproteins Flavodoxin
471 602
Flavodoxin:quinone-reductase
46
fldA gene
602
FldA protein
602
602
flg genes flgB gene
323
flgC gene
323
flgD gene
616
flgDE2 gene
550
616
554
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
flg genes (Cont.) flgE gene
323
flgE2 gene
616
flgG gene
616
flgg2 gene
616
flgH gene
616
flgI gene
616
flgK gene
616
flgL gene
616
flgM gene
616
flgP gene
547
flgQ gene
547
flgR gene
flgS gene
554
616
551
553
574
612
617
621
551
553
621
Flg proteins FlgA protein
548
FlgB protein
548
FlgC protein
548
FlgD protein
548
FlgE protein
548
FlgE2 protein
548
FlgF protein
548
FlgH protein
548
FlgI protein
548
FlgK protein
548
FlgM protein
548
FlgP protein
547
FlgQ protein
547
FlgR protein
548
551
615
FlgS protein
548
551
616
flhA gene
551
553
584
flbB gene
323
551
flhF gene
551
555
552
flh genes
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Flh proteins FlhA protein
548
FlhB protein
548
FlhF protein
549
FlhG protein
549
FlhX protein
549
553
551
554
611
613
fli genes fliA gene
612
fliD gene
616
flip gene
551
fliR gene
551
fliS gene
584
Fli proteins FliA protein
548
FliD protein
548
FliE protein
548
FliF protein
548
FliG protein
548
FliH protein
548
FliI protein
546
FliK protein
548
FliL protein
549
FliM protein
548
FliN protein
548
FliO protein
548
FliR protein
548
FliS protein
548
FliY protein
548
Flies, transmission of Campylobacter
670
Flippase
451
548
538
Fluoroquinolone(s) for Campylobacter enteritis
111
in poultry
183
veterinary use
111
183
653
659
233
This page has been reformatted by Knovel to provide easier navigation.
651
Index Terms
Links
Fluoroquinolone resistance
111
consequences in human infections
656
in different food animal reservoirs
654
emergence
651
interventions to reduce
660
mechanisms
263
source attribution studies
656
Focal infections
108
folP gene
649
182
233
649
Food(s) Campylobacter in
627
cross-contamination during preparation
638
detection of Campylobacter in
628
survival of Campylobacter on
637
Food additives
688
Food animals, transmission of antimicrobial resistance to humans
645
Food chain surveillance
153
Food packaging
638
Foodborne Disease Burden Epidemiology Reference Group Food-borne illness
160 627
common-source outbreaks in industrialized countries
172
identifying source
283
molecular subtyping network
277
prevention
639
sporadic infections
177
FoodNet
279
283
Foreign travel, see Travel-related infections Formate, electron donor
51
Formate dehydrogenase
51
“Founders”
32
FqrB protein
602
frdCAB operon
55
Frozen food
637
Fructose bisphosphate aldolase FspA protein Fumarase
85 320
326
613
86 This page has been reformatted by Knovel to provide easier navigation.
404
Index Terms
Links
Fumarate, electron acceptor
55
Fumarate reductase
52
Fumarate respiration
44
55
87
406
617
fur genes
600
612
Fur proteins
458
600
612
487
566
G G protein-coupled receptors, Campylobacter invasion of intestinal epithelium G+C content
302 74
galE gene
394
Gallinacins
338
Gamma-glutamyl transpeptidase
49
65
Ganglioside(s), see also Antiganglioside antibodies peripheral nerve Ganglioside mimicry C. jejuni-related GBS Gastric acid
252 381
484
508
252 334
576
GBS, see Guillain-Barré syndrome GDla-immunized rabbits
382
Gene regulation, C. jejuni
611
Genetic diversity, C. jejuni
560
Genome diversity, C. jejuni
63
67
3
27
Genome sequence see also Comparative genomics C. fetus
406
C. jejuni
41
identification of flagellar sequences non-jejuni Campylobacter Genomic colinearity, Campylobacter genomes
546 79 79
Genomic islands C. fetus
405
non-jejuni Campylobacter Genomic reduction
82 87
Genomotyping
232
C. jejuni
36
Gentamicin protection assay
320
Gentamicin resistance
648
652
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
ggt gene
49
66
glnA gene
206
219
glpT gene
45
Gluconate, electron donor
52
Gluconeogenesis
44
Glucosyltransferase
493
Glucuronic acid-enriched polysaccharides
672
538
Glutamate transport
42
459
Glutamine synthase
47
49
Glutamine:2-oxoglutarate aminotransferase
47
49
Gluteal abscess
127
glyA gene
206
Glycan free
459
N-linked, see N-linked glycan O-linked, see O-linked glycan Glycomics
447
Glycoproteins, N-linked
451
459
Glycosylation, see N-linked protein glycosylation; O-linked flagellar glycosylation Glycosyltransferase
449
GM1, rabbit model of GBS
381
488
493
gmh genes gmhA gene
487
gmhA2 gene
509
gmhD gene
487
516
gne gene
459
Gne protein
449
Gram stain
228
230
578
9
572
620
groEL/groES gene GroEL/GroES protein
583
Groundwater
635
grpE gene
572
GTPase, Rho
305
Guillain-Barré syndrome (GBS)
245
animal models
254
antecedent infections
249
axonal forms
247
509
620
256
255
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Guillain-Barré syndrome (GBS) (Cont.) burden of illness
154
C. jejuni-related
109
bacterial gene characteristics
251
gene-specific variation in C. jejuni
251
rarity of disease
254
whole-genome polymorphisms in C. jejuni
250
clinical features
245
diagnostic criteria
245
heterogeneity
246
immunoglobulin treatment
395
immunopathology
248
outcome
256
pathogenesis
484
rabbit model
381
complement-mediated nerve injury
394
ganglioside immunization model
382
immune attack on nerve root axons
387
lipooligosaccharide immunization model
391
passive transfer model
396
251
251
subgroups
246
vaccine-related
433
437
517
19
231
264
gyrA gene
661 GyrA protein
404
H HcrA protein
620
hdd genes
509
516
Heat stress C. jejuni
571
heat shock proteins
572
novel and alternative heat stress responses
573
regulatory genes in heat shock response
573
Heat-labile toxin, E. coli, mucosal adjuvant
438
hec genes
92
Helicobacter, taxonomy Helicobacter canis
4 140
This page has been reformatted by Knovel to provide easier navigation.
