bacteriophages

5/26/2018 Bacteriophages

1/297


2/297

Bacteriophages in Health and Disease


3/297

Bacteriophage T4 seen obliquely from the base plate end

Picture created by Steven McQuinn using protein structure data from Rossmann Lab,furnished by the RCSB Protein Databank. The various software utilized includes UCSF

Chimera, Accelrys DS Visualizer, MeshLab and DAZ Carrara.


4/297

Bacteriophages in Health and Disease

Edited by

Paul Hyman

Department of Biology/Toxicology, Ashland University

and

Stephen T. Abedon

Department of Microbiology, The Ohio State University

Advances in Molecular and Cellular Microbiology 24
http://www.cabi.org/


5/297

Advances in Molecular andCellular Microbiology

Through the application of molecular and cellular microbiology we now recognisethe diversity and dominance of microbial life forms on our planet, that exist in all

environments. These microbes have many important planetary roles, but for wehumans a major problem is their ability to colonise our tissues and cause disease. Thesame techniques of molecular and cellular microbiology have been applied to theproblems of human and animal infection during the past two decades and haveproved to be immensely powerful tools in elucidating how microorganisms causehuman pathology. This series has the aim of providing information on the advancesthat have been made in the application of molecular and cellular microbiology tospecific organisms and the diseases that they cause. The series is edited by researchersactive in the application of molecular and cellular microbiology to human disease

states. Each volume focuses on a particular aspect of infectious disease and willenable graduate students and researchers to keep up with the rapidly diversifyingliterature in current microbiological research.

Series Editor

Professor Michael Wilson

University College London


6/297

Titles Available from CABI

17. Helicobacter pyloriin the 21st CenturyEdited by Philip Suton and Hazel M. Mitchell

18. Antimicrobial Peptides: Discovery, Design and Novel Therapeutic StrategiesEdited by Guangshun Wang

19. Stress Response in Pathogenic BacteriaEdited by Stephen P. Kidd

20. Lyme Disease: an Evidence-based ApproachEdited by John J. Halperin

22. Antimicrobial Drug Discovery: Emerging StrategiesEdited by George Tegos and Elefherios Mylonakis

24. Bacteriophages in Health and DiseaseEdited by Paul Hyman and Stephen T. Abedon

Titles Forthcoming from CABI

Tuberculosis: Diagnosis and TreatmentEdited by Timothy McHugh

Microbial MetabolomicsEdited by Silas Villas-Bas and Katya Ruggiero

The Human Microbiota and Microbiome

Edited by Julian Marchesi

Meningitis: Cellular and Molecular BasisEdited by Myron Christodoulides

Earlier titles in the series are available from Cambridge University Press (www.cup.cam.ac.uk).
http://www.cup.cam.ac.uk/http://www.cup.cam.ac.uk/


7/297

CABI is a trading name of CAB International

CABINosworthy WayWallingfordOxfordshire, OX10 8DEUK

Tel: +44 (0)1491 832111Fax: +44 (0)1491 833508E-mail: [email protected]: www.cabi.org

CABI875 Massachusets Avenue

7th FloorCambridge, MA 02139

USA

T: +1 800 552 3083 (toll free)T: +1 (0)617 395 4051

E-mail: [email protected]

CAB International 2012. All rights reserved. No part of this publicationmay be reproduced in any form or by any means, electronically,mechanically, by photocopying, recording orotherwise, without the priorpermission of the copyright owners.

Cover image. The electron micrograph on the cover is of Salmonellaphage bound to cell wall residues forming a phage bouquet. The phagewere negatively stained with uranyl acetate. This image was prepared by

Jochen Klumpp at the Institute of Food, Nutrition and Health, Zurich,

Switzerland.

A catalogue record for this book is available from the British Library,London, UK.

Library of Congress Cataloging-in-Publication Data

Bacteriophages in health and disease / [edited by] Paul Hyman,Stephen T. Abedon.

p. ; cm. -- (Advances in molecular and cellular microbiology ; 24) Includes bibliographical references and index. ISBN 978-1-84593-984-7 (alk. paper) I. Hyman, Paul (Paul Lawrence) II. Abedon, Stephen T. III. C.A.B.International.IV. Series: Advances in molecular and cellular microbiology ; 24.

[DNLM: 1. Bacteriophages. 2. Biological Therapy--methods. QW 161]

579.26--dc23

2012007325

ISBN-13: 978 1 84593 984 7

Commissioning editor: Rachel CutsEditorial assistant: Alexandra LainsburyProduction editor: Simon Hill

Typeset by Columns Design XML, Reading.Printed and bound in the UK by MPG Books Ltd.
http://www.cabi.org/http://www.cabi.org/


8/297

vii

Contents

Contributors ixForeword xiiiPreface xv

1. Phages 1

Stephen T. Abedon*

PartI: Phages, BacterialDiseaseandNormalFlora

2. Bacteriophages as a Part of the Human Microbiome 6Andrey V. Letarov*

3. Diseases Caused by Phages 21Sarah Kuhl*, Stephen T. Abedon and Paul Hyman

4. Prophage-induced Changes in Cellular Cytochemistry and Virulence 33

Gail E. Christie, Heather E. Allison, John Kuzio, W. Michael McShan, Mathew K.Waldor and Andrew M. Kropinski*

5. The Lion and the Mouse: How Bacteriophages Create, Liberate andDecimate Bacterial Pathogens 61Heather Hendrickson*

6. Phages and Bacterial Epidemiology 76Michele L. Williams and Jeffrey T. LeJeune*

PartII: Phage-basedBiomedicalTechnology

7. Phages as Therapeutic Delivery Vehicles 86Jason Clark, Stephen T. Abedon and Paul Hyman*

8. Clinical Applications of Phage Display 101Don L. Siegel*


9/297

viii Contents

9. Phages and Their Hosts: a Web of Interactions Applications to Drug Design 119Jeroen Wagemans and Rob Lavigne*

10. Bacteriophage-based Methods of Bacterial Detection and Identification 134Christopher R. Cox*

11. Phage Detection as an Indication of Faecal Contamination 153Lawrence D. Goodridge*and Travis Steiner

PartIII: Phage-basedAntibacterialStrategies

12. Phage Translocation, Safety and Immunomodulation 168Natasza Olszowska-Zaremba*, Jan Borysowski, Krystyna Dbrowska and Andrzej Grski

13. Phage Therapy of Wounds and Related Purulent Infections 185Catherine Loc-Carrillo*, Sia Wu and James Peter Beck

14. Phage Therapy of Non-wound Infections 203Ben Burrowes and David R. Harper*

15. Phage-based Enzybiotics 217Yang Shen, Michael S. Mitchell, David M. Donovan and Daniel C. Nelson*

16. Role of Phages in the Control of Bacterial Pathogens in Food 240

Yan D. Niu, Kim Stanford, Tim A. McAllister and Todd R. Callaway*

17. Phage-therapy Best Practices 256Stephen T. Abedon*

Index 273

* Corresponding author


10/297

ix

Contributors

Stephen T. Abedon, Ph.D., Department of Microbiology, The Ohio State University, 1680University Drive, Mansfield, OH 44906, USA; [email protected]

Heather E. Allison,Ph.D., Department of Functional and Comparative Genomics, Institute ofIntegrative Biology, University of Liverpool, Liverpool L69 7ZB, UK.

James Peter Beck,M.D., Department of Orthopaedics, The University of Utah, 590 Wakara

Way, Salt Lake City, UT 84108, USA; George E. Wahlen Department of Veterans AffairsMedical Center, VA Salt Lake City Health Care System, Salt Lake City, UT 84148, USA.

Jan Borysowski, M.D., Department of Clinical Immunology, Institute of Transplantology,Medical University of Warsaw, ul. Nowogrodzka 59, 02-006 Warsaw, Poland.

Ben Burrowes,Ph.D., Ampliphi Biosciences Corporation, Colworth Science Park, Sharnbrook,Bedfordshire, MK44 1LQ,UK.

Todd R. Callaway,Ph.D., Food and Feed Safety Research Unit Agricultural Research Service/USDA, TX, USA; [email protected]

Gail E. Christie, Ph.D., Molecular Biology and Genetics, School of Medicine, VirginiaCommonwealth University, Richmond, VA 23298-0678, USA.

Jason Clark,Ph.D., Novolytics Ltd, ITAC-BIO Daresbury Laboratory, Daresbury Science andInnovation Campus, Warrington, WA4 4AD, UK; BigDNA Ltd, Wallace Building, RoslinBioCentre, Roslin, Midlothian EH25 9PP, UK.

Christopher R. Cox, Ph.D., Colorado School of Mines, Department of Chemistry andGeochemistry, Golden, CO 80401, USA; [email protected]

Krystyna Dbrowska,Ph.D., Phage Laboratory and Therapy Unit, Institute of Immunologyand Experimental Therapy, Polish Academy of Sciences, ul. Rudolfa Weigla 12, 53-114Wroclaw, Poland.

David M. Donovan,Ph.D., Animal Biosciences and Biotechnology Lab, ANRI, ARS, USDA,Bldg 230, Room 104, BARC-East 10300 Baltimore Ave, Beltsville, MD 20705, USA.

Lawrence D. Goodridge,Ph.D., Center for Meat Safety and Quality, Department of AnimalSciences, Colorado State University, Fort Collins, CO 80523, USA; [email protected]

Andrzej Grski,M.D., Ph.D., Phage Laboratory and Therapy Unit, Institute of Immunologyand Experimental Therapy, Polish Academy of Sciences, ul. Rudolfa Weigla 12, 53-114Wroclaw, Poland; Department of Clinical Immunology, Institute of Transplantology,Medical University of Warsaw, ul. Nowogrodzka 59, 02-006 Warsaw, Poland.