649
Index Terms
Links
Helicobacter cinaedi infections
139
clinical features
126
130
epidemiology
126
139
microbiology
125
139
pathogenesis
140
strain characterization
139
treatment
140
Helicobacter fennelliae infections
139
clinical features
126
130
epidemiology
126
139
microbiology
125
139
pathogenesis
140
strain characterization
139
treatment
140
Helicobacter pullorum
140
Helicobacter pullorum infections
125
Helicobacter pylori, VBNC state
579
Helicobacter rappini
141
Helicobacter rappini infections
126
Helicobacter westmeadii
141
Helix-turn-helix motif
611
Hemagglutinin, filamentous
139
92
Hematology, Campylobacter enteritis
110
Heme
592
Heme oxygenase
596
Hemoglobin, iron availability
596
Hemolysin
596
Hemolytic-uremic syndrome
131
Hemopexin
596
Hemophore
595
Hepatitis
105
Heptose biosynthesis
509
Heptosyltransferase
487
Hfr strains
139
594
492
84
Highly conserved signaling domain
358
High-resolution melt analysis
194
hipO gene
230
Hippurate hydrolysis test
230
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
HisJ protein
452
Histidine kinase sensor
351
357
406
78
494
621
405
561
551
585 Homoplasy ratio
560
Homopolymeric tracts Horizontal gene transfer Host associations C. coli
200
C. fetus
214
C. jejuni
199
MLST studies
198
Host cell entry
297
see also Adherence, bacterial; Intestinal epithelial cells; Invasion, bacterial Host range, plasmids
77
Housekeeping genes
29
hrcA gene
572
HrcA protein
612
612
620
612
hsd genes
88
hslU gene
572
hspR gene
572
574
HspR protein
573
620
htrA gene
572
574
htrB gene
487
492
572
53
573
htr genes
Hybrid-cluster proteins
86
Hydrogen, electron donor
51
Hydrogen peroxide
50
electron acceptor Hydrogenase
53 51
Hydrogen-requiring campylobacters
87
135
clinical features
136
epidemiology
136
isolation
229
microbiology
125
Hydroxyisourate hydrolase
49
Hydroxyl radicals
53
Hygromycin resistance
77
135
599
This page has been reformatted by Knovel to provide easier navigation.
598
Index Terms
Links
Hyperglycemic bug
66
Hyperosmotic stress
575
Hypogammaglobulinemia
103
Hypo-osmotic stress
575
128
156
I IcIR protein
612
IgA1 protease
620
92
Ileostomy stoma ulceration
105
Imipenem
234
Immune evasion
317
Immune response, see also Innate immunity Campylobacter enteritis
99
Campylobacter infections
155
flagellin as antigen
478
GBS
248
role of N-linked protein glycosylation
456
systemic vs. mucosal antibodies
432
Immunocompromised patient
182
318
432
161
163
430
163
C. fetus
127
C. hyointestinalis
132
C. lari
134
Campylobacter enteritis
103
Incubation period, Campylobacter enteritis
101
Industrialized countries, Campylobacter infections
156
age and sex distribution
170
burden of illness
171
common-source outbreaks
173
comparisons of incidence between countries
169
seasonality
170
surveillance
168
trends
168
INDX Campy-JCL
230
Infective dose
101
Inflammatory bowel disease, postinfectious
154
316
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Innate immunity Campylobacter infections
333
defense against enteric pathogens
333
in gastrointestinal tract
333
in intestinal submucosa
334
340
in systemic circulation
334
340
308
337
334
339
adherence of C. jejuni
298
317
Campylobacter translocation across mucosa
305
318
defense against enteric pathogens
333
invasion by C. jejuni
299
Interleukin-8 Internalization, see Invasion, bacterial Intestinal biota, normal Intestinal epithelial cells
Intestinal hemorrhage
319
315
105
Intestinal mucosa Campylobacter translocation across
305
318
innate immune defenses
334
340
Intestinal tract innate immune defense
333
iron availability
592
Intranasal challenge model, murine
374
Intravenous immunoglobulin, treatment of GBS
395
Invasion, bacterial C. jejuni in intestinal epithelium
299
capsular polysaccharide in
513
cytoskeleton in
299
role of N-linked protein glycosylation
456
signal transduction
301
Invasion receptors
305
Invasion studies
320
Inversion event, sap genes in C. fetus
417
315
Iron functions
591
oxidative stress defense mechanisms
573
598
604
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Iron acquisition
404
406
602
617
ferrous iron
592
heme compounds
595
host iron-binding proteins
343
involvement of Pgl proteins
458
siderophore-mediated
594
TonB and associated genes
597
Iron availability
591
intestinal tract
592
transcriptional responses
601
Iron metabolism
591
biofilms
605
transcriptional profile of iron metabolism genes
604
Iron regulatory proteins
600
Iron stimulon
601
Iron storage
598
Iron stress
598
Iron-binding proteins, host
343
Irradiated food
638
Irritable bowel syndrome, postinfectious
109
IS elements
78
ISCco1
83
ISCjd1
83
non-jejuni Campylobacter
83
ISO method, detection of Campylobacter in food
456
592
596
617
600
596
628
J jlpA gene
320
JlpA protein
91
298
319
76
648
652
340
572
583
603
605
617
KatA protein
598
601
604
Kdo
506
kdtA gene
492
K Kanamycin resistance katA gene
This page has been reformatted by Knovel to provide easier navigation.