11/297

x Contributors

David R. Harper, Ph.D., Ampliphi Biosciences Corporation, Colworth Science Park,Sharnbrook, Bedfordshire, MK44 1LQ,UK; [email protected]

Heather Hendrickson,Ph.D., New Zealand Institute for Advanced Study, Massey University,Private Bag 102 904, North Shore Mail Centre, Auckland, New Zealand; former address:Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Rd,Oxford, OX1 3QU, UK; [email protected]

Paul Hyman, Ph.D., Department of Biology/Toxicology, Ashland University, 401 CollegeAvenue, Ashland, OH 44805, USA; [email protected]

Andrew M. Kropinski,Ph.D., Department of Molecular and Cellular Biology, University ofGuelph, Guelph, ON N1G 2W1, Canada; Public Health Agency of Canada, Laboratory forFoodborne Zoonoses, Guelph, ON N1G 3W4, Canada; [email protected]

Sarah Kuhl,M.D., Ph.D., VA Northern California Health Care System, Martinez, CA 94553,USA; Contra Costa Regional Medical Center, Martinez, CA 94553, USA; sarah.kuhl@va.

govJohn Kuzio, Ph.D., Department of Microbiology and Immunology, Queens University,

Kingston, ON K7L 3N6, Canada.Rob Lavigne, Ph.D., Division of Gene Technology, Department of Biosystems, Katholieke

Universiteit Leuven, Kasteelpark Arenberg 21 Box 2462, B-3001 Leuven, Belgium; [email protected]

Jeffrey T. LeJeune, D.V.M., Ph.D., Food Animal Health Research Program, The OhioAgricultural Research and Development Center, Wooster, OH 44691, USA; [email protected]

Andrey V. Letarov, Ph.D., Winogradsky Institute of Microbiology, Russian Academy of

Sciences, Moscow, Russia; [email protected] Loc-Carrillo, Ph.D., Department of Orthopaedics, The University of Utah, 590

Wakara Way, Salt Lake City, UT 84108, USA; George E. Wahlen Department of VeteransAffairs Medical Center, VA Salt Lake City Health Care System, Salt Lake City, UT 84148,USA; [email protected]

Tim A. McAllister,Ph.D., Agriculture and Agri-Food Canada, Lethbridge, AB, Canada.W. Michael McShan,Ph.D., Department of Pharmaceutical Sciences, College of Pharmacy, The

University of Oklahoma, Oklahoma City, OK 73126-0901, USA.Carl R. Merril,M.D., 6840 Capri Place, Bethesda, MD 20817, USA.Michael S. Mitchell,Ph.D., Institute for Bioscience and Biotechnology Research, University of

Maryland, Rockville, MD 20850, USA.Daniel C. Nelson,Ph.D., Institute for Bioscience and Biotechnology Research, University of

Maryland, Rockville, MD 20850, USA; Department of Veterinary Medicine, University ofMaryland, and Virginia-Maryland Regional College of Veterinary Medicine, College Park,MD 20742, USA; [email protected]

Yan D. Niu,Ph.D., Agriculture and Agri-Food Canada, Lethbridge, AB, Canada.Natasza Olszowska-Zaremba, M.S., Department of Clinical Immunology, Institute of

Transplantology, Medical University of Warsaw, ul. Nowogrodzka 59, 02-006 Warsaw,Poland; [email protected]

Yang Shen,Ph.D., Institute for Bioscience and Biotechnology Research, University of Maryland,

Rockville, MD 20850, USA.Don L. Siegel,Ph.D., M.D., Division of Transfusion Medicine, Department of Pathology and

Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, 510Stellar-Chance Laboratories, 422 Curie Blvd, Philadelphia, PA 19104, USA; [email protected]

Kim Stanford,Ph.D., Alberta Agriculture and Rural Development, Lethbridge, AB, Canada.Travis Steiner, B.S., Center for Meat Safety and Quality, Department of Animal Sciences,

Colorado State University, Fort Collins CO 80523, USA.


12/297

Contributors xi

Jeroen Wagemans,M.S., Division of Gene Technology, Department of Biosystems, KatholiekeUniversiteit Leuven, Kasteelpark Arenberg 21 box 2462, B-3001 Leuven, Belgium.

Mathew K. Waldor,Ph.D., Department of Medicine, Brigham and Womens Hospital, Boston,MA 02115, USA; and, Channing Laboratory and HHMI, Harvard University, Boston, MA02115, USA.

Michele L. Williams, D.V.M., Ph.D., Food Animal Health Research Program, The OhioAgricultural Research and Development Center, Wooster, OH 44691, USA.

Sia Wu,Department of Orthopaedics, The University of Utah, 590 Wakara Way, Salt LakeCity, UT 84108. USA; George E. Wahlen Department of Veterans Affairs Medical Center,VA Salt Lake City Health Care System, Salt Lake City, UT 84148, USA.


13/297

This page intentionally left blank


14/297

xiii

Foreword

It has been personally gratifying to examine the sections and chapters that make up this book,Bacteriophages in Health and Disease. The book represents real progress since the nowalmost half a century when I was first introduced to the bacteriophages at a Cold SpringHarbor Phage Course in 1965. At that time, the course concentrated on just a few phage strains,as these viruses and their bacterial hosts were seen as a potential gateway for the understandingof basic life processes. As I had recently graduated from medical school, I raised the questionas to why these bacterial viruses were not employed in antibacterial therapy. One of the coursementors answered that question by suggesting that I read Sinclair Lewis, 1925, bookArrowsmith. On reading the book, it soon became apparent that the fictionalized clinical trialsdescribed in the book failed not because of deficiencies in the antibacterial phage strains but

because the hero/antihero chose to break the double-blind code, so that he could treat all theinhabitants on the fictional island who were in the midst of a major plague epidemic. In otherwords, the actions of the fictionalized character, Arrowsmith, destroyed the capacity of theclinical trial to determine the efficacy of phage therapy, so the book really could not addresswhether the phage were useful as an antibacterial therapy or not.

At the time of the 1965 course, mouth pipeting represented the state of the art of laboratorytechnology, so I asked whether the bacteriophage could have an adverse effect on the studentsin the course. The reply, by one of the instructors, was that they could not because: they are

bacteriophage, which means bacteria eaters. I found this reply unsatisfactory, as it occurred tome that the viruses might not be restricted in their actions by a name (bacteriophage) that wehad given them. Following the phage course, I returned to my laboratory at the NIH, where,with the help of colleagues, I was able to screen for effects of phage strains on mammalian cellsin culture. We found no clear-cut gross effects with the phage strains we tested, such as cellkilling or alterations in cellular growth paterns. When transducing phages carrying the

bacterial galactose genes were used with human cells defective in certain galactose genes,however, we found evidence for restoration of the defective galactose pathway. In addition, incontrol experiments we discovered that most of our stocks of fetal calf serum used as an

ingredient in mammalian cell cultures were contaminated by bacteriophage. At first wethought that this contamination was limited to our lab, but we soon found similar contaminationin all of the fetal calf serum samples that we were able to sample from our colleagues at theNIH.

As we realized that many human vaccines are produced by mammalian cell cultures grownin fetal calf serum, we tested some vaccines that we purchased at a local drugstore and foundthat they were also contaminated by phages. The FDA confirmed our results. To permit thecontinued use of the vaccines they obtained a Presidential Executive Order permiting the sale


15/297

xiv Foreword

and use of bacteriophage-contaminated vaccines, since bacteriophage were known to bespecific for bacteria. In a subsequent presentation, at the NIH, I noted that phage are notentirely innocuous, such as in the case of diphtheria, a disease which is caused by a toxin genethat is carried into and expressed in the infecting bacteria by a phage. Lest we forget, this toxinhas enzymatic activity that results in the inactivation of a critical mammalian protein synthesisfactor and which has no known effects in bacteria.

We also initiated experiments to study phages in germ-free animals, to simplify the systemand to look for interactions between phage and animals. It was in these experiments that werecognized the immediate ability of animals to respond to phages even before antibodyproduction could be initiated. Most phages are destroyed in the liver while a small numberremained intact in the spleen (they could still infect bacteria when isolated from the spleens ofthese animals). It was these experiments that gave us ideas that we used in subsequentexperiments to select phage strains that could remain in the circulation to optimize their

capacity to serve as antibacterial therapeutic agents.While our carefully controlled studies using phages to treat animals with experimental

systemic infections were very encouraging they did not generate support from ouradministrators, who ofen cited the book, Arrowsmith, which they clearly hadnt read, astheir evidence against phage therapy. In addition, an argument that was raised repeatedly wasthat anyone who worked with phages in a lab should know that phage therapy would never

be useful because the development of phage-resistant bacterial strains is inevitable. I guessthey didnt know that this is true for most antibacterial agents including most antibiotics, buta number of lives can be saved in the meantime. Besides, with our capacity to isolate newphage strains and engineer existing ones it should be possible to overcome most phage-

resistant bacteria. Of course, we should always be aware of possible phage-mediateddetrimental effects such as the ability of phage to carry toxin genes, as in the case of diphtheria.In addition, as noted in the last chapter in this book, on Phage-therapy Best Practices (seeAbedon, Chapter 17, this volume), careful considerations also need to be used in developingphage protocols and formulations.

For the above reasons and experiences I am personally pleased to see that despite thenaysayers, efforts to use phages as therapeutic agents, both as antibacterial agents and astherapeutic delivery vehicles (see Clark et al., Chapter 7, this volume), have substantiallyprogressed. Studies of phages in the human microbiome may facilitate the development ofnew therapeutic phage strains and therapeutic strategies. Most importantly, careful clinical

trials are being designed and conducted as described in several chapters to help us avoidfalling into the trap that undid Arrowsmith. In time, truth will overcome the shadow of thatfiction.

Carl R. Merril


16/297

xv

Preface

If all things have to have a beginning, then the beginning of this monograph, arguably, can betraced to the 1980s in the laboratory of Harris Bernstein (Department of Microbiology andImmunology, University of Arizona). That was the lab that both of us joined to do our Ph.D.research and where we met. Though we worked on separate projects, both involved phagesand this led to many discussions. Thus the team began. Afer graduating we went our separateways. While S.T.A. dabbled in what eventually would be all things phage ecological, P.H.followed a much more molecular, then medical, then applied route, but ultimately returned tophage biology and what would be a series of collaboration with S.T.A., starting in 2001. Thestory of how this monograph came to be is slightly more complicated, however.

S.T.A. at the time at the beginning of his career at the Ohio State University, in 1995 founded what would become the Bacteriophage Ecology Group. His management of theassociated web site (phage.org, though see also archaealviruses.org) began to lead to chapterinvitations in the early 2000s, resulting in part in his contributing to the editing of RichCalendars The Bacteriophages 2/e, which was published in 2006. Very soon afer S.T.A.scontribution to that monograph ended he was invited to edit a monograph on phage ecology,which would become the 2008 Bacteriophage Ecology(Cambridge University Press), included inwhich was a chapter we coauthored.

Meanwhile, speculators brought down the worlds economy, hiting academic publishersquite hard. The series that Bacteriophage Ecologyhad been a part of,Advances in Molecular andCellular Microbiology, was sold to CABI Press (cabi.org), a not-for-profit international organizationthat improves peoples lives by providing information and applying scientific expertise to solve problems

in agriculture and the environment. It was 100% their idea to do the current monograph and,based on his experience with the Bacteriophage Ecologyvolume, S.T.A. was recruited to edit it.S.T.A., though, was busy at the time, pulling together an edited volume on phage therapy (seeCurrent Pharmaceutical Biotechnology, volume 11, issue 1), so recruited P.H. to join in on yetanother collaboration. We developed a formal proposal including recruiting the authors of themany chapters found herein and the result is this monograph.

Bacteriophages in Health and Disease, an early working title that stuck, is an effort to providean introduction to the breadth of roles that phages play or can play in our everyday lives. Theycan serve as causes of disease, treatments of disease, indicators of the potential for disease,preventers of disease, and even contribute in various ways to the evolution of bacterialpathogens. To capture this variety of phage roles in human conditions, both natural andapplied, we have divided the text into three parts. Weve also provided a brief introduction tovarious concepts and terminology associated with phages (chapter 1); for a glossary coveringthese basics of phage biology, and much more, see phage.org/terms.


17/297

xvi Preface

Part I considers the role of phages in the natural state. That is, where phages are, how theycontribute directly to disease, the underlying mechanism by which phages do this, and then,especially, how they can contribute indirectly to disease, that is, to pathogen evolution. Thethree basic themes are ones of phage presence, phage genes, and phage-mediated transductionof bacterial DNA between bacteria, though a common thread is that they touch upon, invarious ways, issues of lysogeny, which is the long-term incorporation of phages as prophagesinto the genomes of their bacterial hosts. These issues are covered by chapters 2 through 6.

Part II, chapters 7 through 11, considers various phage-based technologies other than theuse of whole phages to combat bacterial infections (i.e., besides phage therapy). This includesin particular the use of both modified and disembodied phage parts. Phages thus can serve ascarriers and delivery vehicles of especially DNA but also of other chemicals, including servingas vectors for either gene therapy or DNA vaccines. The potential of phages to serve in thesefunctions stems greatly from their relative safety as well as their genetic malleability, such as

one sees for example in phage display technologies. Phage antibacterial properties too can beused as a means of discovering novel targets for action by traditional small-moleculeantibacterial chemotherapeutics. In addition, phages can be employed in bacterial detectionand identification including as indicators of faecal contamination.

Phages typically in a relatively unaltered state can be employed directly as antibacterials,that is, in phage therapy, a theme covered by Part III and chapters 12 through 17. Here weprovide a chapter introducing various phage properties as medicinals, including the relativesafety associated with phage application to bodies. Phage therapy of humans is then consideredin two chapters, which we have divided up roughly into treatment of wounds and non-woundinfections. Certain phage parts can be used, on their own, as antibacterial drugs, most notably

phage lysins which are covered in a subsequent chapter. The potential phage role in foodsafety is considered, and we then end the monograph with a chapter targeted to would-bephage therapy experimentalists, one that considers, in light of phage properties, especiallyhow phage therapy protocols may be developed in terms of the use of animal models of

bacterial disease.If your primary interest is in learning about the myriad roles that phages can and do play in

human health and disease, read all of the chapters perhaps up to this last one. If you alreadyknow all there is to know about phages such that much of this book is superfluous, you maystill want to at least read this last chapter! In any case, we see this monograph as appealing tothe needs of two not-so-disparate groups, those with a primary interest in phage biology,

though from an especially applied perspective, and those who are curious as to how phagebiology can impact the practice of medicine. In particular, we provide a snap shot of the currentstate of the field of phage biology, applied as well as basic, concentrating on the roles thatphages can play in both health and disease.

We thank our editors at CAB International, Rachel Cuts and Alex Lainsbury, for theirsupport and help bringing this volume to fruition and dedicate this volume to our common

but nonetheless not-so-common mentor, Harris Bernstein.


18/297

CAB International 2012. Bacteriophages in Health and Disease(eds P. Hyman and S.T. Abedon) 1

1 Phages

Stephen T. Abedon11Department of Microbiology, The Ohio State University

In a phrase such as Bacteriophages in Healthand Disease, the least familiar term to anyonewill be that of bacteriophages. It is thepurpose of this introductory chapter to makesure that this term is less unfamiliar even to

those working in thefi

eld so that greateremphasis may be made, in this monograph,on issues of health and disease. In thischapter, I thus will provide a brief and alsonot terribly molecular overview of the virusesof bacteria. Greater detail as applicable toparticular topics will be found in appropriatechapters. For additional discussion of con-cepts, see www.phage.org/terms/, as well asthe resources list presented by www.ISVM.

org.For general references, see www.phage.org/terms/phages.htmland www.phage.org/terms/phage_history.html.

Bacteriophages, or phages for short, areviruses that infect members of what is knownas the domain Bacteria, which are thecommon prokaryotic organisms associatedwith, for example, the human body. To helporient readers towards understanding therole that phages can play as part of the humanmicrobiome (see Letarov, Chapter 2, thisvolume), in health as well as disease (seeKuhl et al., Chapter 3, this volume), in thischapter I provide an overview of basic phageproperties. These include, in particular, what

phage virions look like, the basic character-istics of phage infections and, because it canplay such a large role in bacterial evolution,the potential for phages to move DNA

between bacteria, a process known as

transduction. I begin, however, with a briefglimpse at the history of phage research.

History

Although a number of publications in the late19th and early 20th century may hint at theobservation of phage-like phenomena, thediscovery of phages is unambiguously traced

to the work of Twort and, independently, thatof dHrelle (see the first volume of thejournal, Bacteriophage, 2011, particularly issues1 and 3, for discussion and references). Earlyphage work by necessity addressed issues of

basic phage biology, although these effortswere based on both primitive techniques andminimal understanding of just what thephage phenomenon actually entailed. Thepromise of phage therapy, with the use ofphages to combat bacterial infections inparticular, was an important driver of earlyphage research (see Burrowes and Harper,Chapter 14, this volume, for additionalhistorical consideration of this issue).

S.T. Abedon
http://www.phage.org/terms/http://www.isvm.org/http://www.isvm.org/http://www.phage.org/terms/phages.htmlhttp://www.phage.org/terms/phages.htmlhttp://www.phage.org/terms/phage_history.htmlhttp://www.phage.org/terms/phage_history.htmlhttp://www.phage.org/terms/phage_history.htmlhttp://www.phage.org/terms/phage_history.htmlhttp://www.isvm.org/http://www.isvm.org/http://www.phage.org/terms/phages.htmlhttp://www.phage.org/terms/phages.htmlhttp://www.phage.org/terms/


19/297

2 S.T. Abedon

It was in the late 1930s and early 1940sthat phage research began to make whattoday we view as its most significant impacton basic biological research. These efforts areassociated in large part with the physicist-by-training Max Delbrck and associatedmembers of what became known as thePhage Group. It was within this context thatthe field of molecular genetics and the relateddiscipline of molecular biology were forged.

James Watson, for example, was a member ofthe Phage Group, and Francis Crick too

became an important contributor to phage

research following their co-discovery of thestructure of DNA.

By contributing to the development oftechniques that ultimately would greatlysimplify the study of the biology of organismsfar more complicated than either phages ortheir bacterial hosts, in a sense phage researchset the stage for its own eclipse, perhapsparticularly in terms of funding opportunities.Changes in this situation involved an

increased focus on areas other than purelythe molecular aspects of phages: considerationof aquatic phage ecology, growing appre-ciation of the role of phages in horizontalgene transfer and increasing emphasis on theapplication of phage-based biotechnologies.The later includes in particular techniquesknown as phage display (see Siegel, Chapter8, this volume), the use of phages in bacterialdetection as well as identification (see

Williams and LeJeune, Cox, and Goodridgeand Steiner, Chapters 6, 10 and 11, thisvolume) and, of course, phage therapy (seeOlszowska-Zaremba et al., Loc-Carrillo et al.,Burrowes and Harper, Shen et al., Niu et al.and Abedon, Chapters 1217, this volume).All of these issues are considered in thismonograph, particularly in the last two parts.The association of phages with bacterialdisease is a primary emphasis of Part I (see

Kuhl et al., Christie et al. and Hendrickson,Chapters 35, this volume).