599
Index Terms
Links
Killed whole cell vaccines
437
517
kps genes
506
512
genetic markers for diagnosis and epidemiology
516
genetic rearrangements within gene clusters
515
kpsC gene
508
kpsE gene
513
kpsM gene
508
kpsS gene
508
515
512
Kps proteins KpsM protein
508
KpsT protein
508
L Lactate dehydrogenase
52
Lactoferrin
592
Lambs
634
594
597
Lateral gene exchange, lipooligosaccharide biosynthesis genes
490
Lateral gene transfer (localized sex)
30
Lauroyl acyltransferase
573
Lawsonia intracellularis
13
206
Laying hens
630
Lectins, mannose-binding
334
343
Legionaminic acid and derivatives
473
555
biosynthesis
534
Leg5Ac7Ac
533
Leg5Am7Ac
473
477
532
Leg5AmNMe7Ac
473
477
533
Levofloxacin resistance
263
lic gene
206
Lipid A
407
biosynthesis
492
structure
483
Lipid A acyltransferase
483
487
Lipid rafts, Campylobacter invasion of intestinal epithelium Lipid-linked oligosaccharide (LLO)
303
305
309
449
451
459
This page has been reformatted by Knovel to provide easier navigation.
538
Index Terms
Links
Lipooligosaccharide (LOS) biosynthesis loci genetic variation
486
phase variation
494
recombination and lateral exchange between
490
biosynthesis pathways O-acetylation of terminal D-2,8 NeuSAc
498
class A and B cores with nonsialylated inner Gal
497
class A cores with sialylated inner Gal
496
class B cores extended from Glc-β(1,2)-HepII
498
class C cores
496
extensions from HepII in sialylated outer cores
495
generating core region variation
494
inner core
492
lipid A
492
outer core
492
C. jejuni
66
501
500
251
381
507 ganglioside mimicry
484
outer core structure
485
structure
484
in transformation
566
Lipopolysaccharide (LPS)
342
C. fetus
407
C. jejuni
339
GBS in immunized rabbits
391
structure
483
Lipopolysaccharide-binding protein
343
Live attenuated vaccines
437
483
517
LLO, see Lipid-linked oligosaccharide lon gene
572
LOS, see Lipooligosaccharide LPS, see Lipopolysaccharide lpx genes
231
Lung abscess
127
luxS gene
585
LysR protein
612
492
619
This page has been reformatted by Knovel to provide easier navigation.
483
Index Terms
Links
M Macrolide resistance
111
emergence
651
interventions to reduce
659
mechanisms
265
Macrophages
649
318
innate immune defenses
334
production of nitric oxide
341
production of reactive nitrogen species
341
maf genes
340
554
Malate oxidoreductase
46
MALDI-TOF MS
16
mapA gene
231
MarR protein
620
Meat, red
634
MEDPeD (Murine Enteric Disease Phenome Database)
375
Megaplasmids
52
638
76
Membrane attack complex Menaquinones
394 12
20
Meningitis
126
403
MerR protein
612
619
Metabolomics
523
future directions
539
in vitro analysis by NMR spectroscopy
534
Methyl-accepting chemotaxis proteins
352
Methylation, DNA
562
Methyl-directed mismatch repair Methyltransferases C. upsaliensis
47
546
548
46
49
50
79 267 90
mez gene
46
Microaerophily
44
54
562 Microfilaments, Campylobacter invasion of intestinal epithelium
299
320
299
309
320
124
132
173
429
627
633
Microtubules, Campylobacter invasion of intestinal epithelium Milk, raw (inadequately pasteurized)
This page has been reformatted by Knovel to provide easier navigation.
180
Index Terms
Links
Miller Fisher syndrome
245
Minimal inhibitory concentration
645
Mismatch repair, non-jejuni Campylobacter Mitogen-activated protein kinase pathway
250
397
78 305
308
27
192
197
C. coli
29
36
204
C. fetus
29
37
204
218
222
29
37
204
MLST (multilocus sequence typing)
C. helveticus C. insulaenigrae
29
32
38
C. lari
29
37
204
C. upsaliensis
29
37
204
317
319
epidemiologic typing
232
host association studies
198
national subtyping network
283
non-jejuni Campylobacter
76 204
modE gene
612
ModE protein
619
Modified atmosphere packaging
638
Molecular mimicry, see Ganglioside mimicry Molecular typing
216
204
C. jejuni
non-jejuni, non-coli Campylobacter
484
191
genotyping of non-jejuni, non-coli Campylobacter
202
methods
192
outbreak investigations
201
source tracking
198
Molybdenum homeostasis
619
Monkey model, C. jejuni infections
436
Morbidity
152
Mot proteins
548
Motility bacteria within biofilms
583
Campylobacter adherence to intestinal epithelium
298
chemosensory signal transduction pathway
351
This page has been reformatted by Knovel to provide easier navigation.