Types of phages

Virus particles consist essentially of twocomponents, what is on the inside and whatis on the outside. What is on the inside

primarily is nucleic acid and what is on theoutside is responsible for transporting thatnucleic acid between cells. The nucleic acidcomponent varies substantially betweenphage types, both in terms of size andstructure, and can consist of either RNAor DNA, but not both, with DNA beingmore common. It can be single-stranded ordouble-stranded, although for most phagesit is double-stranded. It also can be multi-segmented, particularly tripartite (e.g.Pseudomonas phage 6), although the vastmajority of phage genomes are monopartite.

The outside portion of virions consists ofa combination of the capsid, which surroundsthe nucleic acid, and various appendages,most of which are involved in virionadsorption to bacteria. While some phageshave capsids that contain lipids (again suchas phage 6), for most phages capsids consistsolely of multiple units of proteins known ascapsomeres. For simple phage virions, cap-sids are either helical/filamentous or ico-

sahedral, which in both cases surroundsingle-stranded nucleic acids (usually DNA,

but not always). These phages fall into thephage families Leviviridae, Microviridae andInoviridae. More specifically, these are single-stranded RNA icosahedral, single-strandedDNA icosahedral and ssDNA filamentousphages (e.g. phages MS2, X174 and M13,respectively). A very small number of phageshave dsRNA genomes (such as 6).

The vast majority of phages by contrastpossess complex rather than simple virionmorphologies; dsDNA, monopartite genomes;and protein capsids lacking lipids. Thedefining feature of these phages all membersof phage order Caudovirales are their tails,however. Tails are multi-protein appendagesinvolved in virion adsorption (see the coverand prelims for illustration). Members of thephage family Podoviridae, such as coliphage

T7, possess short, non-contractile tails. Thetails of members of the family Siphoviridae,including coliphage , can be quite long andare also non-contractile. Finally, the tails ofmembers of the familyMyoviridae, coliphageT4 being the most familiar, are approximatelyintermediate in length while possessing anability to contract in the course of virionadsorption to bacteria. All tailed phages are


20/297

Phages 3

lytic and some are also temperate. Discussionof basic phage morphologies in greater detailcan be found in a number of reviews thathave been published by Hans-WolfgangAckermann.

Virus-like particles

The concept of virus-like particle, or VLP,has at least two distinct meanings as used inthis monograph. First are entities that appearto be virus-like as viewed in an electron

micrograph (see Letarov, Chapter 2, thisvolume). Secondly, the term is used todescribe phages that lack genomes and whichtherefore are virus-like rather than actuallyviruses (see Clark et al., Chapter 7, thisvolume). The term ghost has also beenemployed in the phage literature to describethese later entities, including those generatedfrom particles that initially possess genomes.

Types of infection

A phage infection begins with virionatachment to a potential host cell, theculmination of a multi-step process known asadsorption. It then proceeds through thetranslocation of phage nucleic acid into thecell variously described as both ejection andinjection, as well as uptake but begins in

earnest only once the phage genome hasmade its way into the bacterial cytoplasm.The results of phage adsorption can varydepending on the characteristics of the phage,the bacterium and circumstances. At its most

basic level, the phage may either live or die(that is, produce or not produce replicativeproducts afer infecting a host cell), and thesame is true for the infected bacterium.Furthermore, and depending on infection

type, all four combinations of living versusdying are possible. See Cox (Chapter 10, thisvolume) for additional discussion of phageinfections.

In an abortive infection, both the infectingphage and infected bacterium die. In a lyticinfection, the infecting phage lives, producingphage virions, while the infected bacterium

both dies and is lysed. In various other

circumstances, the infecting phage can die butnot the infected bacterium, which is seenparticularly when bacteria carry restrictionendonucleases, although also when they carrythe CRISPR/cassystems. Lastly, under certaincircumstances both phage and bacterium canlive. While one means by which this co-survival can occur is seen with the chronicinfections of members of the phage familyInoviridae, most commonly simultaneoussurvival of both infecting phage and infected

bacterium is seen with phage exhibition oflysogeny (below). Productive infections, those

quickly leading to the release of infectiousviral progeny, normally occur afer infection

by both lytic and chronic phages. Overall, themajority of phages are lytic phages.

Once the virion genome of a lytic virusgets into the bacterial cytoplasm, the phagegenes are expressed. This gene expressionhas the effect of taking over the bacterialmetabolism so that phage virion particles areproduced, the proteins of which are encoded

by additional phage genes (see Wagemansand Lavigne, Chapter 9, this volume, forgreater consideration of what goes on, at amolecular level, within a phage-infected

bacterium). The products of yet more phagegenes contribute, at the end of what is knownas the phage latent period, to the destructionof the bacterial cell envelope so that phageprogeny can leak into the extracellularenvironment. For tailed phages, this lysis

process involves coordinated action by atleast two phage gene products, the holin andendolysin or lysin proteins. In particular, theholin protein is responsible for controllingthe timing of host lysis, stimulating the actionof the lysin, and, as a by-product, shutingdown infection and therefore host meta-

bolism. Lysin is the enzyme that is responsiblefor degradation of the bacterial cell wall. SeeShen et al. (Chapter 15, this volume) for

additional discussion of phage lysins.

Lysogeny

The other major state that involves bothsuccessful phage infection and bacterialsurvival is known as the lysogenic cycle, aphenomenon, as noted in the preface, that


21/297

4 S.T. Abedon

serves to unite the chapters found in Part I ofthis volume. All phages that can displaylysogeny are described as temperate, and themajority of temperate phages are tailed(although a small number are insteadfilamentous, i.e. members of the phage familyInoviridae; see Christie et al., Chapter 4, thisvolume, for discussion of a prominentmember of the later, Vibrio cholerae phageCTX). Lysogenic cycles are characterized bytwo features. First, the phage genome, nowcalled a prophage, is replicated sufficientlyrapidly within infected bacteria that daughter

bacteria, following binaryfission, each inheritat least one prophage copy. Secondly, theinfections are not productive, that is, novirions are produced.

The prophage can exist either integratedinto the bacterial chromosome or as aplasmid. Integration is typically accomplished

by the action of a phage protein termed anintegrase. Integrases generally bind to aspecific site on the bacterial genome and a

corresponding, partially homologous site inthe phage genome. The result is site-specificrecombination to integrate the phage genomeinto the bacterial genome.

A bacterium that can undergo a lysogeniccycle is described as a lysogen, the process oflysogen formation is called lysogenizationand the conversion of a lysogenic infectioninto a productive one (typically a lytic one) iscalled induction. In some lysogens, theprophage only expresses genes whose pro-teins are needed to prevent induction or,instead, trigger induction upon receiving anappropriate signal. Expression of the re-pressor proteins in particular prevents suchinduction but also has the effect of blockingthe infection of lysogens by similar (known ashomoimmune) phages, a process known assuperinfection immunity. Other prophagesalso express genes that can alter the phenotypeof the lysogenic bacterium, a process called

lysogenic conversion (see Kuhl et al. andChristie et al., Chapters 3 and 4, this volume).

Common terms

To avoid ambiguity when consideringdifferent types of phage infection, I provideexplicit definitions of the following terms:

Temperate description of a phage that iscapable of displaying a lysogenic cycle; alltemperate phages and indeed all phagesalso display productive cycles at somepoint in their life cycles. Note thatlysogenic phage is not a synonym fortemperate phage but instead is amisnomer (bacteria can be lysogenic,while phages are temperate).

Lytic description of a phage that lyses itshost in the course of productive infection;note that most temperate as well as mostnon-temperate phages are lytic phages.

Chronic description of a phage that doesnot lyse its host in the course of produc-tive infection; these phages are released bycrossing relatively intact bacterial cellenvelopes.

Obligately lytic description of a func-tional phage that is not capable of display-ing either lysogenic or chronic infections.

Professionally lytic description of a phagethat is both obligately lytic and not recently

descended from temperate phages. Virulent a common synonym of obli-

gately lytic, although it can also describethe potential for a phage isolate to bring apopulation of target bacteria undercontrol, particularly through infection thatis followed by bacterial lysis and associ-ated phage population growth.

Phage titre (or just titre) a measure of thenumber of phages per millilitre in a liquid

stock, and typically a measure of viablephages as determined via plaque countsrather than of virion particles as deter-mined by various forms of microscopy.

Plaque a visible clearing on a bacteriallawn, growing in or on agar found in aPetri dish, which is the result of localizedinhibition of bacterial growth such as can

be mediated by phages.

To this list one might add additionalterms such as PFU standing for plaque-forming unit (just as CFU stands for colony-forming unit in bacteria). Note that thefilamentous coliphage M13 is a prominentchronically infecting phage. See Clark et al.and Siegel (Chapters 7 and 8, this volume) fordiscussion of the utility especially of thisphage to biotechnology and also Goodridge


22/297

Phages 5

and Steiner (Chapter 11, this volume)for further discussion of chronically infectingphages (i.e. members of the family Inoviridae).

Transduction

Although covered in greater detail by Kuhl etal., Christie et al.and Hendrickson (Chapters35, this volume), I provide here a briefintroduction to the idea of phage-mediatedmovement of bacterial DNA. Any time thatphages are identified within an environment,

this means that a potential exists for phage-mediated horizontal gene transfer of bacterialDNA between bacteria. Generally, we candifferentiate phage-mediated horizontal genetransfer into four categories: generalizedtransduction, specialized transduction, phagemorons and a category of gene movementthat technically is not transduction at all butinstead is the movement of phage genes bytemperate phages. I will discuss these briefly.

Generalized transduction is the move-ment of bacterial DNA that has been packedinto phage capsids without accompanyingphage DNA. Generalized transduction hasthe property of being able to transfer largesegments of DNA, that is, many tens ofthousands of base pairs, such as thoseassociated with bacterial pathogenicityislands (see Hendrickson, Chapter 5, thisvolume). Specialized transduction, by con-

trast and as narrowly defined, is the in-corporation of bacterial genes that are foundadjacent to prophage integration sites into the

bacterial chromosome. Typically, onlyrelatively few genes are transferred. As phagegenomes tend to be limited in their size byconstraints on their packaging into phagecapsids (heads), this transduction of evenrelatively few bacterial genes can result inimpairment of phage functioning. This type

of specialized transduction is by definitionlimited to temperate phages that integratetheir genomes into the host chromosome inthe course of infection.

Specialized transduction, as morebroadly defined, involves simply the inte-

gration of bacterial genes into phage genomes.An aspect of this form of transduction has

become associated with the term moron,meaning more DNA. Morons are typicallyconsidered to be bacterial genes that have

become incorporated into phage genomes viaprocesses of illegitimate recombination andwhich do not encode a mechanism for theirremoval. Lastly are seemingly legitimatephage genes, ones associated with temperatephages that modify the phenotypes oflysogens. Included under this heading are anumber of virulence factor genes. The dif-

ference between a moron and these lysogenicconvertinggenes is one of degree of integrationinto the phages genetic structure, withmorons more evidently newly acquired bythe transferring phage. See Christie et al.(Chapter 4, this volume) for consideration oflysogenic conversion in general as it appliesto the encoding of bacterial virulence factors

by phages.

Conclusion

It is important to keep in mind that phagesare highly diverse. This diversity is seen interms of genotype, phenotype, the proteinsproduced and interactions with hosts.Phages also vary in terms of their host range(which bacteria they infect), their trans-

ducing ability, their virion morphology, andalso with respect to their general infectioncharacteristics. Thus, whenever mention ismade of a specific phage, it should be kept inmind that substantial effort may be necessaryto elucidate the specific properties, especiallyphenotypic, that are associated with thatphage or phagehost combination. On theother hand, various generalizations mayalso be made. This monograph will present a

mix of both generalizations and specifics inconsidering phage presence in bodieswithout disease, their role in both bacterialdisease and pathogen evolution, and howphages can be employed to combat infectiousdiseases, in particular of humans.


23/297

CAB International 2012. Bacteriophages in Health and Disease6 (eds P. Hyman and S.T. Abedon)

Human beings, our problems, our triumphsand our everyday lives, have through theages been central to philosophy, literature,arts and religion. This anthropocentriccomprehension of the Universe has led to

numerous, dramatic and sometimes quitecontroversial considerations in the naturalsciences as well. These include, for example,resistance to acceptation of the heliocentricmodel of the Universe (16th century), thediffi culties that the theory of evolution met

before gaining wide acceptance (19th century)and, more recently, the numerous problems ranging from underestimation to over-estimation in interpreting the mental

capacities of different animal species, rangingfrom bees to dogs, dolphins, or even to our-selves (Alcock, 2001).

The anthropocentric view has led to aperception by many modern biologists,whose work is not directly aimed at species-specific aspects of animal biology, to regard atfirst glance the human animal as a validmodel for mammals in general or vice versa.Indeed, almost everyone, scientist as well as

non-scientist, is biased towards the idea thathumans or indeed any animal or plant consists exclusively of cells, which can bedescribed unambiguously as animal or plantcells. Such a perspective, however, is not

even close to accurate, as each macroscopicorganism hosts many types of microorganismsas part of its normal state: the microbiome ofthe organism. In this chapter, I emphasizeessentially this aspect of the human body

that traditionally one wouldn't regard as thehuman body. Furthermore, my emphasis willnot even be on all of the microbiome butinstead on the viruses that infect most ofthose non-human cells, the bacteriophages orphages. See Loc-Carrillo et al. (Chapter 13,this volume) for a discussion focusing in partinstead on the bacterial aspects of the humannormal flora.

This discussion serves as the beginning of

an exploration of copious connectionsbetween humanity, the most powerful oforganisms, and phages, which are the mostnumerous. Indeed, in this monograph thenumerous roles that phages can play inhealthy humans, in contributing to humandisease (see Kuhl et al., Christie et al. andHendrickson, Chapters 35, this volume), inthe treatment of human disease (see Siegel,Wagemans and Lavigne, Olszowska-Zaremba

et al., Loc-Carrillo et al., Burrowes and Harper,Shen et al.and Abedon, Chapters 8, 9, 1215and 17, this volume) and in the prevention ofhuman disease (see Clark et al., Cox,Goodridge and Steiner, and Niu et al.,

2 Bacteriophages as a Part of the HumanMicrobiome

Andrey V. Letarov11Winogradsky Institute of Microbiology, Russian Academy of Sciences

A.V. Letarov


24/297

Bacteriophages as a Part of the Human Microbiome 7

Chapters 7, 10, 11 and 16; see also Williamsand LeJeune, Chapter 6, this volume) allwill be discussed. First, and the emphasis ofthis chapter, will be the normal state of affairs,which is the phage contribution to thecollection of cells and microorganisms thattogether make up the human body; that is,ourselves and our microbiome, especially asseen during the normal, healthy state.

Phages and the Human Microbiome:an Overview

From the point of view of the microorganism,the human or animal body is merely a systemof connected colonizable ecotopes, that is,ecologically distinct features of environments.Each ecotope varies in terms of host factors(the colon versus the lung, for example), interms of its history (dictating in part whatorganisms can be present), in terms ofinteractions between the microorganisms

that are there and as a consequence of feed-back mechanisms between microorganismsand host. The result is a high potential forvariation in microorganism types, includingvirus types, going from organ to organ, tissueto tissue, individual to individual, and alsoover time both within individuals andthrough the generations (both humans andour ancestors). Adding further to thesecomplications is the potential for at least

some microorganisms to move betweenspecies. For phages, we can also add anability to modify their bacterial hosts (seeChristie et al. and Hendrickson (Chapters 4and 5, this volume), including in terms ofphage susceptibility (see Williams andLeJeune, Chapter 6, this volume).

The result of all of these factors is thepotential for a high degree of individualvariation in the composition of, for example,

intestinal bacterial populations in humans, aswas reported by Costello et al. (2009). Theindividuality of the associated phagepopulations can be even higher than that of

bacteria (Reyes et al., 2010; Caporaso et al.,2011). Consistently, divergence of both

bacterial and phage communities at the samebody sites but in different individuals werefound to be higher than in the same

individuals over time (Costello et al., 2009;Reyes et al., 2010; Caporaso et al., 2011). Onthe other hand, communities of gut bacteriawere much more related in closely connectedpeople such as monozygotic twins and theirmothers, although this consistency appliesless so to the phage component, whichappears to be highly individual but none theless quite stable over time (Reyes et al., 2010).

These data highlight the effect of amplifi-cation of slight differences in physiology (orconditions) by the complex events of micro-

bial interactions as well as exposure history,

resulting in significantly different states ofmicrobial systems in different subjects, andindicate that phage communities may bemore sensitive to colonization history than

bacterial microflora (Costello et al., 2009; seealso Nemergut et al., 2011). The data on thephage prevalence and activity in differentsites of the human body, analysed in thischapter, thus may not represent any para-digmal model for phages in animal-associated

systems but could instead reflect particularfeatures of our species, and maybe even ofindividuals or subpopulations included inthe studies cited. By comparison, we can con-sider the current understanding of bac-teriophage ecology in other animal-associatedsystems, as recently reviewed elsewhere(Letarov and Kulikov, 2009).

Despite early work with phages wherethere was a strong emphasis on the impact of

phages on the human antibacterial immunity,along with a strong medical orientation of

basic research in microbiology and virologyin recent decades, the phage ecology of thehuman body was substantially neglecteduntil recently. Even now it is poorly under-stood if compared, for example, with phageecology in aquatic systems. The firstobservations of human-associated bacterio-phages, however, were published by one of

the discoverers of these viruses, FelixdHrelle, who demonstrated phages lysingenterobacteria in faeces (dHrelle, 1921).Nevertheless, and despite two periods ofhigh interest in phage therapy in the 1920sto 1930s and from the 1990s until now(Abedon, 2011) these endogenous phageshave been subject to systemic research onlyover the past few years. Thus, our knowledge


25/297

8 A.V. Letarov

of bacteriophage impact on microbial ecologyand on macro-host homeostasis in humans(and, more widely, in animals) is highlymosaic, with many important parts of thispuzzle still missing.

Here, I focus exclusively on the availabledata on phages in human symbioticmicrobiota, presenting a limited subset ofdata for other species only for comparison. Iconsider the phage component followingstandard biogeographical principles, that is,

based on body site. As has been evident sinceearly electron-microscopy based studies of

non-cultured viral communities from humanfaeces and other body sites, and later con-firmed by metagenomic analysis (reviewed

below), phages in fact appear to dominate thehuman body virome, being much moreprevalent than any eukaryotic viruses.