204
Index Terms
Links
Motility (Cont.) flagellar components involved
545
phase variation
553
requirements revealed by mutagenesis screens
546
Mouse model C. fetus infections
410
C. jejuni infections C3H SCID limited flora mice C57BL/6 IL-10
+/+
mice
373 370
colonization models
369
diet effects
374
disease models
370
early studies
368
experimental design for mouse studies
374
future goals
375
immunocompetent mice refractory to C. jejuni
369
individual housing
374
-/-
NF-KB mice
371
quantifying clinical signs of disease and pathology
375
screening immunodeficient mice for pathogen susceptibility specific-pathogen-free status
372 368
C. jejuni vaccine development
373
intranasal challenge model
374
MEDPeD
436
375
Mucins
336
Mucosal translocation
305
318
Multidrug efflux pump
270
336
Multidrug resistance
234
270
3
9
Multilocus sequence analysis
650
Multilocus sequence typing, see MLST Multispecies farming
671
Murine model, see Mouse model Mushrooms MviN protein Myocarditis
637 91 108
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
N NAD(P)H oxidation Nalidixic acid resistance
nap genes
51 233
235
404
650
652
654
55
85
231
574
51
87
406
488
491
603 NARMS (National Antimicrobial Resistance Monitoring System)
279
283
NARTC group (nalidixic acid-resistant thermophilic campylobacters) NASC strains (nalidixic acid-susceptible strains) Natural killer cells, Siglecs NDH-1 complex
15 15 341 47
Neomycin resistance
649
Neonatal infections
103
Nephritis
107
Nerve root axons, immune attack in AMAN
387
neu genes neuA gene
474
neuA1 gene
394
neuB gene
474
neuB1 gene
394
neuC gene
488
Neu proteins
488
493
Neutrophil(s) innate immune defenses
334
production of antimicrobial peptides/proteins
340
production of reactive oxygen species
340
Neutrophil-activating protein
598
Nf-KB
337 -/-
Nf-KB mice NfrA protein
340
339
371 617
Nitrate electron acceptor salivary Nitrate reductase
55 334
344
55
85
335
604 This page has been reformatted by Knovel to provide easier navigation.
576
Index Terms
Links
Nitric oxide
576
antimicrobial property
335
produced by macrophages
341
Nitric oxide detoxification
55
Nitric oxide reductase
86
Nitric oxide synthase
335
339
341
604
Nitrite, electron acceptor
55
Nitrite reductase
55
86
Nitrogen metabolism
47
86
572
576
Nitrosative stress Nitrous oxide reductase
600
86
NJC, see Non-jejuni Campylobacter NK-KB
308
N-linked glycan
567
biological effects of disrupting N-glycan pathway
456
biosynthesis
448
effect on protein structure and function
453
structure
448
N-linked glycoproteins
451
N-linked protein glycosylation
447
476
see also pgl genes in bacterial adherence
298
biological effects of disrupting N-glycan pathway
456
C. jejuni
66
in chicken colonization
457
effect on protein structure and function
453
in immune response
456
in invasion
456
pathways in Proteobacteria
464
transformation and
567
NMR spectroscopy, in vitro metabolomics
534
NOD proteins, defense against enteric pathogens
334
Nodes of Ranvier
394
Nonclonal population Nonhuman primate models, C. jejuni infections
456
448
537
338
344
31 436
This page has been reformatted by Knovel to provide easier navigation.
617
Index Terms
Links
Non-jejuni Campylobacter (NJC) bacteriophage
81
bacteriophage Mu
79
comparative genomics
73
contingency genes
78
CRISPR elements
82
features of genomes
74
genome structure
79
genomic islands
82
genomic reduction
87
homopolymeric tracts
78
IS elements
83
metabolism
85
mismatch repair
78
MLST
76
plasmids
76
proteomes
74
pseudogenes
78
respiration
86
restrictionlmodification systems
88
transposons
83
virulence and pathogenicity
90
Novobiocin resistance
82
270
N-oxides, electron acceptors
56
nrd genes
87
nrf genes
55
572
nssR gene
572
612
NssR protein
576
618
NssR regulatory system
600
NtrC protein family
612
nuo genes
574
603
615
O O-antigen
407
Oligosaccharides, breast milk
334
344
Oligosaccharyltransferase
448
451
459
This page has been reformatted by Knovel to provide easier navigation.
538
Index Terms
Links
O-linked flagellar glycosylation
447
C. coli
530
C. fetus
406
C. jejuni
66
flagellar pathway elucidation
475
glycan structure
472
glycosylation locus
474
mechanism of glycosylation
477
structure of flagellin glycoproteins
471
459
471
524
O-linked glycan flagellar
471
biological role structure determination OmpR protein
477 472 612
Organic acid(s), electron donors
52
Organic acid transport
44
Osmotic stress, C. jejuni
572
Osteitis
108
Osteomyelitis
127
OutbreakNet
279
Outcome tree
154
614
575
Outer membrane proteins novel, detection and identification
434
vaccine target
433
Oxidase test
230
Oxidative stress
617
C. jejuni
572
iron and
598
602
2-Oxoglutarate:acceptor oxidoreductase
46
Oxygen sensing
50
357
p19 system
595
604
pan genes
603
Pan proteins
603
Pancreatitis
106
par genes
264
604
P
Partially clonal population
617
31
This page has been reformatted by Knovel to provide easier navigation.