Skin

The skin is the largest organ of the humanbody and is a reservoir for multiple andhighly diverse habitats for symbiotic bacterialcommunities. The diversity of bacteria on theskin is comparable to that of the gut, althoughthe total microbial biomass is much lower.The variation in species composition of theskin microflora of individual people appearsto be highest among all the body sites thathave been analysed (Costello et al., 2009). The

data on skin-associated bacteriophages,however, is almost non-existent for humans,as well as for other animals.

Alternatively, phages potentially associ-ated with skin microflora have been isolatedfrom downstream habitats such as Staphy-lococcus aureusphages isolated from sewage.Generally, such occurrence of phages in-fecting human-associated (and animal-associated) hosts from sewage and other

downstream sources is a well-knownphenomenon. Moreover, this kind of samp-ling is widely used in many works on theisolation of phages of pathogenic bacteria forphage therapy or diagnostic applications (seeGill and Hyman, 2010, for review). There isalways a question, however, of whether thesephages come from human or animal micro-

biomes or instead are indigenous for specific

systems (waste-water drains, for example). Innumerous cases, it has been difficult to isolatephages directly from animals, but they areeasily found in farm waste water. Further-more, T-even-related bacteriophages areseldom isolated from healthy humans oranimals (see below), but they can frequently

be isolated from sewage samples (where T4and T6 were originally found; Abedon, 2000).The details of the life strategy that make aphage beter adapted to macro-host-associated or downstream habitats are notyet clear.

Respiratory tract

The respiratory tract of healthy subjects isbelieved to be poorly colonized by anymicroflora, thus suggesting that no stable

bacteriophage population should be present.In good agreement with this conclusion arethe results of a recent metagenomic study of

the viruses contained in the sputum samplesfrom five cystic fibrosis (CF) patients and fivenon-CF subjects (Willner et al., 2009). Thediversity of both viral communities was at thelevel of about 175 viral genotypes in bothmetagenomes. The metagenomes of CFpatients shared significant similarity and itwas possible therefore to evaluate the core ofthe CF-specific phage community (Willner etal., 2009; Willner and Furlan, 2010). These

phages were proposed to play significantroles in the pathological microbial ecology inCF patients lungs, encoding numerous

bacterial virulence factors including adhesins,biofilm-formation genes and quorum-sensinggenes (Willner and Furlan, 2010; see Christieet al., Chapter 4, this volume, for a generaloverview of phage encoding of bacterialvirulence factors). At the same time, thephages present in the lungs of non-CF

persons were found to vary considerably andmost likely represented the random samplesof environmental viruses. Two non-CFsputum metagenomes collected from peoplesharing the environment with CF patientswere more similar to the CF profile than theother three non-CF samples, and thisobservation was in good agreement with theassumption that the viral community of the


26/297


respiratory tract in healthy people is repre-sented by transitionally captured particlesoriginating from the air.

Gastrointestinal tract

The human gastrointestinal tract includes themouth cavity, throat, oesophagus, stomachand gut, the later comprising the smallintestine, large intestine and rectum. Due tothe bactericidal activity of high acidity alongwith the proteolytic enzymes found in gastric

juices, the stomach is poorly colonized bymicroorganisms and contains only about 103

bacterial cells ml1. In the small intestine, thebacterial population is also limited by therapid peristalsis combined with the action ofthe bile and pancreatic secretions and reachesabout 105 cells ml1 (up to 108 ml1 in theileum; Baranovsky and Kondrashina, 2008).The main reservoir of the gut microbial

biomass is the large intestine, harbouring upto a total of 1014bacterial cells per individual(Savage, 1977) and corresponding tohundreds of grams of bacterial biomass.Below, I consider the phage populations inthe oral cavity and in the large intestine thatrepresent the major ecotopes of the humangastrointestinal tract colonized by bacteria.

Oral cavity and pharynx

Studies assessing phage presence in thehuman oral cavity and pharynx are notextensive. Direct electron microscopic ob-servations indicate the presence of a largenumber of VLPs in some but not in allsamples of the dental plaque material (Bradyet al., 1977). The presence of VLPs in the

matrix of complex multi-species biofilmsseems logical, as the data suggest that insome bacterial species, such as Pseudomonasaeruginosa, induction of the prophage andproduction of viral particles may be a normalprogrammed stage of the biofilm develop-ment (Rice et al., 2009).

These observations of high concentrationsof VLPs in dental plaque to my knowledge

have not been corroborated. Nevertheless,the existence of bacteriophages in dentalplaque material was recently confirmed by asmall viral metagenomic survey of the dentalplaque uncultured viral community collectedfrom a single individual (Al-Jarbou, 2012). Ofthe 80 sequences obtained, only a total of 21phage-related sequences were discovered.Despite the paucity of the data set obtained inthis study, it indicates clearly the lowcomplexity of the dental plaque viral com-munity.

Another recently published metagenomic

study of viral communities of pooled oro-pharyngeal swabs from 19 healthy individuals(Willner et al., 2011) also indicated that theyare formed almost exclusively by phages amix of virulent and temperate phages. Thesole exception was the eukaryotic EpsteinBarr virus. Of interest, the complete genomesof three phages were assembled and amongthem was Escherichia coli phage T3. Thenatural reservoir for this virus was never

identified, although the data of Willner et al.(2011) indicated that it may be a normalinhabitant of the human pharynx. Theauthors also detected a number of phages ofPropionobacterium, Lactobacillus and otherlactic acid bacteria, streptococci and otherhosts as well as non-host-attributed

bacteriophage-related sequences. The SM1bacteriophages encoding the platelet-bindingfactors of Streptococcus mitis were also found,

suggesting possible involvement of theoropharyngeal microbiota in development ofendocarditis. The total abundance of the

bacteriophages was not directly determinedin this study but would appear not to be veryhigh, as the authors applied an amplificationprocedure prior to sequencing. The diversitywas estimated to be about 236 different viralgenomes from the 19 pooled samples.

A much more profound metagenomic

study of human saliva viromes has recentlybeen published by Pride et al. (2011). Theauthors collected saliva samples from severalhealthy subjects at different time points. Theviruses were quantified by epifluorescencemicroscopy and the morphology of VLPswas investigated by electron microscopy.Profound sequencing of the virome meta-genomes was performed accompanied by


27/297

10 A.V. Letarov

sequencing of bacterial 16S rRNA genelibraries from the same samples. This studyconfirmed the presence of a robust indigenousphage population dominating the salivamicrobiome, comprising up to 108VLPs ml1of saliva. The composition of the viromes indifferent subjects was highly individual, buttwo viromes collected from members of thesame household showed much greaterrelatedness than the others, indicating theimpact of externally acquired phages on thissystem (which makes a striking contrast tothe data of Reyes et al., 2010, which demon-

strated that faecal viromes are markedlydistinct in closely connected subjects such asmonozygotic twins and their mothers see

below). The composition of the viromeswithin subjects at different time points washighly related but nevertheless exhibitedsignificant changes over time, especially at6090-day time intervals. The saliva viruspopulation thus appears to be quite denseand at the same time dynamic. The presence

of phage integrases in 10% of all contigs andthe identification of virus contigs matchingdistinct regions in sequenced bacterialgenomes indicated a high prevalence oflysogeny. At the present time, however, thedata are insuffi cient to estimate the relativesignificance of virulent and temperate phages(or of phage multiplication in the lytic cycleversus lysogen induction in this system).Overall, the substantial numbers of phage

particles detected by Brady et al. (1977), Al-Jarbou (2012) and Pride et al. (2011) in oralcavity-derived samples is inconsistent withpublished negative results of atempts toisolate phages from the human oral cavityactive against normal indigenous bacteriapresent in that site (Hitch et al., 2004); theseauthors were, however, able to isolate a phageof a non-oral pathogen, Proteus mirabilis.

A very limited number of successful

phage isolations from the oral cavity havebeen described. For example, Tylenda et al.(1985) reported the isolation of actinobacterialphages from ten out of 336 samples of humandental plaque material. The isolation ofVeillonella bacteriophage from the oralsamples was also published by Hiroki et al.(1976). More recently, phages for Enterococcus

faecalis were cultured from human saliva

(Bachrach et al., 2003), but it should be notedthat in this later study the authors failed todetect bacteriophages for a number of otherspecies of bacteria of the normal oralmicrobiota. Temperate E. faecalis phages werealso successfully induced from root canalisolates of this bacterium (Stevens et al., 2009).

At the moment, it is difficult to build acomprehensive concept of phage ecology inthe oral cavity. Summarizing the data, onecould conclude that the bacterial communityof the human oral cavity and probably also ofthe pharynx are not substantially impacted

by phages. The metagenomic data of Pride etal. (2011), however, suggests that the viralcommunity in this site is dynamic and is ableto incorporate externally acquired phagestrains and to maintain quite an elevateddensity. No coherent explanation of thesecontradictions was suggested. Given thelimited amount of data on oral bacteriophagesfrom non-human species that has beenpublished, it is difficult to speculate on howthese human traits compare with the phageecology of the oral cavity in mammals.

The gut

The gut, especially the lower intestine, isbelieved to be the main habitat of the human-associated microbiota including bacterio-phages. Being the natural habitat of the

worlds best-studied bacterium E. coli theintestinal microbial system of humans andanimals has served as a subject of multiplestudies of coliphage ecology (see below),starting from dHrelles pioneering work in1921.