555
Index Terms
Links
Passive immunization, chickens
673
Passive transfer model, GBS
396
Pasteurization process
633
Pathogen-associated molecular patterns
318
Pathogenicity islands, C. jejuni Paxillin
637
65 320
pckA gene
46
PCR methods C. fetus
216
Campylobacter in food
629
epidemiologic typing
232
ERIC-PCR
194
identification of campylobacters
230
Rep-PCR
194
peb1 gene
218
320
PEB proteins PEB1 protein
42
298
319
321
517
583
PEBla protein
42
91
407
459
PEB3 protein
448
450
452
Penicillin binding proteins
269
648
Penicillin resistance
234
269
Penner serotyping
507
515
44
406
Pentose phosphate pathway Peptidoglycan
579
Pericarditis
403
Perinatal infections
107
Periodontal disease
126
Peripheral nerve gangliosides
252
Perirectal abscess
105
Peritonitis
108
Peroxidases
599
Peroxide stress defense
599
perR gene
648
135
128
404
572
599
601
612
617
PerR protein
599
617
Pets
181
This page has been reformatted by Knovel to provide easier navigation.
603
Index Terms
Links
PFGE (pulsed-field gel electrophoresis)
192
C. coli
282
C. concisus
202
C. fetus
202
C. hyointestinalis
203
C. jejuni
195
C. sputorum
203
C. upsaliensis
203
epidemiologic typing
232
non-jejuni, non-coli Campylobacter
202
PulseNet
277
pflA gene
547
PflA protein
548
pfr gene
598
Pfr protein
598
pgl genes
447
autoregulation
457
gene expression in C. jejuni
457
mutations
456
orthologs in Proteobacteria
464
pglB gene
448
195
222
404
280
537
567
453
456
538
567
567 pglC gene
449
pglD gene
539
pglE gene
449
453
456
pglF gene
449
457
477
pglH gene
457
538
pglI gene
538
pglJ gene
538
pglK gene
538
transcriptional profiling
457
460
Pgl proteins amino acid transport and biosynthesis
458
iron acquisition
458
PglA protein
449
PglB protein
448
PglC protein
449
PglD protein
449
459
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Pgl proteins (Cont.) PglE protein
449
476
535
PglF protein
449
476
535
PglH protein
450
PglI protein
449
PglJ protein
450
PglK protein
449
Phage, see Bacteriophage Phagocytes innate immune defenses
340
Siglecs
341
Phase variation
621
C. jejuni
65
capsular polysaccharide
515
flagellar motility
553
lipooligosaccharide biosynthesis genes
494
Phenotypic characteristics
15
pho box
615
phoA gene
616
Phosphate acetyltransferase
584
Phosphoenolpyruvate carboxykinase
46
Phosphoglucose isomerase
85
Phosphoinositol-3 kinase
304
Phospholipase
514
Phos proteins
614
phosR gene
612
Phylogenic tree
5
344
215
596
30
Pili, type IV
564
Pilins
560
564
Piperacillin resistance
234
648
Planktonic bacteria
583
Plasmids
434
C. jejuni
64
cryptic
76
drug resistance
267
host range
77
incompatibility groups
77
integrated
82
230
651
270
This page has been reformatted by Knovel to provide easier navigation.
84
Index Terms
Links
Plasmids (Cont.) megaplasmids
76
non-jejuni Campylobacter
76
pVir
82
565
Plasticity regions
621
pldA gene
515
596
PldA protein
515
596
Pleuritis
404
Polyphosphate
575
584
Polysaccharides C. jejuni
514
capsular, see Capsular polysaccharide zwitterinonic
516
neutral
514
Polysulfide reductase Population studies
87 191
C. coli
36
C. fetus
37
C. jejuni
32
Campylobucter
27
analysis of populations
28
variation within genus
28
interpretation of biological variation
28
measuring variation
27
sampling the population
27
porA gene
320
PorA protein
319
Porins
269
Pork
634
Poultry
627 see also Chickens colonization
629
bacteriophage therapy
679
broilers
667
eliminating Campylobacter from colonized flock
674
epidemiology of flock colonization
668
lag phase
668
phage effects
688
prevalence rates between countries
668
559
319
405
667
679
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Poultry (Cont.) prevention of introduction of Campylobacter into flock
670
risk factors
669
seasonal variation
669
sources
669
drug-resistant Campylobacter emergence
654 651
genetic resistance to Campylobacter
673
sporadic infections and
177
Poultry house, biosecurity
670
Poultry meat
629
fresh products
629
marinated
630
retail level
630
slaughterhouse level
630
180
632
ppk1 gene
572
575
Prepilin
560
564
Prepilin peptidase
565
Probiotics
672
Proctitis
132
Prodrome, Campylobucter enteritis
101
Proinflammatory factors, C. jejuni-induced
308
Proline dehydrogenase Promoters
584
139
47 611
pgl operon
457
ProSpecT Campylobacter immunoassay
228
Prostatitis
108
Prosthetic hip sepsis
108
127
Protein glycosylation, see N-linked protein glycosylation; O-linked flagellar glycosylation Protein kinase, Campylobacter invasion of host cells
304
Protein kinase C
303
305
307
Protein secretion Campylobacter invasion of host cells
315
identification of secreted proteins
325
Proteobacteria, N-glycosylation pathways
464
Proteomes, non-jejuni Campylobacter
74
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Protozoa, Cumpylobacter survival in
516
pse genes
555
581
pseA gene
474
477
527
534
pseB gene
475
477
526
534
pseC gene
475
477
527
pseD gene
475
477
pseE gene
475
554
pseF gene
526
pseI gene
526
Pse proteins PseA protein
549
PseB protein
449
459
476
535
476
535
549
549 PseC protein
459
PseD protein
549
PseE protein
549
554
PseF protein
475
535
549
PseG protein
476
535
549
PseH protein
476
535
549
PseI protein
476
535
549
449
459
473
biosynthesis
524
530
534
Pse5Ac7Ac
474
524
530
Pse5Ac7Am
473
533
PseAm
530
555
Pseudaminic acid and derivatives
Pseudogenes C. jejuni
65
non-jejuni Campylobacter
78
Pseudomonas, flagellar genes
550
Pseudopilins
560
pst genes
616
PSTC proteins
563
ptm genes
477
ptmA gene
474
ptmB gene
474
ptmC gene
475
ptmD gene
475
564
530
532
477
534
This page has been reformatted by Knovel to provide easier navigation.