Total viral counts

Transmission electron microscopy-based

studies have repeatedly demonstrated a highabundance of VLPs in the faeces and intestinalcontents of humans (Flewet et al., 1974), aswell as in other species. The later includecatle and sheep (Paynter et al., 1969;Hoogenraad et al., 1967), the rumen ofreindeer (Tarakanov, 1971), the forestomachsof Australian marsupials (Hoogenraad andHird, 1970; Klieve, 1991) and the large


28/297


intestine contents and faeces of horses(Alexander et al., 1970, Kulikov et al., 2007). Inall of these cases, the vast majority of observedVLPs belonged to tailed bacteriophages.No quantitative data characterizing themorphological diversity of the human gut

bacteriophages, however, has been published(in contrast to some animal-associated com-munities, as recently reviewed by Letarovand Kulikov, 2009).

Metagenomic studiesCurrently available data on the diversity ofthe non-cultured viral community of thehuman gut has been based mainly onmetagenomic data. The first metagenomicanalysis of the virome of a single specimen ofhuman faeces, collected from a 30-year-oldmale subject, was published by Breitbart et al.(2003). In their study, about two-thirds of thesequences obtained were database orphans;

among the rest, known viral sequencesconstituted 27%. The predominant viralgroup, judged by database hits, in humanfaeces were siphoviruses (bacteriophageswith long, non-contractile tails), which areprobably most prevalent in the majority ofnatural habitats (Weinbauer, 2004). Theestimated diversity of bacteriophages wasabout 1200 viral genotypes present in thesample.

Sequences related to eukaryotic virusescomprised only a minor fraction. This isconsistent with the fact that such particles arerarely seen in transmission electron micro-scopy images of faecal viral communities. It isinteresting that, in metagenome analyses ofRNA-containing VLPs extracted from humanfaeces, the sequences of plant viruses werehighly predominant (Zhang et al., 2006),indicating that ingestion of these particles

with food is much higher than the internalproduction of RNA-containing phages. Thelater metagenomic analysis of multiplesamples of human faeces used high-throughput sequencing technology (pyro-sequencing) and revealed very interestingfeatures of these communities.

In order to determine the impact of thegenetic background of the macro host on the

composition of the human intestinal viromeand the stability of the individual phagepopulations, Reyes et al. (2010) collectedsamples of faeces from four pairs of twinsand their mothers over a 1-year period. VLPswere extracted from these samples (32viromes in total were characterized) and theassociated viral metagenomes were se-quenced. About 85% of the sequences fromthe VLP metagenomes did not correspond toany known viruses, while most of the restmatched various known prophages andtemperate phages.

The bacterial diversity of the same set ofsamples was analysed by 16S rRNA genelibrary sequencing. The diversity of bacterialcommunities was estimated to be about 800species-level bacterial phylotypes, while thecomplexity of the phage community wasmeasured by two different approaches as 522773 (median 346) or 10984 (median 35)predicted virotypes. Both the abundance ofphages and the number of phage species per

bacterial species was therefore quite low incomparison with known, free-living bacterialcommunities. The idea of a temperate natureof predominant phage types in the analysedviromes was strongly supported by theidentification of a significant number ofsequences related to known bacteriophageintegrases, the phage-encoded site-specificrecombinases responsible for integration ofthe temperate phage genome into the host

chromosome.The similarities seen in the bacterial

communities studied by Reyes et al. (2010)strongly correlated with family links betweenthe subjects. In contrast, the VLP metagenomecomposition was highly individual andvaried considerably between the subjects,regardless of family relationships. At thesame time, the variability was low withinindividuals: over the time of the study, the

sequences of the VLP metagenomes werealmost stable. Moreover, the authors wereable to detect the dominant phage thatpersisted at high levels in one of theindividuals for an extended period of time

but showed no significant divergence ormutations in its genome.

The proposed low dependence of thephage populations in this environment on


29/297

12 A.V. Letarov

their success in competition for host bacteriamight facilitate long-term persistence of thosephage populations that happen to colonize aparticular niche first. Alternatively, if theassumption of high prevalence of temperatephages made in the above-cited studies iscorrect, then those temperate phages thatcolonized the genomes of bacteria that haveestablished their populations in an individualgut ecosystem may predominate. The highindividuality of the phage community indifferent subjects therefore may reflect thehistory of colonization of the infant gut

by bacteria, phages and phage lysogens(Breitbart et al., 2008).

What does not seem to be present issubstantial selective pressure acting on either

bacteria or phages, at least over the intervalsanalysed by metagenomic studies. The dataof Reyes et al.(2010), in particular, providedno indication of Red Queen dynamics(continuing change and adaptation of hostand parasite; Weitz et al., 2005) in human

faecal microbial communities, althoughperhaps the resolution of metagenomicstudies could be unable to provide such asignal. Thus, the bacteriophages in thisenvironment do not seem to exert a sufficientinfluence on the dynamics of bacterialpopulations in the gut to result in bacterialresistance. Abedon (2011) suggested that thissort of low phage pressure despite ongoingphage presence may be the norm, given

bacterial persistence in environments pre-dominantly as biofilms (see also MacFarlaneet al., 2011).

This concept of human intestinal phageecology is in good agreement with the data ofCaporaso et al.(2011) who used a somewhatdifferent approach for analysis of themetagenomic data. These researchers studied26 viral metagenomes of human faecescollected from 12 individuals (one to four

samples per individual). They compared thecomposition of these viromes with theviromes obtained from a variety of free-livingcommunities. The authors found that thedistances between the individual viromes ofhuman faeces were higher than between thesamples of related free-living communities.They also confirmed the predominance oftemperate phages and the absence of

observable phage/bacteria co-evolution in thehuman gut.

In agreement with these observations,Minot et al. (2011) reported that, followinganalysis of metagenomic sequences of humanfaecal viromes collected from five peopleover a 1-week period, they did not detect anysigns of Red Queen dynamics in thesesystems. The viromes studied also showedhigh individual variability; however, 1-weekdiet interventions (as low fat/high fibre orhigh fat/low fibre diets) led to an increase insimilarity of the viromes between subjects fed

the same diet. This may indicate that, inhumans, alterations in virome compositionfollow the composition of the microbiome.This may also be interpreted as an argumentin favour of the hypothesis of a high impactof lysogen induction in generations of phageVLPs in the human gut. Minot et al. (2011)analysed the viral (phage) contigs assembledfrom their data for genetic relatedness (usingvarious criteria, described in detail in the

paper) for temperate phages or prophagesand found about 14% of them to be potentiallytemperate. However, this value was only thelowest estimate, as not all fragments of thetemperate phage genomes would fit thecriteria applied.

Culture-based analyses

When considering the culture-based datadealing with phage indication and quanti-fication in natural environments, one mustalways remember that the majority of theseviruses tend to be specific for only a subset ofthe strains found within bacterial species(Hyman and Abedon, 2010). Thus, the phagetitre obtained with any given indicator

bacterium reflects only a fraction of the phageparticles, that is, those able to infect that

particular bacterial strain (the phenomenon isknown in phage ecology as the great plaquecount anomaly; see Weinbauer, 2004).Moreover, some of the phages that are able toinfect the bacterial strain used neverthelessmay display reduced plating efficiency, atleast during the first passage on a host strain.This effect may be due to the action ofrestriction-modification systems, as well as of


30/297


many other host resistance mechanisms(Labrie et al., 2010), and can lead to serious several orders of magnitude under-estimations of the real abundance of aparticular phage strain in a sample. In eachcase, one therefore has to consider the choiceof the bacterial host used for phagequantification, especially if comparison ofphage titres in samples collected fromdifferent subjects is involved.

Notwithstanding these caveats, the dataof culture-based studies are in generalagreement with the hypothesis of a low

phage impact on bacterial populations inhuman intestinal ecology. Furuse et al.(1983),for example, found that faecal coliphage titresin healthy humans are low, and the pools offree virions in faeces are represented mainly

by temperate phages. In contrast to healthypeople, the phage populations in somepatients with internal and leukaemic diseasescontained a substantial fraction of virulentphages (including T-even-related phages), as

well as an increased faecal coliphage back-ground (Furuse et al., 1983). In severalpatients, phage titres increased when theseverity of the clinical symptoms increased.

Consistent results of phage isolationfrom Bangladesh paediatric patients withdiarrhoea were reported by Chibani-Chennoufiet al.(2004b). About 19% of acutediarrhoeal stools yielded quite divergent T4-related phages infecting the E. coli K803

laboratory indicator strain. The detection ofphages in the stools from convalescentpatients was less frequent. It is interestingthat other E. coli strains used for phageisolation from the same set of samplesyielded completely different Siphoviridaecoliphages. The occasional presence of bothtemperate and virulent coliphages as well assome culturable phages infecting other hostshas also been reported in the literature. In

most cases, these viruses were present inlow titres in healthy subjects (Dhillon et al.,1976; Havelaar et al., 1986; Cornax et al.,1994; Grabow et al., 1995; Calci et al., 1998;Gantzer et al., 2002; Schaper et al., 2002; Coleet al., 2003; Lusiak-Szelachowska et al., 2006;see also Letarov and Kulikov, 2009, forreview).

Temperate phages in the gut

The results of metagenomic analysis per-

formed by Reyes et al.(2010) suggested thatthe main source of free phage particles in thehuman intestine are not productive cyclesimmediately following phage adsorption, asseen in the majority of other naturalcommunities (see Weinbauer, 2004 forreview), but instead are the induction of

bacterial lysogens that can occur due tostarvation, such as of bacteria in faeces. Thisconclusion is similar to that of Furuse et al.

(1983) who first suggested that, in healthypeople, most of the released phage particlesare produced by induced lysogenic cells andtherefore that phage multiplication may havea limited impact on the intestinal coliformmicroflora.

Interestingly, the vast majority of tem-perate coliphages isolated by Dhillon et al.(1980) from free phage particles of humanand animal faeces belonged to the lambdoid

group, while phages obtained from culturedbacterial lysogens were all immunologicallyP2-related. These data may be explained by ahigher frequency of induction of the lambdoidprophages present in the studied microbiomesthan that of P2-like prophages. Alternatively,conditions favourable for phage multi-plication in the lytic cycle may occur in thehealthy human gut, perhaps in spatiallylimited sites or over short periods of time,

and this multiplication could allow thereproduction of some phages independent ofthe induction of lysogenic bacteria. Theabove-cited results could thus reflect thepreferential success of the lambdoid phageswithin hypothetical windows for lytic cyclemultiplication.

F-specific phages

RNA-containing F-pilus-specific (F-RNA)

coliphages (members of the familyLeviviridae)have been found in human as well as in someanimal wastes and show certain speciesspecificities. These phages can be subdividedinto multiple genetic groups that can be alsodistinguished serologically. The incidence ofthese serotypes varies significantly amongspecies: horse faeces, for example, rarely


31/297

14 A.V. Letarov

contain them, while the faeces of more than70% of chickens contain high titres of thesephages (105107 plaque-forming units (PFU)g1). Only about 1020% of human faecescontains F-RNA coliphages, but the occurrenceof group II and III F-RNA phages in thesesamples is much higher (80% of all isolates)than in animals, where groups I and IV areprevalent at the same level (Furuse et al., 1978;Havelaar et al., 1986; Schaper et al., 2002; Coleet al., 2003). No coherent explanation for thisgroup specificity has yet been suggested. SeeGoodridge and Steiner (Chapter 11, this

volume) for further discussion.

Possible limitations on phage replication in

the gut

Among the environmental factors that maycontribute to the protection of bacteria fromphage atack in the human gut are chemicalsthat can inhibit phage infection as well as thespecific physiological state of bacterial

populations. Bile salts and carbohydrateshave been shown to inhibit the adsorption ofa variety of coliphages (Gabig et al., 2002).This effect is suppressed if Ag43 protein,mediating cell aggregation and atachment, ispresent on the surface of E. coli cells. Theexpression of Ag43 is regulated in a phase-variation manner. Thus, phages may, undersome circumstances, select against increased

biofilm formation. For Bacteroidesphages, the

addition of bile salts to the medium had anopposite effect, that is, its addition improvedphage plating efficiency (Araujo et al., 2001).Some food-derived phage inhibitors may also

be present in the intestine, at least occasionally.For example, Swain et al.(1996) demonstratedthat tannic acid at physiological concen-trations may inhibit bacteriophage replicationin the rumen of ruminants. This is consistentwith the observations of de Siqueira et al.

(2006) that tea infusions can inactivatephages. Similar compounds are also ingested

by humans with certain types of food andcould therefore contribute to low phage lyticactivity in their gastrointestinal tract.

Growth in biofilms and on the surfacesof mucosa and food particles may alsocontribute to bacterial anti-phage protection,although this potential has at best only been

demonstrated inconclusively (Abedon, 2011).There is also some evidence that the E. colipopulation in the mouse gut lumen may bestarving and the actively replicating popu-lation in fact may be limited to microcoloniesfound on the mucosal surface (see Chibani-Chennoufi et al., 2004a, and referencestherein). It is not clear, however, if this modelof E. coliecology in the gut can be extended tohumans and other large animals (Letarov andKulikov, 2009).

Phage propagation predominantly vialysogenic cycles or associated prophage

induction, resulting in a reduced impact ofphage infections on the microbial ecology ofthe human gut, may not be unique amongmammalian species (reviewed in detail byLetarov and Kulikov, 2009; Clokie et al., 2011).For example, atempts to isolate bacterio-phages from dog faeces using indigenouscoliform strains (Ricca and Cooney, 2000)were largely unsuccessful. Over 500 in-digenous coliform strains isolated from six

specimens of dog kept in private homes didnot detect phages in the same samples, andonly one of these samples yielded phages onthe laboratory E. coliC strain. In 16 dogs froma kennel, however, coliphages were detectedat variable titres from 0 to 107 PFU g-1. Theauthors suggested that a low abundance ofcoliphages in home-kept dogs may be due toisolation from other dogs and too cleanliving conditions. Recontamination by faecal

microbes is also limited in humans, whichcould suggest that phages able to overcomeexisting barriers for replication in the gut areseldom acquired by humans. Perhaps theoccurrence and possible impact of phages inhumans would be higher if we lived in a lesscivilized manner. It would be interesting tocompare the phage prevalence in humansubpopulations living in the same area butdiffering in their quotidian hygienic practice

(for example, in Bangladeshi state employeesand peasants).

The opposite situation to that of highlydomesticated animals seems to exist in thehorse intestinal ecosystem. In our studies(Golomidova et al., 2007; Kulikov et al., 2007;reviewed by Letarov and Kulikov, 2009), thestructure of the indigenous coliform com-munity was compatible with high phage


32/297


pressure. Of the dozens of horse coliphageisolates that we were able to identify, all

belonged to known groups of virulent phages(T-even-related, T5-related, rv5-related, N-related, similar to Salmonella SETP3 phage,Felix 01-related, similar to Caulobacter phageCd1, coliphage K1F-related and others;Golomidova et al., 2007, and our unpublisheddata). The investigation of morphologicaldiversity of uncultured horse faecal VLPs(Kulikov et al., 2007) has demonstrated thehigh abundances of a diversity of large

bacteriophages that are very likely to be

virulent. The ecology of intestinal bacterio-phages thus appears to be highly dependenton host-species-specific features of digestivetract physiology along with the generalenvironment.

The vagina

The vaginal microbiota of healthy, fertile

women is normally dominated by vaginallactobacilli (Hillier, 2008), usually comprisingmore than 70% of the total bacterial count inthis environment. These bacteria are believedto play an essential role in colonizationresistance against pathogenic bacteria andfungi by contributing to the establishment ofa low vaginal fluid pH (Servin, 2008; Linhareset al., 2011), along with frequently producinghydrogen peroxide (Martin and Suarez,

2010). Lactobacilli also possess more-specificantagonistic activities against pathogenic

bacteria (Servin, 2008). Partial loss of theindigenous vaginal lactobacilli in combinationwith polymicrobial anaerobic overgrowth onthe vaginal mucosa are characteristic featuresof bacterial vaginosis (BV) one of the mostcommon reasons women seek medical help(Sobel, 2000). The density of Lactobacilluscolonization of the vaginal mucosa is quite

high at 106107CFU per vaginal swab. Thus,one can assume that episodes of mass killingof lactobacilli by phages may occur, althoughatempts to detect free bacteriophages invaginal swabs have been unsuccessful (Kiliet al., 2001). At the same time, however,lysogenic strains of lactobacilli were shown to

be prevalent in this environment (Kili et al.,2001).

Although the vagina is subject torelatively low levels of exchange of bacteriaand viruses with the external environment,except as mediated by sexual activity, asudden breakdown of Lactobacillus popu-lations is frequently observed in examinationsof vaginal swabs from clinically healthywomen. Some individuals also developanaerobic bacterial vaginosis syndrome, whenlactobacilli are replaced by anaerobic bacteriasuch as Gardnerella vaginalis, Prevotella,Porphyromonas and Mobiluncus species. Thiscondition has the epidemiology of a sexually

transmited disease (Verstraelen et al., 2010). Itwas proposed by Blackwell (1999) that thecausative agent triggering the breakdown ofthe normal vaginal microbiota might be a

bacteriophage atack. I speculate that suchevents may take place if a lysogenic strain ableto produce a phage infectious for the majorresident strain(s) is acquired (Letarov andKulikov, 2009). A similar scenario wasmodelled on in vitroE. coli populations both

experimentally and mathematically by Brownet al.(2006).

The results of our recent study of thediversity of individual populations of vaginallactobacilli at the strain level indicate thatindividual populations of these bacteria arenormally dominated by a single strain, whichcould be resolved by repetitive element PCRfingerprinting (A. Isaeva, E. Ilina, A.Bordovskaya, A. Ankirskaya, V. Muraviouva

and A. Letarov, unpublished data). This iscompatible with the proposed mechanism ofthe rapid drop of lactobacilli counts due tomass phage-mediated lysis, as almost all ofthe cells in individual populations would(presumably) represent the same phage-susceptibility type. In such a case, a singlephage liberated by an invading lysogenicstrain potentially could reduce the residentpopulation density and make enough room

for a newly acquired lysogenic strain (andsecondary lysogens formed in some cells ofthe resident strain). In contrast to the resultsof Kili et al.(2001) described above, however,the search for inducible lysogens in ourcollection of vaginal isolates yielded nostrains producing a phage viable on any otherisolate tested (A. Isaeva and A. Letarov,unpublished data). An observation similar to


33/297

16 A.V. Letarov

ours was recently reported for a Spanishsubpopulation of women (Martin et al., 2009).These authors suggested that high con-centrations of H2O2 produced by vaginallactobacilli select for prophage-cured lineagesthat do not suffer from the mortality induced

by elevated phage induction due topermanent oxidative stress characteristic ofthis environment. Thus, the vaginal eco-system may be considered as one in whichphages exert, ecologically, a low impact onthe resident bacterial populations. Thepossible exceptions, however, require further

investigation.

Direct Interaction of the Virome withthe Macro Host

The phage ecology of human-associatedmicrobial systems generally should beconsidered as a tripartite interplay

bacteriophages

Documents