555
533
Index Terms
Links
ptm genes (Cont.) ptmE gene
475
ptmF gene
475
ptmG gene
533
PubMLST website
30
36
Pulsed-field gel electrophoresis, see PFGE PulseNet
277
Campylobacter
280
next-generation subtyping methods
279
partnerships
279
PFGE subtyping methods
277
PulseNet International
279
PulseNet USA
277
Purine/pyrimidine phosphoribosyltransferase
565
pyc genes
45
pyk gene
44
pyrC gene
84
Pyruvate carboxylase
45
Pyruvate dehydrogenase
86
Pyruvate kinase
44
Pyruvate:acceptor oxidoreductase
46
Pyruvate: flavodoxin oxidoreductase
86
Q QacR transcriptional factor
271
QALY (quality-adjusted life-year)
152
D-QuiNAdNAc
528
Quinol dehydrogenase
55
Quinol oxidase
53
Quorum sensing, AI-2-mediated
535
585
R Rabbit model C. jejuni infections
436
Guillain-Barré syndrome
381
complement-mediated nerve injury
394
ganglioside immunization model
382
immune attack on nerve root axons
387
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Rabbit model (Cont.) immunoglobulin treatment
395
lipooligosaccharide immunization model
391
passive transfer model
396
rac genes racR gene
572
612
racS gene
572
615
RacR protein
573
614
RacS protein
615
615
Rac proteins
Radiography, in Campylobacter enteritis
110
RAPD (random amplification of polymorphic DNA)
193
C. concisus
204
C. fetus
218
C. jejuni
204
non-jejuni, non-coli Campylobacter
204
Rash, Campylobacter enteritis
222
103
Reactive nitrogen species gastrointestinal
334
produced by macrophages
341
Reactive oxygen species
50
gastric
334
produced by neutrophils
340
recA gene
566
RecA protein
566
Recombination
688
DprA-mediated
566
flagellin genes
561
intergenomic
561
interspecies
561
interstrain
621
intragenomic
561
lipooligosaccharide biosynthesis genes
490
sap genes in C. fetus
418
Rectal swab
227
Refrigeration
637
Regulons
611
Reiter’s syndrome
109
339
572
591
621
This page has been reformatted by Knovel to provide easier navigation.
598
Index Terms
Links
Removable intestinal tie adult rabbit diarrhea
436
Renal disease, with Campylobucter enteritis
107
Rep-PCR
194
Reptiles, C. fetus
221
Respiration, non-jejuni Campylobacter Response regulatory protein
Restriction/modification systems
86 351
355
551
614
406
553
612
617
614
616
562
C. jejuni
64
non-jejuni Campylobacter
88
type I
88
type II/III
90
RFLP, Fla-RFLP
192
Rhodotorulic acid
594
Ribonucleoside-diphosphate reductase
358
602
87
Riboprinting
193
Ribosomal protection protein
266
Ribosomal proteins
265
650
Ribotyping
193
232
Ridascreen Campylobacter
228
Rifampicin resistance
621
Rifampin resistance
270
Risk assessment, antimicrobial resistance
657
rml genes
491
RNA polymerase
611
rpmA gene
603
RpmA protein
604
621
rpo genes rpoB gene
9
rpoD gene
612
rpoN gene
550
Rpo proteins RpoD protein
611
RpoN protein
548
611
Rrf2 protein
612
620
rRNA methylases
265
267
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Ruminants C. fetus
214
C. jejuni
199
S 16S rRNA Arcobacter
19
C. fetus
222
Campylobacter
4
Sulfurospirillum
20
16S/23S RNA
231
antimicrobial resistance
16
265
650
Salad vegetables
636
Salmonella, flagellar genes
547
Salt tolerance
575
Sanitizer resistance
581
637
sap genes C. fetus
411
gene expression
415
genetic organization (clustering)
413
sap island
416
sapA gene
412
sapB gene
413
sapCDEF genes
420
sap typing
222
Sarafloxacin
651
sat4 gene
649
Scheduled slaughtering
674
421
416
sda genes sdaA gene
47
sdaC gene
43
sdh genes
85
572
627
635
Campylobacter infections
167
170
poultry colonization
669
Secondary transmission
181
Seafood
574
Seasonality 198
This page has been reformatted by Knovel to provide easier navigation.
603
Index Terms
Links
Secreted proteins S-layer proteins of C. fetus
419
vaccine target
433
Secretins
560
Sedoheptulose-7-phosphate isometase
487
Sensor kinase
319
563
614
see also Histidine kinase sensor Sensory receptors, E. coli
352
Septicemia
126
Sequence types
29
204
C. coli
36
200
C. fetus
37
C. jejuni
32
Serine dehydratase
36
199
222
232
47
Serine hydratase
406
Serine protease
92
Serine transport
43
Seroepidemiology
153
Serotyping
191
Sessile bacteria
583
Sex distribution, Campylobacter infections
166
Sheep
634
C. fetus infections
214
C. jejuni infections
199
170
403
Shellfish
636
Short variable regions, fla-SVR sequencing
192
Sialic acid synthase
488
Sialic acid transferase
251
Sialyated O-acetyltransferase
498
501
Sialyltransferase
488
493
Siderophore(s)
592
594
Siderophore piracy
594
Siglecs
341
Sigma factors
611
C. jejuni
612
FliA
611
RpoD
611
RPoN
611
423
493
498
613
614
616
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Sigma factors (Cont.) Sigma-58
550
regulation of flaA
552
613
552
sigma-54
554
sigma-54 dependent flagellar genes
550
sigma-70
612
614
Signal transduction Campylobacter invasion of intestinal epithelium
301
chemosensory signal transduction pathway
351
Slaughtering process
629
634
S-layer proteins, C. fetus
409
414
antigenic diversity in ovine immune responses
423
antigenic variation
415
in ovine abortion
423
regulation of production
422
sap genes
411
secretion
419
Slipped-strand mispairing
621
SNP (single nucleotide polymorphism) binary typing
194
sodB gene
340
SodB protein
598
Sodium channels
394
Solute transport
41
Source tracking, molecular methods
599
198
S-oxides, electron acceptors
56
Speciation, bacterial
30
Species tree
572
638
5
Spectinomycin resistance
649
Spices
637
Splenic rupture
108
Sporadic infections
627
contact with animals and
181
cross-contamination and
180
industrialized countries
174
poultry and
177
raw milk and
180
180
This page has been reformatted by Knovel to provide easier navigation.
674
Index Terms
Links
Sporadic infections (Cont.) secondary transmission
181
travel-associated
182
waterborne
180
SPOT gene
572
574
579
584
621 Starvation
573
detriments and possible benefits
575
resistance
575
Stationary phase stress, C. jejuni
572
Statutory disease
401
Stool antigen tests
228
Streptomycin resistance
648
Streptothricin resistance
648
Stress response, C. jejuni
515
571
Stringent response
575
584
Subunit vaccines
439
517
Succinate dehydrogenase
52
85
Sulfite, electron donor
52
Sulfonamide resistance
649
651
5
19
Sulfurospirillum Sulfurospirillum arcachonense
19
Sulfurospirillum arsenophilum
19
Sulfurospirillum barnesii
19
Sulfurospirillum cavolei
19
Sulfurospirillum deleyianum
19
Sulfurospirillum halorespirans
19
Sulfurospirillum multivorans
19
Superoxide dismutase Superoxide radicals
574
654
573
599
50
53
Superoxide stress defense
599
Surface recognition
514
Surface water
635
621
604
573
Surveillance Campylobacter infections industrialized countries
168
United States
164
emerging antimicrobial resistance
182
This page has been reformatted by Knovel to provide easier navigation.
598
Index Terms
Links
Surveillance pyramid
152
Swine
634
drug-resistant Campylobacter
656
piglet model, C. jejuni infections
436
167
T TAT signal peptides
51
56
Taxonomy bacteriophage
680
C. fetus
401
Campylobacter aceae
3
MLSA approach
9
taxonomic history
3
whole-genome
5
Tbp proteins
597
Telithromycin resistance
265
tet(O) gene
649
Tet(O) protein
266
TetR family
271
Tetracycline(s), clinical indications
114
Tetracycline resistance
mechanisms
213
651
612
618
76
183
234
404
649
651
654
661
603
618
266
Thinning, poultry
672
Thiol peroxidase
604
Thioredoxin
602
604
Thioredoxin reductase
599
602
Ticarcillin resistance
234
648
Tigecycline, clinical indications
128
Tight junctions
306
tkt gene
206
Tlp proteins
352
group A receptors
355
groups B, C, and Aer receptors
356
TMAO reductase
56
Toll-like receptors
478
defense against enteric pathogens tonB gene
318
355
86
334
339
595
597
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
TonB protein
594
596
Topoisomerase
263
650
Toxic megacolon
104
Toxins, host damage caused by
307
tpx gene
603
Tpx protein
583
Transamination
602
604
319
604
47
Transcription
611
Transcription factors
611
614
Transcriptional profiling iron metabolism genes
604
pgl genes
457
460
Transcriptional regulation, flagellar genes
550
Transcriptional repressors
618
Transferrin
592
594
596
77
319
406
453
658
Transformation
559 candidate gene analyses to identify proteins
566
conserved proteins associated with
560
DNA discrimination
562
N-linked protein glycosylation and
567
physiology
562
type IV secretion system
565
Translocation, Campylobacter across intestinal mucosa
305
Transport media
228
Transposon(s), non-jejuni Campylobacter
318
83
Transposon mutagenesis
546
551
Traveler’s diarrhea
429
431
Travel-related infections
111
182
430
Trimethoprim resistance
270
649
651
Trimethoprim-sulfamethoxazole resistance
236
599
603
601
604
trx genes trxA gene
603
trxB gene
583
Trx proteins TrxA protein
602
TrxB protein
599
This page has been reformatted by Knovel to provide easier navigation.
618
Index Terms
Links
“Twin arginine” translocation system, see TAT signal peptides Tylosin
652
Type I secretion system
615
C. fetus
659
420
Type II secretion system
563
Type III secretion system
546
C. jejuni
315
evolution
327
recognition and export of CiaB
323
Type IV secretion system pVir-encoded Tyrosine protein kinase
434
453
456
565 304
U UDP sugars
525
UDP-6-deoxy-β-L-AltNAc4N
526
UDP-diacetamido-trideoxyhexose
525
UDP-GlcNAc
449
UDP-GlcNAc dehydratase
449
UDP-GlcNAc/ Glc-4-epimerase
487
UDP-GlcNAc3N
492
UDP-monoacetamido-trideoxyhexose
525
UDP-α-D-QuiNAc4NAc
528
Undecaprenol
450
Undecaprenylpyrophosphate
449
Undecaprenylpyrophosphate-heptasaccharide
538
538
538
United States, Campylobacter infections age and sex distribution
166
burden of illness
167
common-source outbreaks
172
regional differences in incidence
166
seasonality
167
sporadic
175
surveillance
164
trends
164
UPTC strains (urease-producing)
15
Urea catabolism
49
133
This page has been reformatted by Knovel to provide easier navigation.
563
Index Terms
Links
Urinary tract infections
107
V Vaccine Campylobacter, see Campylobacter vaccine chickens
673
Vacuum packaging
638
Vaginosis
403
Vascular disease
127
136
404
VBNC state, see Viable but nonculturable state Vegetable food types
636
VetNet
279
Viability count
577
Viable but nonculturable (VBNC) state
571
resuscitation
283
577
579
VBNC formation changes in cell wall
578
morphology and viability changes
577
Vibrio, flagellar genes Vibrio fecalis
550 13
VirB10 protein
453
Virulence factors/virulence phenotypes C. jejuni
316
non-jejuni Campylobacter
90
W waa genes waaC gene
487
waaD gene
583
waaF gene
487
493
Waterborne illness
627
635
common-source outbreaks in United States
172
sporadic infections
180
White cells, fecal
492
228
Whole-genome sequence, see Genome sequence Wild birds
181
199
Wildlife animals
630
636
Wolinella
630
5 This page has been reformatted by Knovel to provide easier navigation.
635
Index Terms
Links
X XylS protein
612
620
Y YLD (years lived with disability)
151
YLL (years of life lost to premature death)
151
This page has been reformatted by Knovel to provide easier navigation.
Address editorial correspondence to ASM Press, 1752 N St. NW, Washington, DC 20036-2904, USA Send orders to ASM Press, P.O. Box 605, Herndon, VA 20172, USA Phone: (800) 546-2416 or (703) 661-1593 Fax: (703) 661-1501 E-mail:
[email protected] Online: estore.asm.org Copyright 0 2008 ASM Press American Society for Microbiology 1752 N Street NW Washington, DC 20036-2904 Library of Congress Cataloging-in-Publication Data Campylobacter / editors, Irving Nachamkin, Christine M. Szymanski, and Martin J. Blaser.-3rd ed. p. ; cm. Includes bibliographical references and index. ISBN 978-1-55581-437-3 (alk. paper) 1. Campylobacter infections. I. Nachamkin, Irving. 11. Szymanski, Christine M. 111. Blaser, Martin J. [DNLM: 1. Campylobacter-genetics. 2. Campylobacter-pathogenicity. 3. Campylobacter Infectionsepidemiology. 4. Campylobacter Infections-physiopathology. 5. Food Contamination-prevention & control. QW 154 C1991 20081 QR201.C25C36 2008 6 16.9’20 1-dc22 2008007610 10987654321
All Rights Reserved Printed in the United States of America Cover: Scanning electron micrograph showing the apical surface of differentiated Caco-2 human intestinal epithelial cells containing typical, densely packed, microvillus extensions of the host plasma membrane, with knobs at their tips. Several spiral-shaped cells of Cumpylobacter jejuni 81-176 are tethered specifically to Caco2 cell microvillus tips via interactions with the sides of bacterial flagella. These binding events cause the flagella to appear bent at angles greater than or equal to 90 degrees at point of contact. The early Campylobacter flagellum-host cell interactions are considered to be a major mechanism of adherence to intestinal cells in the gut lumen. This initial adherence event can sometimes lead to subsequent yet uncharacterized specific bacterial invasion ligand-host receptor binding, triggering host signal transduction events that cause host cell internalization of C. jejuni prior to bacterial transcytosis across the intestinal epithelial mucosa. Photo courtesy of Dennis J. Kopeck0 and Han Lu, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Md.
